Robo-Advisory - Investing in The Digital Age (2021 Peter Scholz)

PALGRAVE STUDIES IN
FINANCIAL SERVICES TECHNOLOGY
Robo-Advisory
Investing in
the Digital Age
Edited by
Peter Scholz
Palgrave Studies in Financial Services Technology
Series Editor
Bernardo Nicoletti
Rome, Roma, Italy
The Palgrave Studies in Financial Services Technology series features orig-
inal research from leading and emerging scholars on contemporary issues
and developments in financial services technology. Falling into 4 broad
categories: channels, payments, credit, and governance; topics covered
include payments, mobile payments, trading and foreign transactions, big
data, risk, compliance, and business intelligence to support consumer and
commercial financial services. Covering all topics within the life cycle of
financial services, from channels to risk management, from security to
advanced applications, from information systems to automation, the series
also covers the full range of sectors: retail banking, private banking, cor-
porate banking, custody and brokerage, wholesale banking, and insurance
companies. Titles within the series will be of value to both academics and
those working in the management of financial services.
More information about this series at

http://www.palgrave.com/gp/series/14627
Peter Scholz
Editor
Robo-Advisory
Investing in the Digital Age
Editor
Peter Scholz
Hamburg School of Business
Administration
Hamburg, Germany
ISSN 2662-5083 ISSN 2662-5091 (electronic)

Palgrave Studies in Financial Services Technology
ISBN 978-3-030-40817-6 ISBN 978-3-030-40818-3 (eBook)
https://doi.org/10.1007/978-3-030-40818-3
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher,
whether the whole or part of the material is concerned, specifically the rights of translation,
reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in
any other physical way, and transmission or information storage and retrieval, electronic
adaptation, computer software, or by similar or dissimilar methodology now known or
hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information
in this book are believed to be true and accurate at the date of publication. Neither the
publisher nor the authors or the editors give a warranty, expressed or implied, with respect
to the material contained herein or for any errors or omissions that may have been made.
The publisher remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
This Palgrave Macmillan imprint is published by the registered company Springer Nature
Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
For Gwendolyn and Nola
FOREWORD
Man or machine?
This is one of the key questions of our time. Albeit not new in its
historical context, it is once again being heavily discussed—now in the
context of the “labor market”. Machines make certain tasks superfluous,
whilst at the same time creating new professions and thus new jobs. And
yet, one question remains: Can the number of jobs created compensate
for the ones lost? What will the future look like for our work and our
lives? When it comes to assembly line robots, automation technology or
artificial intelligence, people today tend to have a somewhat negative bias
of discomfort, mistrust and perhaps even fear toward machines. And on top
of it all, they are now asked to entrust their financial security to a machine.
As always, though, there are two sides to the coin: As a matter of fact,
we all enthusiastically use smart machines every day. We can no longer
imagine life without a laptop, smartphone, Thermomix or an automated
vehicle kitted out with driver-assisting technologies to increase road safety
and convenience. So why not entrust our money to smart machines?
Especially when they offer professional asset management with little
effort for even small investments. “Professional increase of money for
everyone”—such robo-advisor solutions may at first sound a bit like a
socialistic approach attempting to penetrate the exclusive world of asset
management, until then only available to people with at least six- or
seven-figure investment sums. When they entered the market a few years
ago though, providers announced no less than a technical and content
revolution, the disruption of traditional asset management and financial
advice. Investments would grow in a cost-effective, simple and automated
vii
viii FOREWORD
way, globally diversified and secured 24/7, with a state-of-the-art risk

management.
The “financial advisor in the machine” acts without emotions, solely
following an algorithm. In theory, this should eliminate classical behavioral
finance errors, which private and professional investors alike are often
subjected to. Ideally, ruling out the principle of the herd instinct—buying
too late and selling too early, panicking, and above all falling at risk
of emotionally charged decision making. “Simply ingenious”, as many a
financial blog dubbed this new option. Sounding too much like an out-and-
out benefit for mankind? As it goes, there is of course a concise business
strategy behind it: using modern technology to gain market shares from
traditional asset managers.
The robo-advisor technology in principal consists of systems profes-
sional asset managers have been using for more than 20 years to recognize
market movements early on and thus act on them in a timely manner.
At its core, it is about gaining a competitive advantage of information.
Financial market professionals have far more options than private investors
when it comes to observing and analyzing markets and the occasionally
occurring, irrational behavior of their participants. Computer programs
play an essential role in this. They recognize patterns and anticipate
developments, thus making it possible to act at once, instead of having
to react as victims of market movements.
The underlying principle is the same as that which was practiced
centuries ago on the London Stock Exchange. Back in those days, there
was always an uncertainty as to whether cargo vessels would make their
way to England. If, for example, they had passed the Cape of Good
Hope unscathed, observers would report this back home. The information
advantage thus gained had a direct impact on share prices. Today, satellite
images and data centers are used for this purpose. Then as now, the correct
evaluation of data is fundamental.
This may well be the reason why even professionals can by no means
always beat the market, even if such computer systems have been at their
disposal for a long time. This gives rise to questions that are of course of
essential importance to the entire financial sector: Can machines alone be
considered a cure-all? Are purely technologically controlled investments
in fact a more lucrative option? Can they reduce risks and increase
opportunities? Will bank advisors and asset managers become redundant
in the foreseeable future? Are robo-advisors really the revolution? And if
FOREWORD ix
so, why have only a handful of robots been able to establish themselves on
the market so far?
This book is the first to provide comprehensive answers to these
questions in a fundamental, decisive, detailed and nuanced way. It clarifies
the basics, the technology and the tactics behind those clever, financial
machines, gives insights into their previous track record to date and much
more. Looking ahead, it provides a preview of what is and may be yet
to come. As a matter of fact, so far only a relatively small percentage
of the global investment community have more or less relied on robo-
advisors, depending on their respective culture. It is also a fact that we
are only at the beginning of development. We have all borne witness to
how exponentially fast things can move forward. One such example is the
evolution of smartphones—which by the way have been around for just a
little longer than robo-advisors.
The business model of banks and asset managers has remained
unchanged and immovable for centuries. Given the possibilities arising
from new technologies, FinTechs such as robo-advisors have created a
completely new situation for themselves, in which no stone seems to
have been left unturned. The financial business world of today is clearly a
hugely exciting place and by no means solely for professionals working in
or dealing with it.
“Man or machine?”—This question really does concern everyone. We
live in times of rapidly exponential and radical change. It is a truth
universally acknowledged that fear is a bad advisor. Confidence, on the
other hand, makes all progress possible and tangible.
Therefore, the answer to this fundamental question can only be:
“Man with machine.”
Munich, Germany Jochen Werne

March 2020
PREFACE
Digitalization—it is a megatrend that has accompanied our economy for

many decades now. Although one might think that the smartphone and
mobile internet have created digitalization, it roots back to Konrad Zuse,
who invented the first working computer in 1941. Since then, compu-
tational power has developed at a rapid pace, like predicted by Moore’s
Law, and accelerated the automatization tendency of industrialization.
The introduction of personal computer in our daily working routine is
considered as industry 3.0 and was strongly driven by Apple and Microsoft.
Only a few years later, during New Economy and internet bubble around
the turn of the millennium, the internet conquered the markets and showed
what potential lies within computer networks. This has been the starting
point for the industry 4.0. Today, smartphones and mobile internet allow
users to enjoy the benefits of global computer networks anytime and
anywhere. The FAANG companies, Facebook, Amazon, Apple, Netflix,
and Google hence are dominating the economy and have disrupted many
industry sectors. Latest since the financial crisis from 2008, due to the loss
in trust in the financial system, the banking sector as well cannot hide from
the digital age anymore and needs to adopt to the new consumer behavior,
which expects mobile access to banking services 24/7 without restrictions
in location. Fintechs already have discovered this field and create innovative
products and services.
Since the New Economy, I am closely watching the digital transforma-
tion of the banking sector. From the first pure online banks in the nineties
until today, when Fintechs and cryptocurrencies are trying to establish at
financial markets: the development is tremendous and it is a privilege to be
xi
xii PREFACE
a part of it. But the digital evolution also comes with many challenges.
In my role as a professor for banking and financial markets, I have to
prepare and enthuse my students for new requirements in the field, which
includes the natural application of software and basic programming skills.
This is crucial, not because they are expected to really code but to be able
to explain to the programmers what to implement and to understand at
least the core of the problems the IT experts are facing, like data integrity
and the configuration of interfaces. IT projects will be an essential part of
banks in future: the application of more and more software solutions needs
more bankers who are able to precisely define and describe processes: for
example, how a machine can provide excellent investment advice.
The digitalization of the banking sector also opens new and interesting
research opportunities. Together with my MBA alumnus Michael Tertilt,
we have analyzed how robo-advisors measure the risk profile of investors
and how they derive a portfolio recommendation from this information.
We have been surprised that they seem to ask way more questions than they
really use for the process of deriving a portfolio for the investors. In sum,
we think that the quality of the digital advice is not worse than the average
level in the market, but we still see much potential for developments in
the future. As I presented our findings at the 54th Meeting of the Eastern
Finance Association in Philadelphia (USA) in April 2018, I was approached
by Palgrave Macmillan and asked if I want to write a book about robo-
advice. I was immediately thrilled by the idea, but I also recognized the
challenge to write a book about such a complex topic since robo-advice is
an interdisciplinary subject. Hence, the idea was born to act as an editor and
to invite different experts in the field to contribute to the book. Since robo-
advisory is a topic that was born in the business life and not in academia,
the current knowledge is widely distributed; therefore, we need to build a
bridge in the book from descriptions of best practice to considerations of
applied sciences. This is why I divided the book in four different parts.
The first part of the book explores the current status quo of robo-
advisory: in Chap. 1 we define the term robo-advice and analyze its clients
and market (potential); we describe the wealth management process in
which robo-advisory aims to participate; we briefly look back in the history
of robo-advisory; and we analyze the past performance of robo-advisors—
probably the ultimate long-term criteria for clients. Furthermore, the
current status quo based on the academic literature is surveyed in Chap. 2.
The second part of the book deals more with the implementation of
robo-advisory. Risk profiling is one of the crucial elements of the advisory
PREFACE xiii
process and a component which can really add value for clients. But as
we learned from our paper (Tertilt and Scholz 2018), it is difficult to
efficiently measure the investor’s risk profile. Hence, we include Chap. 3,
which explains how the insights from psychometrics can help to improve
risk profiling of robo-advisors. As we will learn in Chap. 1, the portfolio
management process of robo-advice still relies widely on investment com-
mittees and there seems generally more potential for human interference
than expected. In Chap. 4, we will discuss and observe the potential
implications and emotional biases that may occur based on not wholly
automatized investment process. But we will also see that, especially US
robo-advisors, apply the concept of goal-based investing and thus seem
to have a more client-centered angle. Chapter 5 introduces the concept
of rule-based investing and will show how human impact in the portfolio
management can be further reduced since robo-advice is privileged to apply
fully automated processes. Another important aspect is the acceptance
of human clients to be advised by machines, which will be addressed in
Chap. 6 Moreover, regulation is also an important part of the investment
advisory business. Especially since the financial crisis, the regulation has
been severely tightened and the development of robo-advisory has further
created the need to adapt regulation; meanwhile regulation has become a
field for true experts. Since the regulation can be very different, we include
two chapters on the topic: one for our US readers (Chap. 7) and one for our
European/German readers (Chap. 8). Although the European regulations
can differ in its intricacies from country to country, many rules are defined
on European level and have been transferred to national law.
The third part of the book contains case studies of best practice. The
first contribution in this part (Chap. 9) is a whitepaper by BearingPoint
and describes from the perspective of experienced consultants the most
important elements which need to be considered if a robo-advisor is
incorporated. The second chapter (Chap. 10) is an extensive report by the
traditional bank Hauck & Aufhaeuser and explains the different challenges
they faced and the solutions they found in their project to introduce a
robo-advisor for their clients. Both contributions give valuable insights in
the practical process of digital transformation.
The fourth part looks ahead to the near future of robo-advisory. It
seems to be a natural assumption that robo-advisory is a perfect field for
big data and artificial intelligence (AI). Today, however, these technologies
are not widely spread amongst robo-advisors. Chapter 11 introduces the
topic of AI and describes the requirements and possibilities of robo-advice
xiv PREFACE
in this field. Whereas the client communication has large potential for AI
applications, the improvement of the portfolio management, especially in
forecasting, seems to be rather limited. In this regard, Chap. 12 analyzes
the marketing impact and data gathering potential of social networks for
the robo-advisory business. Finally, Chap. concludes by summing up the
most important factors, which seem to determine the success of a robo-
advisor in the future.
We created the book for a diverse readership: from students, who want
to speed up their knowledge in digital transformation and investment
advisory topics, to academics, who want to gather a profound overview
about robo-advice, to practitioners, who may seek intellectual support for
how to implement a robo-advisor, and last but not least, to those investors,
who are highly interested in the topic and want to learn more about the
technology and process of robo-advice. I am grateful for working together
with all contributing authors and I am confident that we created a book
which is enriching for our readers.
Hamburg, Germany Peter Scholz

May 4th, 2020
REFERENCE
Tertilt and Scholz (2018) Michael Tertilt and Peter Scholz. 2018. “To
Advise, or Not to Advise—How Robo-Advisors Evaluate the Risk
Preferences of Private Investors”. The Journal of Wealth Management
21(2): 70–84.
ACKNOWLEDGEMENTS
A large project like this book on robo-advisory, where so many perspectives

are brought together, can never be achieved alone. Therefore, I am
grateful to all authors that contributed their expertise to create such a
differentiated picture of such a complex topic. Dear Alexander D. Beck,
Ana-Maria Climescu, Melanie L. Fein, Joachim Goldberg, Goetz Greve,
David Grossmann, Christian Hammer, Christian von Keitz, Sinan Kruecke-
berg, Frederike Meyer, Monika Mueller, Yvonne Quint, Paul Resnik, Jan
Rocholl, Thorsten Ruehl, Madeleine Sander, Craig Saunders, Theodor
Schabicki, Soeren Schroeder, and Michael Tertilt: thank you so much
for your valuable contributions! I really appreciate your effort and time!
Please allow me to extend a special thank you for the smooth teamwork
to my co-authors Joachim Goldberg, David Grossmann, and Michael
Tertilt. Another big and special thank you goes to Theodor Schabicki from
BearingPoint, who permitted to use their whitepaper as a case study in this
book! I am really looking forward to further cooperation with you!
But support is also necessary behind the scenes. I would like to thank
Jochen Werne in particular, since he made contacts with other authors, who
then contributed. Moreover, I would like to thank you for your inspiring
foreword. To know such a highly sophisticated networker is a real privilege.
Another privilege I really appreciate is the cooperation with my for-
mer Ph.D. students David Grossmann and Sinan Krueckeberg. We went
together through many research projects in the past and yet teamed up
again. Both of you have contributed as authors, and both of you have
provided constructive and valuable feedback to my contributions. Thank
you very much for your friendship and support!
xv
xvi ACKNOWLEDGEMENTS
Furthermore, I owe a big thank you to my student assistant Laura

Andrade Pardo, who assisted me in the LATEX transfer and proper refer-
encing of the bib-file. My colleague Franziska Buttler provided assistance
in organization and administrative work: thank you for keeping me free of
many administrative ties! The book is written in LATEX: here, another big
thank you goes to my friend Lars Geiger for his support in troubleshooting
if the code did not work and the error seemed to be impossible to find. He
really is a lender of last resort!
I am also grateful to my publisher Palgrave Macmillan, who contacted
me and gave me the opportunity to edit a whole book on such an
interesting and up-to-date topic. Thank you in particular for the patience
and support during this project. Although it has been a true peak in
workload, I really enjoyed to deep-dive into the field of robo-advice!
Support from my employer HSBA is gratefully acknowledged.
Finally, and most importantly, I am thankful to my wife Britta for her
support and understanding. Above all, I am most grateful to my wonderful
daughters Gwendolyn and Nola for being the light in my life.
CONTENTS
Part I Status Quo of Robo-Advisory 1
1 Robo-Advisory: The Rise of the Investment Machines 3

Peter Scholz and Michael Tertilt
1.1 Introduction 4
1.2 Robo-Advisory as Part of the Wealth Management
Process 4
1.3 A Brief History of Robo-Advisory 7
1.4 Clients and Market for Robo-Advice 8
1.5 Performance of Robo-Advisors 12
1.6 Robo-Advice in a Nutshell 15
References 16
2 Situating Robo-Advisory 21
Sinan Krueckeberg
2.1 Introduction 21
2.2 What Are Fintech and Robo-Advisory, Really? 22
2.3 Differentiation: Strengths and Weaknesses 23
2.4 Impact on Incumbents 26
2.5 The Path Ahead 28
References 29
xvii
xviii CONTENTS
Part II Implementation of Robo-Advisory 33
3 Risk Preferences of Investors 35

Monika Mueller, Paul Resnik, and Craig Saunders
3.1 Introduction 36
3.2 Risk Profiling Explained 37
3.2.1 Financial Risk Tolerance 38
3.2.2 Risk Capacity 39
3.2.3 Risk Required 40
3.2.4 Risk Profiling Is Part of Knowing Your Client 40
3.3 Choosing a Valid and Reliable Risk-Tolerance Test 40
3.3.1 Psychometrics Versus Gambles-Based Testing 41
3.3.2 The Science of Prospect Theory 42
3.3.3 The Science of Psychometrics 43
3.4 Risk Profiling Is Harder for Robo-Advisors 45
3.5 Mapping Risk Profiles to Investments 47
3.6 Conclusion 48
References 49
4 Robo Economicus? The Impact of Behavioral Biases on

Robo-Advisory 53
Peter Scholz, David Grossmann, and Joachim Goldberg
4.1 Introduction 54
4.2 Home Bias 55
4.2.1 Does Robo-Advisory Fall for Home Bias? 56
4.3 Mental Accounting 58
4.3.1 Does Robo-Advisory Care About Mental
Accounts? 59
4.4 Overconfidence 61
4.4.1 How Confident Are Robo-Advisors? 63
4.5 Summary 65
References 66
5 Quant Models for Robo-Advisors 71

Thorsten Ruehl
5.1 Introduction 72
5.2 What Strategies Are Suitable for Robo-Advisory? 73
CONTENTS xix
5.3 What Quantitative Approaches Does the Robo-Advisory

Model Offer? 75
5.3.1 Maximization of the Diversification Effect 76
5.3.2 Equal Distribution of Risks to the Investment
Instruments Contained in the Portfolio 79
5.3.3 Risk Minimization 80
5.3.4 Methods Based on Return Forecasts 81
5.4 Dealing with Risk Targets 84
5.4.1 Adherence to Lower Value Limits 84
5.4.2 Specification of a Risk Preference by Choosing
a Target Investment Period 85
5.5 Return Targets and Risk-Bearing Capacity: Need for
Information 86
5.6 Requirements for the Investment Universe and
Instruments 88
5.7 Customization by Investors 89
5.8 Summary 91
References 91
6 Analysis of the Use of Robo-Advisors as a Replacement

for Personal Selling 93
Goetz Greve and Frederike Meyer
6.1 Introduction 94
6.2 Robo-Advisors as Customer–Salesperson Interaction
Technologies 94
6.3 Theoretical Background and Research Propositions 96
6.3.1 Trust 96
6.3.2 Perceived Risk 97
6.3.3 Psychological Reactance 97
6.3.4 Perceived Use 98
6.4 Experimental Design and Data Collection 99
6.5 Results 99
6.5.1 Data Preparation and Manipulation Checks 99
6.5.2 Hypotheses Test 100
6.6 Managerial Implications and Limitations 101
References 102
xx CONTENTS
7 Regulation of Robo-Advisers in the United States 105

Melanie L. Fein
7.1 Introduction 106
7.2 Investment Advisers Act of 1940 106
7.2.1 Registration of Robo-Advisers 106
7.2.2 Anti-fraud Provisions 108
7.2.3 Fiduciary Duty 108
7.2.4 SEC Staff Guidance 112
7.2.5 SEC Investor Alert 118
7.2.6 Supervision 121
7.3 Securities Exchange Act of 1934 121
7.4 Investment Company Act of 1940 122
7.5 Employee Retirement Income Security Act of 1974 125
7.6 State Regulation 129
7.7 Conclusion 129
References 130
8 Regulation of Robo-Advisory in Europe and Germany 133

Christian Hammer
8.1 What Are the Activities for Which a Provider Needs a
License? 134
8.1.1 Product-Related 135
8.1.2 Service-Related 137
8.1.3 Liability Umbrella as an Alternative to
Holding a License 143
8.2 Regulatory Requirements for Fintech Firms and
Robo-Advisors: Practical Issues 144
8.2.1 Information Obligations 145
8.2.2 Investment Broking and Execution Only
Transactions 146
8.2.3 Duty of Care Obligations in Investment
Advice and Portfolio Management 148
8.2.4 Practical Importance and Border Cases in
the Robo-Advisory Universe 149
8.2.5 Reallocation 152
8.3 Onboarding Obligations Under GWG 153
8.3.1 Organizational Requirements 153
8.3.2 General Due Diligence Requirements 154
CONTENTS xxi
8.4 Conclusion of Contract 155

8.5 Data Management 156
8.5.1 Profiling 157
8.5.2 Use of Cloud Services 158
8.6 Conclusion 159
References 159
Part III Case Studies of Robo-Advisory 161
9 (Re-)Launching a Robo-Advisor as a Bank 163

Theodor Schabicki, Yvonne Quint, and Soeren Schroeder
9.2 Background 165
9.3 Key Challenges and Considerations 166
9.4 Implementation of a Robo-Advisor 168
9.4.1 Strategic Framework and Parameters 168
9.4.2 Robo-Advisory Specification 170
9.4.3 Evaluation Criteria for Provider Selection 173
9.4.4 Implementation of the Robo-Advisor 176
9.5 Evaluation of Existing Robo-Advisors 176
9.5.1 Evaluation Criteria 178
9.5.2 Options for Action 182
9.6 Future Trends 183
9.6.1 Digitalisation Requires Individualisation 184
9.6.2 Marketplaces and Ecosystems 184
9.6.3 Start Small to Go Big 185
9.6.4 Once a Competitor, Now an Ecosystem
Partner 185
10 How Can Robo-Advisory be Implemented and

Integrated into Existing Banks? 187
Ana-Maria Climescu, Christian von Keitz, Jan Rocholl,
and Madeleine Sander
10.1 Background 188
10.1.1 Introduction 188
10.1.2 Client Needs 190
xxii CONTENTS
10.1.3 Market Analysis 193

10.2 The Implementation and Integration Process 197
10.2.1 Strategy 199
10.2.2 Product 205
10.2.3 Marketing and Sales 206
10.2.4 Regulatory Requirements 213
10.2.5 Technical Platform 217
10.3 The Real Question Is: “How Can the Bank Use the
Latest Technology to Expand its Range of Services in
the Interest of the Client?” 219
References 221
Part IV The Future of Robo-Advisory 225
11 The Role of Artificial Intelligence in Robo-Advisory 227

Alexander D. Beck
11.1 Introduction to Artificial Intelligence 228
11.1.1 Historical Development of AI in the
Investment Management Industry 229
11.1.2 What Can be Expected from AI Today? 230
11.2 Value Potential for AI Along the Core Processes of
Robo-Advisory 231
11.2.1 Customer Data and Collection Strategies 231
11.2.2 Marketing, Sales, and Service 233
11.2.3 Matching Customers to Portfolios 236
11.3 AI Supporting Portfolio Management 237
11.4 The Way Forward 241
References 242
12 What Role Do Social Media Play for Robo-Advisors? 245

Ana-Maria Climescu
12.2 Social Media vs. Robo-Advisor 247
12.3 Robo-Advisor Feat. Social Media 249
12.4 Social Advisory 251
12.4.1 Social Media as Marketing and Sales Support 252
CONTENTS xxiii
12.4.2 Social Media as an Alternative to

Robo-Advisory 253
12.4.3 Social Media as Potential Cooperation
Partner for Robo-Advisors 253
References 254
13 Success Factors for Robo-Advisory: Now and Then 257

Madeleine Sander
13.2 Growth Factors Identified 260
13.2.1 No. 1: Overall Market Growth 260
13.2.2 No. 2: Replacing Existing Products 260
13.2.3 No. 3: Attracting Previously “Uninvested”
People to the Financial Markets 261
13.3 Success Factors for Achieving the Identified
Potentials and Increasing the Market Penetration of
Robo-Advisors 265
13.3.1 Extending the Reach 265
13.3.2 Improving Financial Literacy 266
13.3.3 Greater Transparency 266
13.3.4 Targeting Women 266
13.3.5 Hybrid Access 267
13.3.6 Expanding and Improving the Product
Range 267
13.4 Summary 267
References 268
Glossary 271
Index 277
NOTES OF CONTRIBUTORS
Alexander Beck is an experienced technology expert in the field of

financial markets forecasting. During his academic career in physics and
economics, he adapted methods of statistics and machine learning from
the field of high energy physics to understanding and forecasting financial
markets. He holds a master’s degree in physics and a Ph.D. in economics.
During his career as a consultant, he advised several financial services com-
panies in the application of machine learning and artificial intelligence. He
is the CTO of the quantitative asset manager Quoniam Asset Management.
Ana-Maria Climescu is Senior Associate at Hauck & Aufhäuser and is

responsible for marketing. She has worked for the bank since 2015 and
during this time, she has developed marketing campaigns for various busi-
ness areas and accompanied projects. She is also responsible for the entire
external appearance of H&A’s digital wealth channel Zeedin. Since 2019,
she has also worked as a lecturer for external corporate communications.
Climescu holds an M.A. in Global Marketing Management from Accadis
University Bad Homburg.
Melanie L. Fein is an attorney who advises financial institutions on

regulatory matters. She formerly was senior counsel to the Board of
Governors of the Federal Reserve System and has served on the adjunct
faculty of Yale Law School. Portions of this chapter have been previously
included in papers by Fein and are available at the Social Sciences Research
Network (SSRN).
xxv
xxvi NOTES OF CONTRIBUTORS
Joachim Goldberg has analyzed the interaction between investors and

markets for more than 40 years. The banker and former currency trader was
inspired by the insight that psychological factors drive financial markets.
Since then, he took a deep look into behavioral finance, a discipline which
analyzes the emotional decision-making process of market participants.
In 1999, together with Professor Rüdiger Nitzsch of RTHW Aachen, he
published the book Behavioral Finance – Gewinnen mit Kompetenz, which
is one of the first books on the subject of the German market. Through
many appearances on radio and TV, press reports, internet columns as
well as numerous events for behavior-oriented market analysis, Goldberg
has established himself as a leading expert in this field in Germany and is
considered one of its best-known representatives.
Goetz Greve is Professor of marketing and sales at the HSBA Hamburg

School of Business Administration, Germany. In his research, he focuses on
the areas of customer relationship management, sales management, and
online marketing. His research findings have been published in interna-
tional journals such as the International Journal of Research in Marketing
and Marketing Review St. Gallen. Goetz has published four books in the
field of CRM and Online Marketing. He is on the editorial boards of
the International Journal of Internet Marketing and Advertising and the
International Journal of Marketing Studies and serves as an ad-hoc reviewer
for other journals, such as the European Journal of Marketing and the
Journal of Marketing Theory and Practice. Before returning to academia,
he worked for Accenture and OC&C Strategy Consultants.
David Grossmann earned his Ph.D. from the Andrássy University

Budapest—University of National Excellence—in cooperation with the
Claussen-Simon Graduate Centre at HSBA. He is the Head of Global
Bank Management at the Otto M. Schröder Bank AG. Formerly, he was
with the Hamburger Sparkasse AG, a significant European institution,
where he worked as a professional in Real Estate Finance, Enterprise Risk
Management, and Performance Management.
Christian Hammer worked as a savings bank manager for private and

corporate clients for 10 years before moving to NFS in 2005. As a
graduated bank manager and financial planner (HfB) he looked for the
dynamic and innovative challenge in the banking landscape and found
NOTES OF CONTRIBUTORS xxvii
his home as an Executive Partner at NFS. By writing Basic knowledge for

investment funds, securities and alternative investment funds, he passed on
his many years of experience in the capital markets to a larger audience.
Christian von Keitz is Senior Project Manager in Hauck & Aufhäuser’s

Inhouse Consulting department. He has more than 12 years of profound
experience in project management within the banking business (special-
ized in IT driven projects with a private wealth, regulatory or business
development related background). von Keitz joined Hauck & Aufhäuser
in 2017, and within the Zeedin project team, he has been responsible
for the conception of the customer self-assessment, the mechanisms for
the resulting investment proposal as well as for several aspects of the
onboarding process. He started his professional career with a vocational
training in banking before pursuing an international management degree
(B.A.).
Sinan Krueckeberg has more than 10 years of experience in conventional

and alternative investments in practice, research, and teaching across the
US, Europe, and Asia.
As CEO of the Krueckeberg Family Office, he is responsible for the
family’s investment strategy which has allowed him to gain a deep under-
standing of the dynamics and challenges of both public and private markets.
He also leads the Private Capital practice of Etribes, a leading European
digital strategy consultancy that helps the private equity industry increase
portfolio company valuations through digital levers.
As executive board member of a large mid-sized family business oper-
ating in the paper trading, packaging and screen & sign industries, he
has first-hand experience of the threats and opportunities that digitization
creates for legacy portfolio companies.
He complements his practical experience with academic research into
alternatives and his work has most recently been published in the interna-
tionally renowned CFA Institute’s Financial Analysts Journal. Sinan is also
a senior lecturer at the Hamburg School of Business Administration where
he lectures on Entrepreneurial Finance in the M.Sc. Finance program.
Frederike Meyer lives in Hamburg and works in a wealth management

team of a regional financial institute where she is responsible for corporate
banking. She finished her bachelor’s degrees in educational science and
xxviii NOTES OF CONTRIBUTORS
business administration. She is pursuing her master’s degree in business

psychology.
Monika Mueller founded FCM Finanz Coaching in 1999 and works as an

Executive Coach mainly with CEOs, directors and financial sales, portfolio-
and management teams of German and international companies in the field
of financial services. She offers coaching in the areas of organizational and
personal transformation and effectiveness, team intelligence, development
of conscious leadership competencies. She also offers training and coaching
in financial decision making for portfolio managers, financial advisors,
and HNWI. She is an expert in risk profiling and brain-friendly decision
making.
Mueller is considered as the pioneer of finance coaching and its devel-
opment in Germany and other European countries. She was one of the first
coaches in Germany to obtain certification with ICF (International Coach
Federation). She is a Master Certified Coach since 2003. She did her coach
training at several institutes in Europe starting from 1989. She was trained
as systemic coach (WISPO), ontological coach (Newfield), EMDR coach
(Sam Foster) and has also conducted studies in body therapy, energy work
and organizational development. Her commitment to the dissemination
of coaching led her to be a founding member of the German Chapter of
ICF in 2000. Since then she is engaged in mentor coaching and the ICF
assessment of coaches wordwide. In 2013, she started the FCM Finanz
Coach® training program in Wiesbaden.
With a Diploma in Psychology from the Johannnes-Gutenberg Univer-
sity (Mainz, Germany), Mueller was head of a training and coaching agency
of a German training institute and founded FCM Finanz Coaching 1999,
a company that is focused on high performance in (financial) decision
making.
Yvonne Quint is a partner in the segment Banking & Capital Markets

at BearingPoint, and advises banks and investment companies regarding
diverse capital markets and investment management topics. Her core areas
of consulting are regulation of capital markets/capital markets compliance
and optimization/digitalization of front-to-back-office processes for capi-
tal markets products. Her client base is broad and encompasses state banks,
private banks, large international banks as well as investment companies.
Over the last 4 years, Quint supported various clients in implementing
NOTES OF CONTRIBUTORS xxix
MiFID II and PRIIPs. SFTR is one of the crucial topics on her agenda.
Before starting at BearingPoint in mid-2010, Quint had gained practical
experience over several years in banks. She holds an Executive M.B.A. dual-
degree from EBS Business School and Durham University Business School
in the UK.
Paul Resnik is a founder of the world’s first psychometric financial risk

tolerance tool for financial advisers. The FinaMetrica Risk Profiler has been
used in more than 20 countries for nearly 25 years and is now part of the
Morningstar group.
Resnik is CEO of the Suitable Advice Institute which advances financial
advice that is customized for the client and founded on their informed
consent, based on their full understanding of risk. Find out more at
suitableadviceinstitute.com.
Jan Rocholl is Director at Hauck & Aufhäuser and heads the Inhouse
Consulting. He has more than 18 years of project and consulting experi-
ence in banking, industry and information technology. Rocholl has been
with Hauck & Aufhäuser since 2011 and has managed and implemented
a wide range of strategic and procedural projects. He was co-project
manager for H&A’s new Online Banking platform and for the development
of H&A’s digital wealth channel Zeedin with focus on all IT related
matters. Prior to Hauck & Aufhäuser, he gained his project competence
as a consultant for Accenture and in various positions within the SAP
Group. Rocholl holds a master’s degree in business administration from
the University of Wuerzburg.
Thorsten Ruehl is partner and head of research and portfolio manage-

ment at CSR Beratungsgesellschaft in Hofheim/Ts since 2014. Earlier,
he was heading the quant asset allocation team at Deka Investment
in Frankfurt for more than 10 years where he was responsible for the
development of quant asset allocation strategies and their implementation
in funds.
Ruehl studied Math and Physics at the Justus-Liebig-University in
Gießen and at the University of Washington in Seattle. He completed his
doctorate in applied solid state physics at the Institute of Applied Physics
at the Justus Liebig University.
xxx NOTES OF CONTRIBUTORS
Ruehl has two decades of experience in quantitative portfolio manage-

ment and is an expert for the implementation of the Black-Litterman-
algorithm and portfolio insurance strategies. He gives lectures in math-
ematics and finance at the HS-RM in Wiesbaden and the HSBA in
Hamburg.
Madeleine Sander is Managing Director and Chief Financial Officer at

Hauck & Aufhaeuser. In this function, she is responsible for the depart-
ments of Corporate Development, Finance & Tax as well as H&A’s digital
wealth channel Zeedin. She joined Hauck & Aufhäuser in 2017 as Head
of Corporate Development and was co-project manager for H&A’s new
Online Banking Platform and the development of H&A’s digital wealth
channel Zeedin with focus on all strategy related matters. Prior to Hauck
& Aufhäuser, she worked more than 6 years for Deutsche Bank in the
Corporate Development (AfK) and Inhouse Consulting department. This
math and business administration graduate started her career at DekaBank
as Business Manager for Corporates & Markets.
Craig Saunders is a communications and marketing consultant who

specializes in telling stories of personal finance and investment. He has
an extensive background as a financial journalist with Australia’s national
TV broadcaster and created and delivered an ongoing professional devel-
opment program to an Australia-wide audience of financial advisers. Find
out more at cbkcommunications.com.au.
Theodor Schabicki has worked in financial services for 24 years, of which

the last 15 years have been at leading consulting engagements at Bearing-
Point, where he became Partner in 2016. As a former securities trader,
his project activities cover capital market products and processes. With the
BearingPoint competence team “Capital Markets – Securities” his focus
lies on E2E processes, Trading, Clearing and Settlement Platforms, Cost
Reduction Programs, Market Data Management and new technologies
such as Distributed Ledgers.
This technical focus is complemented by his activities around process
management. Since 2007, Schabicki is the leader of the competence team
“Process Optimization & Design” with its main focus on activities in
Process Strategy and optimization methodologies, Business Process Man-
agement (BPM), Process Mining, Robotics & Process Automation (RPA)
NOTES OF CONTRIBUTORS xxxi
and Artificial Intelligence (AI), Business Process Outsourcing (BPO) and

holistic concepts like Data Driven Operations in Banking.His international
experience spreads over engagements in Tokyo, London, Paris, Milan,
Vienna, Geneva and Zurich.
Peter Scholz is Professor of Banking and Financial Markets at Hamburg

School of Business Administration (HSBA) where he has been a faculty
member since 2013. He is the academic head of the MSc Finance program
and responsible for developing online teaching methods at HSBA. Scholz
got his PhD at Frankfurt School of Finance and Management where he
worked in the Centre for Practical Quantitative Finance. His research
interest is focused on digitalization aspects of financial markets such as
fintechs and cryptocurrencies. Furthermore, he is active in the fields of
behavioral finance and active asset management. His work is published in
academic journals such as the Financial Analysts Journal and the Journal
of Wealth Management.
Moreover, Scholz has practical experience in the financial sector. He
started his career with Deutsche Bank where he learned the business from
the scratch: from investment advisor to equity derivatives sales trader. He
has also worked in the quantitative equity division as senior portfolio
manager for Deka Investments.
Soeren Schroeder is a manager at BearingPoint with 10 years of experi-

ence in Financial Services. He advises banks regarding transformation and
process optimization topics and is part of the BearingPoint competence
team “Capital Marketes – Securities”. His core areas of expertise are digi-
talization of front-office processes and management of IT-Transformation
projects. Over the last 6 years Schroeder supported various clients in
implementing new product/service offerings and transforming their IT-
Infrastructure.
Before starting at BearingPoint in 2013, Schroeder worked for a large
retail bank. He graduated from ESADE Business School and is a Chartered
Financial Analyst (CFA).
Michael Tertilt holds a bachelor’s degree in Technology and Business and

worked since 2004 in several positions at Diebold Nixdorf in the field of
IT outsourcing services for financial industry. He completed an M.B.A. at
the Hamburg School of Business Administration in 2016, specializing in
xxxii NOTES OF CONTRIBUTORS
bank management. He has been following the developments in the Fintech

industry for several years and wrote his M.B.A. thesis and publications on
Robo-Advisors.
Jochen Werne is a German manager with a proven track record in business

& digital transformation and in the field of People’s Diplomacy. He is
a member of the Executive Committee of PROSEGUR Cash Services
and is in charge of Business Development, Innovation, Business Process
Outsourcing, Solutions, New Products and International Sales. He is
a member of the Federal Ministry of Education & Research artificial
intelligence initiative “Learning Systems”; member of one of the world’s
leading think tanks, the Royal Institute of International Affairs “Chatham
House”. He is a recognised keynote speaker and co-author of several
books about innovation, leadership, AI, FinTech, Cyber Security and so
on. He has been awarded several times for his engagement in the support
of international relations. He is recognised by Focus-Magazine as one of
the leading AI experts in Germany. He is the ambassador of the Peter
Tamm sen. Foundation, the founder of the Global offshore Sailing Team
(GOST) and co-founder of the NGO Mission4Peace, which is dedicated
to historic research, the establishment of international, diplomatic relations
and the encouragement of international dialogue. For his efforts, Werne
has repeatedly received international recognition.
ACRONYMS
AI Artificial Inteligence
AISP Account Information Service Providers
API Application Programming Interface
ARC Asset Risk Consultants
AuM Assets under Management
B2B Business to Business
BaFin Bundesanstalt für Finanzdienstleistungsaufsicht (Federal
Agency for Financial Services Supervision)
CAC Customer Acquisition Costs
CAGR Compound Annual Growth Rate
CAPM Capital Asset Pricing Model
CEO Chief Executive Officer
CD Corporate Design
CLV Customer Lifetime Value
CSIA Charles Schwab Investment Advisory
CSIT Customer-Specific Interaction Technologies
DIY Do It Yourself
DOL Department of Labor
DR Diversification Ratio
ERISA Employee Retirement Income Security Act of 1974
ESMA European Securities and Markets Authority
ETC Exchange Traded Commodities
ETF Exchange-Traded Fund
ETN Exchange Traded Notes
xxxiii
xxxiv ACRONYMS
FINRA Financial Industry Regulatory Authority

FinTechs Financial Technologies
FinVermV Investment Brokerage Regulation
FTSE Financial Times Stock Exchange
GewO German Trade Regulation Act
GDPR General Data Protection Regulation
GRNN General Regression Neural Network
GWG(AMLA) Geldwäschegesetz (Anti-Money Laundering Act)
H&A Hauck & Aufhäuser Privatbankiers AG
HNWI High-Net-Worth-Individual
ICBC Industrial and Commercial Bank of China
IRA Individual Retirement Account
IT Information Technology
KAGB German Capital Investment Act
KWG Kreditwesengesetz (Banking Act)
KYC Know your customer (aka client)
MBTI Myers-Briggs Type Indicator
MiFID Markets in Financial Instruments Directive
MiFIR Markets in Financial Instruments Regulation
MVO Mean Variance Optimization
MVP Minimum Viable Product
PNN Probabilistic Neural Network
PRIIP Packaged Retail Investment and Insurance-based
Products
PSD2 Payment Services Directive 2
REST Representational State Transfer
S&P Standard and Poor’s
SaaS Software as a Service
SEC Securities and Exchange Commission
SOAP Simple Object Access Protocol
SSL Secure Sockets Layer
TER Total Expense Ratio
TLS Transport Layer Security
UCITS Undertakings for the Collective Investment
in Transferable Securities
UHNWI Ultra High-Net-Worth Individual
USP Unique Selling Proposition
ACRONYMS xxxv
VaR Value at Risk

VAT Value Added Tax
VermAnlG German Investment Products Act
WpHG Wertpapierhandelsgesetz (Securities Trading Act)
WpPG German Securities Prospectus Act
LIST OF FIGURES
Fig. 1.1 Wealth management process adapted from Evensky et al.

(2011) 5
Fig. 3.1 Financial risk tolerance is a stable psychological trait
which follows a normal distribution pattern, where most
people cluster around the middle peak (the “norm”) with
reducing numbers of outliers at each end (the very risk
averse and extreme risk seekers) 45
Fig. 5.1 Diversification effect 78
Fig. 5.2 Diversification ratio shows maximum at 2.0 78
Fig. 5.3 Risk profile of a portfolio with 3% volatility and 2.5%
expected return 88
Fig. 9.1 Challenges of implementing/introducing a robo-advisor 166
Fig. 9.2 Dimensions for specification of the robo-advisor in the
context of the overall securities strategy 171
Fig. 9.3 Evaluation dimensions for provider selection 174
Fig. 9.4 Introduction of a robo-advisor 177
Fig. 9.5 Evaluation criteria for existing robo-advisors 179
Fig. 9.6 Evaluation of existing robo-advisors 180
Fig. 10.1 H&A strategy 2020 (taken from presentation slide) 189
Fig. 10.2 Private banking client segments in Germany 2018 (based
on zeb.research) 194
Fig. 10.3 Products and services offered by existing market providers
in the robo-advisory segment 198
Fig. 10.4 Strategic objective of an integrated omni-channel 199
Fig. 10.5 Intelligent interaction between people and technology 203
Fig. 10.6 Zeedin product range (H&A webpage) 207
Fig. 10.7 Customer journey with Zeedin 208
xxxvii
xxxviii LIST OF FIGURES
Fig. 10.8 Analysis of customer needs 209

Fig. 10.9 Zeedin logo 210
Fig. 10.10 Layout of Zeedin’s investment process 211
Fig. 10.11 Visualizations—personae and messages 211
Fig. 12.1 Active users of social media worldwide (Own illustration
based on Kontor4 2019) 247
Fig. 12.2 Areas of influence of social media on robo-advisors
(authors’ own image) 250
Fig. 13.1 IAI Q1 2018 USA Barometer on awareness of robo-advisor
based on Antoniotti et al. (2018) 262
Fig. 13.2 Ebase survey Q1 2019 on the awareness of the term
robo-advisor based on Nicolaisen (2019) 262
Fig. 13.3 Ebase survey Q1 2019 on the likelihood of using a
robo-advisor to invest in the next 12 months Nicolaisen
(2019) 263
Fig. 13.4 Investopedia affluent millennials study shows the wealth
distribution amongst the different investor generations
(Gobell 2019) 264
LIST OF TABLES
Table 6.1 Results of Kruskal–Wallis H tests 100

Table 8.1 At a glance: which asset classes could be advised and brokered 144
Table 8.2 At a glance: which services are related with which license 144
xxxix
PART I
Status Quo of Robo-Advisory

CHAPTER 1
Robo-Advisory: The Rise of the Investment

Machines
Peter Scholz and Michael Tertilt
Contents
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Robo-Advisory as Part of the Wealth Management Process.. . . . . . . . . . . . . 4
1.3 A Brief History of Robo-Advisory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4 Clients and Market for Robo-Advice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.5 Performance of Robo-Advisors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.6 Robo-Advice in a Nutshell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
P. Scholz ()
Professor for Banking and Financial Markets, Hamburg School of Business
Administration, Hamburg, Germany
e-mail: [email protected]
M. Tertilt
Research Affiliate, Hamburg School of Business Administration, Hamburg,
Germany
© The Author(s) 2021 3

P. Scholz (ed.), Robo-Advisory, Palgrave Studies in Financial Services
Technology, https://doi.org/10.1007/978-3-030-40818-3_1
4 P. SCHOLZ AND M. TERTILT
1.1 INTRODUCTION
The Rise of the Machines—these words sound more like the title of a movie
or a video game rather than being related to finance. But more than 30
years ago, the story was on the cover of TIME magazine:1 For the first
time, the idea that robots and digital solutions may take over the economy
was discussed in the public domain. Meanwhile, the digital revolution has
disrupted many sectors like the music and publishing industry, imaging
technology, retail business, and so on. And since the financial crisis of 2008,
the financial sector has got increasingly into the crosshairs of disruptors.
It is noteworthy, however, that the use of technology is nothing unusual
for the financial sector. Capital has always been creative through the
application of innovations to capture individual advantage. For example,
one of the first applications of the telegraph was the long-distance trans-
mission of news which was relevant for trading in distant marketplaces. The
invention of the smartphone has been a comparable technical revolution
that has had a huge impact on consumer behavior and on the expectations
toward banking services. Therefore, it should not come as a surprise that
FinTechs came up with the idea to digitalize the advisory process in asset
management.
1.2 ROBO-ADVISORY AS PART OF THE WEALTH

MANAGEMENT PROCESS
The term “robo-advisor” is a blend of the two words “robot” and “advisor”.
It basically describes that the client-interface of the advisory process is
not a human anymore but a machine. Based on the wealth management
process, in our version adapted from Evensky et al. (2011) (see Fig. 1.1),
the question is, which exact process steps are covered by the robo-advisory?
In contrast to common perception, robo-advice does not necessarily mean
that all these process steps are fully automatized. In fact, the advisory
contribution of the robot may be quite diverse. A very common feature
of the robo-advisory is the onboarding process of new customers and the
determination of the client’s risk profile. The management of the sample
portfolios, however, is not necessarily a part of robo-advice.
1 TIME Magazine, May 30, 1983.

ROBO-ADVISORY: THE RISE OF THE INVESTMENT MACHINES 5
Fig. 1.1 Wealth management process adapted from Evensky et al. (2011)
If we take the largest robo-advisor, Vanguard Personal Advisor Services,

as an example, we find a description explaining that the “proprietary
algorithm uses [the clients’ investment profile] data to recommend a
particular investing track” (Vanguard Advisers, Inc. 2019, p.8). So, there
seems to be a machine that matches customer data to a specific portfolio.
However, there is always a human advisor as well if the client feels the
need for any discussion (Vanguard Advisers, Inc. 2019). The dominant
robo-advisor, therefore, relies more on a hybrid form of advisory. If we
dive deeper into the wealth management process, we find that model
portfolios are not generated by a robot but an investment committee:
“We may propose the addition, removal, or adjustment of sub-asset class
exposures based on continuing portfolio construction research performed
by Vanguard Investment Strategy Group” (Vanguard Advisers, Inc. 2019,
p. 8). On the other hand, there seems to be an algorithm managing risk:
“When recommending, setting, and adjusting your asset allocation, we
weigh shortfall risk—the possibility that a financial plan or Portfolio will
fail to meet longer-term financial goals—against market risk” (Vanguard
Advisers, Inc. 2019, p.8). But there is no active market timing for
investments: “The algorithms don’t consider prevailing market conditions

when making recommendations to you” (Vanguard Advisers, Inc. 2019,
p. 10).
If we consider the pursuer, Schwab Intelligent Portfolios, they seem
to follow a similar process. The asset allocation apparently is not created
by an algorithm but an investment committee: “Using asset allocations
and ETF selection parameters determined by Schwab, CSIA has created
a number of investment strategies for the Programs” (Charles Schwab &
Co., Inc. 2019, p. 2).2 The robo-advisor algorithm itself is applied to select
a portfolio recommendation based on the client’s investment preferences
for rebalancing, for tax optimization, and for triggering orders. In con-
trast to the two market leaders, the first robo-advisors, Betterment and
Wealthfront, seem to rely more on rule-based allocations. Both explicitly
refer to an asset allocation process that is driven inside the robo-advisor
and relies on the Modern Portfolio Theory (Wealthfront 2020a; Grealish
2019). In Europe, Nutmeg and Liqid seem to have a committee approach
for asset allocation as well (Port 2013; Nutmeg 2020b), whereas Scalable,
comdirect, and Quirion operate an explicit asset allocation process as part
of their robo-advisory services. A rather lean approach is the service of
WeltInvest (Weltsparen 2020), where only sample portfolios with a fixed
asset allocation are presented and the investor must choose between the
four ETF portfolios. However, there is no advice on which portfolio to
select or how to identify risk tolerance, risk budget, and so on. Hence,
the investor is basically left on his own. In return, the service is rather
competitive with an approximate 0.5% p.a. fee, including product costs.
In summary, it is a bit difficult to describe what really defines a robo-
advisor. From our perspective, the very core is the advisory process—the
risk profile of the investor is determined, and a portfolio recommendation
is given. Hence, we would not denote the offer from WeltInvest as
robo-advisor. Within the group of robo-investors, there could be many
subgroups. At least, we would form a group based on those robos with a
quantitative, rule-based allocation process, and another group that relies
on investment committees or external advisors. In conclusion, based
on Peter’s experience as an investment advisor, the essence of robo-
advisory is the replacement of the human advisor, or maybe better, the
human salesperson, with an online interface by granting benefits like 24/7
2 CSIA is an affiliate of Schwab, the Charles Schwab Investment Advisory, Inc.

availability, potentially lower costs, scaling, and automatization of at least

some aspects of the portfolio management like rebalancing. Consequently,
a robo-advisor just substitutes and eases certain steps within the asset
management process, but it certainly does not create a whole new process
on its own.
1.3 A BRIEF HISTORY OF ROBO-ADVISORY

The first robo-advisory services, which were designed for retail investors,
came into life in the US after the financial crisis of 2008, probably
also as an answer to the increasing distrust in the investment advisory
business (Becchi et al. 2018). Betterment, founded by a team including
Jon Stein, and Wealthfront, launched by Andy Rachleff and Dan Carroll,
are considered as the first robo-advisors in the retail market (Fisch et al.
2018). Only a few years later, the first robo-advisory solutions came up in
Europe: The first one founded in the UK was Nutmeg in 2011 (Lielacher
2016), and Quirion was established in Germany in 2013 (Kümpel 2016).
Interestingly, the roots of robo-advisory can be traced back either to classic
start-ups or to asset managers, whereas the banks by and large did not
belong to the early adaptors. Especially for asset managers, robo-advisory
services offer a huge leveraging potential (Becchi et al. 2018)—they can
easily shift parts of their clients’ capital to digital advisory services and lower
the entry barrier for new customers. It comes as no surprise that currently
the largest robo-advisor, measured in assets-under-management, is owned
by Vanguard. The initial idea of robo-advisory has been to disrupt the
human advisory services of banks by offering affordable investment advice
to widen the customer base by including retail clients and help them make
good investment decisions (Nguyen 2018). Today, we can see that robo-
advisory is mostly cooperating with financial institutions like banks instead
of disrupting them. Based on institutional economics in the FinTech sector
(Scholz 2018), it can be shown that mainly due to regulation and trust
issues, such as moral hazards, investors are hesitant: What does the robo-
advisor do with the money the moment it is invested? Despite the financial
crisis, banks still enjoy a kind of bonus trust over start-ups. Moreover, banks
are trying to mitigate the distrust issue regarding innovation by applying
hybrid models where humans and robo-advisors are combined.
While the early business models have been rather simple—they were
concerned with determining the risk preferences of the client and rec-
ommending an ETF-based portfolio of stocks and bonds—the advisory

tends to become more complex over time, for example, by including more
asset classes, active funds, tax optimization, and so on. However, based on
our analysis from 2018, we found that the investment process by robo-
advisors has been very simple at that time (Tertilt and Scholz 2018). The
measurement of risk preferences typically relied only on a few questions,
but it does not seem enough to provide proper risk classification and
hence to derive adequate investment advice. From that perspective, we
assume that the robo-advisory business is still in fledgling stages and further
development seems absolutely necessary.
1.4 CLIENTS AND MARKET FOR ROBO-ADVICE

Since its early beginning, the market volume of robo-advisory has grown
significantly. However, it is still small with respect to the global investment
industry. The 2018 global assets-under-management (AuM) of robo-
advice has been valued at around $ 500 billion (Statista 2020c). To put
this number into perspective, the global market for ETFs is ca. nine
times larger and accounts for approximately $ 4.7 trillion (Statista 2020a).
As one could expect, the US as the country of origin is the dominant
market for robo-advice with approximately $ 377 billion AuM in 2018
(Statista 2020b). The biggest player by far is currently Vanguard Personal
Advisor Services, which belongs to the Vanguard Group, a privately held
investment management company. According to Willis Towers Watson
(2018), Vanguard is the second largest asset manager worldwide. This
seems to confirm the potential leveraging by shifting parts of clients to
digital advisory services. Based on their overall AuM of ca. $ 5.2 trillion,
Vanguard (2020a) only needs to put a small portion of approximately
$ 115 billion (Friedberg 2019) in their robo-advisors to place first. That
corresponds to a mere 2.2% of their total AuM. The next rival is Schwab
Intelligent Portfolios, the robo-advisor of the Charles Schwab Corporation,
a large stockbroker company. Compared to Vanguard’s robo-advisor, they
only manage about a third of the assets with $ 33.4 billion (Friedberg
2019), demonstrating the dominance in market shares by the market
leader. This picture becomes even clearer if we consider the two first robo-
advisors on the market, Betterment and Wealthfront, which are comparably
dwarfed with ca. $ 16 billion and ca. $ 11 billion, respectively, even though
they are ranked third and fourth, respectively, in the list of the US AuM.
The European markets for robo-advice are clearly smaller and just
account for around $ 16 billion. Within the European market, the UK (ca.
$ 6.1 billion) and Germany (ca. $ 4.5 billion) had the largest market share
in 2018 (Statista 2017).3 Since not all robo-advisors publish reports on a
regular basis and most of them have grown at a rapid pace, it is a bit difficult
to find current and reliable numbers. Recently, Nutmeg is supposed to be
the largest robo-advisor in the UK with approximately £ 1.5 billion in AuM
(Jones 2019). In Germany, Scalable, which is affiliated with Blackrock, is
the only robo to hit the e 1 billion barrier (Scalable Capital 2020b) and
is now estimated to manage around e 2 billion (Scalable Capital 2020c).
These numbers show that the European market still has large potential for
development: the estimated growth until 2023 is ca. 41% p.a. This number
is in line with the growth of the German market; for the UK market, the
estimated growth is ca. 38% p.a. and hence a bit smaller than the European
average. On a global level, the growth is estimated to be around 31%
p.a., which is also the result of the comparatively slow expansion in the
largest market. The US is only estimated to increase the AuM by ca. 25%
p.a. The largest driver of global growth is supposed to be China. With
an approximately AuM of $90 billion in 2018, the Chinese market is
estimated to increase to $ 880 billion in 2023, which corresponds to an
annual growth rate of ca. 46% p.a. (all growth estimates calculated with
Statista (2017) data). However, at least for 2023, the US is still seen as
being the largest market for robo-advice based on Statista (2017) forecasts.
Breaking down from markets to customers, around 6.6 million clients
currently use robo-advisory services in the US. This number is supposed to
grow by ca. 15% p.a. until 2023. Looking at the per capita investment, an
average US robo-advice investor in 2018 put approximately $ 65,000 on
his or her account. This number is assumed to grow quickly until 2020 to a
bit more than $ 100,000 and then levels there. In China, the second biggest
market in terms of AuM, around 17.5 million people used robo-advice in
2018. And this number is estimated to grow by ca. 39% p.a. to ca. 124.5
million people in 2023. The per capita investment, however, is significantly
smaller than in the US: the average Chinese investor in 2018 placed only
around $ 5100 in robo-advice. And this number probably increases only
3 The numbers in the Statista dossier include data until end of 2017 and forecasts for 2018
and beyond. To have a fair standard of comparison, the 2018 numbers are used which may
include forecasts sometimes.
by ca. 7% p.a. over time. The clear driver in China seems to be the number
of clients and not the average investment amount. In Europe, there are
currently only around 900,000 robo investors, but the forecast sees a ca.
28% p.a. expansion. Regarding the per capita investment, the Europeans
are in-between the US and China: currently, on average, ca. e 15,500 are
placed on an account with a growth perspective of about 13% p.a., which
gives an average investment of ca. e 29,000 in 2023. Since we do not
have exact information about how the investments are distributed among
investors, the interpretation of these numbers gives some leeway. One
potential explanation could be that especially in Europe and China, robo-
advice is indeed used for retail investors that use the digital service to benefit
from professional investment advice that they could not afford otherwise.
In the US, by contrast, middle-class investors seem to use robo services.
Another explanation attempt points to a trust issue. Maybe investors in
Europe and China do not fully trust the robo-advice and therefore use
only a smaller portion of their wealth; they also distribute their investments
among different suppliers and use human advisors as well. This could be
especially true for Europe, which is often seen as more skeptical about
digitalization. At least for Germany and Italy, this assumption is backed by
two studies of Kaya (2017) and Kaya (2019), which find that the average
investor possesses medium or high wealth (result for Italy), has around
e 4000 as monthly net income, robo-invests between e 1000 and e 1500
annually (Germany), and typically has a university degree (both countries).
At least, this points more toward private banking than a retail banking
clientele. In China, by contrast, people tend to be very technology-friendly,
but maybe the trust issue here points more at the direction of institutions—
this may explain the lower investment amounts. In the US, people tend
to be more open-minded regarding technology and hence may rely more
easily on the robo-advisor than on the bank (all forecasts in this paragraph
are based on Statista (2017) data).
It seems to be a natural assumption that the typical investors of
robo-advice are the millennials who are aged 24 to 35 years (Zeldis
Research Associates 2018). However, according to Zeldis Research Asso-
ciates (2018), millennials do not share exuberance for robo-advisory, but
they have “limited awareness” of these services. Kaya (2017) states that
the millennials widely belonged to the early adopters, but this has shifted
significantly. Based on Kaya (2017) and Kaya (2019) statistics, the age of
the average investor is rather between 45 and 48 years—it holds for both
US and German investors. Moreover, based on estimates from Germany,
the overwhelming number of investors is male (Kaya 2019). In Italy, the

distribution is a bit more balanced, but there are still significantly more
male than female investors applying digital investment advice (Kaya 2017).
The dominance of male investors is probably also a fair assumption for most
other robo-advisory markets.
Despite the relatively optimistic development predictions, trees do not
grow in the sky in the robo-advisory business as well. To be profitable,
the provider of robo-services needs to raise their AuM above a certain
threshold (Jones 2019). Currently, there are more than 200 robo-advisors
in the US, 30 in Germany, and 20 in the UK. More and more financial
institutions try to establish a robo-advisor, maybe as an open platform
or solely as a distribution channel for their own products—for example,
ICBC, one of the largest Chinese banks, has recently tried to launch a
robo-advisor for their clients. In a report from March, HSBC estimates
that a robo service in North America, which charges 0.25% as fee, needs
approximately between $ 11.3 billion and $ 21.5 billion AuM to break even
(Jones 2019).4 At present, only the top four providers in the US reach
these numbers. In Europe, where HSBC observed higher fees at around
0.45%, at least more than $ 3.5 billion AuM seems to be necessary. Given
that the scope for increasing fees seems to be limited, most robo-advisors
need to collect more money and acquire more clients to become profitable.
Hence, a consolidation of the sector seems to be inevitable. At least in
Germany, some well-known financial institutions have abandoned their
plans with robo-advisory—for example, Commerzbank, the second largest
bank in Germany, stopped their plans launching its own robo-advisor and
now relies on the service of their affiliate comdirect. A similar case is
the Spanish bank Santander, which recently stopped their robo-advisor.
While banks typically have a customer base, which they could convince of
the robo-advice, for start-ups it is harder to gain trust and attract more
clients. Therefore, a cooperation between FinTechs and banks seems to be
a reasonable way to address consolidation. While there seems still to be
plenty of room for innovative providers with reasonable fees, it will play
a key role in matching customer needs and expectations with respect to
performance and quality of advice. However, for those asset managers who
do not have a clear and precise philosophy, or those who charge rather high
fees, it will probably become a tough market.
4 Original study of HSBC could not be obtained.

1.5 PERFORMANCE OF ROBO-ADVISORS

To create a competitive service, it comes back to the performance at the
end. Irrespective of the elaboration of the service, if the (risk-adjusted)
returns do not convince, the robo-advisor will not be successful in the long
run. To measure the historical performance of robo-advisors can be a little
tricky, since, to the best of our knowledge, there is no need to publish
data nor to create a database to collect the performance of different robo-
advisors internationally. Hence, we can only provide some statistics with
varying granularity and quality of data. A rather current development is
AISP platforms, which allow for individual comparisons between different
robo-advisors.
The largest robo-advisor from Vanguard does not seem to publish
detailed historical performance data. It states, however, that approximately
“85% of our funds have performed better than their peer-group averages
over the last 10 years” (Vanguard 2020b). Moreover, on a website that
compares different robo-advisors, we found information that “Vanguard’
claims its US stock funds offer a 7.51% rate of return in line with S&P
500 avg returns” (Moore 2019). So, generously spoken, an investor seems
to earn approximately market returns while investing in Vanguard’s robo-
advisor. It remains unclear if this holds before or after fees. For Schwab’s
robo-advisor, the situation seems to be pretty similar: Schwab does not
publish charts or return data but a quarterly written report (Charles Schwab
& Co., Inc. 2019). In the section “How did Schwab Intelligent Portfolios
do?”, the management gives brief information about the development of
the different strategies in a rather qualitative way—there are hardly any
numbers presented. Wealthfront at least provides some return numbers
for their different sample portfolios (Wealthfront 2020b). Even if the
information is not very granular, Wealthfront states returns based on
different frequencies: annual, three years, five years, and since inception.
This allows the investor to get at least a certain feeling about the past
performance. However, the risk figures are not displayed. Looking at the
tax-optimized portfolio with the highest risk score, Wealthfront has yielded
8.05% p.a. since inception (December 15, 2011, until July 31, 2019). This
is significantly lower than an investment in the S&P 500 index, which
would have earned an annual 14.4% if dividends had been reinvested
(DQYDJ 2020). Even if we consider that this number does not include
fees for an ETF or tax abbreviations, it seems to be significantly higher
than any robo-advice-based investment with Wealthfront at first glance.
Still, we should keep in mind that the Wealthfront portfolio is supposedly

better diversified and may therefore carry less risk. This comparison can
only deliver a first indication about the “real performance” of the robo-
advisor. The most detailed performance information from the big four
US robo-advisors is provided by Betterment: they display a chart with
monthly returns from their different portfolio strategies which dates back
to January 2004, well before their official launch (Egan 2014). If we
take a similar investment period (here: December 2011 until February
2019), a 100% stock portfolio yielded an annual return of ca. 9.64%.
During the same period, the S&P 500 increased by ca. 14% p.a., if one
includes dividends, but before fees and taxation. The gap seems to be
smaller for Betterment than for Wealthfront. But again, since we have
only little information other than returns, the comparison is limited in its
value. Another source of performance information is Statista (2019), which
claims to provide a global robo-advisory comparison for 2018 based on
a report by BackendBenchmarking (2018). In their analysis, however, all
the robo-advisors are based in the US. In this ranking, the largest robo-
advisor from Vanguard is placed first with a 2018 return of 6.56%, followed
by WiseBanyan (6.25%). Other robos with large AuM are ranked #7th
(Schwab with 4.77%) and #8th (Betterment with 4.17%). In the Statista
analysis, it is not clearly stated which portfolios they exactly use or how
they exactly perform the analysis. Hence, this should also be taken with a
grain of salt.
In the UK, Nutmeg provides charts for their different portfolios:
monthly returns, which can be seen in the chart, daily returns, which solely
are given by the chart, and an average competitor comparison based on
Asset Risk Consultants (ARC) data (Nutmeg 2020a). They declare their
returns after fees and with reinvested dividend payments. Their “all-time”
data contains the period from September 30, 2012, until July 31, 2019,
and reports a 9.2% annual return for the portfolio with the highest risk
compared to the ARC average competitor return of 8.4% p.a. Looking at
the calendar years, Nutmeg outperformed the average in each year with
positive returns. In 2018, however, the first year with negative returns of
9.9%, they failed to beat their benchmark (−6.5%). Also, in the current
year, they are behind their benchmark (2.4% compared to 5.4%). So, it
seems that they have fallen behind the average during the past one and a
half years. Trying to analyze the chart, the maximum drawdown happened
in mid-2018 and recovered until April 2019 with an approximate 25%
downturn, which seems a bit high for the period of evaluation. Since the
equity portfolio of Nutmeg has rather high stakes in both US (ca. 43%)
and UK (25%) stocks, we look at the FTSE and S&P500 performance
as a standard of comparison. For the S&P500, the performance shows
an annual return of 13.38%; for the FTSE, it is 7.58%. If we weight the
returns of both indices accordingly, we receive an annual return of around
11.25%, which is comparable to the 9.2% return of Nutmeg, keeping in
mind that index data does not contain fees and is not as diversified as
the Nutmeg portfolio. Comparing Nutmeg to its peers, there is a UK
website providing performance data for different suppliers from July 1,
2017, to June 30, 2018 (Boring Money Business 2020). For the high-risk
portfolios, the Nutmeg 10 portfolio showed the highest return (7.92%),
although it falls behind the FTSE benchmark (8.4%). The other robs-
advisors followed within a range of 7.46% (IG 5 Aggressive) to 4.15%
(Scalable 25% VaR). The question remains: how they would compare in
risk-adjusted returns as the level of diversification is probably not the same.
Still it is an interesting result to compare the highest risk portfolios of
different robotic managers, assuming that the suppliers offering the highest
returns probably also follow the most aggressive investment styles.
In Germany, robo-advisors tend to publish only little performance
information. The market leader in AuM, Scalable, displays a graph of
different value-at-risk strategies that they offer (Scalable Capital 2020a). It
is important to note that even the most aggressive strategies are not 100%
invested in equities but in other asset classes like bonds, real estate, and
so on. Hence, they compare to a mixed-funds benchmark from Morgan
Stanley. Since their inception in January 2016, they show similar returns.
However, the deflections and hence the risk seem to be somewhat lower
for the robo-advisor than for the mixed-fund. Liqid also provides a chart
for their performance, which is not very granular (Liqid 2020). The
riskiest portfolio has currently an equity quota of around 95%. Liqid uses
the average of German wealth managers as the benchmark. Since April
2016, they performed significantly higher than their benchmark (ca. 33%
vs. 15.5% in total). Quirion delivers a table with annual performance
information (from June to June of the corresponding years), but they
do not compare to a benchmark (Quirion 2020). In their 100% equity
strategy, Quirion achieved rather high returns in four years (19.9%, 16.4%,
15.8%, and 6.6%) and a negative return in one year (−5.6%). Owing to the
granularity of data, no risk information is available. For the robo-advisor of
comdirect, no public available performance information was found. For the
German market, we found a website that tests the robo-advisors with real
money and publishes the performance data (Broker Vergleich 2020).5 In

this test, comdirect yields the highest returns in 2019, followed by smaller
robo-advisors. The next big player is Scalable, which is ranked seventh.
We would not like to interpret too much into these results. However,
it becomes clear that portfolios that are made for a certain type of investor
can show a wide range of different return and probably risk characteristics.
Robo-services as rational agents do not seem as homogenous as one
might expect. Hence, it seems to be a wise recommendation to check
carefully the information provided by suppliers while exploring the offers
of different vendors. Another key takeaway is our impression that by and
large the performance of robo-advisors seems to be reasonable and that
most of them do a decent job of managing the money of their clients. The
ongoing Corona crash is most certainly a true stress test for robo-advisors.
It will be interesting to learn which robo-advisor philosophy—algorithms
or committees—will be more successful in managing the extremely high
volatilities in the market.
1.6 ROBO-ADVICE IN A NUTSHELL

As of today, banks and asset managers widely see robo-advisory as a
promising new sales channel for investment management. Depending on
the specific robo-solution, the capital requirements to access these services
can be rather low, at least lower than in the traditional wealth management
industry. Hence, it is tempting for many established asset managers to
introduce robo-advisory to select clients between different segments.
However, it would be a pity if robo-advisory would be interpreted only as
a low-budget solution for the “not so rich” investors in the future. Maybe,
the true potential lies in the integration of both human and robo-advice,
which is also known as hybrid model. Some tasks like data collection most
certainly can be done more efficiently by a robo-advisor. In other aspects,
such as displaying empathy in response to the individual client’s situation,
the human is superior. At least, this is currently especially the case for
the matter of trust. By and large, robo-advisory emerged after 2008 and
hence did not have to face a real bear market so far. It will be interesting
to see how robo-advisors would perform in a plummeting market. It is
an essential question if purely digital advice and online interfaces would
5 Another website that compares information from robo-advisors is biallo.de (2020).

be enough in such a scenario to cover the larger information need of

clients and maybe even to provide emotional support them. Undoubtedly,
these requirements of human investors will have their impact on the
further development of robo-advisory services and may accelerate the
tendency to hybrid approaches. Another important factor will be the issue
of automatization within the wealth management process itself. Robo-
advice is predestined for rule-based investments and algorithms watching
over portfolios for rebalancing and risk management purposes. And finally,
technologies like predictive banking and artificial intelligence may find their
way into robo-advisory. So, there is plenty of room for future development.
In fact, it will be interesting to see which innovations and interpretations of
the automated service will thrive the advisory business in the near future.
REFERENCES
BackendBenchmarking. 2018. The Robo Report. Third Quarter 2018. Last
Accessed: March 20, 2020. https://storage.googleapis.com/gcs.backendb.
com/1/2019/04/2018_3Q.pdf .
Scott M. Becchi, Ugur Hamaloglu, Taroon Aggarwal, and Samit Panchal. 2018.
The Evolution of Robo-advisors and Advisor 2.0 Model: The Future of Investment
Management and Financial Advisory.
biallo.de. 2020. Die Besten Robo-Advisor im Performance-Vergleich. Fonds.
Last Accessed: March 20, 2020. https://new.biallo.de/vergleiche/etf-fonds-
aktien/robo-advisor/.
Boring Money Business. 2020. Robo Adviser Performance—12 Months to 30th June
2018. Last Accessed: March 20, 2020. https://www.boringmoneybusiness.co.
uk/research/robo-report-q2-2018/.
Broker Vergleich. 2020. Daten zur Performance aller Robo Advisors im Test. Online-
Broker. Last Updated: March 31, 2020. https://www.brokervergleich.de/
robo-advisor/echtgeld-test/.
Charles Schwab & Co. Inc. 2019. Schwab Intelligent Portfolios and Q2 2019 Market
Performance. Last Accessed: March 20, 2020. https://intelligent.schwab.com/
public/intelligent/insights/blog/q22019-performance-review.html.
Charles Schwab & Co., Inc. 2019. Schwab Intelligent Portfolios Solutions, Disclosure
Brochure. Last Accessed: September 29, 2019. https://www.schwab.com/
public/file/CMS-BDL100049.
DQYDJ. 2020. S&P 500 Return Calculator with Dividend Reinvestment. Last
Accessed: March 20, 2020. https://dqydj.com/sp-500-return-calculator/.
Dan Egan. 2014. Explore Betterment’s Historical Performance. Last Accessed:
March 20, 2020. https://www.betterment.com/resources/betterment-
historical-performance/.
Harold Evensky Stephen M. Horan, and Thomas R. Robinson. 2011. The New
Wealth Management: The Financial Advisor’s Guide to Managing and Investing
Client Assets. CFA Institute Investment Series. Wiley.
Jill E. Fisch, Marion Laboure, and John A. Turner. 2018. “The Emergence of the
Robo-Advisor”. Wharton University of Pennsylvania Pension Research Council.
Working Paper WP2018-12. https://repository.upenn.edu/prc_papers/10/.
Barbara A. Friedberg. 2019. Robo-Advisors With the Most Assets Under
Management—2019. Last Accessed: March 20, 2020. https://www.
roboadvisorpros.com/robo-advisors-with-most-aum-assets-under-
management/.
Adam Grealish. 2019. The Betterment Portfolio Strategy. Last Accessed: March
20, 2020. https://www.betterment.com/resources/betterment-portfolio-
strategy/.
Alexander Jones. 2019. Why Robo-Advisors are Struggling to Break Even. Last
Accessed: March 20, 2020. https://internationalbanker.com/brokerage/why-
robo-advisors-are-struggling-to-break-even/.
Orçun Kaya. 2019. German Robo-Advisors: Rapid Growth, Robust Performance,
High Cost. Deutsche Bank Research. https://www.dbresearch.com/PROD/
RPS_EN-PROD/PROD0000000000487351/German_robo-advisors%3A_
Rapid_growth%2C_robust_perform.pdf .
Orçun Kaya. 2017. Robo-Advice—A True Innovation in Asset Management.
Deutsche Bank Research. https://wwwdbresearchcom/PROD/RPS_
ENPROD/PROD0000000000449125/Roboadvice_%E2%8%93_a_true_
innovation_in_asset_managemen.pdf .
Anna-Lena Kümpel. 2016. Fintech: Bei der Geldanlage geht der Trend zum Robo
Advisor. Last Accessed: March 20, 2020. https://berlinvalley.com/fintech-
geldanlage-trend-robo-advisor/.
Alex Lielacher. 2016. The Complete Guide To UK Robo Advisors. Last
Accessed: March 20, 2020. https://smartmoneysmartliving.com/2016/10/
12/complete-guide-uk-robo-advisors/.
Liqid. 2020. LIQID Wealth. Last Accessed: March 20, 2020. https://www.liqid.
de/liqid-wealth.
Barry D. Moore. 2019. What Is A Robo Advisor? Hybrid vs AI vs Robo Advisors.
Last Accessed: March 20, 2020. https://www.liberatedstocktrader.com/what-
is-a-robo-advisor/.
Phuong Nguyen. 2018. Betterment: Jon Stein. Last Accessed: March 20,
2020. https://www.npr.org/2018/1/26/661049018/betterment-jon-stein?
t=1564478275377&t=1584701835665.
Nutmeg. 2020. Fully Managed Portfolios. Last Accessed: March 20, 2020. https://
www.nutmeg.com/fully-managed-portfolios.
Nutmeg. 2020. How We Invest. Last Accessed: March 20, 2020. https://www.
nutmeg.com/how-we-invest#portfolio-performance.
Shaun Port. 2013. An Inside View of the Nutmeg Investment Process: Asset
Allocation. Last Accessed: March 20, 2020. https://www.nutmeg.com/
nutmegonomics/an-inside-view-of-the-nutmeg-investment-process/.
Quirion. 2020. Über Uns. Last Accessed: March 20, 2020. https://www.quirion.
de/geldanlage/.
Scalable Capital. 2020. Erzielte Rendite. Last Accessed: March 20, 2020. https://
de.scalable.capital/erzielte-rendite.
Scalable Capital. 2020. Robo-Advisor. Last Accessed: March 20, 2020. https://de.
scalable.capital.
Scalable Capital. 2020. Über Uns. Last Accessed: March 20, 2020. https://de.
scalable.capital/ueber-uns.
Peter Scholz. 2018. “Hostile Takeovers: Are Banks Threatened by FinTechs?”
Corporate Finance 01-02: 3–7.
Statista. 2020. Entwicklung des weltweit in ETFs verwalteten Vermögens von 2005
bis 2019. Last Accessed: March 20, 2020. https://de.statista.com/statistik/
daten/studie/219372/umfrage/weltweit-in-etfs-verwaltetes-vermoegen-
seit-1997/.
Statista. 2019. Performance of Selected Robo-Advisors 2018. Last Accessed: Septem-
ber 29, 2019. https://www.statista.com/statistics/755599/performance-of-
selected-robo-advisors/.
Statista. 2020. Prognose zur Entwicklung des verwalteten Vermögens der Robo-
Advisors in den USA von 2017 bis 2023. Last Accessed: March 20, 2020. https://
destatista.com/statistik/daten/studie/740573/umfrage/entwicklung-des-
verwalteten-vermoegens-der-robo-advisors-in-den-usa/.
Statista. 2017. Robo-Advice: Dossier. https://de.statista.com/statistik/studie/id/
48179/dokument/robo-advice/.
Statista. 2020. Robo-Advisors. Worldwide. Last Accessed: March 20, 2020. https://
www.statista.com/outlook/337/100/robo-advisors/worldwide.
Michael Tertilt and Peter Scholz. 2018. “To Advise, or Not to Advise—How Robo-
Advisors Evaluate the Risk Preferences of Private Investors”. The Journal of
Wealth Management 21(2): 70–84.
Vanguard. 2020. Fast Facts About Vanguard. Last Accessed: March 20, 2020.
https://about.vanguard.com/who-we-are/fast-facts/.
Vanguard. 2020. Why Vanguard. Last Accessed: March 20, 2020. https://investor.
vanguard.com/why-vanguard.
Vanguard Advisers, Inc. 2019. Vanguard Personal Advisor Services Brochure. Last
Accessed: September 29, 2019. https://personal.vanguard.com/pdf/vpabroc.
pdf .
Wealthfront. 2020. Wealthfront Investment Methodology White Paper. Last
Accessed: March 20, 2020. https://research.wealthfront.com/whitepapers/
investment-methodology/.
Wealthfront. 2020. Wealthfront’s Average Annual Net-of-Fees, Pre-tax Returns.

Our Historical Returns. Last Updated: March 31, 2020. https://www.
wealthfront.com/historical-performance.
Weltsparen. 2020. WeltInvest. Das ETFPortfolio für clevere Anleger. Last Accessed:
March 20, 2020. https://www.weltsparen.de/etf-portfolio/.
Willis Towers Watson. 2018. The World’s Largest 500 Asset Managers. Thinking
Ahead Institute and Pensions & Investments Joint Research. Last Accessed:
March 20, 2020. https://willistowerswatson.pl/aktualnosci/pdf/2018.10.29-
TAI-PI-500_2018.pdf .
Zeldis Research Associates. 2018. Uncertain Futures: 7 Myths about Millennials
and Investing. Tech. rep. Last Accessed: September 29, 2019. FINRA Investor
Education Foundation and CFA Institute. https://www.cfainstitute.org/en/
research/survey-reports/millennials-and-markets-2018.
CHAPTER 2
Situating Robo-Advisory
Sinan Krueckeberg
Contents
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2 What Are Fintech and Robo-Advisory, Really? . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3 Differentiation: Strengths and Weaknesses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4 Impact on Incumbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.5 The Path Ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.1 INTRODUCTION
Academic research into robo-advisory is still nascent but developing
dynamically, especially in light of an increasing amount of attention
directed toward what is doubtlessly one of the ‘hot-topics’ in the field
of finance today. Inquiry can to date be clustered into four distinct streams
that are sequentially linked to each other: (1) A definitorial cluster that aims
to situate robo-advisory in the broader context of the financial industry
in general and financial technology in particular. (2) Analyses of strengths
and weaknesses of robo-advisory, especially in terms of differentiation
S. Krueckeberg ()
Krueckeberg Family Office, Hamburg, Germany

22 S. KRUECKEBERG
from each other and from legacy products and services. (3) Building on
analyses regarding strengths and weaknesses, a third strand of research asks
the question of what impact robo-advisory will thus have on incumbents.
(4) The fourth strand of inquiry, finally, aims to shed light onto the path
ahead for this emerging field. This chapter will synthesize and discuss
extant literature emanating from academic research into robo-advisory
and cluster key insights, according to the four streams of inquiry outlined
above, within distinct sections.
2.2 WHAT ARE FINTECH AND ROBO-ADVISORY,

REALLY?
Puschmann (2017) describes financial technologies as offshoots of an
ongoing macro process of economy-wide digitalization. The reasons why
the process of digitalization through the implementation of information
technology is developing a pronounced force within the realm of financial
services is seen in two characteristics. Firstly, in that financial products are
intrinsically based on information which, through the use of information
technology, can be channeled and used in countless new fashions. Secondly,
in that most processes within the financial industry can be implemented
almost entirely without physical contact. This latter argument regarding
physical interaction is particularly interesting within the scope of this
book, as the field of advisory presents a unique exception—a phenomenon
that will be discussed in many of the following chapters. The spread
of information technology within the financial industry is expected to
lead not only to the automation of certain elements of the value chain,
but to a fundamental reorganization of entire chains altogether. This
reorganization is observed to be driven by two dynamics: on the one hand,
by the market entry of new actors into financial services that are digital-
native at their core, such as Amazon, Apple and the like; on the other
hand, through the introduction of innovative new business models driven
by emergent technological solutions, such as robo-advisory. Ferrari (2016)
and Puschmann (2017) seem to agree that two of the key drivers behind
this trend are (1) changing consumer behavior, showing an increased
quantity and quality of interaction of consumers via digital channels,
and (2) changing regulation following the subprime crisis that imposed
significant financial and compliance burdens onto legacy players, putting
them at a disadvantage versus flexible digital challengers. Connecting to
SITUATING ROBO-ADVISORY 23
Puschmann, Ferrari (2016) expects the financial industry to transition

from a universal banking model to verticalization that allows focused
challengers to develop highly specialized niche capabilities in verticals such
as credit products, transactions, or robo-advisory. Evidence provided by
Dorfleitner et al. (2017) supports this proposition, showing that verticals
driven by financial technologies have shown average annual growth rates
of 150% over the years 2010 to 2016 in Germany alone. How, then,
is the scope of robo-advisory delineated within the academic discourse?
Sironi (2016) frames robo-advisory as an information technology–based,
automated investment advisory service. Following Maedche et al. (2016),
robo-advisors can be classified as based on interactive user-assisted systems
that possess a certain degree of context awareness. Taking these elements
together, robo-advisors help to create and manage investment accounts on
a fully automated basis (Jung et al. 2018). Reflecting on these foundational
concepts leads Jung et al. (2018) to conclude that robo-advisory embodies
the capability to transform the advisory process from a traditional human-
to-human setting into one that can potentially be entirely human-to-
machine-based. Key metrics that traditionally were compiled by a human
advisor are translated into digital routines for data gathering and translated
into investment advice along the typical investment dimensions of risk,
return, liquidity, and so on. This is seen as a catalyst for significant
cost reductions in advisory services, in effect lowering the barriers to
participation for individual investors—robo-advisory as a force toward
the democratization of asset management. While certainly an interesting
thought, challenging the structure of the established financial order will not
be an easy feat. Above all else, clear differentiation of the offering based on
leveraging strengths and compensating weaknesses will be a decisive factor
for the success or failure of challengers.
2.3 DIFFERENTIATION: STRENGTHS AND

WEAKNESSES
Robo-advisory as a new fintech enabled banking vertical is without doubt
an enticing proposition. However, new vertically specialized contenders
will need to differentiate themselves both internally from each other as
well as externally from traditional advisory to develop a unique value
proposition based on their strengths that can provide increased value to
customers. Turning first to internal differentiation among robo-advisors,
24 S. KRUECKEBERG
Sironi (2016) provides a concise overview that arrives at four key factors
along which propositions align on the competitive landscape. The first
factor is the degree of passive management. Some robo-advisors might
choose to manage portfolios entirely passively with a limited preset of broad
index products and very infrequent rebalancing by which portfolios are
recalibrated only rarely in reaction to movements in the market. Others
might take a more active approach that involves both more granular
picking of securities and more frequent rebalancing. This connects further
to the second element of differentiation, namely that of the depth of
investment automation. To what degree are investors involved in the
ongoing investment process, if at all? Some robo-advisors might involve
customers more actively in the investment process, while others might
advertise a hands-off solution. Third, such an investment process itself
will be guided and differentiated by some form of assessment mechanism
that aims to translate preferences and circumstances of the individual
investor into a tailor-made asset allocation. Thereby, precision and fidelity
of investor classification could provide a source for product differentiation.
This in turn connects to the fourth element, the target clientele. Some
robo-advisors might choose to focus on the tech-savvy Millenials that
value convenience above all else, others on the affluent baby boomers
which might have a different set of preferences and needs altogether. Some
might value virtual-only experiences, others require a hybrid approach—a
question that will be discussed in detail in Sect. II of this book. Wherever
robo-advisors choose to align along these four factors regarding internal
differentiation, common characteristics arise that can be isolated as factors
for external competition versus traditional advisory.
Analyzing competitive strengths, Jung et al. (2018) point out the
cost-competitiveness that digital services in general and robo-advisors in
particular can use to their advantage, a point that is driven particularly
by strong scalability of the service offering, which can be available 24/7.
Sironi (2016) concurs by emphasizing cost-efficiencies on the part of
such challengers as a key strength. Without having to finance costly real-
estate and an expansive network of advisory staff that is limited in both
accessibility and total hours of availability, robo-advisors are found to be
able to offer their services both at lower fees and with a lower minimum
threshold of invested capital (Jung et al. 2018). Meola (2017) shows
that minimum invested capital can be as low as US$ 0. In times of
increasing regulation and compliance requirements, customer on-boarding
has become a focal point for financial services companies. Sironi (2016)
finds advantages on the side of robo-advisors versus incumbents, not

least carried by technological innovations such as customer-not-present
video identification and intuitive self-profiling. This ties into the point of
overall simplicity, which is argued to be supportive for both the customer
journey and the overall innovation process in robo-advisory (Sironi 2016).
Jung et al. (2018) point out two further strengths that are somewhat
linked: automated rebalancing and tax-loss harvesting. Where customers of
traditional advisory might either engage themselves in active rebalancing
with the assistance of various digital and analogue tools or pay for human
advisors to coordinate rebalancing, robo-advisors are seen as reliable
executors of predefined granular rebalancing strategies. This ties into the
automation of tax-loss harvesting, through which automated algorithms
optimize rebalancing with the aim of harvesting advantages based on tax-
rate arbitrage.
Turning to current weaknesses, Fein (2015) studies the user agreements
and/or disclosure brochures of three leading US robo-advisors in detail
and paints a more critical picture of the robo-advisory landscape. A coun-
terpoint to the cost-efficiency argument is supplied in that while customers
might not be required to explicitly pay for advisory services consumed as
such, costs associated with these services are later recovered and thereby
paid for by the consumer through third-party fees such as brokerage costs,
transaction fees, and other expenses. This might occur through outright
fees, all-in fees, or revenue-sharing arrangements with third parties. This
links to a related point of criticism regarding the independence from
conflicts of interest on the part of robo-advisors. Fein lays out that within
the scope of her study, explicit mention was made on the parts of robo-
advisory firms about being tied to affiliated brokers that might not provide
best execution of clients’ trading orders. The robo-advisor’s own profit
motive might stand against best execution of its customers. Furthermore,
some robo-advisors disclose and request consent for proprietary trading
and reserve the right to recommend investments to customers that are
being held in proprietary portfolios. This can lead to biased investment
recommendations especially paired with a further point of criticism, namely
some robo-advisors reserving the right to aggregate and pool client orders
as well as to execute in the form of cross transactions. All these points in
conjunction do paint a picture of potential for conflicts of interest. Two
final points of criticism are that robo-advisors do not truly provide tailor-
made investment advice, which is tied to further criticism regarding the
assessment of individual risk tolerance and the level of overall personaliza-
26 S. KRUECKEBERG
tion of the user experience (Fein 2015). Tertilt and Scholz (2018) connect
to Fein’s research and supply supporting evidence with an analysis of the
robo-advisory market, specifically, how robo-advisors translate individual
customers’ risk tolerance into equity exposures. Generating a proprietary
data sample comprising robo-advisors from Germany, the UK, and the US,
Tertilt and Scholz (2018) show that robo-advisors ask only relatively few
questions, and, of these, only about 60% have any impact on the actual
risk categorization. Customers seem to receive only little guidance in the
process of assessing their idiosyncratic risk budget and risk appetite (Tertilt
and Scholz 2018). Faloon and Scherer (2017) provide further support for
this thesis by summarizing questions typically used by most robo-advisors
and thereby concluding that most of the generated advice is generic and
does not qualify as personalized advice. In light of current strengths and
weaknesses, what can the potential impact of robo-advisory be on legacy
players and how can established advisory providers react to challengers?
2.4 IMPACT ON INCUMBENTS

Who are the early adopters making use of these new digital service offer-
ings? Woodyard and Grable (2018) study the characteristics of customer
segments that exclusively use either robo-advisory or traditional advisory.
As could be expected, they find that individuals who exclusively use robo-
advisors as their service provider are commonly young, confident in their
own abilities, and distrustful of traditional channels of financial advice.
Epperson et al. (2015), based on their study among 4000 US consumers,
support these results by reaching a similar conclusion. Whether this consti-
tutes the beginning of a broader trend for future generations to increasingly
transition toward purely digital service offerings or whether preferences will
shift when today’s youth become tomorrow’s seniors remains to be seen.
Whichever the trajectory will be, incumbents are today confronted with
the emergence of competing value propositions. How should incumbents
react? Cocca (2016b) studies the impact of virtual advisory models onto
traditional advisory, based on a survey among wealth management clients
in Austria, Germany, and Switzerland. Considering what he calls the
potential virtualization of advisory, Cocca arrives at the conclusion that
the competition between traditional advisory incumbents and digital chal-
lengers will play out along a complexity versus standardization spectrum.
According to this approach, challengers will migrate some service offerings
into digital processes that are highly standardized and automated, such as
risk-profiling and/or portfolio management. Traditional advisors in turn
are expected to be forced into more complex advisory services such as
international asset structuring and/or tax advisory (Cocca 2016a). Jung
et al. (2018) provide findings that can be connected to Cocca, in that robo-
advisors are expected to focus on individuals with below-average to average
income as traditional advisory is expected to focus on the upper end of the
spectrum. Summarizing both sources, this would mean that robo-advisory
might find success in focusing on standardized and automated low-cost
services, while traditional advisory might focus on higher-cost/higher-
margin complex advisory services. Cocca (2016a) adds to this, concluding
that wealth management specifically might find a winning proposition in a
hybrid approach. To be able to engage customers with a hybrid strategy,
Cocca recommends incumbents not to understand fintechs as competitors
but as potential integration partners (Cocca 2016b). Gold and Kursh
(2017) concur with the idea of integrating fintechs into incumbent organi-
zations and add three further strategic choices, using D’Aveni’s framework
of strategic options for incumbents as their reference (D’Aveni 2002).
Besides acquiring challengers, incumbents might launch their own digital
offerings, leveraging internal resources and thereby building on established
customer relationships. Or they might partner with challengers using
their value proposition as a complement to traditional services. Finally,
incumbents can also choose to match parts of the offering of challengers,
that is, by lowering fees, thereby eliminating an important incentive for
customers to switch service providers. Overall, Gold and Kursh (2017)
conclude that taking a wait-and-see approach might be the most favorable
course of action in light of regulatory circumstances and lock-in effects for
incumbents due to proprietary customer data. Whether proprietary data in
and of itself will remain a competitive factor in the emerging Application
Programming Interface (API) economy, which is increasingly based on
the free exchange of information, remains to be seen. Gold and Kursh
recognize this, especially in light of recent European legislation, not least
the Payment Services Directive II (European Parliament and the Council
of the European Union 2015), mandating banks to grant fintechs access
to their customer data via API.
28 S. KRUECKEBERG
2.5 THE PATH AHEAD

Which developments and success factors does the academic literature antic-
ipate for the future? Arwas and Soleil (2016) anticipate the next-generation
Robo-Advice 2.0 and expect success factors not to lie in technology itself
but rather in the creation of personalized and differentiated offerings
that are geared toward serving the entirety of the customer life cycle.
Among the factors that will play a crucial role in robo-advisory being
able to transition into the mainstream over the coming years, Arwas and
Soleil (2016) highlight simplicity in both language and process. Moreover,
emphasis is put particularly on transparency regarding costs—a thought
that can be seen as linking back to Fein’s 2015 critique of robo-advisory,
as discussed above. Key emphasis on transparency is a thought that Jung
et al. (2018) echo with the results of their study. Designing a robo-advisor
within a laboratory environment, Jung et al. analyze the requirements of
both experts and potential customers for robo-advisory solutions. To do
this, first, design principles are extracted from existing literature, which are
then, second, validated via laboratory experiments. Four design principles
are summarized that have in the past been suggested as fundamental
to digital financial advisory success. These are ease of interaction, work
efficiency, information processing, and transparency. Sources for these
principles are studies conducted by Koerner and Zimmermann (2000),
Moewes et al. (2011), Nussbaumer (2012), Nueesch et al. (2014), Kilic
et al. (2015), and Ruf et al. (2015). Jung et al. (2018) break down the
four design principles into design requirements and thereafter analyze the
appropriateness of each factor in the so-called ‘design cycles’. The first cycle
includes expert interviews, the second and third cycles comprise a mixed-
method experimental usability study in the laboratory, where prospective
customers are analyzed for their likelihood to adopt the prototypical
solution. By doing this, Jung et al. (2018) find evidence that all four
design principles indeed are appropriate guiding elements to build future
digital advisory services around and that transparency can be identified
as the single most important aspect. Regarding the specific variables to
structure the advisory process itself around, Scherer (2017) conducts a
panel survey with German households and finds two factors that strongly
seem to influence portfolio choice. The first factor relates to the household
balance sheet variables, that is, net worth. The second factor revolves
around household personal characteristics such as the overall level of risk
aversion. Scherer concludes by recommending future development of the
robo-advisory process to be centered around these two key factors (Scherer

2017).
Which direction will future research into robo-advisory take? Glaser
et al. (2019) are setting out to conduct a laboratory experiment with 200
individuals to gain insight into the optimal design of robo-advisors for the
purposes of determining appropriate ways to present risk to investors. It is
the aim to generate an understanding of how different types of individuals
understand the concept of risk. Individuals will be tested regarding their
grasp of statistical properties and their capability to regulate their own
emotions. To allow conclusions, Glaser et al. (2019) hope to measure the
confidence, decision times ,and heart rates of participants and to measure
changes depending on how and how much information is presented.
This volume provides research that connects to many of the strands of
inquiry outlined in the review above and develops additional specialized
fields that go to the heart of what characterizes robo-advisory. Starting with
a review of the performance of robo-advisory solutions, insights are then
provided into goal-based investing and rule-based investing in the context
of automated digital services. How can robo-advisory devise and execute
appropriate strategies in a context of goals and rules? Contributions are also
made to the question of how robo-advisors can enable customers’ own-
ership of the decision-making process when structuring portfolios. This
question is linked to the following contribution shedding light onto the
topic of portfolio recommendations. Which asset classes lend themselves to
be included in robo-advisory recommendations how granular should asset
classes be sub-segmented ,and which are appropriate rebalancing intervals?
Finally, forays are made into potential future regulation of the advisory
space, sales strategies along the human versus machine continuum, as well
as bleeding-edge aspects such as artificial intelligence, big data, and social
networks.
REFERENCES
Andrew Arwas and Katie Soleil. 2016. “Robo-Advice 2.0: The Next Generation”.
Journal of Financial Transformation 43: 30–36.
Teodoro Cocca. 2016. LGT Private Banking Report 2016. Tech. rep. LGT Group
Foundation.
Teodoro Cocca. 2016. “Potential and Limitations of Virtual Advice in Wealth
Management”. Journal of Financial Transformation 44: 45–57.
30 S. KRUECKEBERG
Richard D’Aveni. 2002. “The Empire Strikes Back. Counterrevolutionary Strate-

gies for Industry Leaders”. Harvard Business Review 80(11): 66–74.
Gregor Dorfleitner Lars Hornuf, Matthias Schmitt, and Martina Weber. 2017. Fin-
Tech in Germany. Springer.
Teresa Epperson, Bob Hedges, Uday Singh, and Monica Gabel. 2015. Hype
vs. Reality: The Coming Waves of “Robo” Adoption. Research rep. AT Kear-
ney. https://www.atkearney.de/financial-services/article?/a/hype-vs-reality-
the-coming-waves-of-robo-advisors.
European Parliament and the Council of the European Union. 2015. Directive
(EU) 2015/2366 of the European Parliament and of the Council of 25 November
2015. Official Journal of the European Union. https://eur-lex.europa.eu/
legal-content/EN/TXT/PDF/?uri=CELEX:32015L2366&from=EN.
Michael Faloon and Bernd Scherer. 2017. “Individualization of Robo-Advice”. The
Journal of Wealth Management 20(1): 30–36. https://doi.org/10.3905/jwm.
2017.20.1.030.
Melanie L. Fein. 2015. “Robo-Advisors: A Closer Look”. SSRN Electronic Journal.
https://doi.org/10.2139/ssrn.2658701.
Roberto Ferrari. 2016. “FinTech Impact on Retail Banking—From a Universal
Bank ing Model to Banking Verticalization”. In The FinTech Book: The Financial
Technology Handbook for Investors, Entrepreneurs and Visionaries. ed. by Susanne
Chishti and Janos Barberis. Wiley Online Library, pp. 248–252. https://doi.
org/10.1002/9781119218906.ch65.
Florian Glaser, Zwetelina Iliewa, Dominik Jung, and Martin Weber. 2019.
“Towards Designing Robo-advisors for Unexperienced Investors with Experi-
ence Sampling of Time-Series Data”. In Information Systems and Neuroscience,
pp. 133–138. Springer.
Natalia A. Gold and Steven R. Kursh. 2017. “Counterrevolutionaries in the Finan-
cial Services Industry: Teaching Disruption—A Case Study of RoboAdvisors and
Incumbent Responses”. Business Education Innovation Journal 9(1): 139–146.
Dominik Jung, Verena Dorner, Florian Glaser, and Stefan Morana. 2018. “Robo-
Advisory: Digitalization and Automation of Financial Advisory”. Business &
Information Systems Engineering 60(1): 81–86.
Dominik Jung, Verena Dorner, Christof Weinhardt, and Hakan Pusmaz.
2018. “Designing a Robo-Advisor for Risk-Averse, Low-Budget Consumers”.
Electronic Markets 28(3): 367–380. https://doi.org/10.1007/s12525-017-
0279-9.
Dominik Jung, Edgar Erdfelder and Florian Glaser. 2018. “Nudged to Win–
Designing Robo-Advisory to Overcome Decision Inertia”. In Proceedings of
the 26th European Conference on Information Systems (ECIS 2018), pp. 23–28.
https://aisel.aisnet.org/ecis2018%5C_rip/19.
Mehmet Kilic, Peter Heinrich, and Gerhard Schwabe. 2015. “Coercing into Com-
pleteness in Financial Advisory Service Encounters”. In Proceedings of the 18th
ACM Conference on Computer Supported Cooperative Work & Social Computing,
pp. 1324–1335. ACM. https://doi.org/10.1145/2675133.2675289.
Veith Koerner and Hans-Dieter Zimmermann. 2000. “Management of Customer
Relationship in Business Media—The Case of the Financial Industry”. In HICSS
’00 Proceedings of the 33rd Hawaii International Conference on System Sciences,
Vol. 6, p. 6045. IEEE. https://doi.org/1.1109/HICSS.2000.926866.
Alexander Maedche, Stefan Morana, Silvia Schacht, Dirk Werth, and Julian Krume-
ich. 2016. “Advanced User Assistance Systems”. Business & Information Systems
Engineering: The International Journal of WIRTSCHAFTSINFOR MATIK
58(5): 367–370. https://doi.org/10.1007/s12599-016-0444-2.
A. Meola. 2017. Robo Investing Reviews: Performance & Fees Comparison. Tech.
rep. Business Insider.
Till Moewes, Thomas Puschmann, and Rainer Alt. 2011. “Service-Based Integra-
tion of IT-innovations in Customer-Bank-Interaction”. In Wirtshaftinformatik
Proceedings 2011. Paper 102, ed. Bernstein, A., Schwabe, G. http://aisel.aisnet.
org/wi2011/102.
Rebecca Nueesch, Thomas Puschmann, and Rainer Alt. 2014. “Realizing Value
From Tablet-Supported Customer Advisory: Cases From the Banking Indus-
try”. In BLED 2014 Proceedings, p. 34. https://aisel.aisnet.org/bled2014/34.
Philipp Nussbaumer Inu Matter Gian Reto a Porta, and Gerhard Schwabe. 2012.
“Designing for Cost Transparency in Investment Advisory Service Encoun-
ters”. Business & Information Systems Engineering: The International Journal
of WIRTSCHAFTINFORMATIK 4(6): 347–361. https://doi.org/10.1007/
s12599-012-0237-1.
Thomas Puschmann. 2017. “Fintech”. Business & Information Systems Engi-
neering: The International Journal of WIRTSCHAFTSINFORMATIK 59(1):
69–76. https://doi.org/10.1007/s12599-017-0464-6.
Christian Ruf, Andrea Back, Richard Bergmann, and Michael Schlegel. 2015.
“Elicitation of Requirements for the Design of Mobile Financial Advisory
Services–Instantiation and Validation of the Requirement Data Model with
a Multi-Method Approach”. In 2015 48th Hawaii International Conference
on System Sciences, pp. 1169–1178. IEEE. https://doi.org/10.1109/HICSS.
2015.142.
Bernd Scherer. 2017. “Algorithmic Portfolio Choice: Lessons from Panel Survey
Data”. Financial Markets and Portfolio Management 31(1): 49–67. https://
doi.org/10.1007/s11408-016-0282-8.
Paolo Sironi. 2016. FinTech Innovation: From Robo-Advisors to Goal Based Investing
and Gamification. John Wiley & Sons.
32 S. KRUECKEBERG
Michael Tertilt and Peter Scholz. 2018. “To Advise, or Not to Advise—How Robo-
Advisors Evaluate the Risk Preferences of Private Investors”. The Journal of
Wealth Management 21(2): 70–84.
Ann S. Woodyard and John E. Grable. 2018. “Insights into the Users of Robo-
Advisory Firms.” Journal of Financial Service Professionals 72(5): 56–66.
PART II
Implementation of Robo-Advisory
CHAPTER 3
Risk Preferences of Investors
Monika Mueller, Paul Resnik, and Craig Saunders
Contents
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2 Risk Profiling Explained . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.1 Financial Risk Tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.2 Risk Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2.3 Risk Required . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.2.4 Risk Profiling Is Part of Knowing Your Client . . . . . . . . . . . . . . . . . . . . 40
3.3 Choosing a Valid and Reliable Risk-Tolerance Test . . . . . . . . . . . . . . . . . . . . . . 40
3.3.1 Psychometrics Versus Gambles-Based Testing . . . . . . . . . . . . . . . . . . . . 41
3.3.2 The Science of Prospect Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3.3 The Science of Psychometrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.4 Risk Profiling Is Harder for Robo-Advisors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.5 Mapping Risk Profiles to Investments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
M. Mueller ()
Founder, FCM Finanz Coaching, Wiesbaden, Germany
P. Resnik
Co-Founder, FinaMetrica Risk Profiler, Caulfield North, VIC, Australia
C. Saunders
Editor, Financial Suitability Forum, Caulfield North, VIC, Australia

36 M. MUELLER ET AL.
3.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.1 INTRODUCTION
“Risk preferences”, as a term, are an apparition. They permeate many
discussions about financial advice all around the world. Yet, as you approach
them, they turn to mist. There is no consensus on a single definition of risk
preferences and therefore they do not have a real meaning—each author
shapes them to fit their own purpose. This problem of having a poor
definition arises because “risk preferences” have emerged as an explanation
of why something is not, rather than to define something that is. The
problem that risk preferences need to solve is the failure of some retail
investors to act rationally when it comes to risk, as the dominant “expected
utility” theory said they should (Cohn et al. 1975). Sometimes, they
would be more risk-averse than the expected utility model would predict.
Clearly, there is something else at play. That something else has evolved
to become known as risk preferences. But the term has become a catch-
all and melting pot of various explanations, motivations, and observations
that might help to explain the investor’s aversion to maximize returns
by accepting risk. Today, the melting pot still contains the same basic
ingredients, which include both psychological and financial elements. But
its name has changed. In commercial settings, the term risk profile has
largely superseded and replaced risk preferences in the discussion; it also
accommodates factors that might influence and/or constrain the investor’s
willingness to accept risk (Klement 2015). However, while the name has
changed, the problems of having a poor definition and inconsistent usage
persist, with risk profile and risk profiling commonly meaning different
things to different people (Moore 2017). Even regulators, who require
risk profiling to be conducted, are inconsistent in their rules about what
should be included and how profiling should be performed. For example,
even though Australian regulations require risk profiling, the Financial
Ombudsman service recognizes that experienced advisors may be able
to secure a client’s informed consent even without using a risk-profiling
tool (Financial Ombudsman Service 2011). We, therefore, think it useful
to provide our definitions of the key risk terms that will be used in our
RISK PREFERENCES OF INVESTORS 37
discussions of risk profiling and align with those presented in a review of

global risk profiling best practices (Brayman et al. 2015):
Risk Tolerance
Risk tolerance is the foundation stone of a risk profile. Risk tolerance
describes the willingness or aversion of the client to take on investment risk.
It is a constant psychological trait that can be measured using psychometric
testing.
Risk Capacity
Risk capacity is an accounting measurement of the investor’s financial
ability to endure investment losses. The question of what is “endurable” is
usually measured against whether losses would result in a requirement for
any adverse lifestyle changes or abandonment of goals.
Risk Required
Risk required, also referred to as “risk needed”, is a mathematical cal-
culation of the level of risk necessary to produce the investment returns
necessary to achieve goals. A person with large goals but small assets may
require larger amounts of risk than another with the same goals who is
starting with more substantial assets.
Risk Composure
This is an estimation of the likelihood that a client perceiving a crisis
might act differently than their rational selves, leading them to liquidate
investment positions and crystallize losses.
Risk Profile
The risk profile aggregates the preceding factors, along with other sec-
ondary factors, to determine the optimal level of risk that maximizes the
client’s potential to reach their goals while remaining consistent with the
level of risk they can financially afford to take and are psychologically willing
to accept.
3.2 RISK PROFILING EXPLAINED

Broadly, a person’s risk profile should measure and record their emotional
and financial capacity to take investment risks. Preparing an accurate profile
is a multidimensional process by combing both objective and subjective

elements. An investor’s risk profile plays a critical role in making investment
decisions and directly influences asset allocations (Droms and Strauss
2003). It is, therefore, of primary importance in making investment rec-
ommendations and giving financial advice. A risk-tolerance measurement
on its own does not tell the full story of a person’s risk preferences which
might better be called their “risk profile”. There are three critical core
components that intersect to form a person’s risk profile:
1. Risk tolerance
2. Risk capacity
3. Risk required
A person’s risk profile emerges from the interaction between these core
aspects, together with other considerations. These include the investment
time horizon, which can impact risk capacity, or the knowledge of financial
products. The knowledge factor is controversial as it is unconnected with
risk tolerance or capacity; it is difficult to quantify and can be readily
ameliorated by instruction and explanation.
3.2.1 Financial Risk Tolerance

Financial risk tolerance is a measure of a person’s ability to remain invested
in assets despite sharp changes in value. Remaining invested throughout
volatility is important as it keeps exposure to markets and avoids material-
izing losses. Financial risk tolerance is a psychological trait that is stable,
constant, and predictable; it is largely beyond our own direct control.
People do not generally change between high-risk and low-risk tolerance,
just as today’s extrovert is unlikely to become tomorrow’s introvert or vice
versa. Risk tolerance tends to be constant over time, except when dramatic
events lead to a change in the core personality (Van de Venter et al. 2012).
A European study of risk preference in 1507 adults subjected them to
39 different risk-profiling test instruments, with a retest of a sample of
109 participants six months later, thereby confirming the stability of risk
tolerance when revealed by psychometric testing (Frey et al. 2017).
A risk-tolerance score is an important starting point for mapping risk
tolerance to investment decisions. But ultimately, a person may quite
legitimately make investments that depart from those suggested by their
risk tolerance score. For example, someone with a lower risk-tolerance

score might decide to take on a higher level of risk than they would
normally be comfortable with because not doing so would mean they
cannot possibly reach their goals. Conversely, a person with high-risk
tolerance might hold low-risk investments if they already have enough
money to meet their needs and goals. Meanwhile, a person, regardless of
risk tolerance, might not be able to afford to bear any financial loss, so they
restrict themselves to low-risk, capital-guaranteed investments. There are
several approaches to determine a risk score, with some of the methods
subject to highly critical concerns about their validity and reliability. These
concerns will be explored when we discuss the criteria for selecting a risk-
tolerance assessment methodology.
3.2.2 Risk Capacity

Risk capacity is the amount of loss a person can incur without jeopardizing
the achievement of important financial goals. It is an accounting measure-
ment that is uncorrelated with risk tolerance—a person with a high level
of risk capacity may have a low-risk tolerance and vice versa. Factors which
influence risk capacity include the following:
• Age and work status, which can directly impact investment time
frames
• Current assets
• Current income
• Future income requirements
• Future inflows and outflows of cash
• Future resources such as pensions or inheritances
• Future liabilities such as taxation, aged care, and family care
There is a lot of interactivity and discretion within these variables. The

investor has a lot of direct control over their future requirements and
situation. For example, they may be able to adjust down their spending
expectations or increase income by choosing to delay their retirement for
a number of years.
3.2.3 Risk Required

Risk required is the amount of investment risk needed to generate the
returns required to meet the investor’s objectives, given their starting
position. It is an analytical calculation. A simple example demonstrates
the principle more easily than a lengthy explanation. Assume a person
wants to have $ 110 a year from now, but today they only have $ 100.
So, they require a risk/return of 10% to achieve their objective. But if their
starting point is only $ 95, the risk/return required increases to 15.8%.
Risk required is dynamic and controllable as different variables can be
manipulated by the investor. Returning to our simple example, this investor
could reduce the risk required by lowering their return expectations from
$ 110 to $ 108, or they can simply extend their investment horizon from
one year to two.
3.2.4 Risk Profiling Is Part of Knowing Your Client

Risk profiling sits within the “know-your-client” (KYC) requirements for
financial advice and product sales that are well established in the United
Kingdom (Financial Conduct Authority 2006), Europe (European Securi-
ties and Markets Authority 2018), the United States (Financial Industry
Regulatory Authority 2011), Canada (Fung et al. 2013), and Australia
(Australian Securities & Investments Commission 2017). The KYC rules
are universally based on the premise that providers who know certain details
about a client will be better able to provide advice and products that are
suitable for that client, given their unique personality, circumstances and
goals. However, the regulations are principle-based and lacking in detail.
They speak of the outcomes or standard that should be met but generally
say very little about the methodologies or processes that might lead to
those goals. Risk tolerance is a key example. All jurisdictions require it to be
assessed, yet none explains how to measure it, how to interpret the results,
or how it should be “taken into account” while making a recommendation.
3.3 CHOOSING A VALID AND RELIABLE

RISK-TOLERANCE TEST
Firms using risk-tolerance tests must ensure that the testing instrument
they are using is fit for the purpose—it should be both valid and reliable. To
be valid, a test must be true to the label, which means it must measure what
it claims to measure—in this case, financial risk tolerance. To be reliable,

the test results must be largely consistent regardless of how, where, or
when the test is administered (Grable and Lytton 1999). Many of the risk-
tolerance tests in use today fail to meet either of these standards. A 2015
review of global best practices found that only 16.7% of the questionnaires
reviewed would be considered fit for the purpose. Around 80% of the tests
had been developed in-house or were left to individual advisors’ discretion.
Regulators typically require that advice-givers conduct initial and ongoing
reviews of the tools used in the advice process (Financial Industry Regula-
tory Authority 2016). That review should include questions covering the
following:
• Whether methodologies and assumptions are well suited to the task

• If methodologies are tested by independent third parties
• What data points will be used and how
• How outputs will be tested to ensure they conform with expectations
• Whether the models used remain appropriate as market conditions
change
• How the tool identifies and deals with exceptions and inconsistencies
Very few in-house risk-tolerance test instruments have been developed

according to such strict criteria. The 2015 global review (Brayman et al.
2015) found that less than 10% of firms knew if their test had been validated
by a statistical check of reliability. Less than 20% had engaged any outside
expertise to assist in developing their test instruments.
Many in-house questionnaires fail to meet the criteria for validity as they
contain questions that pertain to something else other than risk tolerance,
such as risk capacity. Reliability is also a problem as many tests do not ask
enough questions to produce meaningful results (Roszkowski et al. 2005).
3.3.1 Psychometrics Versus Gambles-Based Testing

Two dominant methodologies are used by commercial providers of risk-
tolerance tests:
1. Psychometric testing of risk tolerance as a personality trait, drawing

on an extensive body of psychological research around psychometrics
(Rust and Golombok 2009);
2. Gambles-based questionnaires about wins versus losses, drawn from

the prospect theory (Kahneman and Tversky 1979).
The divide is sharp—providers’ products tend to be rooted in either

psychometric or gambles-based methodologies, but generally not both.
To express a belief in one methodology is essentially to express disbelief
in the other. This does, in fact, make intuitive sense. Supporters of
psychometrics argue that the task tests a stable psychological trait, while
gambles-based supporters argue that the task determines preferences from
a series of choices. They are very different approaches. The debate over
which methodology is effective and/or superior is just as sharp with
somewhat of a “tribal” flavor. There is little overlap between the two, and
the debate tends toward polarization. This debate has been going on for
many decades. A series of due diligence questions about a risk-tolerance
test has been developed by a team, including psychometrics specialists,
from the London School of Economics (Resnik 2016). It puts forward
questions to be asked about the face validity of the test; expertise of the
people who created it; independent expert scrutiny and testing; reliability
of the assumptions relating to constructs underpinning the test; and the
usability of the test for the client (Erskine 2016).
3.3.2 The Science of Prospect Theory

Kahneman’s and Tversky’s Nobel prize-winning prospect theory is a
groundbreaking work that has amassed many hundreds of thousands of
academic citations. Prospect theory describes how people choose between
probabilistic alternatives involving risk where the probabilities of outcomes
are uncertain. It says that people use heuristics or mental shortcuts to evalu-
ate gains and losses, which they value more highly than the final outcome.
The theory was developed partly to address the failure of the expected
utility theory to accurately describe consumer behavior. It is seminal in
the field of behavioral economics and widely respected. But the use of
prospect theory in the assessment of financial risk tolerance is controversial,
with concerns about both validity and reliability. A key validity concern is
the integrity of data gathered through gambles-related questions, given
the difficulty that many people face in providing a meaningful answer to a
question involving complex mathematical topics such as probability.
Much of the population lack the capacity to complete a calculation of

this type, while others will avoid the question because of the “math anxiety”
it provokes (Ashcraft 2002). In the absence of a meaningful response,
the investor may resort to guessing an answer, thereby undermining the
efficacy of the test (Baloğlu 2004). Research has also shown that even
in simple lotteries people tend to rely on heuristics, rather than making
the necessary calculations, to understand the risk and reward relationship
(Cokely and Kelley 2009). The hypothetical nature of the gamble is also a
concern for the validity. Research has shown that people can behave very
differently when stakes are real and meaningful compared to stakes that are
hypothetical and immaterial (Barreda-Tarrazona et al. 2011)—this raises
doubts over the answers to gambles-related questions (M. Levy and H.
Levy 2002). Reliability is also an issue for gambles-based risk-tolerance
tests, with anecdotal evidence of variances in results upon retest.
Meanwhile, some in the financial services sector feel some kind of basic
discomfort in forming an association between gambling and investing in
their clients’ minds. Many have fought hard to distance financial advice
from the “stock-market is a casino” idiom, and they have been reluctant to
introduce “gambles” into a disciplined investment process.
3.3.3 The Science of Psychometrics

Psychometric testing of risk tolerance has been subject to extensive inde-
pendent reviews and testing; it has been shown to be a valid and reliable
methodology (Hallahan et al. 2003). Psychometrics is an amalgam of
psychology and statistics that can quantify and assess psychological traits
and constructs. While its heritage has some controversy, it is today a well-
established and accepted science with validation tools to determine the
technical quality of psychological assessment tools like questionnaires.
However, many questionnaire tools that claim to be psychometric are
not so—indeed, they may not have any scientific basis at all. For example,
the Myers–Briggs Type Indicator (MBTI) is a self-report questionnaire
which is the most widely used personality test in the world, with around
two million people a year undertaking the questionnaire. It is commonly
used during recruitment processes to identify “personality types” that
explain how people perceive as well as respond to the world around them.
Unfortunately, this test consistently fails the key scientific requirements
of psychometrics as it displays poor validity and reliability. Poor validity
manifests as not measuring what the test proposes to measure, not having
any predictive power, or not having any items that can be generalized. Poor
reliability means that the test can produce different results for a person
taking the test at different times and not being comprehensive (Grant
2013). These problems can be traced to the development for the MBTI
test, which did not use any scientific processes. This test was developed by
Katharine Cook Briggs and her daughter, Isabel Briggs Myers, who based
their work on the theories of the psychiatrist Carl Jung, whose work was
observational and deductive. Katharine began developing her personality-
types framework in 1917 based on her own reading of the literature as
an introspection. However, her work then languished for more than 20
years. Katharine and her daughter did not create a test until 1944, when
it found traction as a tool to help place women into appropriate industrial
jobs in America during World War II. True psychometric testing follows a
scientific development process, which includes extensive testing of the test
instrument (the questions) and the “proving” of results for validity and
reliability.
Psychometric testing of financial risk tolerance emerged in 1999 with
the development and launch of FinaMetrica (T. G. Davey 1999)—the
first custom-built psychometric test for risk tolerance that was launched
in Australia and is now used globally in more than 20 countries. Other
providers of psychometric tests have since entered the global market. The
FinaMetrica test was developed using a scientific process. More than 100
questions were tested in rigorous academic environments and processes,
with most being discarded as they were not dependable in producing valid
and reliable results. The final test was reduced to 25 questions where
confidence about validity and reliability was very high. The FinaMetrica
psychometric test of financial risk tolerance produces a “score”, which
places the investor with a continuum of range from no risk at all at one
extreme to 100% risk appetite at the other. But in reality, almost no one
ever appears at those extremes. Most people are fairly similar and tend
to cluster around the middle, with only a few outliers showing up near
the very conservative or very risky ends. This is a “bell curve” or normal
distribution pattern (Katsikatsou and Erskine 2018). The curve shows
that 86% of people fall within one standard deviation of the “middle”
(Fig. 3.1). But while the differences between people may appear small in
statistical terms, they can nonetheless become very significant in real life—
a person with a score of 37 will be very different to one with a score of
57, even though both are equally close to the middle. Almost 1.5 million
Fig. 3.1 Financial risk tolerance is a stable psychological trait which follows a
normal distribution pattern, where most people cluster around the middle peak
(the “norm”) with reducing numbers of outliers at each end (the very risk averse
and extreme risk seekers)
people have had their risk tolerance measured using the FinaMetrica test
instrument. This data, after being deidentified, has been made available for
academic analysis, testing, and review. It has been verified for validity and
reliability.
3.4 RISK PROFILING IS HARDER FOR ROBO-ADVISORS

Robo-advisors must comply with the same rules and regulations as other
financial advice systems, with all major regulators having made statements
that the rules governing financial advice are the same regardless of how it is
delivered. This means human-driven advice systems, cyborg advice systems
(where a human is augmented by, or augments, technology), and robo-
advisors must all meet the same standards for risk profiling. Everyone faces
the same initial hurdle: selecting a fit-for-purpose risk-tolerance assessment
methodology and instrument on which to build the profiling process. But
the risk-profiling job is harder for robo-advisors, and some have failed to
meet the required standards (Smith 2016).
Challenge 1: Understanding and Empathy

The first challenge the robo faces is the “warm body effect” (Fisch et al.
2018). The robo suffers from the lack of insight and flexibility that a
human can bring. A human advisor brings their presence, expert judgment,
and emotional intelligence to an encounter with a client. They “read”
the client’s speech, actions, and body language and inevitably make many
minor adjustments in how the meeting proceeds. Based on what they see
and hear, they might pause to spend extra time on helping clients with a
matter of financial literacy, consider adjusting goals or attitudes, or address
apparent inconsistencies in the facts before them. A robo-advisor, however,
is the opposite in virtually every respect. A human knows a client as a
person. But a robo-advisor only knows a client as a set of data points.
For a robo-system, the only inputs from the client are predefined and
narrow. Often, they must select one of four prewritten answers as their
“best fit” answer to a question. The robo cannot “see” or otherwise intuit
how the client looks, sounds, or seems as they provide the answer. The
only input is that single “best-fit” data point. So, the robo-advisor cannot
respond dynamically as a human might. A human might sense that help is
needed and accordingly offer it. But a robo-advisor will only offer help
when the user specifically asks for it or a pre-coded alarm is triggered
by the answer to a particular question. The immediate problem is not
how to make robo-advisors more human because research has shown that
this can actually diminish the client’s confidence in the advice given by
the robo (Hodge et al. 2018). Intriguingly, the research found people
are more prepared to follow advice from a robo when it exhibits fewer
human characteristics. Rather, the immediate problem is how to equip
robo-advisors to understand and deal with the subtleties and nuances that
a client might experience during an automated encounter. A research study
(Glaser et al. 2019) was announced in 2018, focusing on robo-advisors that
can “speak” to investors in plain language in an intuitive way and “listen” to
the investor by monitoring their emotional reactions. The answers will lie
in creating “smarter” robos with sophisticated algorithms that can create
a “view” of the investor as a person whose sum is greater than the parts of
their financial data points.
Challenge 2: Building More Comprehensive Algorithms

Building more comprehensive algorithms is hard. It means dealing with
known knowns, known unknowns, and unknown unknowns to be con-
sidered and accommodated. The mapping can quickly become like a
spider web. These expanded algorithms will almost inevitably require extra
questions, inputs, or steps. To the designers of online experiences, this is
anathema. The tech mantra is to simplify and reduce rather than expand-
ing to add complexity. However, financial advice is an area that defies
simplification. Risk-tolerance tests are a good example. A psychometric
risk-tolerance test may contain as many as 25 questions. The test is designed
to capture the required data, but it will also ensure internal consistency
and identify unusual or atypical responses. It requires a large number of
questions for completion of all that work. But robo-advice risk-tolerance
questionnaires can be as short as just five questions. It is difficult to see
that the robo’s algorithm can be sufficiently informed by a dataset so
small. The problem is not one of intent—the designers of robo-advisors
are responding to solid evidence that users of online services want the
quickest possible outcomes with the least possible work. “Friction” is to
be eliminated. But in a highly regulated, closely monitored industry like
financial advice, or aviation, or medicine, there are often no shortcuts.
A good process requires a good process, which often requires numerous
steps.
Challenge 3: Applying “Professional Judgment”

A human advisor uses their presence, insight, and emotional intelligence to
inform the exercise of their professional judgment—often on a case-by-case
basis. This approach can introduce problems of inconsistency. But it also
creates a greater opportunity to shape the advice to the individual receiving
it. How a robo-advisor might approach and apply professional judgment
is a key issue to be resolved. Will it be done at all? Should it be done at a
“house” level where all advice prioritizes the one variable over others? Or
should it be a highly individualized experience, reflecting a sample of only
one? Once the approach has been decided, it can be codified into a robo-
advisor, which should then produce consistent outcomes based on the
algorithm. This can overcome the problems of variability and inconsistency
that may arise among human advisors.
3.5 MAPPING RISK PROFILES TO INVESTMENTS

A risk profile’s purpose is to guide investment decisions. However, great
care must be taken when mapping a particular risk “score” to appropriate
investments. The most immediate danger is that the labels and terminology
around risk cannot be trusted or relied upon, as most of them have no

defined or agreed meaning. Terms like these can mean almost anything at
all—it all depends on how the particular product provider defines it:
• Conservative
• Aggressive
• Defensive
• Balanced
• Low risk
• Medium risk
• High risk
A 70% exposure to growth assets, such as equities, might be called

“balanced” by one provider, “aggressive” by another, and “medium risk”
by yet another. There is no regulation anywhere in the world governing
how or when terms like these should be used.
3.6 CONCLUSION
Robo-advisors operate in a highly regulated marketplace that is difficult
to “disrupt” as robos are obligated to follow the same rules that apply
to human advisors. Nonetheless, they are exciting for the opportunities
they present to scale financial advice, particularly in sectors that might
otherwise be unprofitable to serve. However, robo-advisors face extra
challenges. They have fewer dimensions than a human advisor in terms
of interaction and dexterity. They must be taught in advance about every
possible situation that can be encountered. The underlying algorithms are
both the strength and weakness for a robo-advisor. When they are robust,
the robo can scale good advice across large numbers of people. But when
they are inadequate, the resulting advice can become flawed. Risk tolerance
is a particularly good example of a challenge for robo-advisors who face
pressure to reduce the number of inputs required from a user.
REFERENCES
Mark H. Ashcraft. 2002. “Math Anxiety: Personal, Educational and Cognitive
Consequences”. Current Directions in Psychological Science 11(5): 181–185.
Australian Securities & Investments Commission. 2017. Regulatory Guide 175—
Licensing: Financial Product Advisers—Conduct and Disclosure. https://
download.asic.gov.au/media/4698465/rg175-published-10-april-2018.pdf .
Mustafa Baloğlu. 2004. “Statistics Anxiety and Mathematics Anxiety: Some Inter-
esting Differences I”. Educational Research Quarterly 27(3): 38–48.
Ivan Barreda-Tarrazona, Ainhoa Jaramillo-Gutierrez, Daniel Navarro-Martinez,
and Gerardo Sabater-Grande. 2011. “Risk Attitude Elicitation Using a Multi-
Lottery Choice Task: Real vs. Hypothetical Incentives”. Spanish Journal of
Finance and Accounting 40: 609–624. https://doi.org/10.108/02102412.
2011.10779713.
Shawn Brayman, Michael Finke, Ellen Bessner, John E. Grable, Paul Grif-
fin, and Rebecca Clement. 2015. Current Practices for Risk Profiling in
Canada and Review of Global Best Practices. Last Accessed: September 13,
2019. https://www.osc.gov.on.ca/documents/en/Investors/iap_20151112_
risk-profiling-report.pdf .
Richard A. Cohn, Wilbur G. Lewellen, Ronald C. Lease, and Gary G. Schlar
baum. 1975. “Individual Investor Risk Aversion and Investment Portfolio
Composition”. The Journal of Finance 30(2): 605–620. https://doi.org/10.
2307/2978738.
Edward T Cokely and Colleen M. Kelley. 2009. “Cognitive Abilities and Superior
Decision Making Under Risk: A Protocol Analysis and Process Model Evalua-
tion”. Judgment and Decision Making 4(1): 20–33.
Thomas G. Davey. 1999. Automated Assessment of Personal Financial Risk
Tolerance. U.S. Patent US6859788B1. https://patents.google.com/patent/
US6859788B1/en.
William G. Droms and Steven N Strauss. 2003. “Assessing Risk Tolerance for Asset
Allocation”. Journal of Financial Planning 16(3): 72–77.
Stuart Erskine. 2016. Due Diligence Questions About a Risk Tolerance Test.
Last Accessed: September 15, 2019. http://www.riskprofiling.com/WWW_
RISKP/media/RiskProfiling/Downloads/Due_Diligence_Checklist.pdf .
European Securities and Markets Authority. 2018. Guidelines on Certain Aspects
of the MiFID II Suitability Requirements. https://www.esma.europa.eu/sites/
default/files/library/esma35-43-1163_guidelines_on_certain_aspects_of_
mifid_ii_suitability_requirements_0.pdf .
Financial Conduct Authority. 2006. FCA Handbook: COB 5.2 Know Your Cus-
tomer. Last Accessed: September 13, 2019. https://www.handbook.fca.org.
uk/handbook/COB/5/2.html?date=2006-03-01.
Financial Industry Regulatory Authority. 2011. Regulatory Notice: Know

Your Customer and Suitability. https://www.finra.org/sites/default/files/
NoticeDocument/p122778.pdf .
Financial Industry Regulatory Authority. 2016. Report on Digital Investment
Advice. https://www.finra.org/sites/default/files/digital-investment-advice-
report.pdf .
Financial Ombudsman Service. 2011. Risk Profiling in Financial Advice Disputes.
Last Accessed: September 13, 2019. https://www.fos.org.au/the-circular-6-
home/risk-profiling-in-financial-advice-disputes/.
Jill E. Fisch, Marion Laboure, and John A. Turner. 2018. “The Emergence of the
Robo-Advisor”. In Wharton University of Pennsylvania Pension Research Coun-
cil. Working Paper WP2018-12. https://repository.upenn.edu/prc_papers/
10/.
Renato Frey, Andreas Pedroni, Rui Mata, Jorg Rieskamp, and Ralph Hertwig.
2017. “Risk Preference Shares the Psychometric Structure of Major Psy-
chological Traits”. Science Advances 3(10). https://doi.org/10.1126/sciadv.
1701381.
Carlin Fung, Paul Hayward, and Mark Skuce. 2013. Know-Your-Client and
Suitability Obligations. https://www.osc.gov.on.ca/documents/en/Dealers/
ro_20131204_kyc-suitability-%09obligations.pdf .
Florian Glaser, Zwetelina Iliewa, Dominik Jung, and Martin Weber. 2019.
“Towards Designing Robo-advisors for Unexperienced Investors with Experi-
ence Sampling of Time-Series Data”. In Information Systems and Neuroscience,
pp. 133–138. Springer.
John E. Grable and Ruth H. Lytton. 1999. “Financial Risk Tolerance Revisited:
The Development of a Risk Assessment Instrument”. Financial Services Review
8(3): 163–181. https://doi.org/10.1016/S1057-0810(99)00041-4.
Adam Grant. 2013. Goodbye to the MBTI, the Fad that Won’t Die. Psychology Today.
Last Accessed: September 15, 2019. https://www.psychologytoday.com/au/
blog/give-and-take/201309/goodbye-mbti-the-fad-won-t-die.
Terrence A. Hallahan, Robert W. Faff, and Michael D. McKenzie. 2003. “An
Empirical Investigation of Personal Financial Risk Tolerance”. Financial Services
Review 13: 57–78.
Frank D. Hodge, Kim I. Mendoza, and Roshan K. Sinha. 2018. “The Effect of
Humanizing Robo-Advisors on Investor Judgments”. SSRN Electronic Journal.
https://doi.org/10.2139/ssrn.3158004.
Daniel Kahneman and Amos Tversky. 1979. “Prospect Theory: An Analysis of
Decision Under Risk”. Econometrica, Econometric Society 47(2): 263–291.
Myrsini Katsikatsou and Stuart Erskine. 2018. FinaMetrica Technical Manual.
https://www.riskprofiling.com/WWW_RISKP/media/RiskProfiling/
Downloads/Technic%09al_Manual_2018.pdf .
Joachim Klement. 2015. “Investor Risk Profiling: An Overview”. The CFA Insti-
tute Research Foundation. https://www.cfainstitute.org//media/documents/
article/rf-brief/rfbr-v1-n1-1-pdf.ashx.
Moshe Levy and Haim Levy. 2002. “Prospect Theory: Much Ado About Noth-
ing?” Management Science 48(10): 1334–1349.
Stephen H. Moore. 2017. “Communication Task Difficulty in Investment Risk
Profiling: A Linguistic Perspective”. Financial Planning Research Journal, 33–
52.
Paul Resnik. 2016. How to Test Your Clients’ Risk Tolerance. FTAdviser.com.
Last Accessed: September 13, 2019. https://www.ftadviser.com/investments/
2016/9/28/how-to-test-your-clients-risk-tolerance/.
Michael J. Roszkowski, Geoff Davey, and John E. Grable. 2005. “Insights from
Psychology and Psychometrics on Measuring Risk Tolerance”. Journal of Finan-
cial Planning 18(4): 66–77.
John Rust and Susan Golombok. 2009. Modern Psychometrics: The Science of
Psychological Assessment. Taylor & Francis Group. Isbn:9780415442152.
Matthew Smith. 2016. How Robo-Advice Could Go Bad. FINSIA InFinance.
https://finsia.com/insights/news/news-article/2016/08/11/how-robo-
advice-could-go-bad.
Gerhard Van de Venter, David Michayluk, and Geoff Davey. 2012. “A Longitudinal
Study of Financial Risk Tolerance”. Journal of Economic Psychology 33(4): 794–
800. https://doi.org/10.1016/j.joep.2012.03.001.
CHAPTER 4
Robo Economicus? The Impact of Behavioral

Biases on Robo-Advisory
Peter Scholz, David Grossmann, and Joachim Goldberg
Contents
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2 Home Bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2.1 Does Robo-Advisory Fall for Home Bias? . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3 Mental Accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.3.1 Does Robo-Advisory Care About Mental Accounts? . . . . . . . . . . . . . 59
4.4 Overconfidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4.1 How Confident Are Robo-Advisors?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
P. Scholz ()
Professor for Banking and Financial Markets, Hamburg School of Business
Administration, Hamburg, Germany
D. Grossmann
Head of Global Bank Management, Otto M. Schroeder Bank AG, Hamburg,
Germany
J. Goldberg
Founder, Goldberg & Goldberg, Frankfurt am Main, Germany

54 P. SCHOLZ ET AL.
4.1 INTRODUCTION
“Robo-advisors advertise investments without emotional bias, because the
algorithm gives rational suggestions. The potential of human failure should
be minimized.”1 One important element of the robo-advisory business is
indeed to avoid the emotional components like greed, fear, or doubt in the
investment process. For example, the largest robo-advisor based on assets
under management claims that “[w]hen it comes to investing, your natural
reactions can get in the way. It’s human nature to overthink, overreact,
and, at times, be overwhelmed. With Vanguard Personal Advisor Services,
an advisor serves as an emotional circuit breaker so you don’t abandon a
well-thought-out plan” (Vanguard 2020a).
The idea to evolve machine support for investment advisory is not
new: more than 15 years ago, cognitrend, a company specialized in
behavioral finance analysis and trading, created an “artificial salesman,”
who would assist the human advisors based on behavioral economics. The
artificial salesman was planned as a web-connected algorithm capable of
rational decision-making. Whereas humans tend to be influenced by past
successes of failures and are biased by imperfect information processing,
machines seem to be unmoved by these pitfalls. But at that time, trust
in human financial advisors and skepticism against internet and computer
were probably too pronounced; and because of that the artificial salesman
failed to catch the fancy of common investors. After the financial crisis, the
preferences underwent a major shift and the first robo-advisors appeared
in the private and retail segment.
Although robo-advisors keep their focus on the various aspects of
investment such as transparency, cost-efficiency, or usability, it is the
unerring rationality as well in which they seem to be ahead of their human
colleagues. In the area of financial decision-making, there is an almost
infinite number of scientific publications available on behavioral finance,
which describe the impact of emotions on human decisions. Authors such
as Daniel Kahneman (“Thinking, Fast and Slow”) or Nassim N. Taleb
(“The Black Swan”) present numerous examples as to how investors can
be trapped by their emotions and limitation in information processing.
Infamous shortcomings include anchoring, availability heuristic, disposi-
tion effect, gambler’s fallacy, and selective perception, to name only a few.
1 Translated from Grzanna (2018). A similar description can be found on Rixse (2018).
ROBO ECONOMICUS? THE IMPACT OF BEHAVIORAL BIASES ON ROBO-ADVISORY 55
Since machines are created and programmed by humans, it would be

interesting to explore if robo-advisors really corresponded to the ideal of
Homo economicus, the fully rational decision-maker—a robo economicus,
for instance—or if the robots were more similar to the humans who
invented them. For this analysis, we took a deeper look into three different
behavioral biases: home bias, mental accounting, and overconfidence.
Even if we believe that the robo-advisors probably make the investment
process less vulnerable to emotional influences, we are interested to see to
what degree the improvement evolves. Each section briefly describes the
respective bias and then analyzes the status quo as well as the potential
exposure of robo-advisory to the bias. Here, we focus primarily on the big
players in the US, the UK, and Germany, selected in terms of assets under
management, who currently drive the market.
4.2 HOME BIAS

“Why seek far afield when the good could not be any closer by?”2 is
a famous saying attributed to Johann Wolfgang von Goethe. Ambiguity
aversion, that is, to prefer the familiar and abandon the unknown, also
seems to hold true on capital markets: as French and Poterba (1991)
discovered, investors tend to overweight domestic stocks and neglect the
benefit of international diversification. In their sample, the bias seems to
be exceptionally evident: for US investors, they found a 0.938 weight in
domestic assets, for Japanese investors 0.9811, and for UK investors 0.820.
They explain the relatively low weight of UK stocks in UK portfolios
compared to the US and Japanese cases with the smaller market value
of UK assets and the consequential need to invest in foreign equity. In
another paper, Tesar and Werner (1995) analyze home bias in Canada,
Germany, Japan, the UK, and the US for stocks and bonds. They emphasize
that “there is strong evidence of a home bias in national investment
portfolios despite the potential gains from international diversification.”
Both analyses have been performed with data sets, in which international
deregulation and international trading just started to accelerate a few
years ago, which might be an explanation for the poor international
diversification at that time. But later studies as well, for example Coval and
Moskowitz (1999), Ahearne et al. (2004), Fidora et al. (2007), and Lütje
2 Translation from German to English by dict.leo.org.

56 P. SCHOLZ ET AL.
and Menkhoff (2007), confirm the persistence of home bias. Furthermore,

there is evidence in some of these studies that even professional investment
managers are prone to it. This raises the question if overweighting domestic
assets is really undesired or if factors like higher transaction costs for foreign
assets, double taxation, currency risk, and information asymmetries might
explain the rationality of home bias. But many studies show that, in general,
poor international diversification seems to reduce a portfolio’s performance
significantly (e.g., French and Poterba (1991), Tesar and Werner (1995),
Lütje and Menkhoff (2007), Seasholes and Zhu (2010)). Lewis (1999)
tries to quantify the cost of insufficient international diversification and
states that “the costs of holding foreign stocks must be extremely large
to dissuade an efficient domestic investor from foreign diversification.”
Hence, it is more likely that the explanation of home bias traces back to
different behavioral biases such as overconfidence and regret aversion, or
factors such as patriotism and excess investments in own-companies’ shares
(Foad 2011).
4.2.1 Does Robo-Advisory Fall for Home Bias?

Clients of robo-advisory services might assume that the portfolio recom-
mendation of the machine is free from home bias. However, since many
robo-advisors rely more on investment committees rather than rule-based
asset allocation, the home bias could also be evident in portfolios based on
robo-advice. As we have detected in our paper (Tertilt and Scholz 2018),
robo-advisors seem to absorb certain other culture differences, since US
robos typically recommend higher equity quotas than what German robos
do. Therefore, we check the recommended model portfolio of the largest
U.S., U.K., and German robo-advisors to verify the distribution between
domestic and foreign equity. Hence, for our analysis, we solely consider
model portfolios with a close to 100% investment in equity for maximal
visibility of the effect.
Considering the largest robo-advisory in the world, Vanguard Personal
Advisor Services, the analysis starts with a bit difficult since it is necessary
to open an account before a model portfolio is displayed. Furthermore,
Vanguard’s robo is more a hybrid advice model since the human advisor as
well as a human investment committee play a crucial role in the investment
process. If we take the Vanguard ETF strategic model portfolio as the
next best guess,3 we find that in an 100% equity portfolio the share of
domestic stocks is 58.8% compared to 39.2% of foreign stocks. If we
compare these numbers to the other large US robo-advisors, then we see
that the figures seem to be in line with the model portfolios of the other
robos. Charles Schwab Intelligent Portfolios has a domestic equity quota
of 53.2%. Furthermore, Schwab’s robo-advisor, as well as Vanguard’s,
relies on human investment committees. The two robo-advisors with the
longest track record, Betterment and Wealthfront, however, use rule-based
algorithms for their portfolio allocation. Still, their equity quotas are not
so different from those of Vanguard and Schwab: they allocate 48.2% and
46.9% in domestic stocks, respectively. Finally, Personal Capital, which is a
digital wealth management service, also relies on an investment committee
and, additionally, on a team of experts, including Nobel laureate Harry
Markowitz, who is the “father of modern portfolio theory,” and Shlomo
Benartzi, a distinguished expert on behavioral finance. Interestingly, their
quota on domestic stocks is the highest amongst all robo-advisors: they
invest 67% in domestic stocks, which is even larger than the North
American equity share in the MSCI World Index.
It is doubtless a tricky thing to clearly determine home bias. But if we
take the MSCI World Index as reference, which is a market cap-weighted
index, then they currently allocate around 56.7% to North American
stocks. Hence, Personal Capital would have a clear home bias in their
portfolio. Vanguard and Schwab are close to the MSCI benchmark, and
even Betterment and Wealthfront are not far away from this reference
level. If we consider the share of global gross domestic product (GDP) as
a benchmark, which attributes approximately 25.9% to North America, all
US robos would show a distinct home bias. Further, it is interesting to note
that nearly all tested European robo-advisors show lower North American
equity quotas than those of their US peers, which also strengthens the
suspicion of a home bias of US robo-advisors.
If we analyze the UK robo-advisors, we find that, especially, Nutmeg
seems to have a rather high domestic stock quota: they hold close to
25% in UK stocks. Compared to a weight of 5.7% in the MSCI World
Index or 3.3% in share of global GDP, this is comparatively high and
seems to qualify for a home bias. Moneyfarm has a lower domestic stock
3 Based on the assumption that the investment committee makes similar decisions in their
ETF portfolio and the robo portfolio.
58 P. SCHOLZ ET AL.
allocation with around 11.6%, but still this seems to be high in comparison
to the significance of British markets or economy in the world. For the
German robo-advisors, we find a similar picture. Since they are part of the
Eurozone, it makes sense to widen the comparison to European stocks
instead. With respect to the MSCI World Index (14.4%), Scalable (24.9%),
Weltinvest (19.7%), and Liqid (22.7%) overweight European stocks. If
we take the share of the global GDP as benchmark (21.9%), the picture
changes and the domestic stock of German robo-advisors seems to be in
line. Hence, for German robo-advisors, the picture is not clear regarding
home bias.
4.3 MENTAL ACCOUNTING

“29,086 measures barley 37 months Kushim” dates back around 5000
years and is one of the oldest records from Mesopotamia that our forefa-
thers left for us (Harari 2015). So, the oldest messages from our past are
not philosophical or religious, and they do not contain some kind of deeper
wisdom; as it seems, they are simply accounting statements (Harari 2015).
Hence, the compulsion of keeping track of one’s own belongings seems to
be deep-rooted in human culture. Today, in our highly organized world,
accounting is self-evident. As Richard Thaler (1985, p. 199) states: “All
organizations, from General Motors down to single person households,
have explicit and/or implicit accounting systems.” From a certain per-
spective, this is quite logical: Accounting allows organizing even complex
capital flows and was one important factor behind the emergence of North
Italian economy in the medieval period. But, as Thaler (1985, p. 199) also
notes, “[t]he accounting systems often influence decisions in unexpected
ways.” Based on Kahneman and Tversky (1979) Prospect Theory, and his
considerations of human decisions involving, for example, sunk costs or
opportunity costs (Thaler 1980), he found a human behavioral trait that
he named mental accounting (Thaler 1985). The first example he presented
in his 1985 paper describes the bias (p.199):
Mr. and Mrs. L and Mr. and Mrs. H went on a fishing trip in the northwest
and caught some salmon. They packed the fish and sent it home on an airline,
but the fish were lost in transit. They received $ 300 from the airline. The
couples take the money, go out to dinner and spend $ 225. They had never
spent that much at a restaurant before.
Thaler (1985) explains:
Example 1 violates the principle of fungibility. Money is not supposed to

have labels attached to it. Yet the couples behaved the way they did because
the $ 300 was put into both “wind fall gain” and “food” accounts. The
extravagant dinner would not have occurred had each couple received a yearly
salary increase of $ 150, even though that would have been worth more in
present value terms.
The similar effect is described in Tversky and Kahneman (1981). They

present the result from an experiment, in which the test persons had to
answer two questions. Question 1 is:
Imagine that you have decided to see a play where admission is $ 10 per
ticket. As you enter the theater you discover that you have lost a $ 10 bill.
Would you still pay $ 10 for a ticket for the play?
Question 2 asks:
Imagine that you have decided to see a play and paid the admission price of
$ 10 per ticket. As you enter the theater you discover that you have lost the
ticket. The seat was not marked and the ticket cannot be recovered. Would
you pay $ 10 for another ticket?
Although both questions relate to the same economic situation, the

test persons react differently: whereas in the first scenario almost everyone
(88%) tends to buy a ticket again, in the second situation only 46% decide
to do so. They assign this bias in behavior to the effect of “psychological
accounting,” which is just another expression for mental accounting. So,
in a nutshell, people have the tendency to create imaginary accounts when
making decisions. Since people are affected more by losses than by profits,
according to the Prospect Theory, the transfer of money from one to
another imaginary account may severely bias decision-making in the same
economic framework.
4.3.1 Does Robo-Advisory Care About Mental Accounts?

It is a valid question if mental accounting is always an unfavorable heuristic
or not. As Shefrin (2000) points out, there are two different types of
investors: On the one side, there are “[m]ean-variance investors [who]
60 P. SCHOLZ ET AL.
care only about the expected returns and variance of the overall port-
folio [and...] have consistent attitudes towards risk.” On the other side,
there are “[b]ehavioral investors [who] build portfolios as pyramids of
assets, layer by layer, where layers are associated with particular goals
and particular attitudes towards risk.” Many investment advisors follow
the mean-variance investor’s approach and try to measure the total risk
tolerance of an investor, which will be applied to the total assets. However,
as Pan and Statman (2012) claim, “each investor has a multitude of risk
tolerances. Probing for one global risk tolerance misses that multitude.” By
and large, investors may pursue different goals with their investments, such
as pension, reserves, education for their children, or to make a dream come
true (c.f. also Pan and Statman (2012)). It seems to be a fair assumption
that “investors consider their portfolios as collections of mental accounts,
each devoted to a goal” (Pan and Statman 2012). Depending on the
goal’s time frame, different risk capacities are possible: The longer the
investment period, the higher the risk class of the assets can be selected. The
German Institute for Equity publishes the “Return-Triangle” in which for
different investment periods the resulting returns are displayed (Deutsches
Aktieninstitut 2020). They show that the longer the investment period,
the lower is the risk of losing money with stock investments. From this
perspective, it makes sense to distinguish between different investment
goals by different risk tolerances. For example, while the reserves should
show lower risk, long-term pension plans can bear more risky assets. As a
consequence, robo-advisory could prefer goal-based investing as proposed
by Chhabra (2005), Das et al. (2010), and Brunel (2011), among others.
Interestingly, the large US robo-advisors indeed pursue strategies for
goal-based investing. Vanguard Personal Advisor Services proposes dif-
ferent goals as an integral part of its advice and differentiates between
its reports based on various investment goals (Vanguard 2020c). It even
has a website dedicated to different investment goals and appropriate
consideration in individual investment plans (Vanguard 2020b). Also,
the US competitors have a clear focus on goal-based investing: Schwab
Intelligent Portfolios provides a systematic goal-tracking tool (Charles
Schwab & Co. Inc. 2020). The goal-tracker aims to “monitor whether
your goal is ‘on target’ towards achieving a savings or income goal.”
Schwab’s robo-advisor as well provides a website, which contains support
regarding investment goals (Charles Schwab & Co. Inc. 2016). They
explicitly introduced a bucket system with different time horizons: from
less than two years (bucket 1) to three to ten years (bucket 2), and
more than ten years (bucket 3). Each of these buckets pursues different
strategies and seems to allow for different risk tolerance. Betterment also
follows the idea of goal-based investing. Like their peers, it has a detailed
website introducing investors to the concept. Betterment clearly states that
“[w]ithin your Betterment account, every investment goal you set has a
target amount and target date(s) for which you desire to meet you goal [...
and] [e]ach of these investment goals requires a different strategy—that is
the quintessence of smarter investing” (Egan 2019). Wealthfront aims to
go even a step beyond its peers and plans to introduce “self-driving money.”
As Wealthfront CEO Andy Rachleff stated: “Our vision is to deliver a
service where you direct deposit your paycheck with us. We automatically
pay your bills. We automatically top off your emergency fund, and then
route money to whatever account is the most ideal for your particular goals,
whether they’re at Wealthfront or elsewhere” (Thiagarajan 2018).
In contrast to their US peers, the European robo-advisors seem to focus
less on the goal-based approach. Nutmeg, for example, asks for financial
goals in its questionnaire: “When you sign up, we first ask you about your
goals and risk preference. With that information, we help you choose a
portfolio that’s right for you [...]” (Nutmeg 2020). But it does not become
clear on its website, how it exactly uses the information and if multiple
goals are allowed. Scalable claims to use goal-based investing (Scalable
Capital 2018), but on its website we could not find much information
about how this concept was applied. Scalable asks in its questionnaire about
the investor’s goals, but how these goals impact the risk-driven investment
style is not transparent. A similar picture is given by the other large German
robo-advisors: they generally ask in their questionnaires for financial goals,
often linked with the time horizon. But an explicitly goal-based investment
approach cannot be recognized. Weltinvest even does not ask for goals and
does not allow holding multiple portfolios currently.
4.4 OVERCONFIDENCE
“What do you think: Are you better or worse than the average driver?” If
you pose this question to a significant sample of people, most of the time
you will probably obtain a result estimating that much more than 50%
of the participants appraise themselves as better than the average driver—
62 P. SCHOLZ ET AL.
which is, in fact, impossible from a statistical point of view.4 In his book,
Taleb (2007) proposes many different questions as to how to test what he
calls “epistemic arrogance.” With these questions, the issue is not “to gauge
[people’s] knowledge, but rather their evaluation of their knowledge”
(Taleb 2007, p.139). The overwhelming result reported by Taleb (2007)
is that people tend to overestimate their knowledge. This overconfidence
is a form of illusion of control and a well-documented bias in the scientific
literature. It traces back to the works of Tversky and Kahneman (1974),
Fischhoff et al. (1977), and Fischhoff et al. (1980), who observed the
tendency of people to overestimate themselves in different aspects. This
bias is problematic because “overconfidence can keep us from realizing
how little we know and how much additional information is needed about
the various problems and risks we face” (Slovic et al. 1981). Therefore,
overconfidence can be observed in at least three major occurrences (Moore
and Schatz 2017):
Overestimation is thinking that you are better than you are. Overplacement
is the exaggerated belief that you are better than others. Overprecision is the
excessive faith that you know the truth.
Especially on financial markets, participants are not free from the

overconfidence bias: For example, Barber and Odean (2013, p.1547)
describe that overconfidence may “explain the relatively high turnover rates
and poor performance of individual investors.” They refer to papers of
Dorn and Huberman (2005), Glaser and Weber (2007), and Grinblatt and
Keloharju (2009) to support their hypothesis. Furthermore, as Hirshleifer
(2015) points out, on the one hand, overconfident investors tend to reduce
diversification in their portfolios; but on the other, they may show over-
or underreaction toward prices due to misleading information signals. It is
not that private investors alone are prone to overconfidence; professional
investors as well seem to fall for this bias. A study by Puetz and Ruenzi
(2011) shows that even mutual fund managers tend to increase their
turnover rates after experiencing a period of prosperous portfolio perfor-
mance. After excluding factors such as incentives, inflows, or managerial
4 This specific example is taken from the German book by Stock and Goldberg (2013):
Genial einfach entscheiden, Finanzbuch Verlag. The question, that is, the observation, goes
back to Svenson (1981).
learning as potential source for the increased trading activity, they conclude
that overconfidence of fund managers is the most likely explanation for
their findings. In a conference paper, Choi and Lou (2010) show that in
their framework, inexperienced mutual fund managers are more influenced
by the overconfidence bias than their more experienced colleagues. In
general, studies largely report that overconfidence increases costs and,
therefore, decreases portfolio performance, indicating that avoiding over-
confidence is beneficial for investors.
4.4.1 How Confident Are Robo-Advisors?

Based on the claim to provide rational investment advice, robo-advisors
should base their processes on risk profiling of investors, subsequent and
adequate portfolio recommendations, as well as the implementation and
monitoring of the suggested portfolio strategies. Active management,
which includes the sales of actively managed mutual funds, should be
avoided. It is likely that these strategies will only raise costs but will not
improve long-term returns.
The large US robo-advisors typically rely on passive investments by
applying ETFs and/or index funds. Vanguard, for example, writes that its
investment philosophy is based on “benefits of low costs, diversification,
and indexing,” and it is primarily using stocks, bonds, and cash for its
portfolio allocation (Vanguard Advisers, Inc. 2019). Schwab as well prefers
ETFs for its asset allocation; however, it also includes more complex asset
classes, such as real estate and commodities. Schwab explicitly states that
the algorithms its robo-advisor is using are “not designed to actively
manage asset allocations based on short-term market fluctuations” (Charles
Schwab & Co. Inc. 2019). Wealthfront joins the group by stating that
it is “rooted in passive investing, which means we’ll build you a globally
diversified portfolio of low-cost index funds” (Wealthfront 2020). The
asset classes of Wealthfront compare to those of Schwab: stocks, bonds,
real estate, and commodities. Betterment does not only rely on ETFs for
its allocation of assets such as stocks, bonds, and cash, it also commits itself
to “systematic decision-making” (Grealish 2019). Compared to its peers, it
is the only robo-advisor out of the big four that applies a rather consequent
rule-based investing approach. Grealish (2019) states that Betterment’s
portfolio optimization is based on modern portfolio theory and the models
by Fama and French (1992) and Black and Litterman (1992). In sum,
64 P. SCHOLZ ET AL.
it could be assumed that the big four US robo-advisors avoid by and

large the overconfidence bias through the application of passive investment
strategies. But, if we take a closer look, Vanguard, Charles Schwab, and
Wealthfront do not follow a holistic rule-based approach for their asset
allocations but allow expert investment committees to interfere.5 This is
a potential gateway for all kinds of behavioral biases if experts have the
possibility of altering the proposed asset allocations. Also, there is the
risk to deviate from the trajectory, especially in times of distress—not to
mention the biases that could emerge from group decision- making or,
more precisely, from hidden profiles or groupthink.
The large European robo-advisors as well as their US peers widely
rely on passive instruments such as ETFs. Cominvest and Liqid, however,
are exceptions and also offer actively managed mutual funds in their
allocations. Moreover, there is a tendency of European robos to be more
active in investments than their US competitors. Nutmeg offers two
different asset allocation models: fully managed by portfolio managers
or fixed allocation. For the fully managed approach, Nutmeg advertises
its active investment approach: “This is not a static, one-off process. We
continually review the asset allocation for all our customers to decide if
we need to make adjustments” (Port 2013). A similar approach is taken
by Liqid that offers an active investment style in its “Select” model.
Scalable does not call itself “active manager” but applies a mix of passive
investment and active risk management based on Value-at-Risk. This seems
to be comparable to Vanguard’s method, which also promotes active risk
management.6 It is, nonetheless, questionable if portfolio management
by active risk adjustments is more successful than a pure buy-and-hold
(with rebalancing) approach. Quirion, however, follows a rule-based and
passive investment style and, therefore, seems to allow very little or no
5 Vanguard: “When recommending, setting, and adjusting your asset allocation, we weigh
shortfall risk—the possibility that a financial plan or Portfolio will fail to meet longer-term
financial goals—against market risk.” Schwab: “[W]e have dedicated an entire team of Charles
Schwab Investment Advisory (CSIA) experienced analysts to continually use state-of-the-art
research and evolve our approach to creating asset allocations designed to improve outcomes
for individual investors.” Wealthfront: “Wealthfront combines the judgment of its investment
team with state of the art optimization tools to identify efficient portfolios.”
6 Vanguard Advisers, Inc. (2019, p.18): “When recommending, setting, and adjusting your
asset allocation, we weigh shortfall risk—the possibility that a financial plan or Portfolio will
fail to meet longer-term financial goals—against market risk.”
human interference. Weltinvest only has fixed asset allocations and does
not provide advice on risk profiles of assets, which could lead investors to
choose inappropriate portfolios from a risk perspective.
4.5 SUMMARY
Summing up the findings from our analysis, we find that robo-advisory is
not completely free from behavioral biases.
Interestingly, the home bias is relatively strong amongst the robo-
advisors. Although it is not easy to find “airtight” evidence for the home
bias, it is surprising that U.S. robo-advisors clearly have different allocations
between foreign and domestic assets compared to their European peers.
It is quite evident as all robo-advisors have the tendency to overweight
domestic assets. Since many robo-advisors still rely on investment commit-
tees and less on rule-based investing, this seems to abet the home bias.
It is noteworthy that the relatively large “non-automatized” portion
in the asset management process of robo-advisors opens a gateway to
different forms of behavioral biases. For “hybrid” robo-advisors, such as
Vanguard’s approach, the dependence is more on “emotional sensitivity”
than algorithm-based models. With respect to the overconfidence bias,
most robo-advisors primarily rely on passive investment instruments such as
ETFs or index funds. But even if they largely abstain from actively managed
mutual funds, most of them do not seem to believe in simple “buy-
and-hold” with regular rebalancing. In a best-case scenario, they apply a
(relatively) strict rule-based approach, but in general there are investment
committees installed. All of these forms, be it active risk management
or active interference in times of turmoil, indicate the underlying belief
that active intervention improves the performance of the investment. In
sum, the large European robo-advisors do offer more active components
in their asset management process than those of their US peers. If we
consult the performance analysis in Sect. 1.5 of this book, it can be
found that despite the active management elements, robo-advisors do not
frequently outperform reasonable benchmarks. So, at least, it seems that
robo-advisors as well (or maybe better, their creators) are not insensitive
to the overconfidence bias since they charge fees for the additional services
provided.
Regarding mental accounting as well, there seems to be a clear difference
between US and European robo-advisors. Whereas the US robos align
to the goal-based investment approach and hence consider that investors
66 P. SCHOLZ ET AL.
might have different risk tolerances for different investment goals, the
European robos seem to ask for goals as it is obligatory. But it does not
seem to be deep-rooted in their investment philosophies.
So, is robo economicus or robo sapiens managing the investor’s assets?
Although we believe that robo-advisors indeed are capable of helping one
make better investment decisions in general, the present situation shows a
relatively distinct human intervention in the investment processes. There
is, however, a real possibility that completely rational and bias-free robo-
investment will emerge in the (near) future.
REFERENCES
Alan Ahearne, William L. Griever, and Francis Warnock. 2004. “Information Costs
and Home Bias: An Analysis of US Holdings of Foreign Equities”. Journal of
International Economics 62(2): 313–336.
Brad M. Barber and Terrance Odean. 2013. “The Behavior of Individual
Investors”. In Handbook of the Economics of Finance, Vol. 2, Part B, Chap. 22,
pp. 1533–1570. Elsevier B.V. https://doi.org/10.1016/B978-0-44-459406-
8.00022-6.
Fischer Black and Robert Litterman. 1992. “Global Portfolio Optimization”.
Financial Analysts Journal 48(5): 28–43. https://doi.org/10.2469/faj.v48.
n5.28.
Jean L. Brunel. 2011. “Goal-Based Wealth Management in Practice”. The Journal
of Wealth Management 14(3): 17–26. https://doi.org/10.3905/jwm.2011.
14.3.017.
Charles Schwab & Co. Inc. 2020. Goal Tracker for Schwab Intelligent Portfo-
lios. Last Accessed: March 20, 2020. https://intelligent.schwab.com/public/
intelligent/insights/whitepapers/goal.html.
Charles Schwab & Co. Inc. 2016. How Savings “Buckets” Can Help You Reach
Your Goals. Last Accessed: March 20, 2020. https://intelligent.schwab.com/
public/intelligent/insights/blog/savings-buckets.html.
Charles Schwab & Co., Inc. 2019. Schwab Intelligent Portfolios Solutions, Disclosure
Brochure. Last Accessed: September 29, 2019. https://www.schwab.com/
public/file/CMS-BDL100049.
Ashvin B. Chhabra. 2005. “Beyond Markowitz: A Comprehensive Wealth Alloca-
tion Framework for Individual Investors”. The Journal of Wealth Management
7(4): 8–34.
Darwin Choi and Dong Lou. 2010. “A Test of the Self-Serving Attribution Bias:
Evidence from Mutual Funds”. SSRN Electronic Journal. https://doi.org/10.
2139/ssrn.1100786.
Joshua D. Coval and Tobias J. Moskowitz. 1999. “Home Bias at Home: Local
Equity Preference in Domestic Portfolios”. The Journal of Finance 54(6): 2045–
2073. 10.1111/0022-1082.00181.
Sanjiv Das, Harry Markowitz, Jonathan Scheid, and Meir Statman. 2010. “Portfo-
lio Optimization with Mental Accounts”. Journal of Financial and Quantitative
Analysis 45(2): 311–334.
Deutsches Aktieninstitut. 2020. Return Triangles. Last Accessed: March 20,
2020. https://www.dai.de/en/what-we-offer/studies-and-statistics/return-
triangles.html.
Daniel Dorn and Gur Huberman. 2005. “Talk and Action: What Individual
Investors Say and What They Do”. Review of Finance 9(4): 437–481.
Dan Egan. 2019. Using Investment Goals at Betterment. Last Accessed: March 20,
2020. https://www.betterment.com/resources/goals-advice-explained/.
Eugene F. Fama and Kenneth R. French. 1992. “The Cross-Section of Expected
Stock Returns”. The Journal of Finance 47(2): 427–465. https://doi.org/10.
2307/2329112.
Michael Fidora, Marcel Fratzscher, and Christian Thimann. 2007. “Home Bias in
Global Bond and Equity Markets: The Role of Real Exchange Rate Volatility”.
Journal of International Money and Finance 26(4): 631–655.
Baruch Fischhoff, Asher Koriat, and Sarah Lichtenstein. 1980. “Reasons For
Confidence”. Journal of Experimental Psychology: Human Learning and Memory
6: 107–118. https://doi.org/10.1037/0278-7393.6.2.107.
Baruch Fischhoff, Paul Slovic, and Sarah Lichtenstein. 1977. “Knowing with Cer-
tainty: The Appropriateness of Extreme Confidence”. Journal of Experimental
Psychology: Human Perception and Performance 3(4): 552–564. https://doi.
org/10.1037/0096-1523.3.4.552.
Hisham Foad. 2011. “Familiarity Bias”. Behavioral Finance: Investors, Corpora-
tions, and Markets, Chap. 15, pp. 277–294. John Wiley & Sons, Ltd. https://
doi.org/10.1002/9781118258415.ch15.
Kenneth R. French and James M. Poterba. 1991. “Investor Diversification and
International Equity Markets”. American Economic Review 81(2): 222–226.
Markus Glaser and Martin Weber. 2007. “Overconfidence and Trading Volume”.
The Geneva Risk and Insurance Review 32(1): 1–36. https://doi.org/10.
1007/s10713-007-0003-3.
Adam Grealish. 2019. The Betterment Portfolio Strategy. Last Accessed: March
20, 2020. https://www.betterment.com/resources/betterment-portfolio-
strategy/.
Mark Grinblatt and Matti Keloharju. 2009. “Sensation Seeking, Overconfidence,
and Trading Activity”. Journal of Finance 64(2): 549–578. https://doi.org/
10.1111/j.1540-6261.2009.01443.x.
68 P. SCHOLZ ET AL.
Marcel Grzanna. 2018. Robos im Aufwind. Last Accessed: March 20, 2020.
https://www.sueddeutsche.de/wirtschaft/robo-advisors-robos-im-aufwind-
1.4197562.
Yuval N. Harari. 2015. Sapiens: A Brief History of Humankind. Vintage.
David Hirshleifer. 2015. “Behavioral Finance”. Annual Review of Financial Eco-
nomics 7(1): 133–159. https://doi.org/10.1146/annurev-financial-092214-
043752.
Daniel Kahneman and Amos Tversky. 1979. “Prospect Theory: An Analysis of
Decision Under Risk”. Econometrica, Econometric Society 47(2): 263–291.
Karen K. Lewis. 1999. “Trying to Explain Home Bias in Equities and Consump-
tion”. Journal of Economic Literature 37(2): 571–608.
Torben Lütje and Lukas Menkhoff. 2007. “What Drives Home Bias? Evidence
from Fund Managers’ Views”. International Journal of Finance & Economics
12(1): 21–35. https://doi.org/10.1002/ijfe.309.
Don A. Moore and Derek Schatz. 2017. “The Three Faces of Overconfidence”.
Social and Personality Psychology Compass 11(8). https://doi.org/10.1111/
spc3.12331.
Nutmeg. 2020. How We Invest. Last Accessed: March 20, 2020. https://www.
nutmeg.com/how-we-invest#portfolio-performance.
Carrie H. Pan and Meir Statman. 2012. “Questionnaires of Risk Tolerance,
Regret, Overconfidence, and Other Investor Propensities”. SCU Leavey School
of Business Research Paper, 10-05.
Shaun Port. 2013. An Inside View of the Nutmeg Investment Process: Asset
Allocation. Last Accessed: March 20, 2020. https://www.nutmeg.com/
nutmegonomics/an-inside-view-of-the-nutmeg-investment-process/.
Alexander Puetz and Stefan Ruenzi. 2011. “Overconfidence Among Professional
Investors: Evidence from Mutual Fund Managers”. Journal of Business Finance
& Accounting 38(5-6): 684–712. https://doi.org/10.1111/j.1468-5957.
2010.02237.x.
Chad Rixse. 2018. The 4 Advantages of Human vs. Robo-Advisors. Last Accessed:
March 20, 2020. https://www.cnbc.com/2018/4/25/the-4-advantages-of-
human-vs-robo-advisors.html.
Scalable Capital. 2018. Scalable Capital gewinnt Openbank die Digitalbank der
Santander- Gruppe, als Partner in Spanien. Last Accessed: March 20, 2020.
https://de.scalable.capital/presse/scalable-capital-gewinnt-openbank.
Mark S. Seasholes and Ning Zhu. 2010. “Individual Investors and Local Bias”.
Journal of Finance 65(5): 1987–2010.
Hersh Shefrin. 2000. Beyond Greed and Fear: Understanding Behavioral Finance
and the Psychology of Investing. Harvard Business School Press.
Paul Slovic, Baruch Fischhoff, and Sarah Lichtenstein. 1981. “Facts and Fears: Soci-
etal Perception of Risk”. Advances in Consumer Research 8: 497–502. http://
www.acrwebsite.org/search/view-conference-proceedings.aspx?Id=5844.
Christin Stock and Joachim Goldberg. 2013. Genial Einfach Entscheiden:

Besser Denken, Handeln und Investieren im Täglichen Entscheidungsdschungel.
FinanzBuch Verlag.
Ola Svenson. 1981. “Are We All Less Risky and More Skillful than our Fellow
Drivers?” Acta Psychologica 47(2): 143–148. https://doi.org/10.1016/0001-
6918(81)90005-6.
Nassim Taleb. 2007. The Black Swan: The Impact of the Highly Improbable. Random
House Group.
Michael Tertilt and Peter Scholz. 2018. “To Advise, or Not to Advise— How
Robo- Advisors Evaluate the Risk Preferences of Private Investors”. The Journal
of Wealth Management 21(2): 70–84.
Linda L. Tesar and Ingrid M. Werner. 1995. “Home Bias and High Turnover”.
Journal of International Money and Finance 14(4): 467–492. https://doi.org/
10.1016/02615606(95)00023-8.
Richard Thaler. 1985. “Mental Accounting and Consumer Choice”. Marketing
Science 4(3): 199–214.
Richard Thaler. 1980. “Toward a Positive Theory of Consumer Choice”. Journal
of Economic Behavior & Organization 1(1): 39–. https://doi.org/10.1016/
0167-2681(80)90051-7.
Harish Thiagarajan. 2018. Testing the GPS Behind Tomorrow’s Self-Driving Money.
Last Accessed: March 20, 2020. https://eng.wealthfront.com/2018/06/28/
testing-the-gps-behind-tomorrows-self-driving-money/.
Amos Tversky and Daniel Kahneman. 1974. “Judgment under Uncertainty:
Heuristics and Biases”. Science 185(4157): 1124–1131. https://doi.org/10.
1126/science.185.4157.1124.
Amos Tversky and Daniel Kahneman. 1981. “The Framing of Decisions and
the Psychology of Choice”. Science 211: 453–458. https://doi.org/10.1126/
science.7455683.
Vanguard. 2020. Introducing Vanguard Personal Advisor Services. Last Accessed:
March 20, 2020. https://investor.vanguard.com/financial-advisor/personal-
advisor-services.
Vanguard. 2020. Investing Goals. Last Accessed: March 20, 2020. https://investor.
vanguard.com/investing/goals.
Vanguard. 2020. Vanguard Personal Advisor Services. Filmed October 14 2015 at
Youtube. Video 1:52. https://www.youtube.com/watch?v=OYgJhCoAUnA.
Vanguard Advisers, Inc. 2019. Vanguard Personal Advisor Services Brochure. Last
Accessed: September 29, 2019. https://personal.vanguard.com/pdf/vpabroc.
pdf .
Wealthfront. 2020. Investing Isn’t an Art. It’s a Science. Last Accessed: March 20,
2020. https://www.wealthfront.com/investing.
CHAPTER 5
Quant Models for Robo-Advisors
Thorsten Ruehl
Contents
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.2 What Strategies Are Suitable for Robo-Advisory? . . . . . . . . . . . . . . . . . . . . . . . . 73
5.3 What Quantitative Approaches Does the Robo-Advisory Model Offer? . 75
5.3.1 Maximization of the Diversification Effect. . . . . . . . . . . . . . . . . . . . . . . . 76
5.3.2 Equal Distribution of Risks to the Investment Instruments
Contained in the Portfolio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.3.3 Risk Minimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.3.4 Methods Based on Return Forecasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.4 Dealing with Risk Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.4.1 Adherence to Lower Value Limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.4.2 Specification of a Risk Preference by Choosing a Target
Investment Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.5 Return Targets and Risk-Bearing Capacity: Need for Information . . . . . . 86
5.6 Requirements for the Investment Universe and Instruments . . . . . . . . . . . . 88
5.7 Customization by Investors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
T. Ruehl ()
Head of Investment Strategies and Research, CSR Beratungsgesellschaft mbH,
Hofheim am Taunus, Germany

72 T. RUEHL
5.1 INTRODUCTION
One of the main advantages of robo-advisory is the ability to offer a large
number of investors automated and thus cost-efficient asset management
that can still be tailored to the client’s needs to a certain extent. The
easy scalability combined with high individuality (compared to the still-
dominant standard solutions for retail customers) is one of the great
strengths of robo-advisory solutions (Bankenverband 2017). To take full
advantage of these benefits, all the components of a robo-advisory platform
must work together as effectively as possible. This, in turn, places demands
on investment strategies that can be used in this context. It makes sense
to favor concepts that are both automated and rule based and which can
be easily parameterized to reflect individual client requirements. This is
the only way to derive full benefit from the scaling advantages offered by
robotics (Lam 2016).
When comparing robo-advisors with independently managed security
accounts or (conventional) professional investment advisors, the benefits
typically mentioned include the following: low costs, focus on risks, technology
instead of emotions, and transparency. All four points can be attributed
to automation benefits: As in other industries, automation also leads to
a reduction in costs, as labor now represents the largest cost item in many
areas. Risks, in turn, can only be quantified and controlled by financial
mathematical models and calculations. This process is inherently linked to
the use of computers and thus predestined to be part of robo-advisory.
The consistent and systematic adherence to an investment approach
is significantly facilitated by a purely rule-based and thus technically
mappable approach. The typical errors in investor behavior can also be
greatly mitigated by the systematic use of smart, stringent approaches.
One of the frequently observed but avoidable investor mistakes is to
exit from a long-term successful systematic approach “at an inopportune
time”. An investment strategy that is understood and “supported” in its
decisions thus helps investors to stay “on board” even in difficult market
phases and to avoid logging in losses. Experience has shown that this
advantage of a strictly rule-based and transparent investment strategy is
often underestimated. If the investor has understood the basic rules of the
investment strategy, he will be able to understand the strategy’s behavior
(and is outcome) in different market phases and will therefore be able to
stick with the strategy even in difficult times, which in turn is important for
the success of the investment in the long term. All four of these advantages
QUANT MODELS FOR ROBO-ADVISORS 73
originate from a common source and can ultimately be traced back to

a largely automated use of algorithm-based investment decisions. Since
robo-advisory is based on highly automated software platforms, it would
be downright wasteful not to take advantage of the resulting benefits at
the heart of the investment strategy itself (Lam 2016). We will, therefore,
next take a closer look at this aspect.
5.2 WHAT STRATEGIES ARE SUITABLE FOR

ROBO-ADVISORY?
To capitalize on the benefits offered by the technology, investment strate-
gies are required that integrate seamlessly with existing technology and
have the same structural benefits. From this point of view, purely quanti-
tative strategies form part of a robo-advisory as the entire process can then
be designed “from a single source”. In principle, discretionary investment
strategies can also be successfully used in asset management. Discretionary
strategies, on the other hand, move between the following two poles, that
is, hybrids are also possible:
Individual
Every portfolio manager makes investment decisions solely for “his”
portfolios. This can mean that portfolio manager A increases the equity
allocation at a given point in time, for example while his colleague B
reduces it on the same day. From the company’s perspective, this approach
offers a considerable advantage: the diversification resulting from this
organizational structure reduces the likelihood that all portfolios will
perform poorly at the same time and lead to overall client dissatisfaction
with the risk of concentrated cash outflows. However, a certain herd
mentality of the formally independent fund managers cannot be ruled out
even with this form of organization, as it is well known that it feels more
comfortable to wander with the masses than to wander alone. Only purely
quantitative processes are immune to such emotional appraisal processes.
In-House Strategy
An investment committee sets guidelines for the currently supported
investment allocation, right up to uniform model portfolios, which must
be implemented by all in-house fund managers. Diverging performances
74 T. RUEHL
of individual portfolios from the same company are thus avoided, but so
are the benefits of style diversification.
The short outline of the two approaches already shows that in the
context of a robo-advisory, only the second variant would be considered, if
at all: uniform model portfolios that serve to control individual securities
accounts by mapping them one-to-one. Although such an approach can
actually be implemented, it is considerably more expensive than a purely
quantitative solution, since the investment process used for controlling
would have to be set up in an entirely discretionary manner, with all the
associated disadvantages on the cost side. The cost advantages resulting
from a larger number of target portfolios per sample portfolio are already
being used today in asset management for smaller portfolios. The last step
toward automation is no longer being taken here. However, fund-linked
asset management with a small number of discretionarily managed funds
of funds, to which the client portfolios are then allocated based on risk
appetite, are already consistent with the solution outlined above. When
“porting” to a robo-advisor platform, only the front end to the client
would change: the investment advisor who makes the selection on behalf
of the client on the basis of a predefined list of criteria would be replaced
by the robo-platform. A further disadvantage of the discretionary solution
approach lies in the limited transparency: although the investment deci-
sions of the investment committee can be published, a uniform approach
across all times and personnel changes cannot realistically be guaranteed.
This circumstance will sooner or later have a negative impact on portfo-
lios with very long-term horizons, for example for retirement provision
purposes, since the investor’s reasons for deciding on a certain model
portfolio may have become obsolete over the years due to changes in the
discretionary process. The disadvantages of discretionary approaches in the
context of robo-advisory are that the degrees of transparency, continuity,
and cost-efficiency that can be achieved with quantitative approaches can
never be fully achieved. Quantitative approaches with transparent rules
show their strengths precisely here (Satchell 2003): once an algorithm
has been set up, it only requires comparatively inexpensive maintenance
at runtime, while the discretionary approach relies on the ongoing work
of a (cost-intensive) investment committee. As quantitative rules consist
of a fixed, always identical set of rules, they can be made transparent to
the investor to an arbitrary degree. Only copy protection, which is not
achievable by law, will set limits here in practice, but not the investment
strategy per se. An investment strategy that, for example, is decidedly
aimed at seeking the same risk contributions per asset class at all times
will continue to do so even after 10 or 20 years, that is, the advantage in
terms of continuity, in addition to the transparency advantage, results in
a type of “accompanying advantage” over the discretionary approach. We
will therefore deal with purely quantitative approaches and identify those
particularly suitable for use as part of robo-advisory.
5.3 WHAT QUANTITATIVE APPROACHES DOES THE

ROBO-ADVISORY MODEL OFFER?
Robo-advisory services take advantage of automated processes—it is
important to pursue this idea consistently right down to the investment
strategy. However, not every rule-based approach is equally suitable
for use in a fully automated implementation. There are many technical
approaches that evaluate historical price patterns and draw conclusions
about the current market situation. A simple example would be the use of
moving averages to determine entry and exit times for any given market
(Brock et al. 1992). Such approaches can be very successful in practice.
However, they are not based on a strictly scientific basis, but on the
use of a (mathematically formulated) heuristic. In order to do this, the
so-called back tests are carried out, but their prognostic significance or
temporal stability is often not given. Here, too, “post-optimization” must
be carried out on an ongoing basis in the future, at the strategic level
rather than at the portfolio level. These strategies thus come close to
discretionary strategies, with all the advantages and disadvantages already
mentioned above, especially in terms of transparency and continuity.
In order to take full advantage of the aforementioned options that are
available in robo-advisory, the strategies used must therefore also have a
high degree of stability over time (Meucci 2009; Grinold and Kahn 2012).
Risk models satisfy this requirement, while return forecasting models have
to be revised regularly and they are also disadvantageous in terms of the
required transparency.
For this reason, we want to focus primarily on quantitative approaches
that make do with pure risk management and do not include return
forecasts in the optimization process. What strategies fall into this category?
First of all, these are all quasi-stationary strategies in which only a regular
rebalancing (practical adjustment frequencies range from weekly to annual)
is carried out according to a fixed rule. This fixed rule can be, for
example, an equal weighting or a weighting based on market capitalization.
76 T. RUEHL
Such strategies have been devised to be highly transparent and easy to

implement. The latter point, however, ensures that these strategies can also
be easily “replicated”, and thus they will always be subject to increased price
pressure. From the provider’s perspective, more sophisticated approaches
should not only be aimed at benefiting the investor. Thanks to the
automated platforms, however, such more sophisticated approaches can be
implemented with comparatively little additional cost. Added value for the
client can be achieved, for example, through the following objectives:
• Maximization of the diversification effect

• Equal distribution of risks to the investment instruments contained in
the portfolio
• Risk minimization
• Adherence to lower value limits
• Specification of a risk preference by choosing a target investment
period
5.3.1 Maximization of the Diversification Effect

The old stock market wisdom of not putting all your eggs in one basket
is often cited, but too often not consistently followed. To be clear: Many
baskets are also of little use if they are mounted on the same bike rack
and the whole bike tips over. It will be difficult to achieve a noticeable
stabilization of the portfolio through diversification effects with equities
from a single sector. A necessary but not yet sufficient prerequisite for
a well-diversified portfolio is, therefore, an investment universe that not
only consists of highly correlated components, but also makes targeted
use of those with low or even negative correlations. The more the com-
ponents differ, the greater is the chance that even in times of crisis
the portfolio can be effectively hedged through opposing developments.
Such a well-diversified investment universe is, therefore, also a necessary
prerequisite for the construction of risk-controlled portfolios on robo-
advisor platforms. However, this is only half the battle. The main audience
for robo-advisor platforms are private investors who are unlikely to have any
experience with investment mathematics. This is where the robo-advisor
platform can demonstrate its strengths, for example, by determining the
mixing ratio of the components for the available part of the investment
universe, where the diversification effect is greatest. This approach can be
formulated mathematically and transformed into an optimization problem

(Choueifaty and Coignard 2008). All you have to do is maximize the
diversification ratio DR, which can be determined as follows:
N
i=1 wi · σi
max DR = max (5.1)
wi σp
given the boundary conditions
N

wi = 1
i=1
and
wi ≥ 0 ∀ i = 1, . . . , N
with wi as weight of asset i within the portfolio, σi as volatility of asset i, σp

as portfolio volatility, and N as number of assets. Or, put in another way, the
diversification ratio can be expressed as portfolio risk without diversification
divided by portfolio risk with diversification. Hence, the weighted sum
of asset risk divided by the total portfolio risk equals the maximum
diversification ratio DR at its peak. In this type of portfolio optimization,
the ratio of the weighted individual risks of the asset classes (excluding
diversification) to the actual portfolio risk (i.e. including diversification) is
maximized. This process in the two-asset case can be illustrated as displayed
in Fig. 5.1. If the two axes are swapped and the diversification ratio is also
plotted, then the point you are looking for in the graph can be read directly
as the maximum (see Fig. 5.2). The advantage of a portfolio structured
in this way is that it has the highest risk-adjusted diversification effect of
all portfolios that can be built from the investment universe (Choueifaty
et al. 2013). Elements from the investment universe that diversify well
are highly rewarded, even if, in themselves, they might not have been
considered when using other portfolio structuring techniques (such as
variance minimization) due to their volatility, which may be somewhat
higher.
78 T. RUEHL
Fig. 5.1 Diversification effect
Fig. 5.2 Diversification ratio shows maximum at 2.0

5.3.2 Equal Distribution of Risks to the Investment Instruments

Contained in the Portfolio
The maximum diversification approach described above not only optimizes
the allocation weightings based on this target, but also implicitly selects
the investment instruments from the investment universe, that is, not all
available instruments are necessarily included in the portfolio. This can
frustrate some investors who have actively chosen a number of instruments
and are now disappointed not to find them all in their portfolio. In this case,
there is a way to ensure that all previously selected instruments are actually
included in the portfolio, while maintaining a balance between them in
terms of risk. This method therefore assigns the same risk contribution to
all instruments that are to be found in the portfolio. Portfolios built in this
way have become quite popular in recent years and are referred to as “risk
parity” portfolios. With risk parity, the portfolio is optimized in such a way
that all instruments have the same contribution to the total risk (Teiletche
et al. 2010).
ACT Ri 1
P CT Ri = = ∀ i = 1, ..., N (5.2)
σp N
where N is number of assets, σp is portfolio volatility, ACT Ri is absolute

contribution to total risk of asset i, and P CT Ri is percentage contribution
to total risk of asset i. The advantage of a portfolio built in this way is that
it avoids structural cluster risks. The correlations between the asset classes
and thus their diversification potential are explicitly taken into account.
However, during major financial market crises, the effect can be observed
time and again that investors on a large scale close out risky positions across
markets and regions and withdraw liquidity from the market. As a result,
the correlations between these risky instruments rise abruptly during the
crisis (so-called diversification breakdown). In other words, where there
was protection by diversification at least on paper, when it is needed most
urgently, it is gone.
In order to anticipate this effect of the increasing correlations in
the crisis, a modified approach can therefore be chosen in advance, in
which uniform volatility contributions are allocated instead of uniform
risk contributions. This modification not only protects against unpleasant
surprises during market corrections, but also offers the advantage of easier
computation, as the portfolio weights can be calculated directly without
80 T. RUEHL
having to carry out a (more time-consuming) optimization.
1
σi
wi · σi ≡ wj · σj ∀ i, j ⇔ wi = N 1
∀ i = 1, , ..., N
j =1 σj
(5.3)
with N as number of assets, wi as weight of asset i, and σi as volatility

of asset i. The latter modification retains the advantage of taking into
account all preselected instruments from the investment universe. This also
simplifies the calculation compared to the conventional risk parity method
(as no optimization is required) and makes the portfolio less sensitive to a
“diversification breakdown” at times of crisis.
5.3.3 Risk Minimization

The two purely quantitative approaches described above ensure that either
the potential diversification effect is fully exploited for a given investment
universe or that the risks contained in the portfolio are distributed as evenly
as possible. For particularly risk-averse investors, however, it is advisable to
make portfolios available that minimize the overall portfolio risk (Clarke
et al. 2011):
N
N
min σp2 = min wi · wj · σi,j
w i
i=1 j =1
N
N
N
= min wi2 · σi2 + wi · wj · σi,j (5.4)
wi
i=1 i=1 j=1
i=j
single risk part

diversification part
with boundary conditions

N

wi = 1
i=1
and
wi ≥ 0 ∀ i = 1, ..., N
as well as N as number of assets, σp2 as portfolio variance, σi,j as covariance

of asset i and j, and wi as weight of asset i.
Portfolios built in this way (generally based on factor models) have
become popular in recent years due to their outperformance in equities. In
theory, however, this portfolio has a serious disadvantage: it is “below” the
capital market line, that is, the combination of a tangency portfolio and the
risk-free investment leads (theoretically!) to the same low risk with higher
expected returns. In practice, however, this disadvantage can be ignored:
as return forecasts are needed to build a tangency portfolio. However,
it is not possible to use historical data to come up with even short-term
forecasts with the same confidence as it is the case with pure risk indicators.
A tangency portfolio calculated using mean-variance optimization is, there-
fore, fraught with such uncertainty that the minimum variance portfolio,
figuratively speaking, still lies within the error bars. The disadvantages that
are relevant in practice are of a different kind: as with the portfolio with
the maximum diversification ratio, even the minimum variance portfolio
does not ensure that all previously selected instruments are included in
the portfolio. In addition, the minimum variance optimization in mixed
portfolios generally leads to a very high share of bonds, because it takes
into account their low volatility, but not their low continuous yield.
5.3.4 Methods Based on Return Forecasts

The methods presented so far are purely risk based, that is, the expected
return forecasts were deliberately omitted in order to circumvent the
associated forecasting error issues. Even if forecast-free strategies have
very advantageous characteristics despite the complete renunciation of
an assessment of the market movement and have already been able to
hold their own on the market (Clarke et al. 2013), it could, nevertheless,
be argued that this in a way throws the baby out with the bathwater,
because to avoid the problems associated with return forecasts, these
have been completely foregone. In the following, therefore, we show
a viable path to conventional portfolio optimization, in which a mean
variance optimization (MVO), according to Markowitz, is performed using
expected returns (Markowitz 1952):
U = μp − λ · σp2 (5.5)
82 T. RUEHL
where μp is expected portfolio return, σp2 is portfolio variance, and λ is the

risk aversion parameter.
The uncertainty under which the MVO is optimized takes into account
only the dispersion of market returns, reflecting portfolio volatility σ . The
return estimates for the individual portfolio components, which aggregate
the expected portfolio return μp , are, however, implicitly assumed to
be the exact mean of the distribution. This assumption, which is far
removed from practice, leads to some very undesirable effects in portfolios
optimized in this way: for example, in an MVO, highly correlated assets
are considered perfect substitutes and are played “against each other” due
to small differences in the return estimate, although the actual forecast
error may be of the same magnitude as the estimated return spread. In
other words, what at first glance looks like taking advantage of an arbitrage
opportunity may turn out to be merely reinforcing a forecasting error
ex post. Over time, comparatively small changes in the return estimates,
which in reality, are due to forecasting errors, can lead to allocation leaps
that ultimately rely on artifacts. These disadvantages of the MVO can be
mitigated by a suitable transformation of return estimators. The Black-
Litterman model (Litterman 2003) is, for example, very suitable for this
purpose. The Black-Litterman model supplements the pure MVO with a
process step in which the “raw” return estimators are modified as follows:
• Forecasts for highly correlated markets will be aligned based on

this information. This counteracts the MVO’s ability to treat highly
correlated assets as perfect substitutes from a risk perspective and to
“play them off” against each other in the event of diverging forecasts.
• Forecasts with higher confidence are given a greater consideration
than those with lower confidence. This approach is intuitive. Bor-
derline cases are pure MVO (all forecasts are highly reliable) and a
preselected anchor portfolio (e.g. the investor’s long-term benchmark
portfolio) in the event that no reliable forecasts are available. In turn,
the forecast-free approaches outlined above can be used as an anchor
portfolio, so that in the event of high forecasting uncertainty an
allocation that is advantageous from a pure risk perspective can be
targeted.
The forecasting quality (confidence) can be determined by a sliding

measurement of the variance of the forecasting errors, implicitly assuming
a certain persistence in the quality of the estimates.
A fundamental disadvantage of forecast-based models, as mentioned at
the beginning, is that return forecast models are associated with increased
maintenance costs compared with pure risk models. However, this can
be limited by picking the right model. Based on our own experience,
the general regression neural network (GRNN), which is an extension
of the probabilistic neural network (PNN) for non-discrete allocations,
is very well suited, as it allows GRNNs to be used to approximate non-
linear correlations such as price forecasts based on economically relevant
variables. The GRNN is based on a very intuitive basic assumption: the
more similar the past explanatory variables are to the current constellation,
the more likely it is that the following price performance will closely
resemble past performance. Another advantage of the GRNN is that only
a single free parameter needs to be determined by optimization. This is the
size of the neighborhood in the weighted approximation. If the chosen
neighborhood parameter is infinitely large, on the one hand, an arithmetic
averaging overall historical events will result. If the chosen neighborhood
parameter, on the other hand, is infinitely small, the GRNN will simply
act as a nearest neighbor estimator. Realistic neighborhood settings will of
course lie between those two extremes.
The GRNN is ideally suited for adaptive forecasts, as each new input
vector (consisting of the currently measured relevant economic variables)
with the corresponding realized market return can be easily integrated into
the existing database, and thus can be immediately fed into the next return
estimate. This approach, therefore, has considerable advantages for robo-
advisory services discussed here. With conventional regression analysis—if
one wanted to use such an adaptive method—it would be necessary to
reestimate the regression coefficients on an ongoing basis or to redevelop
the regression function for every newly added data set completely from
scratch, which would be even more time consuming. A detailed description
of the procedure can be found, for example, in Specht (1991) and Rühl
(2001).
84 T. RUEHL
5.4 DEALING WITH RISK TARGETS
5.4.1 Adherence to Lower Value Limits

The focus on and assessment of risks is rightly considered to be one
of the benefits of robo-advisory services. The strategies presented above
focus on the risk side: either by avoiding cluster risks, by maximizing the
diversification effect, or by minimizing the overall risk. The possibility of
automating a robo-advisor platform makes it possible to agree an individual
lower value limit at the securities account level. Although lower value
limits in the sense of capital preservation are no longer possible for a
one-year time period due to the current interest rate environment, a
previously accepted loss in value of max. 10%, for example, still represents
a considerable limitation of the loss potential compared with an unsecured
investment.
The maximum loss on the paid-up capital borne by the investor must be
converted into an actual current maximum loss, which takes into account
the previous market performance, that is, a positive market performance
will increase the actual buffer available, while a negative market perfor-
mance will erode part of the buffer. The buffer actually available on the
basis of these two effects (initial buffer + market performance) then defines
the maximum still acceptable value at risk (V aRmax ) of the portfolio. If the
actual value at risk (V aRact ) threatens to exceed the remaining buffer, the
portfolio will have to become more defensive. The decision-making and
control process that must be carried out continuously (and automatically!)
in such a portfolio is as follows:
• As long as V aRakt ≤ V aRmax applies, the current allocation can

be retained or any safety measures can be resolved until V aRakt =
V aRmax again.
• If, however, as a result of a negative performance or an increase in
market risk with V aRakt > V aRmax , a more defensive allocation must
be selected until V aRakt ≤ V aRmax again. The VaR reduction can be
achieved either by adding liquidity or by choosing a more defensive
but still fully invested allocation.
However, this procedure has one distinct disadvantage: the lower the
available risk buffer, the higher is the probability that the portfolio will
have to be completely removed from all risky investments (the so-called
cash lock). Especially with long investment periods, it can happen that
while the investment period, which is still available for value growth, can
no longer be used, the portfolio remains “logged in” to the maximum loss.
In other words, although it is technically feasible and even practicable at
a reasonable cost to protect the value of an individual securities account,
it raises the question of what to do in the case of a cash lock. In the case
of conventional securities accounts, where clients have access to advisors,
a solution can be found through dialogue. However, automated solutions
must be offered as part of robo-advisory services. In the case of a hedging
horizon of one year, a cash lock can, of course, be “paused” until a new
risk buffer is made available again at the beginning of a new calendar
year. If, however, a market reset occurs very early in the year that forces
the portfolio completely out of the market and the market subsequently
recovers, this usually leads to a high disappointment potential, as a negative
portfolio result is offset by a positive annual financial statement at a market
level. Other forms of risk management are available to avoid such potential
conflicts going forward. For a practical implementation of risk targets,
for example, an investment period to be chosen by the investor can be
specified, at the end of which the invested capital is preserved with a
sufficiently high degree of confidence. This takes advantage of the fact that
the risk increase will be sharper than linear in the shorter term, but weaker
in the long term (Danielsson and Zigrand 2006).
5.4.2 Specification of a Risk Preference by Choosing a Target

Investment Period
Instead of working with a maximum loss target, investors can alternatively
choose an investment horizon after which the invested capital is highly
likely to be maintained at least nominally with a specified level of confidence
(e.g. 95% or 99%). Over this period, the expected return of the portfolio
“applies” and the expected value after this period is well above 0%. This type
of risk target means that the investor is not confronted with very technical
specifications such as the choice of a risk aversion parameter λ or a target
volatility σ and can focus on the essential: a savings target in the future.
Compared to a lower value limit for shorter periods, the advantage of this
approach is that it is immune to the cash lock risk.
The mathematical-technical implementation of this specification takes
place where it can be solved with comparatively little (additional) cost due
86 T. RUEHL
to the existing infrastructure: on the robo-advisor platform. Even if short-

term return forecasts are extremely unreliable, estimates of excess returns
at the asset class level can be made, at least in the long term, with sufficient
confidence for the purpose intended here. In order to calculate the required
investment horizon for a certain allocation, one takes advantage of the
effect that the risk√of an investment will increase with the square root of
time, that is, σ ∼ t, and it is thus stronger than linear in the short term,
but weaker in the long term.
If one makes a conservative assumption (neglecting the compound
interest effect) of a linear increase in the expected return over time,
you can determine the point of intersection for each selected allocation
and thus the investment period from which you can expect a capital
preservation. This can be done for any confidence level using the z-factor
of choice. If the investment strategies that can be mapped using the robo-
advisory platform are categorized based on their risk/return profiles, it
is possible to filter out from this strategy universe those that achieve the
sufficiently high confidence point before or at the end of the desired
investment horizon. Alternatively or additionally, in the case of mean-
variance-optimized strategies, those risk aversion parameters which satisfy
this condition can be determined.
5.5 RETURN TARGETS AND RISK-BEARING CAPACITY:

NEED FOR INFORMATION
In discussions with investors, it repeatedly becomes clear that the full impli-
cations of the low interest rate environment for the return expectations
of all asset classes are all too often insufficiently understood. As a result,
there are often unrealistically high expected returns on the one hand and
an inappropriate risk bearing capacity on the other. This dual distortion
of expectations results in a high potential for disappointment. Anyone
who pursues a 5% return target and believes they can do so risk-free in
today’s capital market environment will almost inevitably be disappointed.
It is therefore necessary to provide information about the fundamental
relationship between risk and return, which extends beyond regulatory
requirements. This also includes fundamental cause–effect relationships.
This basic understanding then helps in the selection of the investment
strategies or their risk characteristics that are suitable for one’s own
investment needs.
Because risk premiums are paid relative to risk-free interest, this means
that if risk premiums remain the same, total return expectations must
decrease as the risk-free rate falls across all premium sources. This can
only be compensated by an increased risk premium. According to our own
experience, this is, however, not the case in the current environment. The
low interest rate environment, therefore, not only affects money market-
related forms of investment, but also lowers the realistic expected returns
across all asset classes. Unfortunately, this is only one side of the coin. Due
to the lower expected value, the return distribution as a whole shifts “to
the left”, that is, further into the negative range so that all percentiles in the
negative range are more likely to occur. In other words: while the expected
returns must be adjusted downward, the risk ratios have to be adjusted
upward. These two factors should not be lost on new robo-advisor clients,
who have made their last investment decision “some time ago”.
Actively managed investment strategies may, under favorable circum-
stances, generate up to one percentage point of additional return for each
percentage point of volatility, as a premium to withstand fluctuations. In
the following, we will assess a strategy with an expected return of 2.5% p.a.
and volatility of 3.0%. Even under this optimistic premise, the risk-bearing
capacity required to maintain a confidence level of 99% for a return target
of 2.5% p.a. is just under −4.5% on a one-year view. If a 95% confidence
is sufficient, the risk-bearing capacity drops to −2.4%, but it is statistically
exceeded every 20 years. For example, in the case of a ten-year government
bond, the period during which the risk falls to zero means that after
ten years, the bond is fully repaid. In the case of actively managed asset
management with open maturities, no such risk-free time in the future
can be identified, but as outlined above, one can statistically calculate the
time after which the paid-in capital is retained or available again with a
given confidence. This is done by taking advantage of the risk growing
weaker over time (with normally distributed returns proportional to a root
function of time) and then determining the time when the expected return
will most likely exceed the risk. In order to retain the invested capital with a
high probability and a return target of 2.5%, the investment horizon must
be extended well beyond one year. In the above example, if the probability
of loss over the course of a year is still around 20%, it will fall to just under
12% after two years and to around 3% after five years. After 7.8 years, the
loss probability will only be 1%. However, government bonds with this
residual maturity have a “guaranteed” negative yield. Figure 5.3 shows the
relationship between confidence levels (90%, 95%, and 99%) and minimum
88 T. RUEHL
Fig. 5.3 Risk profile of a portfolio with 3% volatility and 2.5% expected return
investment duration for a defensive investment strategy (assumption: 3.0%

volatility and 2.5% expected return): The solid line intersects with zero
already after 2.4 years, that is, after this time the value at risk has dropped
to 0 at 90% confidence level. It will take the aforementioned 7.8 years for
the dotted line (99% confidence level) to intersect with 0.
To present these relationships to the (potential) investor at an early stage
will also be worthwhile in the long term from the provider’s point of view.
While this may “put off” a few prospective clients in the short term, in the
long term it will ensure a stable client relationship, as this was not entered
into under the premise of unrealistically optimistic assumptions.
5.6 REQUIREMENTS FOR THE INVESTMENT

UNIVERSE AND INSTRUMENTS
The requirements for the investment universe inevitably arise from the
points already discussed. The investment universe must allow for sufficient
diversification so that strategies focused on risk management can leverage
their strengths. In addition, preference should be given to markets that
can be modeled with sufficient precision using the risk models used on the
platform. For volatility-based risk models, this means that market returns
need to be approximated as normally distributed, which can be assumed in
many liquid markets if the data frequency is not too high. Our own calcula-
tions have shown, for example, that in the current environment corporate
bonds can be described with sufficient accuracy using a parametric value-at-
risk, up to a confidence level of about 97%. However, in the case of higher
confidence levels, the risk is increasingly underestimated when assuming
normally distributed returns. In the case of a volatility-adequate mixture
of equities and government bonds, the risk can be adequately estimated
at the same data frequency with a confidence level of 99% with a (normal
distribution-based) parametric value-at-risk.
In principle, it is possible to use simple (plain vanilla) components as part
of a robo-advisory, which are then “refined” by using the relevant invest-
ment strategy. Although this means abandoning alpha at the component
level. However, this potential disadvantage is offset by the fact that passive
components are not exposed to the dangers of a manager change and can
thus be modeled with a greater degree of confidence. In order to avoid
inducing any further avoidable transactions such as rolling transactions
other than those induced by the investment strategy, exchange-traded
funds (ETFs) are preferable to derivatives despite slight cost disadvantages.
5.7 CUSTOMIZATION BY INVESTORS

Robo-advisory requires a certain prior understanding on the part of the
investor, but can use similar questions as conventional asset managers to
lead the investor to the best possible solution. The investor should be able
to select the following features of the portfolio or investment strategy:
• Investment universe: From the portfolio of available markets, the

investor must be able to choose the markets or ETFs to be included
(or excluded). To facilitate the selection process for less experienced
investors, it is appropriate to define standardized solutions, for exam-
ple German, Eurozone, European or world equities (or bonds).
If different strategy concepts are offered, this also applies analogously at

the strategy level. As a general rule: If forecast-based strategies are used,
they will generally operate on a narrower investment universe than the
90 T. RUEHL
so-called forecast-free strategies. For this reason, the entire robo-platform

investment universe will not be available for every investment strategy. To
avoid overwhelming inexperienced investors, reasonable standard solutions
should also be offered here that reflect the risk categories “conservative”
to “aggressive”. This can either take the form of several portfolios graded
by risk (conservative, balanced, aggressive) or two basic portfolios at the
extreme ends of the risk spectrum (conservative and aggressive), which are
then combined to match the investor’s chosen risk profile. This takes us to
the next point.
• The risk appetite (see also “Risk targets over investment horizons”):
Very few investors are able to specify their risk aversion parameters λ
for a Markowitz-based optimization. A large proportion of the target
audience of a robo-advisory platform will find it difficult to name a
concrete target volatility.
Matters are complicated by the following effect, which can be observed

quite often: Depending on the current market environment, strategies
are often preferred that significantly overburden the investor’s actual risk-
bearing capacity in times of crisis. If such a risky strategy is in the immediate
vicinity of a new all-time high, its inherent risk is typically underestimated,
as “everything has always worked out fine”. If the historical drawdowns,
which can be recognized from the graphically visualized time series, are
then experienced in real time, they are perceived as much more threatening:
What if things do not work out fine this time? Often the investment is then
terminated in an untimely manner, and the resulting loss is realized.
It is, therefore, more effective if the investor either specifies the maxi-
mum loss amount or specifies the investment horizon according to which
at least the capital employed is highly likely to be obtained (or recovered).
With the help of these specifications, those strategies can then be presented
for further selection along with their risk characteristics that meet these
conditions with a high degree of confidence. Depending on the complexity
of the platform and the level of professionalism of the investor, a single
standardized solution can be offered at this point, which adheres to both
conditions.
5.8 SUMMARY
Robo-advisory thrives on automation and quant models enable a high
degree of automation on the strategy side. While quant models can work
well in more conventional environments (this will remain the preferred
choice for institutional investors for the foreseeable future), robo-advisory
and quantitative investment strategies represent a very good structural fit.
The cost savings that robo-advisory offers compared to conventional asset
management can largely be passed on to the investor. This can deliver
added value, especially compared to the highly standardized solutions
for small investors that are otherwise customary on the market. Within
the now broad spectrum of quantitative strategies, however, a distinction
must be made: time-stable, low-maintenance models are preferable, which
implicitly amounts to a renunciation of return forecasts. Strategies from
the field of postmodern portfolio theory are optimal in this respect, which
focus specifically on risk budgeting and/or risk minimization.
The strength of the models described here lies, among other things,
in the control of portfolio risks not previously achieved by less affluent
investors, right up to the specification of maximum loss limits or the
specification of an investment horizon, according to which it is highly
probable that the invested capital will be available again at least nominally.
The models can also be set up in such a way that they are customizable by
the investor within the previously defined framework. The complexity at
the level of the selection process must accommodate the investor’s level of
experience. Less experienced investors should therefore continue to have
access to a manageable number of standardized solutions in the future.
While such standardized solutions are still often the state of play for all
investment groups, the approaches shown here can also appeal to more
demanding investors.
REFERENCES
Bankenverband. 2017. Positionspapier des Bankenverbandes zu Robo-Advice. Tech.
rep. https://bankenverband.de/media/files/2017_03_20Positionspapier_
RoboAdvice.pdf .
William A. Brock, Josef Lakonishok, and Blake LeBaron. 1992. “Simple Tech-
nical Trading Rules and the Stochastic Properties of Stock Returns”. Journal
of Finance 47(5): 1731–1764. https://doi.org/10.1111/j1540-6261.1992.
tb04681.x.
92 T. RUEHL
Yves Choueifaty and Yves Coignard. 2008. “Toward Maximum Diversification”.

The Journal of Portfolio Management 35(1): 40–51. https://doi.org/10.3905/
JPM.2008.35.1.40.
Yves Choueifaty, Tristan Froidure, and Julien Reynier. 2013. “Properties of the
Most Diversified Portfolio”. The Journal of Investment Strategies 2: 49–70.
https://doi.org/10.21314/JOIS.2013.033.
Roger G. Clarke, Harindra de Silva, and Steven R. Thorley. 2011. “Minimum
Variance Portfolio Composition”. The Journal of Portfolio Management 37(2):
31–45. https://doi.org/10.2139/ssrn.1549949.
Roger G. Clarke, Harindra de Silva, and Steven R. Thorley. 2013. “Risk Parity,
Maximum Diversification, and Minimum Variance: An Analytic Perspective”.
The Journal of Portfolio Management 39(3): 39–53. https://doi.org/10.2139/
ssrn.1977577.
Jon Danielsson and Jean-Pierre Zigrand. 2006. “On Time-Scaling of Risk and the
Square–Root–of–Time Rule”. Journal of Banking & Finance 30(10): 2701–
2713. https://doi.org/10.1016/j.jbankfin.2005.10.002.
Richard C. Grinold and Ronald N. Kahn. 2012. Active Portfolio Management.
McGraw-Hill Professional.
Jonathan W. Lam. 2016. Robo-Advisors: A Portfolio Management Perspective. New
Haven, CT: Yale College. https://economics.yale.edu/sites/default/files/
files/Undergraduate/Nominated%20Senior%20Essays/2015-16/Jonathan_
Lam_Senior%20Essay%20Revised.pdf .
Bob Litterman. 2003. Modern Investment Management: An Equilibrium Approach.
Hoboken: Jon Wiley & Sons.
Harry M. Markowitz. 1952. “Portfolio Selection”. Journal of Finance 7(1): 77–91.
Attilio Meucci. 2009. Risk and Asset Allocation. Springer Finance.
Thorsten Rühl. 2001. “Studien zur Verbesserten Ausnutzung des Informationsge-
haltes von Multisensorsystemen”. PhD thesis. Institut für Angewandte Physik
Justus-Liebig-Universität Gießen. http://geb.uni-giessen.de/geb/volltexte/
2002/709/pdf/d020009.pdf .
Stephen Satchell. 2003. In Advances in Portfolio Construction and Implementation
(Quantitative Finance), ed. Alan Scowcroft. Butterworth-Heinemannn.
Donald F Specht. 1991. “A General Regression Neural Network”. IEEE Trans-
actions on Neural Networks 2(6): 568–576. https://doi.org/10.1109/72.
97934.
Jérôme Teiletche, Thierry Roncalli, and Sébastien Maillard. 2010. “The Properties
of Equally Weighted Risk Contribution Portfolios”. Journal of Portfolio Man-
agement 36(4): 60–70. https://doi.org/10.3905/jpm.2010.36.4.06.
CHAPTER 6
Analysis of the Use of Robo-Advisors

as a Replacement for Personal Selling
Goetz Greve and Frederike Meyer
Contents
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.2 Robo-Advisors as Customer–Salesperson Interaction Technologies . . . . . 94
6.3 Theoretical Background and Research Propositions. . . . . . . . . . . . . . . . . . . . . . 96
6.3.1 Trust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.3.2 Perceived Risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.3.3 Psychological Reactance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.3.4 Perceived Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.4 Experimental Design and Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.5 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.5.1 Data Preparation and Manipulation Checks . . . . . . . . . . . . . . . . . . . . . . 99
6.5.2 Hypotheses Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.6 Managerial Implications and Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
G. Greve ()
Hamburg School of Business Administration (HSBA), Hamburg, Germany
F. Meyer
Sparkasse Harburg-Buxtehude, Hamburg, Germany

94 G. GREVE AND F. MEYER
6.1 INTRODUCTION
With the development of robo-advisors and their adoption by consumers,
the sales management of financial services is changing rapidly. This devel-
opment may lead to the disintermediation of salespeople, as technologies
emancipate customers and they can inform themselves about offerings.
Consequently, customers may not view the buying process as one that
is necessarily driven by humans. Whereas research has already surveyed
the perspective of salesperson technology adoption (e.g. Ahearne and
Rapp 2010; Verma et al. 2016; Moncrief 2017), little is known about
the customer perspective when it comes to customer–salesperson inter-
action technologies. The goal of this study, therefore, is to compare
robo-advisors with salespersons and hybrid solutions and their impact on
behavioral constructs. Using experiments, we contribute to the literature
by investigating how different forms of customer–salesperson interaction
technologies impact customer perception with respect to trust, perceived
risk, psychological reactance, and perceived use. Second, we contribute to
the growing robo-advisory literature by demonstrating how robo-advisory
services are perceived by customers.
6.2 ROBO-ADVISORS AS CUSTOMER–SALESPERSON

INTERACTION TECHNOLOGIES
Robo-advisors can be defined as automatic, web-based tools that provide
customers financial advice without human intervention using computer
algorithms to manage client portfolios. Robo-advisors are seen as a true
asset management innovation with significant growth potential. Accord-
ingly, fintech start-ups as well as traditional banks and asset managers are
increasingly showing interest in offering robo-advisory services as these
may offer the possibility to replace human salesperson advice by robo-
advice (Wirtz et al. 2018). Traditionally, the salesperson has been the
leading player in a buyer–seller dyad (Zboja et al. 2016). However, driven
by the development of digital technologies and tools, consumers have
changed their buying behavior in accordance with the adoption of interac-
tive new media (Crittenden et al. 2010) as information about products and
services is available online without any salesperson contact. As interactive
new media increases the availability of online information to a maximum,
it is not surprising that, on the one hand, rising customer expectations
and customer avoidance of personal buyer–seller negotiations on the other
ANALYSIS OF THE USE OF ROBO-ADVISORS AS A REPLACEMENT FOR… 95
impact personal selling and sales management (Anderson 1996). Today, for
a large variety of products and services, consumers are more sophisticated
than ever before, and are buying without the aid of sales personnel (Verma
et al. 2016). Ahearne and Rapp (2010) conclude that the more upcoming
technologies enable the consumer to make an educated buying decision
on his/her own, the higher is the probability that disintermediation of
salespeople will occur. On a salesperson–customer interface technology
continuum, they propose on the one end salesperson-specific technologies,
that is, technologies used solely by the sales representatives. They point
out that in most circumstances customers do not see or are not even
aware of technology use. On the other end, they consider customer-specific
technologies as technologies used solely by the customer. They propose
that these technologies may eliminate the role of the salesperson. This
end, therefore, reflects a potential for disintermediation of salespeople. It
can be proposed that consumers may no longer be exclusively driven by
human contact or they even may not require human contact for making
buying decisions anymore (Moncrief 2017; Ahearne and Rapp 2010;
Moore 2015; Marshall et al. 2012). This argumentation may be true for
a large variety of existing products that are already sold via e-commerce
solutions. However, it can be questioned whether this proposition also
holds true for complex financial products. With respect to technology, the
research on sales from the salesperson’s perspective has been extensive so
far (Sharma et al. 2010). Studies have revealed that for consumer self-
service technologies, acceptance is driven by ability and the ease of use
(Meuter et al. 2005; Brown et al. 2008). However, we see a distinct gap
in the analysis of different levels of customer–salesperson interaction tech-
nologies. So far, it remains unknown whether customer-specific interaction
technologies (CSIT) impact the disintermediation of salespeople, especially
within the field of complex financial products. Consequently, we follow
the argumentation of Anderson (1996) and Ahearne and Rapp (2010)
and test the effects of either robo-advisors as a form of selling without
human contact or human-driven personal selling in an investment-advisory
setting. As a third option, we consider a hybrid form of personal selling with
technology support, thereby depicting different positions on the proposed
salesperson–customer interface technology continuum.
6.3 THEORETICAL BACKGROUND AND RESEARCH

PROPOSITIONS
As suggested by the relationship marketing paradigm as well as social
exchange theory, trust (Pavlou 2003), perceived risk (Sheth et al. 1999;
Pavlou 2003), psychological reactance (Brehm 1968), and perceived
use (Thibaut and Kelley 1959) are important concepts of buyer–seller
interactions-especially in the online (e-commerce) context. In accordance
with the conceptualization of Ahearne and Rapp (2010), we consider
personal selling as close relational exchanges with a maximum of social
interaction, thereby describing the maximum of salesperson-oriented
interaction technologies on the proposed salesperson–customer interface
technology continuum. On the other end of the continuum, we consider
robo-advisors as discrete exchanges characterized by nil social interaction.
Robo-advisors are solely used by the customers, and, so, eliminate the role
of the salesperson. Robo-advice can be defined as digital investment advice
tools that match consumers on the basis of their personal preference to
financial products (Ringe and Ruof 2018). It should be noted that a key
characteristic of robo-advisors is the absence of any human contact between
the advisor and the customer (Fisch et al. 2017). Hence, employing these
concepts in the uncertain context of robo-advisors as a new form of
online selling is also reasonable. Therefore, perceptions of trust, risk,
psychological reactance, and use are likely to be important factors in
consumer acceptance of robo-advisors.
6.3.1 Trust
According to social exchange theory (Thibaut and Kelley 1959), people
form exchange relationships on the basis of trust and perceived risk. Trust
is defined as “confidence in the exchange partner’s reliability and integrity”
(Morgan and Hunt 1994). It is viewed as an important component for
successful relationship building (Stewart and Pavlou 2002). However,
exchange relationships that are likely to cost more than the potential
reward will be avoided. On the internet, customers typically perceive
higher risk compared to a conventional shopping environment (Tan 1999)
due to distance, virtual identity, and lack of regulation. Therefore, trust
is the preliminary condition for consumers’ e-commerce participation.
In the context of robo-advisors, it becomes clear that there is a risk of
monetary loss since consumers have to solely rely on online information
and, so, may become more vulnerable to inaccurate, incomplete, or wrong

information provided by robo-advisors. In addition, there is a risk of privacy
loss by providing personal information to robo-advisors. Consequently,
the importance of trust is elevated in e-commerce because of a possible
high degree of uncertainty and risk present in most online transactions
(Jarvenpaa et al. 1999). Hence, we propose Hypothesis 1: A higher level
of CSIT will lead to a lower level of trust.
6.3.2 Perceived Risk

Perceived risk is defined as the consumer’s subjective belief of suffering a
loss in the pursuit of a desired outcome (Bauer 1960; Sheth and Parvatiyar
1995). Within the context of e-commerce, perceived risk is regarded
as an important driver of consumer intentions to buy. In the online
context, the distant and impersonal nature of the online environment
has been associated with environmental uncertainty (e.g. technology) or
behavioral uncertainty (e.g. relational) (Bensaou and Venkatraman 1996).
Environmental uncertainty may arise out of the missing control of the
consumer as regards its information. Although the seller has an important
influence on the security of the transaction medium (e.g. encryption,
authentication, and firewalls), there is still a possibility of third parties
compromising the transaction process. Behavioral uncertainty may arise
because of opportunistic behavior of the seller, including false information,
misleading product presentations, and misleading advertising. Perceived
risk has been shown to negatively influence consumer online buying inten-
tions (Jarvenpaa et al. 1999; Pavlou 2003). The perceived risk associated
with robo-advisors may reduce perceptions of behavioral and environ-
mental control, and this lack of control may negatively influence buying
intentions. Within the context of robo-advisors, we assume that customer-
specific technologies will result in a higher level of customer’s perceived
risk due to technology-driven environmental uncertainty. Therefore, we
derive Hypothesis 2: A higher level of CSIT will lead to a higher level of
perceived risk.
6.3.3 Psychological Reactance

The socio-psychological theory of psychological reactance indicates that
when a perceived freedom is eliminated or threatened with elimination,
the individual will be motivated to reestablish that freedom (Brehm 1968).

Relationship marketing is widely based on commitment (Morgan and Hunt
1994). This commitment can be either formalized by a contractual setting
or not. However, formalization may be interpreted by the customer as a
thread to its freedom of choice as for promotional influence or manipulative
advertisement (Clee and Wicklung 1980). Consequently, customers will
show psychological reactance. Transferred to the context of CSIT, con-
sumers are confronted with many unwanted marketing communications
through various channels. In addition, they may recognize persuasive
tactics of salespersons and their attempt to push the purchase of a specific
product. These forced intrusions are perceived as threats to their freedom
of choice (Martin and Murphy 2017). Consequently, consumers may show
psychological reactance, which, in turn, may motivate consumers to regain
their lost freedom (Edwards et al. 2002). We, therefore, assume that a
higher level of CSIT will lead to a lower level of psychological reactance.
Hence, we state Hypothesis 3: A higher level of CSIT will lead to a lower
level of physiological reactance.
6.3.4 Perceived Use

According to social exchange theory, “perceived use” is a key construct to
explain why consumers continue or complete a social interaction. Thibaut
and Kelley (1959) postulate that a consumer judges the use of an interac-
tion on the basis of a comparison level describing an individual cost-benefit
ratio. Applied to technology, the term “perceived usefulness” is defined as
the individual’s perception that the use of the new technology will enhance
or improve his or her performance (Davis 1989). In the context of robo-
advisors, usefulness refers to the degree to which consumers believe the use
of robo-advisors will improve their performance or productivity, thereby
enhancing the outcome of investment advice. However, in comparison
with investment advice by salespeople, the relationship between robo-
advisors and perceived usefulness is not clear. With decreasing financial
literacy in society (Hastings et al. 2013), we propose an opposite trend
toward salespeople interaction in investment advice. Hence, we propose
Hypothesis 4: A higher level of CSIT will lead to a lower level of perceived
use.
6.4 EXPERIMENTAL DESIGN AND DATA COLLECTION

To test the derived framework, a single factorial between-subjects exper-
iment was conducted with the experimental factors personal salesperson
advice, robo-advice, and hybrid advice (robo-advice + salesperson advice).
For the experiment, three treatment groups of randomly selected cus-
tomers of a German savings bank were formed, and each of them was
exposed to a different stimulus. For the purpose of this research, it was
decided to use real robo-advisors and a real investment advice setting at
a savings bank, including real salespeople and robo-advisors, in order to
have a high degree of realism (Geuens and De Pelsmacker 2017). Any
other influences, such as the local conditions, were reduced, or, at least,
homogenized, for all groups to limit potential biases. The manipulated
stimuli as well as the entire study design were pretested among partic-
ipants. In toto, a convenience sample of 75 participants was collected
and randomly assigned to the three treatment groups, resulting in three
independent samples, with 25 participants each using a parallel design. The
participants were aged between 18 and 35 years. Furthermore, the sample
was equally distributed between women and men. To measure the surveyed
constructs, multi-items were used that were already established by prior
research and are accepted in literature; most items were then measured on
a seven-point Likert scale. The measurement of trust was done by applying
the scale of Lee and Henderson (1992). Perceived risk was measured by
adapting the scale of Sheth and Parvatiyar (1995). Psychological reactance
was measured by adapting the scale of Hong and Page (1989). Perceived
use was measured by adapting the scale of Zaichkowsky (1994).
6.5 RESULTS
6.5.1 Data Preparation and Manipulation Checks

Almost all constructs had alpha scores well above 0.7 (trust = 0.911;
perceived risk = 0.785; psychological reactance = 0.686; perceived
use = 0.889), displaying a good level of reliability (Nunnally and Bernstein
1994). Pearson’s Chi-square test and Kruskal–Wallis H-test were applied
to check for structural equality based on gender and age. The results
indicate no significant difference between the three experimental groups.
Correspondingly, the groups can be compared and used to reliably
and validly investigate the proposed differences in consumer behavioral
constructs under different conditions. Statistical measures show that

the manipulation of salesperson advice, robo-advice, as well as hybrid
advice was successful. To check the manipulations of salesperson advice
(salesperson advice, robo-advice, and hybrid advice), pairwise Kruskal–
Wallis tests were applied, resulting in Chi-square values ranging from
8.967 (psychological reactance) to 30.276 (trust) on a <0.01 significance
level.
6.5.2 Hypotheses Test

To test our propositions, we conducted planned comparisons. Kruskal–
Wallis H-tests for each treatment group and constructs showed that there
are statistically significant differences in the dependent variables between
the different advice treatments. Table 6.1 shows the respective results.
For all surveyed constructs, we can partially confirm our hypotheses. The
comparison for robo-advice between groups indeed revealed that these
forms of customer-specific interaction technology achieve the highest ranks
for perceived risk and psychological reactance and the lowest ranks for trust
Table 6.1 Results of Kruskal–Wallis H tests
Construct Hyp. N Alpha a Mean SD χ2 p Rank
Trust H1 75 0.911 5.71 1.11 30.288

Robo-advice 25 0.00 18.66
Hybrid advice 25 0.00 49.98
Salesperson advice 25 0.00 45.36
Perceived risk H2 75 0.785 5.62 0.91 22.398
Psychological H3 0.686 2.07 0.84 8.967
Perceived use H4 0.889 5.41 1.21 22.340
a Cronbach’s alpha scores

and perceived use compared to the other groups. However, comparing

salesperson advice and hybrid advice, interestingly, hybrid advice showed
higher ranks than salesperson advice for trust and perceived use.
For risk and psychological reactance, the groups indicated lower ranks
compared to salesperson advice. By this, we can conclude that our findings
are somewhat contradictory to the conceptualization of Ahearne and Rapp
(2010). We cannot confirm a linear increase of the measured constructs
with respect to salesperson–customer interaction technologies. Interest-
ingly, the hybrid advice seems to outperform the other two forms of
interaction, emphasizing the point that the human touch in selling still
seems to matter. This might be explained by the fact that in the case of
complex financial products, consumers still prefer a human touch. This can
be concluded by the fact that the hybrid advice (robo-advice + salesperson
advice) scores higher for all researched constructs. Hence, personal selling
for complex financial products can still be the key, even in the presence of
robo-advisors. However, it can be assumed that the detected phenomenon
may not necessarily exist forever. It can be expected that, on the one hand,
robo-advisors will develop their capabilities further on with the help of
artificial intelligence. On the other hand, it can be proposed that consumers
will adapt to the fact that more and more robots will help them with their
buying decision-making, resulting in a higher acceptance rate of robo-
advisors, and also for the financial services industry.
6.6 MANAGERIAL IMPLICATIONS AND LIMITATIONS

Our findings have important implications for the implementation of robo-
advisors in financial services. As the momentary trend of shifting resources
from salesperson-centric technologies to customer-centric technologies is
still growing strong, more and more companies evaluate the implemen-
tation of, for example, robo-advisors as a form of CSIT without any
human touch. However, our findings suggest that companies should not
underestimate the value of personal selling from a consumer’s perspective.
Hence, based on our experimental setting, we can only partially agree to the
conceptualization of Ahearne and Rapp (2010), who suggest disinterme-
diation of salespeople by machines. Our results rather indicate that banking
customers still favor a hybrid of personal selling and robo-advisory services.
This result is possibly linked to the fact that assessment advisory has a high
degree of complexity for the consumer. Based on this fact, we assume that
a disintermediation of the salesperson does not occur in a strong way as

long as the financial product has a high degree of complexity, specialization,
and individualization. Financial products with a lower degree of complexity
have a higher risk of the disintermediation of the salesperson, which can
already be seen in the area of banking accounts (e.g. Paypal). However, this
must not necessarily be the case in the future. The further development of
artificial intelligence and, therefore, the improvement of robo-advisors may
increase the risk for traditional banking services like personal investment
advisors to be disintermediated by robo-advisors. Regarding the possible
limitations of the study, it needs to be said that the external validity of the
study may be limited, as the data collection was made using an experimental
design. Hence, the relationship between CSIT and the surveyed constructs
needs to be examined in real-life settings.
REFERENCES
Ahearne, Michael, and Adam Rapp. 2010. “The Role of Technology at the
Interface Between Salespeople and Consumers.” Journal of Personal Selling
and Sales Management 30 (2): 111–120. https://doi.org/10.2753/PSS0885-
3134300202.
Anderson, Rolph E. 1996. “Personal Selling and Sales Management in the New
Millennium.” Journal of Personal Selling and Sales Management 16 (4): 17–32.
Bauer, Raymond A. 1960. “Consumer Behavior as Risk Taking.” In Dynamic Mar-
keting for a Changing World, ed. R.S. Hancock, 389–398. Chicago: American
Marketing Association.
Bensaou, Ben M., and Venkat N. Venkatraman. 1996. “Inter-Organizational
Relationships and Information Technology: A Conceptual Synthesis and a
Research Framework.” European Journal of Information Systems 5: 84–91.
Brehm, Jack. 1968. “A Theory of Psychological Reactance.” The American Journal
of Psychology 81. https://doi.org/10.2307/1420824.
Brown, Susan A., Viswanath Venkatesh, Jason Kuruzovich, and Anne P. Massey.
2008. “Expectation Confirmation: An Examination of Three Competing Mod-
els.” Organizational Behavior and Human Decision Processes 105 (1): 52–66.
https://doi.org/10.1016/j.obhdp.2006.09.008.
Clee, Mona A., and Robert A. Wicklund. 1980. “Consumer Behavior and Psycho-
logical Reactance.” Journal of Consumer Research 6 (4): 389–405. https://doi.
org/10.1086/208782.
Crittenden, Victoria, Robert Peterson, and Gerald Albaum. 2010. “Technology

and Business-to-Consumer Selling: Contemplating Research and Practice.”
Journal of Personal Selling and Sales Management 30: 103–110. https://doi.
org/10.2753/PSS0885-3134300201.
Davis, Fred D. 1989. “Perceived Usefulness, Perceived Ease of Use, and User
Acceptance of Information Technology.” MIS Quarterly 13 (3): 319–340.
Edwards, Steven, Hairong Li, and Joo-Hyun Lee. 2002. “Forced Exposure and
Psychological Reactance: Antecedents and Consequences of the Perceived
Intrusiveness of Pop-Up Ads.” Journal of Advertising 31 (3): 83–95. https://
doi.org/10.1080/00913367.2002.10673678.
Fisch, Jill E., Marion Laboure, and John A. Turner. 2017. The Economics of
Complex Decision Making: The Emergence of the Robo Adviser. Last Accessed:
25 November 2019. http://www.geog.ox.ac.uk/events/170911/Robo-vs-
Human-Advisers-Aug-28.pdf .
Geuens, Maggie, and Patrick De Pelsmacker. 2017. “Planning and Conduct-
ing Experimental Advertising Research and Questionnaire Design.” Journal
of Advertising 46 (1): 83–100. https://doi.org/10.1080/00913367.2016.
1225233.
Hastings, Justine S., Brigitte C. Madrian, and William L. Skimmyhorn. 2013.
“Financial Literacy Financial Education and Economic Outcomes.” The Annual
Review of Economics 5: 347–373.
Hong, Sung-Mook, and Sandra Page. 1989. “A Psychological Reactance Scale:
Development, Factor Structure and Reliability.” Psychological Reports 64 (3):
1323–1326. https://doi.org/10.2466/pr0.1989.64.3c.1323.
Jarvenpaa, Sirkka L., Noam Tractinsky, and Michael R. Vitale. 1999. “Consumer
Trust in an Internet Store.” Information Technology and Management 1: 45–71.
Lee, Soonchul, and John C. Henderson. 1992. “Managing I/S Design Teams: A
Control Theories Perspective.” Management Science 38 (6): 757–777.
Marshall, Greg, William Moncrief, John Rudd, and Nick Lee. 2012. “Revolution
in Sales: The Impact of Social Media and Related Technology on the Selling
Environment.” Journal of Personal Selling and Sales Management 32: 349–363.
https://doi.org/10.2307/23483286.
Martin, Kelly D., and Patrick E. Murphy. 2017. “The Role of Data Privacy in
Marketing.” Journal of the Academy of Marketing Science 45 (2): 135–155.
https://doi.org/10.1007/s11747-016-0495-4.
Meuter, Matthew, Mary Bitner, Amy Ostrom, and Stephen Brown. 2005. “Choos-
ing Among Alternative Service Delivery Modes: An Investigation of Customer
Trial of Self-Service Technologies.” Journal of Marketing 69 (2): 61–83.
https://doi.org/10.1509/jmkg.69.2.61.60759.
Moncrief, William. 2017. “Are sales As We Know It Dying . . . or Merely Trans-
forming?” Journal of Personal Selling and Sales Management 37: 1–9. https://
doi.org/10.1080/08853134.2017.1386110.
Moore, Sarah G. 2015. “Attitude Predictability and Helpfulness in Online Reviews:

The Role of Explained Actions and Reactions.” Journal of Consumer Research
42 (1): 30–34. https://doi.org/10.1093/jcr/ucv003.
Morgan, Robert M., and Shelby D. Hunt. 1994. “The Commitment-Trust Theory
of Relationship Marketing.” Journal of Marketing 58 (3): 20–38.
Nunnally, Jum C., and Ira H. Bernstein. 1994. Psychometric Theory. 3rd ed. New
York: McGraw-Hill.
Pavlou, Paul A. 2003. “Consumer Acceptance of Electronic Commerce: Integrat-
ing Trust and Risk with the Technology Acceptance Model.” International
Journal of Electronic Commerce 7: 101–134.
Ringe, Wolf-Georg, and Christopher Ruof. 2018. A Regulatory Sandbox for Robo
Advice. eng. ILE Working Paper Series 14. Hamburg. http://hdl.handle.net/
10419/179514.
Sharma, Dheeraj, Jule Gassenheimer, and Bruce Alford. 2010. “Internet Channel
and Cannibalization: An Empirical Assessment of Sales Agents’ Perspective.”
Journal of Personal Selling and Sales Management 30: 209–222. https://doi.
org/10.2753/PSS0885-3134300302.
Sheth, Jagdish N., and Atul Parvatiyar. 1995. “The Evolution of Relationship Mar
keting.” International Business Review 4 (4): 397–418.
Sheth, Jagdish N., Banwari Mittal, and Bruce Newman. 1999. Customer Behavior:
Consumer Behavior and Beyond. Cincinnati: South-Western College Publishing.
Stewart, David W., and Paul A. Pavlou. 2002. “From Consumer Response to Active
Consumer: Measuring the Effectiveness of Interactive Media.” Journal of the
Academy of Marketing Science 30 (4): 376–396. https://doi.org/10.1177/
009207002236912.
Tan, Soo J. 1999. “Strategies for Reducing Consumers’ Risk Aversion in Internet
Shopping.” Journal of Consumer Marketing 16: 163–180.
Thibaut, John W., and Harold H. Kelley. 1959. The Social Psychology of Groups.
New York: Wiley.
Verma, Varsha, Dheeraj Sharma, and Jagdish N. Sheth. 2016. “Does Relationship
Marketing Matter in Online Retailing? A MetaAnalytic Approach.” Journal of
the Academy of Marketing Science 44: 206–217.
Wirtz, Jochen, Paul Patterson, Werner Kunz, Thorsten Gruber, Vinh Lu, Stefanie
Paluch, and Antje Martins. 2018. “Brave New World: Service Robots in the
Frontline.” Journal of Service Management 29 (5): 907–931. https://doi.org/
10.1108/JOSM-04-2018-0119.
Zaichkowsky, Judith L. 1994. “The Personal Involvement Inventory: Reduction,
Revision, and Application to Advertising.” Journal of Advertising 23 (4): 59–
70. https://doi.org/10.1080/00913367.1943.10673459.
Zboja, James J., Ronald A. Clark, and Diana L. Haytko. 2016. “An Offer You Can’t
Refuse: Consumer Perceptions of Sales Pressure.” Journal of the Academy of
Marketing Science 44 (6): 806–821. https://doi.org/10.1007/s11747-015-
0468-z.
CHAPTER 7
Regulation of Robo-Advisers
in the United States
Melanie L. Fein
Contents
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.2 Investment Advisers Act of 1940 .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.2.1 Registration of Robo-Advisers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.2.2 Anti-fraud Provisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
7.2.3 Fiduciary Duty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
7.2.4 SEC Staff Guidance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
7.2.5 SEC Investor Alert. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.2.6 Supervision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.3 Securities Exchange Act of 1934 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.4 Investment Company Act of 1940 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.5 Employee Retirement Income Security Act of 1974 . . . . . . . . . . . . . . . . . . . . . 125
7.6 State Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
7.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
This chapter was written as of July 2019. It does not reflect any changes in the law that
might have occurred or may occur after then.
M. L. Fein ()
Fein Law Offices, Great Falls, VA, USA
106 M. L. FEIN
7.1 INTRODUCTION
Robo-advisers operate in the United States under a regulatory framework
that has evolved over decades to govern the securities industry. This
framework was not designed with robo-advisers in mind but has been
applied by regulators to robo-adviser activities. US regulators have not
adopted regulations specifically aimed at robo-advisers but rather have
applied the existing framework to them.
7.2 INVESTMENT ADVISERS ACT OF 1940

The Investment Advisers Act of 1940 (Advisers Act) is the federal statute
governing investment advisers in the United States. Robo-advisers are
treated as investment advisers under this act, which is administered by the
Securities and Exchange Commission (SEC). The SEC has said that “robo-
advisers are subject to all of the requirements of the Advisers Act.”1
The Advisers Act does not provide a comprehensive regulatory regime
but rather provides a principles-based overlay for investment adviser activ-
ities. As the SEC has explained:
Unlike the laws of many other countries, the US federal securities laws do
not prescribe minimum experience or qualification requirements for persons
providing investment advice. They do not establish maximum fees that
advisers may charge. Nor do they preclude advisers from having substantial
conflicts of interest that might adversely affect the objectivity of the advice
they provide. Rather, investors have the responsibility, based on disclosure
they receive, for selecting their own advisers, negotiating their own fee
arrangements, and evaluating their advisers’ conflicts.2
7.2.1 Registration of Robo-Advisers

Section 203 of the Advisers Act provides that “it shall be unlawful for
any investment adviser, unless registered under this section, to make use
of the mails or any means or instrumentality of interstate commerce in
1 See Division of Investment Management (2017) (“SEC Staff Guidance”), n.8. Because
the SEC uses the spelling “robo-advisers” (as opposed to “robo-advisor”), this chapter will
use the SEC’s spelling for consistency.
2 Investment Advisers Act of 1940, 17 C.F.R. § 275 and 279 (2008).
REGULATION OF ROBO-ADVISERS IN THE UNITED STATES 107
connection with his or its business as an investment adviser.” Robo-advisers

come within the definition of “investment adviser” in the Advisers Act and
thus generally cannot conduct business without registering under the Act.
The Act defines “investment adviser” to mean:
any person who, for compensation, engages in the business of advising

others, either directly or through publications or writings, as to the value
of securities or as to the advisability of investing in, purchasing, or selling
securities, or who, for compensation and as part of a regular business, issues
or promulgates analyses or reports concerning securities ...3
Advisers that meet this definition and have assets under the management
of $100 million or more are required to register with the SEC as investment
advisers.4 Advisers with less than $100 million in assets under management
must register with a state regulator rather than the SEC. State statutes
generally require advisers to register in every state in which the adviser
obtains more than a de minimis number of clients.
Under an SEC rule, investment advisers that provide advice through
interactive websites, such as robo-advisers, may register with the SEC rather
than the states.5 The rule is intended to lessen the burden of multi-state
registration on investment advisers that operate exclusively through the
Internet.6
3 Investment Advisers Act of 1940, Pub. L. No. 115–417 § 202(a)(11) (2019). State
securities laws similarly define “investment adviser”.
4 There are currently over 8000 SEC-registered investment advisers in the United States. A
non-US adviser giving advice to US persons must register with the SEC unless an exemption
from registration is available (in which case it still may be subject to state registration). Advisers
that would otherwise be obligated to register with 15 or more states may register with the
SEC.
5 SEC Rule 203A-2(e) defines an “interactive website” as a website in which computer
software-based models or applications provide investment advice to clients based on personal
information provided by each client through the website. An adviser relying on the exemption
may not use its advisory personnel to elaborate or expand on the investment advice provided
by its interactive website, or otherwise provide investment advice to its Internet clients, except
as permitted by the rule’s de minimis exception.
6 Investment Advisers Act of 1940, 17 C.F.R. § 275.203A-2(e) (2002).
108 M. L. FEIN
7.2.2 Anti-fraud Provisions

Section 206 of the Advisers Act contains the Act’s anti-fraud provisions,
which form the basis of the SEC’s principles-based regulation of investment
advisers. These provisions provide as follows:
It shall be unlawful for any investment adviser by the use of mails or any
means or instrumentality of interstate commerce, directly or indirectly—
1. to employ any device, scheme, or artifice to defraud any client or

prospective client;
2. to engage in any transaction, practice, or course of business which
operates as a fraud or deceit upon any client or prospective client;
3. to act as principal for his own account, knowingly to sell any security
to or purchase any security from a client, or acting as broker for a
person other than such client, knowingly to effect any sale or purchase
of any security for the account of such client, without disclosing to
such client in writing before the completion of such transaction the
capacity in which he is acting and obtaining the consent of the client
to such transaction; or
4. to engage in any act, practice, or course of business which is fraudulent,
deceptive, or manipulative7
These provisions are intended to prevent investment advisers from

defrauding, deceiving, or manipulating their clients. They do not by their
terms impose a fiduciary standard of care on an adviser’s investment rec-
ommendations or require an adviser to act in the best interest of its clients.
However, as discussed below, these provisions have been interpreted by the
courts and the SEC as imposing a fiduciary duty on investment advisers.
7.2.3 Fiduciary Duty

Based on the Adviser Act’s anti-fraud provisions, the courts and the SEC
have stated that investment advisers are “fiduciaries.” 8 However, the SEC
7 Investment Advisers Act of 1940, 15 U.S.C. §80b-6 (2020). The act authorizes the SEC
to adopt rules and grant exemptions from these prohibitions.
8 See SEC v. Capital Gains Research Bureau, Inc., 375 U.S. 180, 194 (1963). Citing
that case, the SEC has stated, “Under federal law, an investment adviser is a fiduciary.” See
Securities and Exchange Commission (2019), at 2.
has not by regulation addressed what the fiduciary duties of investment

advisers are. Rather, the SEC historically has taken a “principles-based”
approach whereby investment advisers are expected to act in their cus-
tomers’ best interest in accordance with common law fiduciary principles.
The SEC recently reaffirmed its principles-based approach in an official
interpretation of the fiduciary duty of investment advisers.9 The SEC’s
“Fiduciary Interpretation” states:
The Advisers Act establishes a federal fiduciary duty for investment advisers.
This fiduciary duty is based on equitable common law principles and is
fundamental to advisers’ relationships with their clients under the Advisers
Act. The investment adviser’s fiduciary duty is broad and applies to the entire
adviser-client relationship.10
An adviser’s fiduciary duty is comprised of a duty of care and a duty of

loyalty:
The fiduciary duty an investment adviser owes to its client under the Advisers
Act [...] comprises a duty of care and a duty of loyalty...11
This fiduciary duty requires an adviser to adopt the principal’s goals, objec-
tives, or ends. This means the adviser must, at all times, serve the best
interest of its client and not subordinate its client’s interest to its own. In
other words, the investment adviser cannot place its own interests ahead of
the interests of its client. This combination of care and loyalty obligations
has been characterized as requiring the investment adviser to act in the
“best interest” of its client at all times. In our view, an investment adviser’s
obligation to act in the best interest of its client is an overarching principle
that encompasses both the duty of care and the duty of loyalty.12
9 See Securities and Exchange Commission (2019) (the “Fiduciary Interpretation”).

10 SEC, Fiduciary Interpretation, 6–8. The SEC acknowledged that the fiduciary duty
is not specifically defined in the Advisers Act or in SEC rules but reflects a Congressional
recognition “of the delicate fiduciary nature of an investment advisory relationship” as well
as a Congressional intent to “eliminate, or at least to expose, all conflicts of interest which
might incline an investment adviser – consciously or unconsciously – to render advice which
was not disinterested.”
11 SEC, Fiduciary Interpretation, 2.
110 M. L. FEIN
The SEC emphasized that the duty to provide investment advice that is
in the “best interest” of the client includes a duty to provide advice that is
“suitable” for the client.13 In order to provide such advice, the SEC said an
adviser must have a reasonable understanding of the client’s objectives. For
retail clients, the SEC said the basis for such a reasonable understanding
generally would include an understanding of the investment profile.14 How
an adviser develops a reasonable understanding “will vary based on the
specific facts and circumstances, including the nature of the client, the
scope of the adviser-client relationship, and the nature and complexity of
the anticipated investment advice.”15 The SEC stated:
In order to develop a reasonable understanding of a retail client’s objectives,

an adviser should, at a minimum, make a reasonable inquiry into the client’s
financial situation, level of financial sophistication, investment experience,
and financial goals (which we refer to collectively as the retail client’s
“investment profile”). For example, an adviser undertaking to formulate a
comprehensive financial plan for a retail client would generally need to obtain
a range of personal and financial information about the client such as current
income, investments, assets and debts, marital status, tax status, insurance
policies, and financial goals.16
This formulation allows flexibility for robo-advisers to comply with the

fiduciary duty to provide suitable investment advice, provided the robo-
adviser also complies with the duty of loyalty. Under the duty of loyalty, an
investment adviser must “eliminate or make full and fair disclosure of all
conflicts of interest which might incline an investment adviser-consciously
or unconsciously-to render advice which is not disinterested, such that a
client can provide informed consent to the conflict.”17
The Fiduciary Interpretation specifically recognizes robo-advisers as
investment advisers, stating that robo-advisers are subject to the fiduciary
obligations that apply to investment advisers under the Investment Advisers
Act of 1940:

16 SEC, Fiduciary Interpretation, 12–13.
17 SEC, Fiduciary Interpretation, 12–13.
This [Fiduciary] Interpretation also applies to automated advisers, which are

often colloquially referred to as “robo-advisers.” Automated advisers, like all
SEC-registered investment advisers, are subject to all of the requirements
of the Advisers Act, including the requirement that they provide advice
consistent with the fiduciary duty they owe to their clients.18
The SEC emphasized that the fiduciary duty “follows the contours of
the relationship between the adviser and its client, and the adviser and
its client may shape that relationship by agreement ... we recognize that
investment advisers provide a wide range of services ... and serve a large
variety of clients.”19 The SEC noted that the fiduciary duty provides
flexibility for different business models:
In our experience, the principles-based fiduciary duty imposed by the Advis-

ers Act has provided sufficient flexibility to serve as an effective standard of
conduct for investment advisers, regardless of the services they provide or
the types of clients they serve.20
Although the SEC did not specifically endorse the robo-adviser business
model, the SEC’s statements implicitly recognize that the fiduciary duty
under the Advisers Act can accommodate robo-advisers as well.
Depending on how the “best interest” standard is interpreted and
applied, however, it could be problematic for some robo-advisers. For
example, the best interest standard can be read to require a robo-adviser
to provide personalized advice in the best interest of each customer
individually rather than its customers collectively. Moreover, it obligates
the robo-adviser, not the customer, to make an affirmative determination
that an investment recommendation is in the best interest of that customer.
It is not clear that all robo-advisers can meet this standard.
The Interpretation requires an investment adviser to provide investment
advice in the best interest of its client “based on the client’s objectives.”21
Thus, the Interpretation appears to endorse “goal-based” investment
advice that robo-advisers are expected to offer. At the same time, however,
the Interpretation would seem to require that algorithms used by robo-
18 SEC, Fiduciary Interpretation, n. 27.

112 M. L. FEIN
advisers be designed to give tailored advice to each individual client, rather

than programmatic advice designed for clients with similar investment
objectives but who may have very different individual circumstances.
Algorithms that generate goal-based investment advice without regard for
each client’s individual circumstances might not be consistent with the
Interpretation.
The Fiduciary Interpretation does not provide specific guidance on how
robo-advisers should fulfill their fiduciary duties. However, the SEC issued
staff guidance on robo-advisers in 2017 and the Fiduciary Interpretation
refers to that guidance.22
7.2.4 SEC Staff Guidance

The SEC’s Division of Investment Management (2017) issued regulatory
guidance on robo-advisers (the Guidance).23 The Guidance states that the
staff has been monitoring and engaging with robo-advisers to evaluate how
they meet their obligations under the Advisers Act.
In keeping with the SEC’s principles-based approach under the Advisers
Act, the Guidance does not impose any obligations or requirements on
robo-advisers but rather makes recommendations that robo-advisers “may
wish to consider” in complying with the act.
The Guidance states that robo-advisers “should keep in mind certain
unique considerations as they seek to meet their legal obligations under
the Advisers Act.” In recognition that not all robo-advisers are alike, the
Guidance states there may be a variety of means for a robo-adviser to meet
its obligations to its clients under the act.
The Guidance focuses on three areas as raising unique issues for robo-
advisers: disclosures, the obligation to obtain information from clients to
support the robo-adviser’s duty to provide suitable advice, and compliance
programs.
7.2.4.1 Disclosures
With respect to disclosures, the Guidance says that robo-advisers may wish
to consider the most effective way to communicate to their clients the
22 SEC, Fiduciary Interpretation, n. 27.

23 “Staff Guidance” or “Guidance”.
limitations, risks, and operational aspects of their advisory services and

should consider providing the following information:
• A statement that an algorithm is used to manage individual client

accounts;
• A description of the algorithmic functions used to manage client
accounts (e.g., that the algorithm generates recommended portfolios;
that individual client accounts are invested and rebalanced by the
algorithm);
• A description of the assumptions and limitations of the algorithm used
to manage client accounts (e.g., if the algorithm is based on modern
portfolio theory, a description of the assumptions behind and the
limitations of that theory);
• A description of the particular risks inherent in the use of an algorithm
to manage client accounts (e.g., that the algorithm might rebalance
client accounts without regard to market conditions or on a more
frequent basis than the client might expect; that the algorithm may
not address prolonged changes in market conditions);
• A description of any circumstances that might cause the robo-adviser
to override the algorithm used to manage client accounts (e.g., that
the robo-adviser might halt trading or take other temporary defensive
measures in stressed market conditions);
• A description of any involvement by a third party in the development,
management, or ownership of the algorithm used to manage client
accounts, including an explanation of any conflicts of interest such an
arrangement may create (e.g., if the third party offers the algorithm to
the robo-adviser at a discount, but the algorithm directs clients into
products from which the third party earns a fee);
• An explanation of any fees the client will be charged directly by
the robo-adviser, and of any other costs that the client may bear
either directly or indirectly (e.g., fees or expenses clients may pay in
connection with the advisory services provided, such as custodian or
mutual fund expenses; brokerage and other transaction costs);
• An explanation of the degree of human involvement in the oversight
and management of individual client accounts (e.g., that investment
advisory personnel oversee the algorithm but may not monitor each
client’s account);
• A description of how the robo-adviser uses the information gathered
from a client to generate a recommended portfolio and any limitations
114 M. L. FEIN
(e.g., if a questionnaire is used, that the responses to the questionnaire

may be the sole basis for the robo-adviser’s advice; if the robo-adviser
has access to other client information or accounts, whether, and if so,
how, that information is used in generating investment advice); and
• An explanation of how and when a client should update information
he or she has provided to the robo-adviser.
The Staff Guidance cautions robo-advisers to avoid misleading clients

about the nature of the services they provide and avoid implying that:
• The robo-adviser is providing a comprehensive financial plan if it is

not in fact doing so (e.g., if the robo-adviser does not take into
consideration a client’s tax situation or debt obligations, or if the
investment advice is only targeted to meet a specific goal—such as
paying for a large purchase or college tuition—without regard to the
client’s broader financial situation);
• A tax-loss harvesting service also provides comprehensive tax advice;
or
• Information other than that collected by the questionnaire (e.g.,
information concerning other client accounts held with the robo-
adviser, its affiliates or third parties; information supplementally sub-
mitted by the client) is considered when generating investment
recommendations if such information is not in fact considered.
Notably, the Guidance acknowledges that robo-advisers may provide

investment advice “without regard to the client’s broader financial
situation,” thus implicitly acknowledging that robo-advisers do not
necessarily give personalized investment advice tailored to each customer’s
complete financial circumstances.
Because robo-advisers rely on online disclosures, the staff said that
unique issues may arise when they communicate key information, risks, and
disclaimers. Accordingly, the Guidance reminds robo-advisers to carefully
consider whether their written disclosures are designed to be effective, e.g.,
“not buried or incomprehensible.” In particular, robo-advisers “may wish
to consider”:
• Whether key disclosures are presented prior to the sign-up process so

that information necessary to make an informed investment decision
is available to clients before they engage, and make any investment

with, the robo-adviser;
• Whether key disclosures are specially emphasized (e.g., through
design features such as pop-up boxes);
• Whether some disclosures should be accompanied by interactive text
(e.g., through design features such as tool-tips or other means to
provide additional details to clients who are seeking more information
(e.g., through a “Frequently Asked Questions” section); and
• Whether the presentation and formatting of disclosure made available
on a mobile platform have been appropriately adapted for that plat-
form.
7.2.4.2 Suitability of Recommendations

The SEC Guidance requires robo-advisers, like other advisers, to pro-
vide investment advice “consistent with the fiduciary duty” they owe
their clients.24 In this regard, the Guidance reiterates that an investment
adviser’s fiduciary duty includes an obligation to act in the best interests of
its clients and to provide only suitable investment advice. As stated in the
Guidance, “an investment adviser must make a reasonable determination
that the investment advice provided is suitable for the client based on the
client’s financial situation and investment objectives.”
However, while acknowledging that robo-advisers do not necessarily
perform this function, the Guidance does not state that robo-advisers
will be found to have violated their fiduciary duty to provide suitable
investment advice if they fail to take into consideration the client’s broader
financial situation.
The Guidance notes that robo-adviser questionnaires vary in the types of
information they request and that some questionnaires are not designed to
provide a client with the opportunity to give additional information or con-
text concerning the client’s selected responses. In addition, the Guidance
notes that robo-advisers may not be designed so that advisory personnel
may ask follow-up or clarifying questions about a client’s responses, address
24 See Division of Investment Management (2017), Staff Guidance, 10. (“As SEC-
registered investment advisers, robo-advisers are subject to all of the requirements of the
Advisers Act, including the requirement that they provide advice consistent with the fiduciary
duty they owe to their clients.”)
116 M. L. FEIN
inconsistencies in client responses, or provide a client with help when filling

out the questionnaire.
In a bulletin to investors who use robo-advisers, the SEC’s staff indi-
cated that robo-advisers might not meet the suitability standard under the
Advisers Act. The staff cautioned that robo-advisers do not necessarily col-
lect sufficient information to make suitable investment recommendations:
[A] robo-adviser’s recommendation is limited by the information it requests

and receives from you, typically through an online questionnaire. It is
important to keep in mind that some robo-advisers may obtain and consider
only limited information about you (Securities and Exchange Commission
2017).
An automated investment tool [i.e., robo-adviser] may not assess all of your
particular circumstances, such as your age, financial situation and needs,
investment experience, other holdings, tax situation, willingness to risk
losing your investment money for potentially higher investment returns, time
horizon for investing, need for cash, and investment goals. Consequently,
some tools may suggest investments (including asset-allocation models) that
may not be right for you.
For example, an automated investment tool may estimate a time horizon for
your investments based only on your age, but not take into account that
you need some of your investment money back in a few years to buy a new
home. In addition, automated tools typically do not take into account that
your financial goals may change (Securities and Exchange Commission and
Financial Industry Regulatory Authority 2015).
Given robo-advisers’ limited interaction with customers, the Guidance

suggests that a robo-adviser, when considering whether its questionnaire is
designed to elicit sufficient information to support its suitability obligation,
“may wish to consider” factors such as:
Whether the questions elicit sufficient information to allow the robo-adviser

to conclude that its initial recommendations and ongoing investment advice
are suitable and appropriate for that client based on his or her financial
situation and investment objectives;
Whether the questions in the questionnaire are sufficiently clear and/or
whether the questionnaire is designed to provide additional clarification or
examples to clients when necessary (e.g., through the use of design features,
such as tool-tips or popup boxes); and
Whether steps have been taken to address inconsistent client responses, such
as: – Incorporating into the questionnaire design features to alert a client
when his or her responses appear internally inconsistent and suggest that the
client may wish to reconsider such responses; or – Implementing systems to
automatically flag apparently inconsistent information provided by a client
for review or follow-up by the robo-adviser.25
The Guidance goes so far as to suggest that robo-advisers may result in

unsuitable investment advice:
Many robo-advisers give clients the opportunity to select portfolios other

than those that they have recommended. Some robo-advisers do not,
however, give a client the opportunity to consult with investment advisory
personnel about how the client-selected portfolio relates to the client’s stated
investment objective and risk profile, and its suitability for that client. This
may result in a client selecting a portfolio that the robo-adviser believes is
not suitable for the investment objective and risk profile the robo-adviser has
generated for the client based on his or her questionnaire responses.
To deal with this problem, the Guidance recommends that a robo-

adviser, consistent with its obligation to act in its client’s best interest,
“should consider providing commentary as to why it believes particular
portfolios may be more appropriate for a given investment objective and
risk profile” and “may wish to consider whether pop-up boxes or other
design features would be useful to alert a client of potential inconsistencies
between the client’s stated objective and the selected portfolio.”
7.2.4.3 Compliance
The SEC Staff Guidance cautions robo-advisers that the provision of advi-
sory services over the Internet with limited human interaction may create
or accentuate risk exposures for the robo-adviser that should be addressed
through written policies and procedures. The Guidance recommends that
robo-advisers consider whether to adopt and implement written policies
and procedures addressing areas such as:
The development, testing, and back-testing of the algorithmic code and the
post-implementation monitoring of its performance (e.g., to ensure that
the code is adequately tested before, and periodically after, it is integrated
25 DIM, Staff Guidance, 6–7.

118 M. L. FEIN
into the robo-advisers’ platform; the code performs as represented; and any
modifications to the code would not adversely affect client accounts);
The questionnaire eliciting sufficient information to allow the robo-adviser
to conclude that its initial recommendations and ongoing investment advice
are suitable and appropriate for that client based on his or her financial
situation and investment objectives;
The disclosure to clients of changes to the algorithmic code that may
materially affect their portfolios;
The appropriate oversight of any third party that develops, owns, or manages
the algorithmic code or software modules utilized by the robo-adviser;
The prevention and detection of, and response to, cyber security threats;
The use of social and other forms of electronic media in connection with
the marketing of advisory services (e.g., websites; Twitter; compensation of
bloggers to publicize services; “refer-a-friend” programs); and
The protection of client accounts and key advisory systems.26
7.2.5 SEC Investor Alert

Concurrent with the issuance of the staff Guidance on robo-advisers, the
SEC’s Office of Investor Education and Advocacy issued an Investor Alert
designed to “provide individual investors with information they may need
to make informed decisions if they consider using robo-advisers” (Secu-
rities and Exchange Commission 2017). The Investor Alert, combined
with the Staff Guidance, indicates that the SEC is largely leaving investors
on their own to act in their own best interests when receiving investment
advice from a robo-adviser. The Investor Alert cautions investors that:
While robo-advisers have similarities to traditional investment advisory pro-

grams, there are also differences. Before making a decision about whether to
invest through a robo-adviser, or in deciding which robo-adviser might be
best for you, you should do your own research. Make sure the robo-adviser
and the investment portfolio it puts together for you are a good match for
your investment needs and goals, and that you understand the potential costs,
risks, and benefits of using that particular robo-adviser.
The Investor Alert highlights certain issues that investors should con-
sider, including:
26 DIM, 8.
• The level of human interaction important to the investor;

• The information the robo-adviser uses in formulating recommenda-
tions;
• The robo-adviser’s approach to investing;
• The fees and charges involved.
With respect to the information used by robo-advisers in formulating rec-

ommendations, the Investor Alert cautions investors that a robo-adviser’s
recommendation is limited by the information it requests and receives and
that some robo-advisers may obtain and consider only limited information
about the customer. The Alert suggests that customers ask:
• Would you use the robo-adviser for a specific financial goal (for
example, retirement, buying a home, or investing for your children’s
education), or to meet your overall financial needs more broadly?
• Does the robo-adviser’s recommendation take into account your
purpose in using the robo-adviser?
• Does the robo-adviser’s recommendation take into account relevant
personal financial information, given your goal? For example, does
the robo-adviser ask for information about high interest credit card
debt or student loans you may have? Does it take into account your
bank and savings accounts? Does it take into account your real estate
holdings, such as your home, or other investments such as retirement
accounts? Does it take into account other assets that you have?
• How does the robo-adviser take into account your tolerance for risk?
How you respond to the robo-adviser’s questions about risk can affect
what portfolio the robo-adviser recommends. In addition to the initial
makeup of your portfolio, how does your risk tolerance impact how
the robo-adviser might rebalance your portfolio (for example, in the
event of a market decline)?
The Investor Alert notes that different robo-advisers have different

approaches to investing and advises investors to take the time to understand
how the robo-adviser develops a portfolio recommendation and to ask the
following:
120 M. L. FEIN
• Does the robo-adviser offer a limited range of investment products,

such as only ETFs? Are the investment products utilized by the robo-
adviser appropriate for your goals?
• Does the robo-adviser only offer certain limited portfolios within
those investment products? How many different portfolios could your
money possibly be invested in? What portfolio does the robo-adviser
recommend for you and why?
• What type of accounts does the robo-adviser manage? For example,
does the robo-adviser manage individual retirement accounts (IRAs)?
Taxable accounts? 401(k) accounts or college savings plans?
• How does the robo-adviser handle volatility? For example, does the
robo-adviser have the ability to freeze sales (i.e., not let you sell your
investments for cash for a certain period of time)?
• How often is your account rebalanced? Rebalancing can have tax
implications, depending on the type of account. What would trigger
a change in the asset allocation or investment categories of your
portfolio?
• Does the robo-adviser utilize tax-loss harvesting?
With respect to fees, the Investor Alert notes that a robo-adviser “may offer
lower-cost investment advice, but if the robo-adviser utilizes investment
products with high costs, your total overall costs could still be high.” The
Alert advises investors to ask:
• What fees would you be charged directly by the robo-adviser? Are

there any other costs (e.g., brokerage fees, management fees for ETFs
purchased for your account) that you would pay directly or indirectly?
• How is the robo-adviser compensated? Does the way it is compensated
create any conflicts of interest with you, the investor? For example, is
the robo-adviser paid to offer particular products or does it offer only
products with which it is affiliated (e.g., mutual funds sponsored by
the robo-adviser or its affiliates)?
• Are there penalties or fees if you want to withdraw your investment,
or transfer or close your account? Liquidating an account may have
tax implications for you as well.
• Does the amount you are charged depend on how much money you
invest?
• Can the costs and fees change over time?
• Does the robo-adviser pay a referral or marketing fee, or other

incentives for finding new clients? Robo-advisers may use different
marketing techniques, such as paying money to others or providing
discounted fees for making client referrals. You should understand if
a robo-adviser has that kind of feature, even if you are not paying a
fee yourself.
It is unclear to what extent the SEC believes that investors will heed this
cautionary advice when using with robo-advisers. Robo-advisers are not
required to give their customers a copy of the Investor Alert and it is unclear
how the SEC anticipates making investors aware of it other than by posting
it on its website. Retail investors do not generally visit the SEC’s website.
7.2.6 Supervision
Investment advisers are not subject to regular or frequent inspection by
SEC examiners. An SEC staff study reported that the average registered
adviser could expect to be examined less than once every 11 years.27 The
SEC has not indicated the extent to which it has conducted supervisory or
compliance inspections of robo-advisers.
No self-regulatory authority exists for investment advisers, unlike
broker-dealers who are required to be members of and are regulated
by the Financial Industry Regulatory Authority (FINRA).
7.3 SECURITIES EXCHANGE ACT OF 1934

The Securities Exchange Act of 1934 (Exchange Act) regulates securities
broker-dealers. Robo-advisers typically offer their advisory services in
conjunction with a broker-dealer, which executes securities transactions
for a robo-adviser’s customer accounts. The broker-dealer may or may
not be an affiliate of the robo-adviser. However, robo-advisers themselves
generally are not broker-dealers and are not required to register as such.
A “broker” is defined in the Exchange Act to mean “any person engaged
in the business of effecting transactions in securities for the account of
27 See Securities and Exchange Commission (2011), 14. In contrast, the study found that
the Financial Industry Regulatory Authority (FINRA) examined 54 percent of its broker-
dealer members in 2009.
122 M. L. FEIN
others.”28 The SEC has taken the position that an investment adviser is
not engaged in “effecting” securities transactions and is not required to
register as a broker-dealer merely because it has discretionary authority to
place orders with brokers and to execute securities transactions for client
accounts without specific compensation for this function.29
7.4 INVESTMENT COMPANY ACT OF 1940

The Investment Company Act of 1940 regulates investment companies,
commonly known as mutual funds. The act also has been interpreted
broadly by the SEC to include advisory programs that provide similar
investment advice to groups of individuals, raising an issue as to whether
robo-advisers are investment companies.
The Investment Company Act defines “investment company” generally
as an issuer of securities that “holds itself out as being engaged primarily
... in the business of investing, reinvesting, or trading in securities.”30 The
SEC interprets this definition to include certain advisory programs that
invest client assets on a group basis. In 1997, the SEC adopted a rule-Rule
3a-4—creating a safe-harbor exclusion for such programs that meet certain
requirements.
Rule 3a-4 provides a nonexclusive safe harbor from the definition of
“investment company” for certain investment advisory programs designed
to provide the same or similar professional portfolio management services
on a discretionary basis to a large number of advisory clients having
relatively small amounts to invest.31 An investment advisory program
that is organized and operated in accordance with the rule’s provisions
is not required to register as an investment company under the Investment
Company Act or to comply with the act’s requirements.32 Robo-advisers
may rely on this rule in seeking to avoid investment company regulation.
28 Investment Advisers Act of 1940, 15 U.S.C. § 78c(a)(4)(A)(2020).

29 50 Fed. Reg. 1 49835, 49839 (1985).
30 Investment Company Act of 1940, Pub. L. No. 111-72 § 3(a)(1) (2009).
31 Investment Company Act of 1940, 17 C.F.R. § 270 (3a)(4) (1997), Status of Investment
Advisory Programs.
32 The rule is intended to be a nonexclusive safe harbor; a program that is not organized
and operated in a manner consistent with the rule does not necessarily meet the Investment
Company Act’s definition of “investment company.”
In order for the safe harbor to apply, Rule 3a-4 provides that:
(i) each client’s account must be managed on the basis of the client’s
financial situation and investment objectives, and in accordance with
any reasonable restrictions imposed by the client on the management
of the account;
(ii) the sponsor of the program must obtain sufficient information from
each client to be able to provide individualized investment advice to
the client;
(iii) the sponsor and portfolio manager must be reasonably available to
consult with each client;
(iv) each client must have the ability to impose reasonable restrictions on
the management of the client’s account;
(v) each client must be provided with a quarterly account statement
containing a description of all activity in the client’s account; and
(vi) each client must retain certain indicia of ownership of all securities
and funds in the account.
In proposing the rule, the SEC stated that it was intended to cover
investment advisory programs where a client’s account is managed on a
discretionary basis in accordance with “pre-selected investment objectives’
and where “clients with similar investment objectives often receive the
same investment advice and may hold the same or substantially the same
securities in their accounts.” The SEC stated that, in light of the similarity
of management, “some of these investment advisory programs may meet
the definition of investment company under the Investment Company
Act.”33
The SEC noted that clients of an investment advisory program with
similar investment objectives may hold substantially the same securities in
their accounts in accordance with a portfolio manager’s model, but that this
“does not necessarily indicate that clients in the program have not received
individualized treatment for purposes of the rule.” The SEC also stated that
it would not be necessary under the rule for a portfolio manager to make
separate determinations regarding the appropriateness of each transaction
for each client prior to effecting the transaction.
33 Company Act, 17 C.F.R. § 270 (3a)(4); Investment Company Act of 1940, 17 C.F.R.
§ 270 (3a)(4) (1995).
124 M. L. FEIN
The SEC also noted that, prior to the rule’s adoption, the Division of
Investment Management had issued no-action letters allowing programs
that allocate client assets in accordance with computerized investment allo-
cation models to operate without being deemed investment companies.34
With respect to the requirement for initial and ongoing client contact,
the SEC stated that a program relying on the rule must provide that
the sponsor or a designated person contact and solicit information when
the client opens the account and at least annually to determine whether
there have been changes in the client’s financial situation or investment
objectives. The SEC stated that the requirement for client contact “is
critical to the provision of individually tailored advice.” However, the SEC
stated that the rule does not dictate the manner in which a sponsor contacts
its clients annually and “contact can be made, for example, in person, by
telephone, or by letter or electronic mail that includes a questionnaire
requesting the client to provide or update relevant information.”
Depending on how a robo-adviser is structured, it may or may not
qualify for the safe harbor from investment company regulation. It is
unclear, for example, whether the safe harbor would apply to programs
that do not permit the customer to purchase and own individual securities.
Moreover, to the extent robo-advisers do not manage client accounts
on the basis of each client’s financial situation and clients do not have
reasonable access to personnel who are available to consult with the client,
the safe harbor might not be available. Robo-advisers are designed to
provide investment advice based on asset allocation formulas and strategies
that result in the same investment recommendations to investors with
broadly similar investment goals. Thus, the advice may not be based on
each client’s individual financial situation. Robo-advisers are designed to
operate with no individual account manager and limit communication
with their clients to an Internet interface. Thus, clients might not be able
to consult with the sponsor or personnel of the manager of the client’s
account who are knowledgeable about the account.
34 The SEC cited no action letters to Qualivest Capital Management Inc. (pub. avail.
July 30, 1990) (sponsor will use computerized investment allocation model to allocate and
reallocate client assets among money managers); Atlantic Bank of New York (pub. avail. June
7, 1991) (sponsor’s asset allocation recommendation will be based on client’s investment
needs and sponsor’s model portfolios).
The SEC’s staff, in issuing its Guidance on robo-advisers in 2017, stated

that robo-advisers should consider whether they meet the conditions of
Rule 3a-4 and contact the staff for further guidance if necessary:
Robo-advisers should consider whether the organization and operation of

their programs raise any issues under the other federal securities laws,
including the Investment Company Act of 1940, and in particular Rule 3a-4
under that act. To the extent that a robo-adviser believes that its organization
and operation raise unique facts or circumstances not addressed by Rule 3a-4,
such adviser may wish to consider contacting the staff for further guidance.35
As of this date, the SEC does not appear to have found that any robo-
adviser is operating as an unregistered investment company.
7.5 EMPLOYEE RETIREMENT INCOME SECURITY ACT

OF 1974
In addition to the statutes discussed above, robo-advisers need to be mind-
ful of the Employee Retirement Income Security Act of 1974 (“ERISA”),
which imposes fiduciary obligations on “fiduciaries” of employee benefit
plans, including 401(k) plans and individual retirement accounts (IRAs)
pursuant to the Internal Revenue Code.
Robo-advisers that provide investment advice to retirement investors
generally are “fiduciaries” under ERISA. A person or firm is a “fiduciary”
to the extent it renders investment advice with respect to assets of a plan or
IRA and receives a fee or other compensation, direct or indirect, for doing
so.36
ERISA imposes a number of duties on fiduciaries, including a duty of
loyalty, a duty to act for the exclusive purposes of providing plan benefits
and defraying reasonable expenses, and a duty of care.37 In addition,
the act prohibits plan fiduciaries from engaging in certain “prohibited”
35 See Division of Investment Management (2017), Staff Guidance, 2.

36 Employee Retirement Income Security Act of 1974, Pub. L. No. 116–136 § 3(21)(A)(ii)
(2020); Internal Revenue Code of 1986, 26 U.S.C. § 4975(e)(3) (2020), as amended
provides a similar definition of “fiduciary” for purposes of Code section 4975 dealing with
individual retirement accounts (IRAs).
37 ERISA, Pub. L. No. 116–136 § 404(a). The DOL has said the duty of care is grounded
in the prudent man standard from trust law.
126 M. L. FEIN
transactions.38 Fiduciaries are personally liable for losses sustained by a plan

which result from a violation of these rules.39
In 2006, Congress created an exemption from the prohibited transac-
tion provisions of ERISA for certain “eligible investment advice arrange-
ments” meeting fee-leveling requirements or using computer models.
Robo-advisers are eligible for this exemption, although not all robo-
advisers can meet its conditions. The exemption applies to advice arrange-
ments that utilize a computer model designed and operated to:
(A) Apply generally accepted investment theories that take into account
the historic risks and returns of different asset classes over defined
periods of time, although nothing herein shall preclude a computer
model from applying generally accepted investment theories that
take into account additional considerations;
(B) Take into account investment management and other fees and
expenses attendant to the recommended investments;
(C) Appropriately weigh the factors used in estimating future returns of
investment options;
(D) Request from a participant or beneficiary and, to the extent fur-
nished, utilize information relating to age, time horizons (e.g., life
expectancy, retirement age), risk tolerance, current investments in
designated investment options, other assets or sources of income,
and investment preferences; provided, however, that nothing herein
shall preclude a computer model from requesting and taking into
account additional information that a plan or a participant or
beneficiary may provide;
(E) Utilize appropriate objective criteria to provide asset-allocation
portfolios comprised of investment options available under the
plan;
(F) Avoid investment recommendations that:
(1) Inappropriately favor investment options offered by the fidu-
ciary adviser or a person with a material affiliation or material
contractual relationship with the fiduciary adviser over other
investment options, if any, available under the plan; or
38 ERISA, Pub. L. No. 116–136 § 406.

39 ERISA, Pub. L. No. 116–136 § 409.
(2) Inappropriately favor investment options that may generate

greater income for the fiduciary adviser or a person with a
material affiliation or material contractual relationship with the
fiduciary adviser; and
(G) (1) take into account all designated investment options ... available
under the plan without giving inappropriate weight to any
investment option40
Prior to using a computer model, the fiduciary adviser must obtain a written
certification from an eligible investment expert that the computer model
meets the requirements of the exemption. Before providing investment
advice, the fiduciary adviser also must disclose to plan participants:
(A) The role of any party that has a material affiliation or material con-
tractual relationship with the fiduciary adviser in the development
of the investment advice program and in the selection of investment
options available under the plan;
(B) The past performance and historical rates of return of the desig-
nated investment options available under the plan, to the extent
that such information is not otherwise provided;
(C) All fees or other compensation that the fiduciary adviser or any
affiliate thereof is to receive (including compensation provided by
any third party) in connection with—
(1) The provision of the advice;
(2) The sale, acquisition, or holding of any security or other
property pursuant to such advice; or
(3) Any rollover or other distribution of plan assets or the invest-
ment of distributed assets in any security or other property
pursuant to such advice;
(D) Any material affiliation or material contractual relationship of the
fiduciary adviser or affiliates thereof in the security or other prop-
erty;
(E) The manner, and under what circumstances, any participant or
beneficiary information provided under the arrangement will be
used or disclosed;
40 29 C.F.R. § 2550.408g-1 (2020).

128 M. L. FEIN
(F) The types of services provided by the fiduciary adviser in connection

with the provision of investment advice by the fiduciary adviser;
(G) The adviser is acting as a fiduciary of the plan in connection with
the provision of the advice; and
(H) That a recipient of the advice may separately arrange for the
provision of advice by another adviser that could have no material
affiliation with and receive no fees or other compensation in
connection with the security or other property.41
The fiduciary adviser also must engage an independent auditor to audit the
advisory arrangement’s compliance with the Department of Labor (DOL)
rule.
It is questionable whether this exemption is available to robo-advisers
that provide only goal-based investment advice. As noted, the exemption
requires that the computer model request information from the user
on current investments and other assets or sources of income and con-
templates that the computer model will generate comprehensive advice,
something robo-advisers typically do not do. Moreover, some robo-
advisers use computer models that favor investment options offered by
the robo-adviser or its affiliates and thus would not be eligible for the
exemption.
Some robo-advisers may rely on another exemption from ERISA’s
prohibited transaction rules for “fee leveling” arrangements. Such an
arrangement is one in which any investment advice is based on “generally
accepted investment theories that take into account the historic risks and
returns of different asset classes over defined periods of time” as well as
other considerations. The advice also must take into account investment
management and other fees and expenses attendant to the recommended
investments and also “to the extent furnished” by a plan, participant, or
beneficiary:
information relating to age, time horizons (e.g., life expectancy, retirement

age), risk tolerance, current investments in designated investment options,
other assets or sources of income, and investment preferences of the partici-
pant or beneficiary.42
41 29 C.F.R. § 2550.408g-1 (2020).

42 29 C.F.R. §2550.408g-1(b)(3).
An adviser relying on the exemption is required to request such infor-

mation but the DOL’s implementing rule specifically provides that noth-
ing requires that the investment advice take into account information
requested but not furnished. The adviser may not receive from any affiliate
or other party, directly or indirectly, any fee or other compensation
(including commissions, salary, bonuses, awards, promotions, or other
things of value) that varies based on the selection of a particular investment
option – the fee “leveling” provision.
The DOL in 2016 adopted a rule interpreting the definition of the
term “fiduciary” as it applies to ERISA plans and individual retirement
accounts (IRAs).43 Under DOL’s so-called fiduciary rule, a person is
generally deemed to be a “fiduciary” if he or she provides an investment
recommendation to a plan or IRA for a fee or other compensation, direct
or indirect. However, the DOL’s rule was vacated by an appellate court in
2018 which ruled that the DOL lacked authority to adopt the rule.44
7.6 STATE REGULATION

Robo-advisers also may be subject to regulation under state securities
or other laws. Massachusetts, through its Securities Division, has been
the most proactive among the fifty states in addressing robo-advisers.
In 2016, the Massachusetts Securities Division issued a policy statement
concluding that “fully automated robo-advisers, as currently structured,
may be inherently unable to carry out the fiduciary obligations of a state-
registered investment adviser.”45
7.7 CONCLUSION
As this chapter has described, robo-advisers are regulated in the United
States under the existing regime that governs financial firms that provide
investment advice to investors. In particular, robo-advisers are treated
43 29 C.F.R. §2510.3–21 (2020); 81 Fed. Reg. 20946 (2016) (final rule).

44 Chamber of Commerce of the United States v. United States Dep’t of Labor, 885 F.3d 360
(5th Cir. 2018).
45 See Massachusetts Securities Division (2016a). See also Massachusetts guidance to state-
registered investment advisers on how to best comply with the fiduciary duties owed to
their clients when they establish sub-advisory relationships with third-party robo-advisers.
Massachusetts Securities Division (2016b).
130 M. L. FEIN
as “investment advisers” for purposes of the Investment Advisers Act of

1940 and as such generally are required to register with the Securities and
Exchange Commission. Robo-advisers also are subject to regulation by the
various states, although most states do not regulate them differently from
other investment advisers.
The Securities and Exchange Commission has identified a number of
unique regulatory issues raised by robo-advisers and has issued guidance
addressing how existing regulations apply to their activities, including with
respect to disclosures, suitability of recommendations, and compliance.
The Commission has also issued an investor alert highlighting issues that
investors should consider before using robo-advisers, including the level of
human interaction important to the investor, information the robo-adviser
uses in formulating recommendations, and the robo-adviser’s approach to
investing.
As reflected in this chapter, US regulators generally have taken the
position that, while robo-advisers may present unique issues, the existing
framework is adequate to regulate their activities and protect investors. No
new regulatory regime has been proposed to address robo-advisers in the
United States nor is any new regulatory framework anticipated.
REFERENCES
Division of Investment Management. 2017. IM Guidance Update–February 2017
No. 2017-02 – Robo-Advisers. Tech. rep. P.8. US Securities and Exchange
Commission.
Massachusetts Securities Division. 2016. Policy Statement: Robo-Advisers and State
Investment Adviser Registration. Last Accessed 1 April 2016. http://www.sec.
state.ma.us/sct/sctpdf/Policy-%20Statement%E2%80%93Robo-Advisers-and-
State-Investment-Adviser-Registration.pdf .
Massachusetts Securities Division. 2016. Policy Statement: State-Registered Invest-
ment Advisers’ Use of Third-Party Robo-Advisers. Last Accessed 1 July 2016.
Securities and Exchange Commission. 2011. Study on Investment Advisers
and Broker-Dealers. Tech. rep. https://www.sec.gov/news/studies/2011/
913studyfinal.pdf .
Securities and Exchange Commission. 2017. Investor Bulletin: Robo-Advisers.
Investor Alerts and Bulletins. Last Accessed 20 March 2020. https://www.sec.
gov/oiea/investor-alerts-bulletins/ib_roboadvisers.html.
Securities and Exchange Commission. 2019. Commission Interpretation Regarding

Standard of Conduct for Investment Advisers–Release No. IA-5248; File No. S7-
07-18. Tech. rep.
Securities and Exchange Commission and Financial Industry Regulatory Authority.
2015. Investor Alert: Automated Investment Tools. Investor Alerts and Bulletins.
Last Accessed 20 March 2020. https://www.sec.gov/oiea/investor-alerts-
bulletins/autolistingtoolshtm.html.
CHAPTER 8
Regulation of Robo-Advisory in Europe

and Germany
Christian Hammer
Contents
8.1 What Are the Activities for Which a Provider Needs a License?. . . . . . . . . . 134
8.1.1 Product-Related . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.1.2 Service-Related . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1.3 Liability Umbrella as an Alternative to Holding a License . . . . . . . 143
8.2 Regulatory Requirements for Fintech Firms and Robo-Advisors:
Practical Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
8.2.1 Information Obligations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
8.2.2 Investment Broking and Execution Only Transactions.. . . . . . . . . . 146
8.2.3 Duty of Care Obligations in Investment Advice and
Portfolio Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
8.2.4 Practical Importance and Border Cases in the
Robo-Advisory Universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
8.2.5 Reallocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
8.3 Onboarding Obligations Under GWG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.3.1 Organizational Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.3.2 General Due Diligence Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8.4 Conclusion of Contract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
8.5 Data Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
C. Hammer ()
Netfonds Financial Service GmbH, Hamburg, Germany

134 C. HAMMER
8.5.1 Profiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

8.5.2 Use of Cloud Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
8.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.1 WHAT ARE THE ACTIVITIES FOR WHICH A

PROVIDER NEEDS A LICENSE?
Hardly any other market is as regulated as the financial services market.
Only a few years ago, a broad spectrum of completely or at least largely
unsupervised products and services were available, but as a result of the
financial crisis and various investment scandals, legislators had no choice
but to tighten the regulations for products and services considerably.
Today, there are hardly any investment products available that would not
require a license or approval.
This has also had a profound impact on the behavior of fintech start-ups.
While only a few years ago they were still trying to position and develop
their companies outside government supervision, various successful start-
ups have shown that strong growth and regulatory compliance are not
mutually exclusive. Compliance with legal requirements is even mandatory
for businesses with exit plans, as hardly any buyers would pay an ambitious
price for a business model operating outside legal requirements. The
resulting civil, administrative and even criminal risks would either cause
buyers to back off or lead to a massive reduction in the purchase price. Both
are in stark contrast to what businesses with exit plans want. In view of the
increased willingness of investors to engage in disputes, compliance with
the rules and regulations is also of enormous importance to all providers of
digital products with capital market ambitions. Many so-called investor law
firms are now scouring the market for failed offers and seeking to represent
a large number of investors in schematically similar cases or potential class-
action lawsuits funded by legal expenses insurance or litigation financing
companies. Irrespective of the personal motivation of the start-ups, it is
clear that compliance with legal requirements is crucial to a successful
business model.
Depending on the business model adopted, fintech firms may also
fall under the supervision of the German Federal Financial Supervisory
REGULATION OF ROBO-ADVISORY IN EUROPE AND GERMANY 135
Authority (BaFin) in accordance with the provisions of the KWG.1 Other

providers, the so-called investment intermediaries, however, are only sub-
ject to the relevant provisions of the Trade Regulation Act. In principle,
we have a regulatory “level playing field” in Germany, that is, the same
financial services and the same risks are subject to the same supervision
and regulation. How does this work in practice? Typically, robo-advisors
operate as so-called intermediaries. Some provide their clients with advisory
and management services and some operate solely as intermediaries. Robo-
advisor is a generic term used to describe various technical systems used
by investors to invest or manage capital themselves.2 When in doubt,
companies need to check with the relevant authorities and/or professionals
to see what licenses or permits they need.
Whether a license will be required or not will depend on the investment
products offered and the activities the company carries out on behalf of its
clients.
8.1.1 Product-Related
Since 2017, all financial investment products are considered to be financial
instruments within the meaning of the German Banking Act (KWG). Only
physical direct investments, such as the purchase of gold or silver bars,
do not fall under the term “financial instrument”. The same applies to
cooperative shares. The term financial instruments thus covers nearly all
the products available to private clients on the capital market. Under the
central provision of Article 32 KWG, anyone wishing to provide advisory
or management services linked to financial instruments or to act as an
intermediary for financial instruments needs a license from BaFin. The
first sentence of Article 2 (6) No. 8 KWG creates a special exception
from this strict principle for companies, which solely provide investment
advice and brokerage services and do not engage in any other financial
services or banking transactions (see below for more details). Although
these companies meet the material requirements to be classified as financial
services institutions in accordance with Article 1 (1a) KWG, they are not
to be regarded as institutions within the meaning of the KWG and are,
1 Details of various products at Bundesanstalt für Finanzdienstleistungsaufsicht (2020).

2 For the definition of typical business models, see also Oppenheim and Lange-Hausstein
(2016).
136 C. HAMMER
therefore, not subject to a license requirement under Article 32 KWG.

The activities of these companies must be limited to the following financial
instruments:
• Shares in investment funds issued by a domestic investment manage-

ment company,
• EU investment assets and foreign AIFs that may be sold under the
German Capital Investment Act (KAGB),
• Investments within the meaning of Article 1 (2) of the German
Investment Products Act (VermAnlG).
If a company adheres to this, it will not require a license from BaFin

in accordance with Article 32 (1) KWG. A business license as a financial
intermediary under Article 34f (1) of the German Trade Regulation Act
(GewO) is then sufficient, an exception that is unique in Europe.
Cross-border provision of services or the establishment of a branch
in another EU member state or in a third country are not covered by
this license under the Trade Regulation Act; this can only be done by
companies licensed under Article 32 KWG. In that regard, investment
intermediaries are regionally restricted to the territory of the Federal
Republic of Germany.
What is hidden behind the three classes of financial instruments men-
tioned above? Generally speaking, one could say that investment interme-
diaries are “fund brokers”, where the term funds refers to so-called UCITS
funds and alternative investment funds such as open-end real estate funds,
but also closed-end products under the KAGB such as ship funds or closed-
end real estate funds. This means that most ETFs licensed under Article 34f
GewO are also a permissible subject of services. However, a more detailed
analysis is needed to provide a concrete response for individual ETFs.
The umbrella term ETFs also covers ETCs (exchange traded commodi-
ties) and ETNs (exchange traded notes). The legal structure is decisive
for the delimitation. Investment intermediaries under Article 34f GewO
may only provide advice or brokerage services for “real” ETFs within the
meaning of the KAGB, which are issued as UCIT funds. Whether or not
they are swap-based does not affect this. In contrast, ETCs and ETNs are
merely securitized claims, i.e. bearer bonds. This is why a KWG license
is required to broker these products; a 34f license will not be sufficient.
To check whether this is a real ETF, we recommend reviewing the sales
prospectus. Under KAGB, products requiring a prospectus are referred

to as “real” ETFs and can be brokered by investment intermediaries.
A license is needed for products requiring a prospectus in accordance
with the German Securities Prospectus Act (WpPG). Other financial
instruments permitted under the license under Article 34f GewO are so-
called financial assets as defined under Article 1 (2) German Investment Act
(VermAnlG). These include investments with low or no fungibility such as
not-securitized shares that grant participation in the results of a company,
for example, trust assets, participatory loans, subordinated loans, profit
participation certificates, silent participations, registered bonds and other
investments that grant or hold the prospect of interest and repayment or
cash settlement assets in exchange for the temporary provision of money.
These are generally less interesting for fintech firms as they can be very
complex, costly, illiquid and not easily tradeable.
In view of the high level of complexity and the resulting need to assist
private investors with information, only a few fintech firms have dared to
approach this product group to date. Digital asset managers are generally
not interested in such non-fungible forms of investment, but investment
advisors appear to be caught in the point of tension between the need
to clarify, on the one hand, and a swift conclusion on the other. Only in
the broking business are there individual providers offering cost-effective
subscription options for investments.
8.1.2 Service-Related
In addition to the products, the execution of buy/sell orders and the
discretionary latitude applied is also intriguing. The investment-related
activities of robo-advisors typically fall under investment advice, investment
and contract broking and/or portfolio management. These activities have
to be kept separate from the others, as they require different licenses and
compliance with different requirements.
While the license under Article 34f GewO is sufficient to offer brokerage
services and investment advice, it is necessary to obtain the license under
Article 32 KWG from BaFin to provide contract broking and portfolio-
management services.
138 C. HAMMER
8.1.2.1 Investment Advice

Under the law, investment advice is defined as providing clients or their
representatives with personal recommendations in respect of transactions
relating to certain financial instruments where the recommendation is
based on an evaluation of the investor’s personal circumstances or is
presented as being suitable for the investor and is not provided exclusively
via information distribution channels or for the general public.3
Behind this lengthy legal definition there are highly relevant questions
for robo-providers. The focus here is primarily on the question of when the
product is directed at clients, what qualifies as a recommendation and when
the necessary check of personal circumstances or suitability is considered
to be suitable for the investor.
Clients are not only persons who actually enter into a business rela-
tionship with the provider but also those who are merely interested in
the product.4 For protection purposes, categorization as a client therefore
begins at an early stage. It is sufficient to use any targeted approach aimed
at concluding a contract at a later date, that is, any advertising measure.5
Every visitor to a homepage is therefore considered to be a prospect and
therefore a “client”.
A “recommendation” is when the client is advised that a particular action
is in his interest.6 A recommendation is a recommendation made to a client
to buy, sell or not to buy a financial instrument.7 It does not matter whether
the client subsequently acts upon that recommendation or not.
A clear distinction must be made between recommendation and infor-
mation where the company merely provides its clients with information
about their assets invested in financial instruments without making any
3 Gesetz über das Kreditwesen, BGBl. I page 881

4 This is clear from the legal definition under which clients are any natural or legal
persons for whom investment firms provide or arrange investment services or ancillary
investment services. The MiFID Directive 2004/39/EC already extended the information
and exploration obligations provided for under Article 19 (2) to “prospective clients”.
5 This has already been set out in the predecessor provision with identical wording, Gesetz
über den Wertpapierhandel, BGBl. I § 543.31(a)(2020), Schwark and Zimmer (2010),
Wertpapierhandel, § WpHG, paragraph 5
6 Cf. Joint information sheet issued by BaFin and the Deutsche Bundesbank on investment
advice, version 2/2019.
7 European Parliament and of the Council, 2006/73/EC § 52 (2006) implementing
Directive 2004/39/EC.
concrete proposals to change the composition of those assets.8 Practical

difficulties usually arise for robo-providers in the context of their online
offerings, for example, when certain financial instruments are presented
and their advantages or opportunities explained or when certain financial
instruments are advertised. The client’s point of view is always decisive
here. If this person has the impression that this goes beyond the mere pro-
vision of information and promotion of the product by suggesting to buy
or sell the product, this will typically be construed as a recommendation
within the meaning of the legal definition.
This recommendation must relate to “transactions in specific financial
instruments”, that is, not just to a particular asset class.9 According to
BaFin, the typical robo-advice fulfills this criterion if the client receives
investment proposals relating to certain financial instruments, such as
certain funds or securities and not just asset classes.10 Where a product
only generates an asset allocation (e.g. European equities 30%, US equities
20%, emerging markets equities 15%, bonds 20%, cash 15%), this does
not qualify as investment advice, as no recommendations are made on
specific financial instruments. However, if the client also receives concrete
recommendations on financial instruments and asset classes, then this
satisfies the investment advice requirements. This also applies if different
product alternatives are recommended, and it is up to the client to make
the selection. In our view, the vast majority of robo-advisors work with
concrete proposals for specific financial instruments.
As a rule, the decisive factor tilting the scales for or against investment
advice is whether the recommendation given reflects the “personal circum-
stances of the investor” or it is presented as “suitable for the investor”.
Few robo-providers would operate today without querying the personal
criteria of the prospective client online and creating an investment proposal
tailored to the client’s communicated needs. In most cases, providers not
only ask for the investment amount and the investment horizon but above
all the risk tolerance of the prospective clients. Information on the financial
9 Entwurf eines Gesetzes zur Umsetzung der Richtlinie über Märkte für Finanzin-
strumente und der Durchführungsrichtlinie der Kommission (Finanzmarkt-Richtlinie-
Umsetzungsgesetz), BT-Drs. 16/4028 § 12 (2007).
140 C. HAMMER
circumstances of the prospective clients is also collected. This means that

these providers undoubtedly fulfill the criterion of “taking into account
personal circumstances” in their recommendations, even if the investment
proposal is generated by an algorithm.
The joint information sheet of BaFin and Deutsche Bundesbank on
investment advice thus focuses on the “financial situation” of the investor.
It is, therefore, sufficient for the provider to obtain only relatively general
information, to the extent that it then recommends concrete financial
instruments on this basis.
In addition, it is also sufficient to be perceived as having taken into
account the personal circumstances of the investor when making the
recommendation. This will normally be the case if the client believes that
the investment proposal reflects the information provided, irrespective of
whether this applies in full or in part.11
This can be counteracted by using clear and easily understandable
disclaimers, which have to be accepted online.12 On this basis, the client
must accept that no “individual” or “tailored” investment proposals will
be provided. Investment advice can be provided with a license in accor-
dance with Article 32 KWG or as an investment intermediary. The main
differences are product-related (see above) but can also be found in some
details relating to their duty of care.
8.1.2.2 Investment Broking

The definition of investment broking is satisfied when companies act as
intermediaries and their business involves the buying and selling of financial
instruments.13 The broker acts as an intermediary that passes on the
investor’s declaration of intent to buy or sell financial instruments to the
person with whom the investor wishes to conclude such a transaction.14 In
this case, it is irrelevant whether the forwarded declaration of intent is an
11 Cf. Tertilt and Scholz (2018). It has been found that robo-advisors often use compara-
tively little information from onboarding to make a portfolio recommendation.
12 CESR’s Technical Advice on Possible Implementing Measures of the Directive
2004/39/EC on Markets in Financial Instruments, CESR 05-024c, 9 (2005).
13 See BaFin’s Guidance Notice on the subject of investment broking at Bundesanstalt für
Finanzdienstleistungsaufsicht (2017b).
14 This administrative practice has also been confirmed by the Federal Court of Justice
(BGH). Bundesgerichtshof, WM 2014, 221/222, III ZR 73/12.
offer by the investor to the seller or the acceptance of an offer by the seller.
The form of transmission also does not matter.
Accordingly, the transmission can also take place electronically, for
example through a computer system. In accordance with the administrative
practice of the supervisory authorities, investment intermediaries are those
who provide an IT system which is used to pass on an investor’s declarations
of intent to buy or sell financial instruments to any contracting parties. If,
as part of the robo-service, the order is placed through the IT system of
the provider, the latter will provide the investment-broking service.
In addition, investment brokerage also involves targeted encouragement
of the investor to get them to conclude a transaction with a third party
to buy or sell financial instruments. Accordingly, investment broking is
also provided by those who consciously and finally influence an investor to
conclude a transaction on the acquisition or sale of financial instruments.15
Anyone who merely brokers contact between the investor and a seller of
financial instruments is not involved in investment broking if he or she
makes said connection in the form of a mere referral service within the
meaning of the German Trade Regulation Act.16 In practice, it is difficult
to distinguish between investment broking requiring a license and referral
broking not requiring a license. The decisive criterion here is whether the
client’s will to conclude a transaction is influenced in any way. On the part
of the authorities, there is certainly the tendency to interpret the notion
as applying to intermediary activities requiring a license at the expense
of activities that do not require a license. This is because (irrespective
of whether this is considered to be investment advice or not) the well-
known robo-providers typically propose very specific transactions; it might
be reasonably expected that the brokerage activity requires a license, even
if the provider’s IT system is not used for the specific orders.
According to the settled case-law, the broking of asset-management
contracts is not considered to be investment broking.17 Asset management
is aimed at buying and selling financial instruments for the account of
the investor on the basis of discretionary investment decisions. Therefore,
both the forwarding of a declaration of intent aimed at the conclusion
of an asset-management agreement and the influencing of an investor
15 See Bundesanstalt für Finanzdienstleistungsaufsicht (2017b).

16 Schaefer, in Boos et al. (2016); Gesetz über das Kreditwesen, § 1 KWG, paragraph 136.
17 European Court of Justice, Z.I.P. 1362 (2017).
142 C. HAMMER
to conclude an asset-management agreement are not considered to be

investment.- broking activities. Similar to investment advice, investment-
broking services can also be provided both under the license pursuant to
Article 32 KWG and Article 34f GewO.
8.1.2.3 Contract Broking

In contrast to investment broking, contract broking requires a license
from Bafin in accordance with Article 32 KWG. Contract broking refers
to buying and selling financial instruments in the name of and for the
account of a third party. A contract broker also acts as an intermediary.
Whereas the investment broker passes on the client’s declaration of intent
to the buyer or seller of the financial instruments in the capacity of an
intermediary, the contract broker submits their own declaration of intent
as their client’s representative. These services are thus mutually exclusive.
Investment broking is more common in practice.
8.1.2.4 Portfolio Management

Under Article 1 (1a) sentence 2 No. 3 KWG, portfolio management is
defined as the management of individual portfolios of financial instruments
for others on a discretionary basis. The distinguishing feature of portfolio
management is that the individual transactions are not based on client
orders. These are initiated by the providers (managers) at their discretion.
The implementation of investment decisions is the responsibility of the
manager rather than the client. In order to manage the securities account,
the client grants the manager so-called discretionary authority. This discre-
tionary authority gives the manager access to the client’s securities account
at the custodian bank.
In the meantime, some market participants, which offer their own
asset-management services, are also involved in the market as fintech
firms or robo-advisors. There are considerable differences between the
products and services on offer. There are conventional providers who offer
digital client onboarding, while at the same time they use algorithms to
select and propose an investment strategy. In contrast to robo-advisors
that offer investment advice, the client does not only receive a one-
off (or continuous) investment recommendation. Instead, the investment
portfolio is managed on an ongoing basis in accordance with investment
guidelines. The transactions are also frequently based on an algorithm.
Crucially, the provider does not only make investment decisions purely
mathematically based on clear rules agreed in advance with the client, but
also has discretionary authority (within the agreed investment guidelines).
Insofar as the investment decisions are purely mathematical and are thus
strictly based on calculations, no discretion is involved. Depending on their
structure, such providers will routinely provide contract-broking services.
They will then also need a license under Article 32 KWG, as contract
broking and investment broking are mutually exclusive.
8.1.3 Liability Umbrella as an Alternative to Holding a License

An alternative to obtaining one’s own license pursuant to Article 32 KWG
or Article 34f GewO is to act as so-called tied agent pursuant to Article
2 (10) KWG. Tied agents provide financial services exclusively for the
account, on behalf of and under the liability of a financial institution that
is in possession of a BaFin license. They use this institution’s license under
Article 32 KWG and are thus not subject to the product-specific restrictions
of investment intermediaries. They are permitted to provide investment
advice and investment-broking services, but not contract-broking and/or
portfolio-management services.
The main prerequisite for acting as a tied agent is the assumption of
liability by a financial institution with its own investment broking and
investment advice license. Liability is assumed under a civil law agreement
between the broker and the liable company. In addition, the broker has
to carry out an activity in the name of and for the account of the liable
company. This is only the case if the broker acts as a representative or
agent of the liable company.18 The broker concludes a transaction on
behalf of the liable company as its representative, which collects the fee.
The companies then agree to share any income in a contractually defined
ratio. To act as tied agents, companies must notify BaFin and have to
register in BaFin’s register of tied agents. By operating under a liability
umbrella, tied agents can focus on their own business model, while the
liable financial institution is responsible for compliance with regulatory and
organizational requirements and its ongoing monitoring.19 This model is
certainly appealing for start-ups. In the meantime, a number of fintech
18 Schaefer in Boos et al. (2016), Article 2, paragraph 128.

19 This obligation is laid down in Article 25e KWG. Gesetz über das Kreditwesen, § 25e
KWG.
144 C. HAMMER
Table 8.1 At a glance: which asset classes could be advised and brokered
Product-related 34f GewO Portfolio manager Bank Liability umbrella

Advice, broking of
Equities No Yes Yes Yes

Bonds No Yes Yes Yes
Loans No Yes Yes Yes
Participation certificates Yes Yes Yes Yes
Investment funds (UCITS) Yes Yes Yes Yes
Open-end alternative Yes Yes Yes Yes
investment funds
Closed-end alternative Yes Yes Yes Yes
investment funds
Certificates No Yes Yes Yes
Table 8.2 At a glance: which services are related with which license
Service-related 34f GewO Portfolio manager Bank Liability umbrella
Investment advice Yes Yes Yes Yes

Investment broking Yes Yes Yes Yes
Contract broking No Yes Yes Yes
Portfolio management No Yes Yes No, only advice
Broking abroad No Yes Yes Yes
Advice abroad No Yes Yes Yes
Referral broking (tipsters) Yes Yes Yes Yes
firms have opted for this solution and operate their sales platforms under
the responsibility of experienced providers. At a glance see Tables 8.1 and
8.2.
8.2 REGULATORY REQUIREMENTS FOR FINTECH

FIRMS AND ROBO-ADVISORS: PRACTICAL ISSUES
Despite political announcements, providers of digital financial services
are offered little relief or special legal regulations that deal with or even
promote the specifics of their business models. Where fintech firms offer
investment services, they must, like analogous providers, comply with
the related duty of care obligations (developed for the latter) under the
German Securities Trading Act (WpHG) or the Investment Brokerage
Regulation (FinVermV). When it comes to investment products, fin-

tech firms must assess the appropriateness or suitability of its potential
clients. The appropriateness assessment determines whether the client is
in a position to evaluate a particular investment product; the suitability
assessment also determines whether a specific product is suitable for this
particular client and financial situation. The information and transparency
requirements are extensive. Products must be tested and evaluated as part
of so-called product governance.
This presents unexpected challenges to some IT-savvy start-ups count-
ing on a quick and easy onboarding process for its clients. Few enthusiastic
entrepreneurs anticipate the regulatory burden in the area of capital
markets.
The decisive factor here is the type of services provided (cf. above
definition). Here are some frequently problematic aspects:
8.2.1 Information Obligations

In investment and contract broking as well as in investment advice, the
provider must provide the client with the necessary information to make
an informed decision about the investment. This obligation exists both
under supervisory and civil law. Article 63 WpHG establishes an obligation
to standardize basic information. Accordingly, abstract information must
be provided relating to the nature of the financial instruments and the
context in which they are provided; this information should form the
basis for an informed investment decision by the client. According to
the wording of the WpHG, the information may also be transmitted in
a standardized form. To meet the regulatory requirements, the use of
brochures, for example, investment funds basics, has become established
in practice, allowing providers to offer cross-product information.
However, the use of standardized information does not meet the
extensive regulatory information obligations of investment firms in full,
in particular, not for every financial product. By handing over the written
standard information material, the financial service provider merely satisfies
its regulatory obligation to provide information. In addition, providers
must give the client a true and fair view (including from civil law per-
spective) of the opportunities and risks involved in the transaction. With
respect to portfolio management, this obligation is limited to information
146 C. HAMMER
relating to the relevant asset class,20 whereas in the case of all other
services, the obligation also extends to information on the risks of the
specific investment object to be purchased. This information must be
made available to the client in good time so that he can take note of it
before making the investment decision;21 subsequent transmission after
the investment decision has been made is thus contrary to regulatory law
and also gives rise to considerable potential liability under civil law.
It is therefore technically feasible to ensure that all relevant information
is available to the client in good time to allow him to take note of it before
making the investment decision. In the case of UCITS, this correlates with
the obligation to provide clients with an information sheet (so-called key
investor information document; KIID). In the context of investment advice
to private clients, the relevant obligation applies to all types of financial
instruments.
It is, therefore, mandatory for firms to provide clients with all the
necessary information before they make their final investment decision. It
is highly recommended to give clients the opportunity to download all
the relevant information. Alternatively, the information can also be sent
by email. For liability purposes, it should, in any case, be verifiable that all
relevant information reached the client in good time to allow the client to
become apprised of it before making the investment decision.
In addition to these risk-related information obligations, there are a
number of other, mostly regulatory information obligations that need to be
considered. These include conflicts of interest policy, best execution policy
and the general information obligations for financial services companies.
8.2.2 Investment Broking and Execution Only Transactions

Investment and contract brokers are required to determine whether the
client is in a position to understand the risks arising from the investment
decision based on the client’s knowledge and/or experience, based on a so-
called appropriateness assessment. To this end, they ask prospective clients
20 Schaefer, in Assmann et al. (2015), Article 23, paragraph 28; the disclosure obligations
refer to share classes, markets and currencies.
21 The question as to when a delivery is “made on time” shall be assessed on a case-by-
case basis; Bundesgerichtshof, WM 2007, 1608; III ZR 145/06; the time interval between
handover and subscription is crucial, as is the client’s background in terms of knowledge and
experience.
about their knowledge and experience with various product classes. Based
on this information, providers must then check whether the investment
decision is consistent with the prior knowledge and experience of investors
and whether, on the basis of their knowledge, investors can accurately assess
the risks associated with the investment (so called appropriateness test).
The investor is informed of the result of the assessment and can then
decide whether or not to continue with the investment—even if the result
of the assessment was negative. This process can be easily designed based on
standardized data; an algorithm then checks the knowledge and experience.
Even if a client has no experience with certain financial instruments,
a transaction may still be appropriate for the client if the information
was provided to the client sufficiently in advance. This certainly places
higher demands on the creativity of fintech firms, as their processes are
usually geared towards a fast and smooth process. Various approaches are
conceivable—for example, confirmation of the client despite the lack of
appropriateness to conclude transactions, but also the termination of the
onboarding process and its continuation after a certain period of time.
In certain circumstances, the appropriateness assessment may even be
waived altogether. In contrast to portfolio management and investment
advice, regulatory law permits a reduction in the duty of care obligations
for so-called execution-only transactions within the scope of investment
broking. If the conditions for the execution-only transaction are met,
the financial service provider’s activity is limited to the mere execution
of the transaction requested by the client and is not associated with
any exploration, appropriateness or suitability assessment. However, the
information obligations set out above must also be fulfilled in this case. The
scope of the execution-only transaction is limited to cases in which all the
following conditions are met: The transaction must involve “non-complex
financial instruments”, it must be initiated by the client who will then be
informed by an explicit statement that no appropriateness assessment will
be carried out.
Complex financial instruments include, for example, subordinated debt
instruments and debt instruments that qualify as PRIIP (packaged retail
investment and insurance-based products) within the meaning of Article
4 No. 1 Regulation (EU) No. 1286/2014 (PRIIP Regulation), as well as
derivatives. Non-complex financial instruments therefore include UCITS
and traditional equities and bonds traded on regulated markets. The
exception for pure execution transactions can therefore be an interesting
148 C. HAMMER
approach for pure broking platforms, allowing the providers to offer

straightforward processes.
Here too, a clever and pragmatic solution will only be successful if
the company has focused on the legal requirements, and the product
selection derived from them. As already mentioned, the clear focus on the
product side is essential. Where an undertaking is required to provide an
appropriateness assessment, it must define the knowledge and experience it
requires so that it might reasonably expect that the client understands the
risks involved in making an investment decision. However, thanks to clearly
structured onboarding processes, this does not present a major hurdle.
8.2.3 Duty of Care Obligations in Investment Advice and Portfolio

Management
Providers are required to provide clients with investment advice minutes
signed by the advisor (replaced now by suitability reports) after they
have received investment advice and before they enter into a securities
transaction. While transmission on a durable medium (e.g. by email or
Postbox) is also possible, it requires the explicit consent of the client, which
must be obtained in advance. In the context of automated investment
advice, there is no personal advisor available to sign such minutes. Since
2018, the minutes must also explain why the recommendations made are
suitable for the client. Investment advisors may only recommend to clients
financial instruments and investment services that are suitable for them.
This means that any recommendation must be suitable for the client. At the
same time, clients must be able to understand the functionality and risks of
the recommendation. For this purpose, the advisor must inquire about the
client’s knowledge and experience, financial circumstances and investment
objectives and review his investment recommendation on the basis of the
information provided. Any gaps in knowledge must be conveyed to ensure
that the client understands how this works, and the opportunities and risks
involved.
The assessment is summarized and explained in writing to the client in
the suitability report. It describes how the recommendation was adapted
to reflect the information provided by the client. For this purpose, all the
information required for the suitability assessment must be incorporated in
the suitability report. The expectation is that not only text modules will be
used for this purpose.22 A certain degree of individualization is therefore

required. This does not quite fit in with the typical service concept of robo-
advisors. In this context, however, BaFin points out that robo-advisors are
not subject to any exceptions from applicable laws.23
The market does not seem to have found any answers to this question
that would allow the recording of investment advice minutes. As ESMA
has highlighted, the suitability report should be created individually and
not be merely put together from readymade text modules. Some providers
have therefore limited themselves to brokering asset-management services
for which no such report is required, others withdrew to the field of pure
broking and product information not involving personal recommendation.
It is still unclear whether a solution that has been technically tailored to the
respective client by the platform would be viable.
Investment advice and portfolio management are subject to far-reaching
exploratory obligations. The exploration in terms of knowledge and expe-
rience, financial circumstances, investment goals and risk tolerance can be
easily automated. However, compared to personal investment advice, a
schematic entry form cannot determine whether further clarifying infor-
mation is required from the client.
8.2.4 Practical Importance and Border Cases in the Robo-Advisory

Universe
In the above text, we raise a number of practically relevant and sometimes
difficult legal issues concerning robo-advisors. Next we will discuss some
other practical issues providers have to deal with:
8.2.4.1 Onboarding Using Online Communication Channels

Client onboarding is subject to far-reaching legal requirements. Above all,
it is crucial for providers to clearly decide which type of investment service
22 Cf. also BaFin expert article, Suitability report: An important document for consumers,
available at https://www.bafin.de/SharedDocs/Veroeffentlichungen/DE/Fachartikel/
2018/fa_bj_1809_Geeignetheitserklaerung.html.
23 BaFin, Robo-advice and auto-trading – platforms for automated investment advice
and automatic trading, BaFin demands this under the so-called Minimum Compliance
Requirements (MaComp). Any inconsistencies and inaccuracies in the information collected
from clients should be identified and pointed out to the investor in order to resolve them 24.
It is, therefore, insufficient to collect only standardized information.
150 C. HAMMER
they wish to offer. This is a typical conflict of interest. While prospective

clients typically look for answers and solutions that meet their individual
needs as far as possible, digital providers usually want IT solutions that
make registration and closing quick and easy. This is the only way they
can easily engage their prospective clients in a business relationship and
keep the acquisition costs low. The client will be interested in receiving
investment advice in the form of investment proposals tailored to his needs.
On the other hand, providers often shy away from the increased expense
and regulatory complexity of investment advice as well as the liability risks
that may arise from unsuitable investment proposals. In particular, the
obligation to hand over suitability reports (formerly advice minutes) is
difficult to integrate.
Common solutions in this respect are mere recommendations of asset
allocation and the submission of recommendations only after the start of
a business relationship, at least in the form of the transmission of concrete
data by the prospective client, but also the use of detailed disclaimers that
make it clear to the client that the presented investments do not cover all
relevant aspects of investment advice, and therefore, no such advice is being
provided. None of the two solutions is perfect, as they both frustrate the
client’s desire for concrete solutions.
8.2.4.2 Hotline and Chat

This also addresses a core problem of digital providers. Many clients attach
importance to human contact before they put their savings into the hands
of a company. This, too, is at odds with the approach of digital providers,
which tend to offer prospective clients support in the form of hotlines or
chats. This in turn raises specific delineation issues and makes compliance
with legal requirements more challenging. Where online communications
can be clearly structured to answer the question of whether investment
advice will be provided or not, this becomes much more difficult in the
context of free communication. As explained, the determining criterion for
or against investment advice is the consideration of the investor’s personal
circumstances.
An “assessment of the investor’s personal circumstances” is answered in
the affirmative if the client merely informs the relevant provider in general
terms of his financial situation and the service provider then recommends
transactions involving specific financial instruments. Alternatively, it is also
sufficient that the recommendation is merely “presented as being suitable
for the investor”.24 This is the case if a client can reasonably expect that
the recommendation given is based on the consideration of the client’s
personal circumstances. The communication of personal circumstances in
the course of a chat or a telephone call and the investment recommendation
made online can thus overall “constitute” investment advice. The staff
employed in this area of client services must therefore receive training
to avoid unwanted legal consequences. This also applies to statements on
products or asset classes. If a financial service provider provides its clients
with incomplete or inaccurate information, the client will be entitled to
claim damages and reverse the transaction. The robo-advisor, therefore,
bears the investment risk. To do this, providers need to select and train
suitable personnel. Nevertheless, the risks from this step into the analogue
world cannot be completely eliminated.
8.2.4.3 Order Placement

The question of how a transaction is triggered again touches on the initial
question of the services the company actually offers. If the order is placed
outside the robo-advisor system, and the robo-advisor does not have any
impact on the client’s investment decision, the actual order placement is
a matter between the client and the custodian. In practice, however, this
will be the absolute exception, as robo-advisors usually also offer order
placement. They then act as investment or contract brokers. The relevant
delimitation has already been explained above.
An important question from the practical standpoint is when so-called
social, copy or mirror trading models will be regarded as broking and/or
management activities. The process of this form of order initiation typically
runs as follows: Clients choose the trading strategy they want to follow.
To copy and execute the individual trading decisions for their portfolio,
they subscribe to information about changes in the respective portfolio
composition. The provider then either forwards the trading decision as
a client order to the client’s custodian bank (Model 1) or transmits the
concrete change to the client, who then orders the transaction (usually
automatically through software) (Model 2).
In Model 1, the investor’s assets are managed in accordance with fintech
decisions. The investor grants the company the appropriate discretionary
24 See Bundesanstalt für Finanzdienstleistungsaufsicht (2016). The custodian then exe-

cutes the submitted order and records it in the securities account.
152 C. HAMMER
authority/powers. This also applies if fintech is backed by another provider

who gives trading signals to them. The activity of this signal generator is
generally attributed to the provider who places the specific orders. As a
result, this platform routinely offers portfolio-management services.
Model 2 is different. Depending on the technical configuration, the
client puts an order through his computer, which is then sent to the
custodian for execution via the robo-advisor’s IT system. By forwarding
the order to the custodian, the provider typically provides the investment
broking service.
If the provider “translates” the investment proposal into a client order
without the client’s own declaration of intent and forwards it to the
custodian, the provider typically acts as the contract broker “in the name
of” the client.
In this respect, the question arises as to whether the automatically
generated declarations of intent are to be understood as the client’s own
declarations or as declarations made by the robo-advisor. Civil case law
also recognizes automatically generated declarations as effective, non-
actionable declarations of intent by the software owner, although sub-
mitted by “auto-reply” if the computer used only executes commands
previously determined by human programming and the declaration, there-
fore, originates in an act initiated by the declaring party and based on its
will .25 The decisive factor is that the client can fundamentally influence
whether an order is generated or not. To what extent these principles are
also recognized by BaFin is uncertain. Ultimately, clarification here can
only be achieved in individual cases, taking into account the respective
product and its technical details.
8.2.5 Reallocation
The same delimitation questions arise in reallocation, that is, when pre-
viously valid weightings for asset classes or weightings due to market
movements result in the initial weighting being restored at certain points
in time.
In contrast to an initial event, asset management can often be excluded
here by agreeing to fixed rules for reallocation and defining the timing and
25 Bundesgerichtshofes in Zivilsachen, BGHZ 195, 126, X ZR 37/12; Oberlandesgericht

Düsseldorf, NJW-RR 1073 (2016).
conditions of the adjustment so clearly that the provider does not have the
conceptual discretion required for portfolio management. But then there
is still the difficult question regarding investment advice or broking. The
answer to this again lies in the individual structure based on the criteria
presented.
8.3 ONBOARDING OBLIGATIONS UNDER GWG

A key issue that has always played an important role for fintech firms is
the client onboarding process, which is far more complex in the financial
industry than in other sectors. We have already pointed out some regula-
tory aspects.
Financial service providers—irrespective of the type of license they
operate under—are subject to the provisions of the GWG and, in particular,
the obligations under Article 10 et seq. GWG (referred to as the know-
your-client or KYC obligation). They are therefore obliged not only to
collect the personal details of their clients but also to verify their identity,
either at the post office counter or, as of recently, by video. This applies
to the initiation of business transactions in general and to the opening
of accounts and securities accounts in particular. This represents another
significant hurdle and disruption to the onboarding process. The GWG
also codifies organizational obligations. Fintech firms should therefore be
careful when picking banking partners to ensure that they are open to
working together on solutions that make onboarding as easy and efficient
as possible, taking into account all legal and regulatory requirements.
8.3.1 Organizational Requirements

To prevent money laundering and terrorist financing, all obliged entities
must have effective risk management systems in place, that is, perform and
document risk analysis for their company and implement adequate internal
safeguards. Article 6 GwG provides an overview of suitable safeguards. In
accordance with Article 8 GwG, companies have a recording and retention
obligation. The obliged entities are thus required to make complete copies
of the identification documents or to record them digitally. For financial
service providers licensed in accordance with the German Banking Act
(KWG), it is also necessary to appoint a money laundering reporting officer
(Article 7 GwG).
154 C. HAMMER
8.3.2 General Due Diligence Requirements

The general due diligence requirements (Article 10 GwG) include in
particular:
• identifying the contracting party and, where applicable, the person

acting on their behalf
• clarifying whether the contracting party is acting on behalf of a
beneficial owner (and, if so, identifying the beneficial owner)
• obtaining and evaluating information on the purpose and intended
nature of the business relationship
• establishing whether a politically exposed person (PEP) is involved
(Article 1 (12) GwG contains a list)
• continuous monitoring of the business relationship, including trans-
actions
With regard to identification, it should be noted that this must be done

before the business relationship is established or the transaction is executed.
This therefore has a direct impact on the onboarding process. In the case of
natural persons, obliged entities are required to collect their name, place of
birth, date of birth, nationality and address based on identity documents.
To establish the identity of legal persons or partnerships, obliged entities
are required to collect information about the company (i.e. name or trading
name), the legal form, commercial register number, where applicable, the
address of the registered office or principal place of business and the names
of the members of its representative bodies or its legal representatives.
Article 12 GwG specifies the documents to be used for identity verification
purposes. The identity of natural persons can be verified on the basis of
conventional ID documents. According to this, providers are required to
verify the identity by examining the documents presented “physically” or
using another procedure suitable for verifying identity in accordance with
anti-money laundering laws and regulations. This also includes the video
identification process and the Postident process. While the latter has long
been recognized, BaFin has only recently specified the requirements for
video identification. The BaFin Circular 3/2017, which is aimed at all
financial institutions and service providers under its supervision, came into
effect on 15 June 2017.26
Under the conditions specified there, the video identification procedure
can now be used by all obliged entities under the Anti-Money Laundering
Act. For example, the employees must receive special training and the
person to be identified must give their explicit consent to the entire
identification process being recorded at the beginning of the video identi-
fication. In addition, the image and sound quality of the communication
must be sufficiently high to ensure unambiguous identification. The entire
identification process shall be recorded and stored in an audiovisual format.
Only end-to-end encrypted video chats are permitted for the necessary
audiovisual communication. Services such as Skype are, therefore, not
suitable for identification purposes. Against the background of these
requirements, it is now possible to outsource video identification to several
providers. At the same time, the requirements laid down in Article 17 GwG
must be observed.
8.4 CONCLUSION OF CONTRACT

The conclusion of contracts online also requires closer examination.
Under MiFID2, a written framework agreement must be concluded
with the client on paper or another durable medium, which sets out the
essential rights and obligations of the investment firm and the client.
Investment firms providing investment advice are only subject to this
obligation where a periodic assessment of the suitability of the financial
instruments or services recommended is performed. Accordingly, mere
intermediaries are not subject to these obligations. However, portfolio
managers and robo-advisors who have an ongoing relationship with their
clients must comply with this requirement. As it is not typically possible to
hand out a hard copy, a copy of the framework agreement must be sent to
the client, for example, in pdf form.
The timing of the conclusion of the relevant contract also poses prob-
lems in this respect. As with any contract, digital contracts are also
concluded through offer and acceptance. While clients routinely submit
their offers online within the scope of robo activities, it is often forgotten
that for the contract to come into effect, the offer needs to be accepted by
26 See Bundesanstalt für Finanzdienstleistungsaufsicht (2017a).

156 C. HAMMER
the provider. This should be integrated into the onboarding process—for

example, by using email.
It should also be borne in mind that digital business transactions are
routinely concluded remotely, which entails special obligations for the
company. Distance contracts are contracts where the company or a person
acting on the company’s behalf and the consumer exclusively use forms of
distance communications to negotiate and conclude the contract. It is clear
that this also applies to digital transactions. On the one hand, this means
that a whole range of information must be made available to the consumer
(cf. Article 246b of the Introductory Act to the Civil Code (EGBGB)). In
the case of distance contracts, however, the business is obliged to provide
the consumer with a confirmation of the contract, on a durable medium, in
which the content of the contract is set out, and to do so within a reasonable
period of time after having concluded the contract, but at the latest before
the performance of the service. Depending on what has been said above
about regulatory law, this gives rise to an obligation to provide the client
with the agreement.
Finally, the consumer’s right to cancel must be taken into account in
distance contracts. If the conditions of a distance contract are satisfied,
the consumer will have the right to cancel the distance contract. The
cancellation period is two weeks. However, in order for the two-week
period to start to run properly, its start is subject to a number of conditions,
which are difficult to meet in practice. This includes the special information
concerning the right of cancellation and the provision of information in
accordance with Article 246b EGBGB. The business must provide the
consumer with the information in written or electronic form at the latest by
the point at which the obligations under the contract have been met in full.
If this does not happen, the cancellation period does not begin to run—
resulting in an infinite cancellation right within the scope of the financial
services.
8.5 DATA MANAGEMENT

Robo-advisors collect and process a variety of sensitive client data. In this
respect, they are subject to the strict requirements set out in the Gen-
eral Data Protection Regulation (GDPR) and the German Federal Data
Protection Act (BDSG). In addition to various organizational require-
ments, robo-advisors have to take into account, in particular, the strict
requirements for the lawfulness of data processing and the far-reaching

information obligations that must be observed during onboarding.
However, there are two aspects which need to be highlighted in relation
to fintech firms and robo-advisors.
8.5.1 Profiling
GDPR specifically focuses on so-called profiling. In a nutshell, profiling
means the automated creation of profiles. In connection with web-based
systems, this mainly involves the creation of user profiles on the basis of
automatically collected personal data, taking personal aspects into account.
This is precisely what robo-advisors usually do for clients as part of their
investment proposals. Providers are not only required to inform their
clients about profiling, but they also have to meet specific requirements
set out in Article 22 GDPR with respect to the lawfulness of this type of
data processing.
Profiling is always prohibited if the processing of personal data is solely
automated, which produces legal effects concerning the data subject or
similarly significantly affects him or her. It should be noted that the
restriction contained in Article 22 (1) GDPR relates to decisions based
“solely” on automated processing of personal data. In addition, Article 22
(2) GDPR sets out the circumstances under which profiling is permitted
after all.
This may be the case, for example, if it is necessary for entering into, or
performance of, a contract between the data subject and a data controller.
This refers to cases in which the conclusion or performance of the contract
reflects the wishes of the data subject, who thus does not consider fully
automated processing to be an infringement of his rights. This exception
should, therefore, routinely work for robo-advisors with regard to the
provision of financial services. It is also important to note that under recital
71, profiling and any other form of automated data processing should not
concern a child. This restriction must be observed at all times.
In addition, there are special rights and obligations, which must be
observed, including the right to contest the decision, to express his or her
point of view and to obtain human intervention.
158 C. HAMMER
8.5.2 Use of Cloud Services

Robo-advisors regularly use cloud services to store data or to meet certain
requirements (such as retention requirements). The use of cloud comput-
ing by credit and financial services institutions may, in individual cases,
involve outsourcing that is material to the execution of financial services or
any of the institution’s other usual services within the meaning of Article
25b KWG. The isolated purchase of software is generally to be classified as
another purchase, and it thus does not constitute outsourcing. However,
this does not apply to software which is used to identify, assess, control,
monitor and communicate risks or which is essential for the performance of
institution-specific tasks. Support services provided for this type of software
are classified as outsourcing.27 Furthermore, the operation of the software
by an external third party is also classified as outsourcing.
As cloud computing software does not work locally on the server of
the robo-advisor but is operated routinely as a cloud-based service, it can
be assumed that this—provided it concerns part of the advice, broking
or management services—qualifies as outsourcing within the meaning of
Article 25b KWG.
In this respect, the qualifying criteria for use of external services have
been significantly tightened by the new version of Minimum Require-
ments for Risk Management (MaRisk) 2017. While under the old rules
the outsourcing of business activities only took place if an institution
engaged another company (outsourcing company) to perform an activity
or function (service) that was material to the activity of the institution in
the long term or at least for a longer period of time, under MaRisk 2017,
the criterion of materiality is no longer applied to the question of whether
cloud computing qualifies as outsourcing.
Accordingly, the use of cloud services must always be assessed to set up
an outsourcing process where appropriate. The first pool audits as part of
the regulatory review of large cloud providers took place in 2018.
27 Helpful for case-by-case assessment: BaFin’s Guidance Notice on outsourcing to cloud

providers, see Bundesanstalt für Finanzdienstleistungsaufsicht (2018).
8.6 CONCLUSION
The regulatory requirements are continually evolving in line with the
developing technical infrastructure. Humans and machines will continue
converging. Standardized technical processes assist clients in their decision-
making and help to eliminate human errors to a large extent. Regulators
will therefore continue to supervise the robo-advisor market and make the
solutions more efficient and forward-looking. MiFID 3 will be the next
major regulatory amendment in the long term.
REFERENCES
Assmann, Heinz-Dieter, Rolf A. Schütze, and Petra Buck-Heeb. 2015. Handbuch
des Kapitalanlagerechts. 4th ed. Munich: C.H.BECK.
Boos, Karl-Heinz, Reinfrid Fisher, and Hermann Schulte-Mattler. 2016. KWG,
CRR VO–Kommentar zu Kreditwesengesetz, VO (EU) Nr 575/2013. 5th ed.
Munich: C.H.BECK.
Bundesanstalt für Finanzdienstleistungsaufsicht. 2016. Robo-Advice und
AutoTrading–Plattformen zur automatisierten Anlageberatung und
automatischem Trading. Last Accessed 20 March 2020. https://www.
bafin.de/DE/Aufsicht/FinTech/Anlageberatung/anlageberatung_node.
html.
Bundesanstalt für Finanzdienstleistungsaufsicht. 2017. Circular 3/2017 (GW) –
Video Identification Procedures. Last Accessed 20 March 2020. https://www.
bafin.de/SharedDocs/Veroeffentlichungen/DE/Rundschreiben/2017/rs_
1703_gw_videoident.html.
Bundesanstalt für Finanzdienstleistungsaufsicht. 2017. Merkblatt
Anlagevermittlung. Last Accessed 20 March 2020. https://www.bafin.de/
SharedDocs/Veroeffentlichungen/DE/Merkblatt/mb_091204_tatbestand_
anlagevermittlung.html.
Bundesanstalt für Finanzdienstleistungsaufsicht. 2018. Merkblatt – Orien-
tierungshilfe zu Auslagerungen an Cloud-Anbieter. Last Accessed 20 March
2020. https://www.bafin.de/SharedDocs/Downloads/DE/Merkblatt/BA/
dl_181108_orientierungshilfe_zu_auslagerungen_an_cloud_anbieter_ba.html.
Bundesanstalt für Finanzdienstleistungsaufsicht. 2020. Gemeinsames Informations-
blatt der Bundesanstalt für Finanzdienstleistungsaufsicht und der Deutschen
Bundesbank zum Tatbestand der Anlageberatung. Last Accessed 20 March 2020.
https://www.bafin.de/DE/Aufsicht/FinTech/fintech_node.html.
Löhr, Andrea, and Daniela Rothe. 2020. Geeignetheitserklärung: Wichtiges
Dokument für Verbraucher. Last Accessed 20 March 2020. https://www.bafin.
de/SharedDocs/Veroeffentlichungen/DE/Fachartikel/2018/fa_bj_1809_
Geeignetheitserklaerung.html.
160 C. HAMMER
Oppenheim, Robert, and Christian Lange-Hausstein. 2016. “Robo Advisor –

Anforderungen an die digitale Kapitalanlage und Vermoegensverwaltung.” WM
Zeitschrift für Wirtschafts- und Bankrecht, 1966–1973.
Schwark, Eberhard, and Daniel Zimmer. 2010. Schwark/Zimmer–
Kapitalmarktrechts-Kommentar. 4th ed. Munich: C.H.BECK.
Tertilt, Michael, and Peter Scholz. 2018. “To Advise, or Not to Advise – How
Robo- Advisors Evaluate the Risk Preferences of Private Investors.” The Journal
of Wealth Management 21 (2): 70–84.
PART III
Case Studies of Robo-Advisory

CHAPTER 9
(Re-)Launching a Robo-Advisor as a Bank
Theodor Schabicki, Yvonne Quint, and Soeren Schroeder
Contents
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
9.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
9.3 Key Challenges and Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
9.4 Implementation of a Robo-Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
9.4.1 Strategic Framework and Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
9.4.2 Robo-Advisory Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
9.4.3 Evaluation Criteria for Provider Selection . . . . . . . . . . . . . . . . . . . . . . . . 173
9.4.4 Implementation of the Robo-Advisor .. . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
9.5 Evaluation of Existing Robo-Advisors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
9.5.1 Evaluation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
9.5.2 Options for Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
This chapter is also published as white paper “Robo-Advisory in the Securities Market” by
BearingPoint.
T. Schabicki () • Y. Quint

BearingPoint GmbH, Frankfurt am Main, Germany
e-mail: [email protected]; [email protected]
S. Schroeder
BearingPoint GmbH, Hamburg, Germany

164 T. SCHABICKI ET AL.
9.6 Future Trends .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

9.6.1 Digitalisation Requires Individualisation. . . . . . . . . . . . . . . . . . . . . . . . . . 184
9.6.2 Marketplaces and Ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
9.6.3 Start Small to Go Big . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
9.6.4 Once a Competitor, Now an Ecosystem Partner . . . . . . . . . . . . . . . . . 185
9.1 INTRODUCTION
Robo-advisors in today’s known form appeared in the UK in 2008 for
the first time and provided customers with a simple and inexpensive way
to invest money. Since their market entry, robo-advisors have developed
rapidly, and by now affect the financial sector’s securities business signifi-
cantly, more specifically:
• In 2018, of the 800,000 new DIY investment accounts opened in the

year to the end of September, one in three were opened with one
of the UK’s main robo-advisors, including Nutmeg and Moneyfarm
(Source: FT, November 2018).
• The number of DIY investment accounts—including robo and plat-
form customers—rose to 4.8 m over the year to September, an
increase of 22%.
• The volume of assets in the DIY investing market, where customers
pick and choose investments without the help of a financial adviser,
grew to £ 224 bn over the 12 months to September and a 15.4%
increase over the year.
In order to participate in the market growth and to be able to remain

competitive in the securities business, banks incrementally cooperate with
fintechs to ensure the implementation of their own robo-advisors. How-
ever, for a successful implementation or further development, some key
considerations must be made.
This chapter addresses all decision-makers in the securities business from
banks to investment service providers, and addresses the following two
questions:
• Firstly, how should robo-advisors be introduced for the first time?

• Secondly, how could the existing robo-advisors’ success be evaluated?
A GUIDE TO OPPORTUNITY ASSESSMENT AND IMPLEMENTATION 165
9.2 BACKGROUND
For more than 10 years, the stock markets have shown an almost constant
growth curve. The sustained stock market development, combined with a
lack of investment alternatives, has led many private investors to invest their
money in securities. Over this period of economic expansion, the quality
of products and services offered by banks was of low importance due to a
lack of alternatives. Driven by continuously increasing share prices, banks
were actively giving generic investment advice, paired with traditionally
expensive products, leading to steadily increasing securities revenues. Banks
were thus able to continuously increase their securities revenue, even with
average investment advice and relatively expensive products .
However, more and more innovative and agile fintechs have entered
the securities business and are challenging established financial institutions
with their own digital product offerings. Robo-advisory has become
increasingly prominent and thus been mentioned as an investment product
in various news and media channels.
The formula for success concerning robo-advisors focuses on trans-
parent, cost-effective—and above all—digital product offerings. Robo-
advisors offer their customers intuitive and easy-to-understand digital asset
management by providing an exclusively digital advisory process. The
portfolio is derived by using a risk profiling process, and is usually mapped
by the robo-advisor via a cost-effective ETF portfolio. This approach
contrasts considerably with the banks’ relatively opaque, expensive and
non-digital securities products.
In order to compete effectively with fintechs, banks increasingly pursue
cooperation agreements with fintechs and offer their own customers robo-
advisors in the form of white-label solutions. The benefit of this approach
is that white-label solutions can be implemented much faster than in-house
developments as they are designed upon the existing expertise of fintechs.
Additionally, many fintechs have also expanded their business model to
include cooperation agreements with banks, with some focusing exclusively
on this.
Offering a robo-advisor itself is no sure formula for success for banks.
The robo-advisors offered by banks in cooperation with fintechs differ only
slightly from sole fintech offerings, and are significantly more expensive
than fintechs on average, positioned as niche products without a strategy
in place. Consequently, the probability of a sustainable product success rate
is low. In order to maximise the chances of success, banks must deal with
the individual institutional challenges of robo-advisor implementation and

create the strategic and/or organisational framework for the digitalisation
of securities advising.
9.3 KEY CHALLENGES AND CONSIDERATIONS

Many banks still process most of their securities business by means of
traditional advice from securities specialists. Digital channels (in terms
of self-service offerings) often only exist for the non-advisory aspect of
business. Implementing a robo-advisor, the original advising business con-
cerning securities can be digitalised for the first time. The main challenges
of such an implementation project lie in the following context (Fig. 9.1):
The key challenge of introducing a robo-advisor is to develop a clear
target vision for the securities business. The robo-advisor does not corre-
spond the introduction of a new product, but rather to the beginning of the
Fig. 9.1 Challenges of implementing/introducing a robo-advisor

digitalisation of the advising business and thereby also leads to a profound

adjustment of the business model. This means the bank must deal with the
following questions: Which distribution types (execution only, investment
advice, asset management) should be offered in the future, how are the
distribution channels to be linked, and how can coexistence between man
and machine take place? Only if the target vision is defined precisely, the
role of the robo-advisor (and therefore the securities advisor) will become
clear.
The introduction of a robo-advisor is inevitably paralleled by fears in
management of possible cannibalisation effects. On the one hand, high
earnings targets are set for the securities business to compensate the low
income from interests. On the other hand, the margin of a robo-advisor is
significantly lower than the margin of traditional securities products. The
response of sales management is to integrate the robo-advisor into the
marketing control with too much care, and to carry out external marketing
in a too controlled manner. Overcoming these cannibalisation fears in
management is critical to success. A clear and well-communicated vision
and support from top management helps to address this challenge and to
embed the robo-advisor in the securities business sustainably.
More significant than the management’s fears of cannibalisation are
the security advisors’ existential fears that arise due to the introduction
of a robo-advisor and the associated lack of acceptance of the solution.
The advisors become sceptical that digital tools such as robo-advisors will
question their position in the securities business value chain. As many
clients maintain long-term and trusting relationships with their advisors, it
is of utmost importance to convince the advisors that robo-advisors bring
additional value. Securities advising is not possible without the advisors,
not only in current terms but in the future too. However, this role must
be reconsidered in regard to the relationship between man and machine—
undoubtedly a major challenge for banks and asset managers. A possible
scenario for the future is for advisors to develop into solely relationship
managers, while the robo-advisor covers all the technical aspects.
Another challenge is to clearly differentiate one’s own product from
the competition. Many robo-advisors from banks differ very little from
the offers of fintechs for instance, profiling, asset allocation and financial
products are all typically similar. The only perceived difference for the
customer is price. Much to the displeasure of banks, in order to achieve
justifiable margins they must offer their robo-advisory services at a much
higher price on average than most fintechs. Consequently, customers prefer
to give their money to the original provider (fintech) instead of a replica

developed by a bank. Therefore, it is crucial for banks to differentiate their
own robo-advisors from fintech products in a way that adds value and is
noticeable to the end client.
The introduction of a robo-advisor is accompanied by technical chal-
lenges. For instance, in a case where the robo-advisor is implemented
with an external partner (instead of in-house development), the degree
of integration within the bank’s existing IT systems must be determined
clearly. Although a deep integration leads to a significantly optimised
customer experience and more efficient processes, it also results in con-
siderably longer project lifetimes, costs, risks and a higher dependency on
the supplier.
9.4 IMPLEMENTATION OF A ROBO-ADVISOR

Banks have recognised the fundamental necessity of digitalising securities
advising in regard to a self-service offering and are starting the digitalisation
process with the introduction of robo-advisory services for asset man-
agement. Since the implementation should happen as quickly as possible
to avoid significant IT impact, an in-house development or extension of
traditional asset management by a digital channel is often not an option.
The fintech market is broad, considering the fact that many firms already
successfully offer white-label robo-advisory solutions in the industry.
However, it is often assumed that robo-advisory services can be intro-
duced as a new product within a lean implementation project. This may
apply to the provision and connection of the technical platform, but not
to the definition of the necessary strategic and organisational framework
required.
For ensuring the sustainable success of a robo-advisor, it is crucial to
define the strategic framework conditions first and commence the robo-
advisory implementation in that context.
9.4.1 Strategic Framework and Parameters

Historically, banks’ securities business is defined by complex advising
processes, extensive financial product portfolios and a rigorous sales man-
agement focused on network marketing. In this network, the advisor
leverages an outstanding position as their advisory services have a signifi-
cant influence on the success and quality of the securities business. In order
to successfully place a robo-advisor within the market, the robo-advisor

should not be regarded as a product but as an enabler to the securities
business. The strategic framework parameters must be defined prior to the
implementation. Three crucial questions must be addressed during that
stage:
• Which forms of distribution will be used to implement the securities

business in the future?
• How should analogue and digital distribution channels interact?
• What should the future role of the (human) securities advisor look
like?
Regarding securities, banks, as a basis, offer three different forms of

distribution; execution only, investment advice, and asset management.
Prior to the start of the digitalisation of the securities business, the
bank needs to make two fundamental decisions. Firstly, it is necessary to
determine which distribution form the bank intends to use in the future for
its securities business. This is crucial, since the complexity of such a project
is greatly increased by the digitalisation of the investment advice service, in
terms of a self-service offering by comparison to asset management. The
fundamental question of whether investment advice should be offered in
the future, or whether the entire advising business should be subsumed
by asset management, is especially interesting. The latter will enable the
securities business to be implemented much more efficiently, as it will
centralise the investment decision, generate returns free of subsidies, and
create less documentation obligations than the investment advice process.
Secondly, it is necessary to determine how the possible duplication of
products in asset management will be dealt with once the robo-advisor
has been implemented. Most banks offer two-asset management services
post the introduction of a robo-advisor—traditional asset management
and digital asset management, provided by the newly implemented robo-
advisor. If the traditional form of asset management is to be continued,
a clear product differentiation is essential. If asset management is to be
consolidated in the future, a consolidation strategy should be developed at
the time when the robo-advisor is introduced.
In addition to the definition of future sales forms, the interaction
between the traditional and the digital sales channel must be defined. The
area of conflict ranges from two independently offered sales channels, to
an omnichannel approach, in which the customer can switch between the

defined channels as and when needed, that is, advisory begins online and is
concluded together with the securities advisor. However, this significantly
increases the requirements for the technical integration of the robo-advisor
into the core banking systems and requires significant adjustments to the
pre-existing sales process to ensure compatibility.
The implementation of a robo-advisor will significantly change the
securities advisor’s profile. So far, the advisor has been responsible for
making asset-allocation decisions and buying/selling securities together
with a customer. In the future, the securities advisor will be released
from these tasks, as the robo-advisor handles the entirety of the technical
side, from risk profiling and the derivation of asset allocation, to the
purchase/sale of financial products. Consequently, the role of the advisers
must be redefined. This is important to prevent a possible threat to
employment, and thus the rejection of the robo-advisor implementation
plans by the advisers at an early stage.
9.4.2 Robo-Advisory Specification

Once the strategic framework conditions have been set, the specification
of the robo-advisor in terms of target group, financial ratios, sales manage-
ment, consulting process and integration depth is carried out (Fig. 9.2).
When determining the target group, age, IT affinity, financial circum-
stances and expected service level of the customers should specifically be
considered. The target group for the robo-advisory service should be
distinguished from the target groups of other securities products to ensure
an optimal customer approach.
The financial ratios are made up of the expectations regarding the
definition of volume and price. Volume planning should be based on the
overall portfolio volume development for the securities business, as well
as the relative allocation of the individual sales forms (execution only,
investment advice and asset management). For the definition of volume
of robo-advisors, the integration with stationary sales, as well as the scope
of the marketing measures, are critical to success. The price of the robo-
advisor consists of the financial product costs and the service fee for the
financial portfolio management, including securities account management.
Externally, the competitors’ prices and the prices of comparable institutions
with similar target groups are to be evaluated. Internally, the price needs
Fig. 9.2 Dimensions for specification of the robo-advisor in the context of the overall securities strategy
A GUIDE TO OPPORTUNITY ASSESSMENT AND IMPLEMENTATION
171
to be considered in the context of the existing financial product portfolio.

Due to the immense pressure of competition, a pricing approach in line
with the market often results in a lower target margin compared to classic
securities products.
In contrast to fintechs, the relationship between advisor and customer
is based on a trusting advisor–customer relationship. In order to integrate
the robo-advisor into such a structure in a useful way, the securities advisors
must be incentivised regarding the robo-advisor’s sales—the robo-advisor
has to be a part of the sales control.
The robo-advisory process mainly consists of five sequential process
steps. Both distribution and regulatory aspects must be considered when
specifying the individual process steps. It is recommended to amalgamate
the advisory process of the robo-advisor manually to the advisory process
of traditional asset management advice from both perspectives, in order to
avoid different securities advising processes, depending on the channel.
The onboarding process involves identifying the customer according
to the KYC (Know-Your-Customer) procedures. The main question in
this respect is who the customer’s contractual partner is. As the customer
relationship is an essential asset of the bank, the customer should close the
financial portfolio contract directly with the bank, and not with the robo-
advisor.
Risk profiling determines the risk-bearing capacity and propensity of
the customer. This is normally processed using questions that are easy to
understand. Questions about behavioural finance are typically becoming
increasingly popular. Therefore, the customer’s risk profile is determined
by “softer” questions about hypothetical decision-making situations. In
addition to risk profiling, the potential customer should also be asked about
their investment objectives and needs, which should be considered in the
context of personal risk profiling.
The product universe defines the financial products with which the
different portfolios are replicated. Most robo-advisors exclusively offer
ETFs due to the favourable cost structure of around 20 basis points. Only
a few robo-advisors additionally mix the portfolios with actively managed
funds or individual securities. The product universe should always be
defined by the bank.
A technical aspect is the depth of integration of the robo-advisor into
the bank’s IT architecture. The depth of integration describes the degree
of integration of the new application within the core banking system, or
within other relevant systems. The extent of the integration defines the
area of conflict between the scope of functions and the implementation

period. The deeper the integration, the higher the functional scope, but the
more complex and time-consuming the implementation. Here, the main
question is whether the securities accounts should be managed by the bank
or an external custodian bank.
9.4.3 Evaluation Criteria for Provider Selection

If a white-label solution is used, the criteria of the strategic framework and
the specification of the robo-advisor are defined for the evaluation and the
subsequent selection of the provider. The criteria should be grouped and
weighed according to their relevance. A possible grouping can be achieved
by looking into the following dimensions (Fig. 9.3):
9.4.3.1 Functional Specification

The basis for the evaluation of the functionality is the defined advising
process. The extent to which the defined requirements are met by possible
providers should be evaluated separately for each process step. The key
factor here is the degree of individualisation of the advisory process. For
instance, is it possible to use an individual risk profiling approach, or is
it necessary to adopt the provider’s profiling approach? A high degree of
individualisation of the solution enables the bank to use the new digital
advising process in line with the existing analogue process and allows
flexible adjustments for the future. In order to interlock robo-advisors and
stationary sales, it must be ensured that the provider also offers special
consultation access points which allow the securities advisors and customers
to use the robo-advisors together.
9.4.3.2 Technical Specification

Considering technical aspects, such as data protection and the integration
of the solution into the bank’s existing systems, is vital to the success of
the technical specification. In regard to data protection, it is necessary
to determine on which IT infrastructure the robo-advisor service will be
operating. Both the location of the data storage and the location of the data
access are relevant and must be adaptable to the needs and requirements
of the bank. In general, banks expect data storage to be in the UK,
and access only available from locations within Europe. Certifications or
regular independent external audits can certify the provider’s compliance
174
T. SCHABICKI ET AL.
Fig. 9.3 Evaluation dimensions for provider selection

of security standards. In addition to security, the legally compliant handling

of customer data must also be ensured.
When it comes to the technical integration of the solution, experience—
or even existing interfaces between the provider and the bank’s core
banking system—banks must ensure diverse advantages during an imple-
mentation. An open architecture in terms of platform banking and the
provision of programming interfaces and so-called APIs, gives third parties
a simple and standardised connection. Such a technological basis offers the
possibility of extending the service through third-party service providers.
9.4.3.3 Profitability
The challenge in evaluating profitability is that providers often have differ-
ent cost structures and are thus not directly comparable with each other.
In addition to the usual platform fees as a percentage of the assets under
management, providers sometimes charge minimum fees and fixed service
and/or maintenance fees. However, in other cases the implementation
costs differ considerably, depending on the provider.
In order to establish comparability between providers, the evaluation
must be carried out based on a business case. For the business case, the bank
makes projections about the development of the custody account volume
and the price for the end customer. Once these projections have been made,
the predicted revenues can be compared with the costs of the providers. In
the case that securities accounts are held by an external custodian bank,
these costs must also be considered in the projection. Since minimum fees
for individual providers can significantly influence the business case, the
case of low volumes and various scenarios for volume development need
to be calculated.
9.4.3.4 Provider-Related Specification

Lastly, the provider analysis will need to take place. The core objective
of this analysis is to evaluate how competitive the provider is and how
efficiently a potential implementation project can be carried out together
with the provider.
The successful implementation of reference projects is probably the
best indicator for a successful implementation. In this case, the provider is
already familiar with the typical challenges, has already implemented best
practices, and can demonstrate sufficient support for the sales strategy.
Other criteria include the accreditation of a financial portfolio manager,

a proven history of years of successful market presence, financial stabil-
ity and sustainability, and financially strong investors. Furthermore, the
provider’s cooperation partners also have an impact on the business model.
For example, in the case of an external custodian bank being commissioned
to manage the client portfolios, it is advantageous if the provider has already
proposed or established a relationship with a cooperation partner.
9.4.4 Implementation of the Robo-Advisor

The considerations for strategic integration, the defined requirements for
the supplier, and the evaluation grid, result in a framework whereby the
market for robo-advisors is viable to explore. This heterogeneous market
can be divided into three segments.
1. Some of the robo-advisory providers only operate within the B2B

market. These companies support banks and asset managers in devel-
oping and providing a robo-advisory service. As software developers,
they do not directly make contact with the investor as an end
customer.
2. Providers are operating in both the B2C and B2B markets. Many
of these companies have originated their business model in the B2C
market and are now expanding by offering white-label solutions for
B2B clients.
3. The market for only B2C providers consists of fintechs and digital
solutions from established financial institutions. The development of
these solutions is partly based on the state of the B2B market.
Provider selection is based on the previously defined evaluation criteria and

should be carried out with a systematic process. The result of the multi-
stage process is the decision for an appropriate cooperation partner for the
implementation of a robo-advisor (Fig. 9.4).
9.5 EVALUATION OF EXISTING ROBO-ADVISORS

Many banks have already implemented their own robo-advisors as white-
label or in-house developments. However, the success of these robo-
advisors is by no means guaranteed. There are already first-hand examples
Fig. 9.4 Introduction of a robo-advisor

of failed partnerships of fintechs and banks on the market. Some com-

mercial banks have also deliberately decided not to offer their own robo-
advisors. In the case of robo-advisors which have been placed on the market
by banks, it can also be assumed that some banks might face problems in
acquiring enough volume, and thus operating profitably.
A reason for this is the strong competitive environment, both on the
market, but also within the bank internally. On the market, robo-advisors
compete with dynamic fintechs, offering products with low margins, excel-
lent customer service and high functionality. Internally, robo-advisors have
to compete with high-margin competitors and old (focused on network
marketing) distribution structures. In addition, the required service and
comfort level for the customers and securities advisors cannot be achieved
due to suboptimal IT integration.
It comes as no surprise that in this competitive environment robo-
advisors fail to achieve their ambitious growth targets. The high market
dynamics paired with the continuous technological progress make it nec-
essary to regularly evaluate one’s own robo-advisor in terms of securities
strategy and to identify fields of action.
9.5.1 Evaluation Criteria

The evaluation of existing robo-advisors must take place in the following
dimensions: target vision, financial key figures, sales control, customer
usage and satisfaction, product design, technical integration and providers
(Figs. 9.5 and 9.6).
9.5.1.1 Target Vision

As mentioned earlier in this chapter, a conclusive target vision supported by
top management represents the foundation for a sustainable process. Only
when the robo-advisor is adequately integrated into the securities strategy
of the bank is it possible to enhance further development. Additionally,
the sustainability of the overall target vision and the specific consideration
of the robo-advisor should be evaluated. On the other hand, it is also
important to validate the communication and transparency of the proposed
strategy within the organisation. The target vision needs to define the
framework conditions for the specific design of the robo-advisor.
Fig. 9.5 Evaluation

criteria for existing
robo-advisors
9.5.1.2 Financial Figures

In the long term, the financial figures are the key measures for evaluating
the robo-advisor’s success. This includes key figures regarding the achieved
volume, the achieved revenues and margins, but also the costs for the
provision of the service itself. The evaluation is carried out as a comparison,
with predicted values, or as a benchmark against the market. The financial
ratios of the robo-advisor should also be compared with the rest of the
bank’s securities products.
9.5.1.3 Sales Control

To ensure the sustainable acquisition of custody account volumes, the
robo-advisor must be carefully considered within sales management. This
includes the overall planning for the securities business, as well as the
specific sales targets. Both nominal targets and the relative relationship to
other control parameters must be evaluated. In addition to the control
parameters, the way the targets are supported by marketing and training
measures must also be evaluated. The acceptance of the robo-advisor
by the securities advisors is also paramount in this dimension. In banks
with an intensive customer–advisor relationship, it is critical to success
to ensure that the advisors regard the robo-advisor as a value-adding
activity, and thus recommend the service to customers alongside traditional
Fig. 9.6 Evaluation of existing robo-advisors

product offerings, bearing in mind not to offer comparable or overlapping

solutions.
9.5.1.4 Customer Usage and Satisfaction

Two preliminary questions to evaluate the success of a robo-advisor detail
the customers general usage and how satisfied they are with the solution. To
answer the first question, information concerning customer characteristics
must be evaluated. The information provided includes age, user behaviour,
usage intensity, education level, investment behavior (amount and time) as
well as other aspects. In principle, the aim of gathering this information is
to assess which group of customers use the robo-advisor in order to address
comparable customers in a targeted way with regard to the product, or to
align sales control appropriately. To answer the second question, the users
of the product must be questioned directly in an online survey. In this
context, customer satisfaction should be explored concerning the entire
end-to-end process. This includes the initial approach to the product, the
advising process, the closure, the reporting and the interaction between
the securities advisor and robo-advisor. These results enable valuable
conclusions to be drawn for further development of the solution.
9.5.1.5 Product Design

In the context of the product design, the aim is to evaluate the degree
to which the robo-advisor meets the customer needs and how the solution
differentiates itself from the competition. The customer survey can be used
to gain insights as to where in the process the robo-advisor should be
optimised, in order to reflect the customer needs. Furthermore, the robo-
advisor should be compared with competing products via benchmarking
processes. In particular, it is important to assess the degree in which the
robo-advisor differentiates itself from the competition, and which features
are specific to the institution.
9.5.1.6 Technical Integration

The degree of technical integration required is derived from the target
vision. However, for long-term success, it is important that the robo-
advisor is connected to the existing system landscape. This dimension
evaluates how well the robo-advisor is integrated into the bank’s existing
systems. In this context, it is essential to establish to what extent media
disruptions occur for customers (e.g. lack of single sign-on or missing
assumptions of existing risk profiles). Another evaluation criterion is the

integration of the custodian bank function, which is currently often
outsourced to external custodian banks, creating additional complexity for
the customer in terms of a second point of contact.
9.5.1.7 Providers
In many robo-advisory solutions, external providers are used for the
financial portfolio management and in part for the custodian bank function
too. Reliability, innovation and service quality are essential evaluation
criteria for successful cooperation with the provider. This is essential if the
robo-advisor’s frontend is to be used for other securities products in the
future (e.g. investment advice), the further development of the securities
business depends significantly on the strength of the provider.
9.5.2 Options for Action

The evaluation result is the basis for a purposeful optimisation of the
existing robo-advisor solution. Three general options can be derived from
the spectrum of possible actions, which can be implemented individually
for each institution:
• Ongoing development of the existing solution

• Fundamental change regarding the existing solution
• Shutdown of the robo-advisor
In the presence of constantly changing customer requirements and mar-

ket changes, further development of the existing solution is inevitable.
By deepening the integration into the bank’s systems, for example, the
customer experience can be significantly optimised by building on already-
existing customer data. When expanding the existing robo-advisor, the
market-related developments, the institute-specific framework conditions
and the innovation potential of the white-label provider must be taken into
account. It is crucial that the robo-advisor is meaningfully integrated into
the bank’s securities strategy.
A fundamental modification of the existing robo-advisor solution must
be considered if the existing white-label provider cannot implement essen-
tial requirements, or if the innovation potential for an expansion is missing.
In these cases, the white-label provider could be replaced, or a separate
solution could be developed. However, it should be established how

the existing customer relationships should be handled. Customer custody
accounts are often held with an external custodian bank and cannot be
transferred without contractual adjustments.
The withdrawal without any alternative option from the robo-advisor
business is an action up for questioning: The combination of demograph-
ically induced shifts in customer behavior and the financial maturity of the
digital natives’ demand for digital solutions in the securities business. In the
short term, a withdrawal might improve financial figures. However, this is
at the cost of a weakening of competitiveness in the securities business.
Digital asset management in the form of a robo-advisor can be regarded as
a fundamental step in ensuring a sustainable digitalisation of the securities
business.
9.6 FUTURE TRENDS

From digital securities advising to digital asset management—technology
trends, which can be observed in other areas of digital banking, are also
finding their way into asset management. In this context already-known
innovations are being used:
• Voice Banking & AI—contactless, voice-based banking, which contin-

uously improves the understanding, interpretation and prediction of
customers’ needs by using Artificial Intelligence (AI) in a self-learning
manner.
• Digital Identities—central management, allocation and safe custody
of uniquely assigned identities and corresponding collateral informa-
tion.
• Blockchain infrastructures—substitutes for traditional and securities
accounts as well as being the basis for trustworthy, highly efficient
(business) transactions, especially in the area of digital identity or
legally secure signatures.
• Open Banking APIs—flexible, rapidly deployable interface technolo-
gies that greatly simplify and accelerate interaction with other market
participants or customers.
• Marketplaces (SaaS)—establishment and operation of a regulatory
and trust-based framework to better serve “supply and demand”.
The future development in the area of digital securities advising will

probably show two different characteristics:
• On the one hand, there will be massive consolidation pressure among

B2C providers, as customer acquisition costs (CAC) and customer
lifetime value (CLV) can only be turned into a positive business model
in larger organisations, by benefitting from the effect of economies
of scale. In this context, the innovation potential of the resulting
organisations, which will be exposed to similar competitive pressure as
existing competitors in the non-digital market, appears questionable.
• On the other hand, we believe that in the B2B segment, models will
prevail, allowing customers easier access to software components and
business processes. Due to the strategic advantage of platform-based
structures, it is expected that for the area of digital asset management,
software-as-a-service (SaaS) marketplaces will emerge, enabling cor-
porate banking customers to implement a business process innovation
or add new digital components to an existing process with a few simple
steps.
9.6.1 Digitalisation Requires Individualisation

Today’s innovation still looks as follows: for each stage of digitalisation,
the best solution for achieving pre-defined goals must be worked out from
a variety of possible providers (internal and external) at first. In larger
IT projects, various integration procedures and connections to different
systems must be developed in order to integrate the corresponding com-
ponents. These usually require an excessive amount of time, money and
resources.
9.6.2 Marketplaces and Ecosystems

The individualisation approach is contradicted by the logic of an SaaS
marketplace, connecting different providers of subcomponents in advance
via standardised interfaces (“APIs”) and making them available to the
respective users individually on demand. In this context, we shape the
term “survival of the API-est”, as we believe that in the future only those
market participants will assume significant ecosystem roles and master
their APIs. The API management systems used for this purpose control
and regulate the connectivity between the providers and their services,
with the corresponding consumers and customers. Therefore, from our
experience, the success of a marketplace operation depends essentially on
the effective use of APIs. The marketplace operator itself is usually only
a provider of SaaS solutions in certain sub-segments—rather, it organises
and orchestrates the market participants on the supply and demand side and
provides the infrastructure for a successful development of the ecosystem
to providers and consumers.
9.6.3 Start Small to Go Big

From the customer’s point of view, based on the collected data from the
platform, a significant increase in efficiency in the areas of software selection
and implementation can be assumed, since the necessary connections for
the implementation of digitalised asset management are already available at
platform level and can be provided with little effort. The establishment of a
SaaS marketplace for digital wealth management also gives providers a high
degree of flexibility regarding the design of their business model and the
necessary infrastructure. It will be much more cost-effective to test certain
innovations first in a smaller, defined target group, as the start-up costs for
digital business models will be a fraction of today’s usual project costs.
9.6.4 Once a Competitor, Now an Ecosystem Partner

From the platform operator’s point of view, the strategic orientation
is lucrative if it is possible to unite a relevant number of partners and
customers on the marketplace and to facilitate simplified access to software
components from different providers with the help of professional business
process management. The participation in such a SaaS market makes
sense for software providers in case the reduced resource requirements for
the acquisition of new customers are invested in further development of
one’s own offering, strengthening the competitive position with respect
to other providers with similar functional scope. Hence, making the
provider attractive for the customers regarding price and/or functional
scope. In order to establish such an ecosystem, significant investments by
the marketplace operator and the SaaS providers are required to create
the necessary technical and procedural prerequisites. The effort seems
reasonable if the ability to innovate, speed and efficiency of innovation can
be materially increased.
CHAPTER 10
How Can Robo-Advisory be Implemented

and Integrated into Existing Banks?
Ana-Maria Climescu, Christian von Keitz, Jan Rocholl,

and Madeleine Sander
Contents
10.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
10.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
10.1.2 Client Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
10.1.3 Market Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
10.2 The Implementation and Integration Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
10.2.1 Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
10.2.2 Product.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
10.2.3 Marketing and Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
10.2.4 Regulatory Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
10.2.5 Technical Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
10.3 The Real Question Is: “How Can the Bank Use the Latest
Technology to Expand its Range of Services in the Interest of the
Client?” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
A.-M. Climescu () • C. von Keitz • J. Rocholl • M. Sander

Hauck & Aufhaeuser, Frankfurt am Main, Germany
e-mail: [email protected];
[email protected]; [email protected];
[email protected]

188 A.-M. CLIMESCU ET AL.
10.1 BACKGROUND
The following case study deals with the question of how robo-advisory can
be implemented and integrated into the existing bank model, using the
example of Hauck & Aufhaeuser Privatbankiers (H&A). In this context,
we understand robo-advisory as an online application that offers automated
financial advice and services.
Firstly, we will look into the bank’s background. As part of its strategy,
H&A has set itself the goal of examining the extent to which the use of
new technologies can improve existing business and unlock new market
potential across all divisions. This gave rise to the following question with
respect to private banking, which is the focus of this chapter: “How can
we use the new technical possibilities to address changing customer needs
as well as generate new business opportunities and revenue streams?” To
answer this, we first analyzed client needs before turning our attention to
the investment market.
10.1.1 Introduction
H&A was created in 1998 as a result of a merger between two well-
established institutions: the Frankfurt-based Bankhaus Georg Hauck
& Sohn (founded in 1796) and the Munich-based Bankinstitut H.
Aufhaeuser (founded in 1870) (Hauck & Aufhaeuser 2020c). H&A is
thus one of the oldest private banks in Germany with a history stretching
over 220 years.1 In addition to serving private clients, the company has
specialized over the past three decades in financial services for professional
investors and businesses. H&A divides its activities into four divisions:
private banking, asset servicing, financial markets and investment banking
(Hauck & Aufhaeuser 2020a).
The bank currently has 722 employees2 across a total of seven locations
in Germany, Luxembourg and Switzerland.3 In addition to its headquarters
in Frankfurt, H&A also has offices in Munich, Hamburg, Duesseldorf and
1 In June 2017, H&A was transformed from a KGaA (partnership limited by shares)
into the legal form of an Aktiengesellschaft (joint-stock company); see Hauck & Aufhaeuser
(2017a, p. 34 ).
2 Annual Report 2019, Hauck & Aufhaeuser (2019a).
3 In addition, H&A has a subsidiary in London and Nanjiing as well as a representative
office in Paris.
HOW CAN ROBO-ADVISORY BE IMPLEMENTED AND INTEGRATED INTO… 189
Cologne. While its German branches mainly focus on private banking, the
Zurich branch specializes exclusively in ethical and sustainable investment
aimed at private and institutional investors. Nevertheless, Luxembourg has
recently become the bank’s largest location. This is due, on the one hand,
to the expansion of both the depositary and fund management business
and, on the other hand, to the acquisition of the Luxembourg-based Sal.
Oppenheim entities in 2017/2018 (Hauck & Aufhaeuser (2016) and
Hauck & Aufhaeuser (2017b)).
This step fits seamlessly into its corporate strategy, “H&A Strategy
2020”, which focuses on growth and digitalization. This was adopted
in 2017 by the Executive Board, the Supervisory Board and the new
Chinese owner Fosun following the takeover (see Fig. 10.1). The takeover
by the Chinese group is one of the most important milestones in the
bank’s recent history. In mid-2015, the bank officially announced that
the Chinese investor group Fosun plans to acquire a majority stake in
the bank. The official transfer of shares took place in the fall of 2016.
Fig. 10.1 H&A strategy 2020 (taken from presentation slide)

H&A is the first German private bank to be majority owned by a Chinese

investor (Mussler 2016). In addition to the analysis and evaluation of
organic and inorganic growth potential, the strategy is to tap into the
cross-selling potential between China and Germany. However, the focus
of the digitization project is on expanding the range of services offered to
customers and increasing customer benefit through intuitive applications.
This also involves modernizing the bank’s internal infrastructure and
automating administrative work processes to spend more time advising
customers. At the same time, concrete measures have been defined for each
business division that contribute to the overarching corporate objective
of developing H&A into a state-of-the-art, market-leading bank in the
German-speaking region.
In the first step, the “H&A Strategy 2020” focused on the evaluation
of strategic, organic growth options for private banking. Initially the
demand of private clients for banking services was analyzed from various
perspectives. For private banking, which is regarded as “people’s business”,
the behaviour and wishes of clients play a particularly important role.
Digitalization has opened up new technological possibilities, some of which
have been accompanied by changes in customer needs (Ernst & Young
2019, pp. 65). Many aspects of the world in which we live and the needs
of the next generation have been through a period of profound change. For
H&A, this aspect is of central importance, as many clients have entrusted
their assets to the bank for the third or fourth generation. In addition, it
was necessary to gain a detailed overview of the existing market in the area
of investment and thus also of the competition. The current minimum
investment amount of H&A is EUR 500,000. There were no suitable
products for clients with lower assets.
10.1.2 Client Needs

Internet portals, newsletters and social media today make it much easier
for private clients to obtain information. In addition, there are intuitive,
simple and entertaining customer applications such as those available in
the travel industry or e-commerce. With just a few clicks you can now
order books, clothes, your next trip and much more online. When you
are shopping for a suit online, you get matching belt recommendations
and when you are browsing for an investment guidebook, you can gain
very interesting insights into the buying behavior of others, for example,
through the “Other customers also bought” window. According to trend

research (Linden and Wittmer 2018, pp. 1, 7, 13, 16), the gigatrend
of digitalization and the resulting megatrends such as individualization,
knowledge culture and mobility will further increase the demand for such
services.
In-house client feedback shows that this is also reflected in the expecta-
tions of private customers when it comes to financial investments—starting
with the desire to have an overview of the performance of their assets at all
times in real time, through to the possibility of being able to compare or
directly conclude individualized proposals during the investment process,
regardless of time or place, and thus avoid going to the branch during
normal business hours.
At the same time, it is interesting to note that studies show that existing
private banking clients, especially those with growing wealth, do not con-
sider the level of digitalization or, surprisingly, the price to be among the
top criteria determining their satisfaction. Instead, they value the service
received from the client advisor, along with asset preservation (Nicolaisen
2018). As a result, it can be noted across the board that personal contact
and proximity to the advisor as well as performance and product variety
remain relevant across all customer segments (Deloitte 2015, pp. 24–25).
In addition, the brand’s reputation is particularly important for traditional
private banking clients (l.c.). While user experience and convenience are
already among the top factors determining corporate customer satisfaction,
existing private banking clients tend to be more reserved in this respect
(l.c.). At the same time, however, the increasing use of technology in
everyday life and the associated stronger affinity for digital applications
obviously mean that having fast and convenient access to a bank account
is becoming even more important to private banking customers (l.c.).
Somewhat surprisingly, reputation and exclusivity of the brand are seen
to be gaining ground in this context (l.c.). This suggests that positioning
in the private banking market will in future range between:
1. convenience and technology

2. service quality
3. brand reputation
The demand for technology and innovation and thus the importance of
the first factor is increasing among clients, whether it is today’s average
client or the inherited wealth client, and is even more pronounced among
the under-40 age group (Nicolaisen 2018). This is where fintechs or
robo-advisors come in. Having its origins in America with pioneers like
Betterment and Wealthfront, the robo-advisory market is growing fast in
China and gaining in importance in Germany, even though the market
forecasts vary somewhat and have to be viewed critically based on the
actual growth rates to date: Statista (2019d) expects the investment volume
in Germany to reach EUR 24 billion by 2022, representing an annual
growth rate of 48% (compound annual growth rate (CAGR) 2019–2021).
Roland Berger, a global consulting firm, anticipates a somewhat more
conservative compound annual growth rate of 20–40% compared to 4–
6% for conventional asset and wealth management (Buess et al. 2018, p.
8). The comparison with the conventional segment clearly underscores the
immense potential of this new market segment.
But what exactly is behind this potential? How should H&A respond to
this development? Offering ease of use (e.g. when opening an account), a
high degree of transparency and, on balance, attractive conditions, robo-
advisors firmly have their finger on the pulse of changing customer behavior
as described above. Specialist studies have also come to the conclusion that
“hardly any of the established asset managers will be able to do without
robo-advice in the future”. At the same time, the average investment
volume, depending on the provider, is around EUR 10,000 (fintego)4
or around EUR 32,000 (Scalable Capital 2018), which is well below the
volume generated by H&A’s current target client segment. The question
here is, therefore: to what extent is this development relevant to H&A and
its target client segment?
These considerations pose major challenges to the banking sector in
general and the private banks such as H&A in particular. As a medium-
sized wealth management boutique steeped in tradition, the bank has
often demonstrated its innovative strength and resistance to crises. Unlike
large corporations, however, private banks do not have multi-billion dollar
budgets available to fund their development. Although open to new ideas,
many bankers were also skeptical about the robo-advisory market at first.
Personal communication and client engagement are key success factors
in sales at H&A. H&A builds long-standing relationships with clients,
knows their families, their goals and interests, and not only looks after
4 Based on ebase report (European Bank for Financial Services GmbH 2019, p. 6).
their investments but also acts as a “sparring partner” for topics such as
career planning, raising children, succession planning or new trends, and
often has an interest in similar hobbies. Acquiring new customers whom
you have never even spoken to before sounds a bit strange, especially in a
private bank.
In addition to these strategic and cultural issues, the growth in infras-
tructure of established private banks over the years presents another
challenge. Without a fundamental transformation, they cannot compete
against the “greenfield” fintechs and their plug-and-play5 approach based
on application programming interfaces (API). Both old core banking
systems and decentralized data repositories are among the biggest obstacles
when it comes to speed. In addition, new methods such as agile project
management or the approval of minimum viable products (MVP), which
are already commonplace in the fintech and robo-advisory sectors, are still
seen as uncharted territory by established banks.
10.1.3 Market Analysis

In addition to customer needs, market potential also played a key role in
H&A’s private banking strategy. Specifically, the question was whether the
use of technology in the existing client segment above EUR 500,000 could
lead to greater customer satisfaction and thus attract a greater number
of new customers, or whether a significant growth potential could be
identified in the other client segments. In order to validate these scenarios,
the first step was to have an in-depth overview of the investment market
(i.e. market volume and medium-term market forecasts) and the existing
providers (or the range of products and services offered by existing robo-
advisory market players) to explore the strategic options for H&A and,
based on this, develop a clear position for the bank’s own approach.
The market potential and volume are determined primarily by the
financial assets of private households. In Germany—the primary target
market for H&A’s private banking activities—household financial assets
amounted to around EUR 6053 billion at the end of the third quarter of
2018 and rose by around EUR 76 billion or 1.3% compared to the second
quarter of 2018 (Statista 2019c). Taking a five-year view, the growth
5 The English term plug-and-play refers to the possibility of connecting new devices to a
computer without having to install additional drivers or changing the settings.
Fig. 10.2 Private banking client segments in Germany 2018 (based on

zeb.research)
rate was around 15%, on the basis of, EUR 5222.6 billion in the third
quarter of 2015 (l.c.). In 2018 (Statista 2019a), German investors favored
conventional savings (39%), followed by fund units (34%), equities (26%),
real estate (26%) and call money (26%). The savings rate, which has been
rising steadily since the last financial crisis, also appears interesting in this
context (Statista 2019b). At this point, it is important to break down the
financial assets of private households even further and to take a closer
look at the relevant market segments in relation to their market potential
(Fig. 10.2).
A private banking study by zeb expects the CAGR or compound annual
growth rate of the private banking market in Germany to be 4%. In more
detail, wealth management clients are forecast to grow by 4–6%, closely
followed by private banking clients with 1–3%. The forecasts for affluent
and retail clients are more restrained: 1–3% for affluent banking clients and
−1 to 1% for retail banking clients. In this context, however, it is important
to consider the market share of the respective client groups. While the
market shares of wealth management clients, private banking clients and
affluent banking clients are 0.16, 1.4, and 1.4%, respectively, retail banking
clients account for the bulk of the market share with 96%. H&A currently
serves private banking clients and wealth management clients. Here it is
necessary to consider how technology can be used to improve or, where
appropriate, expand the existing range. It is therefore necessary to take

a closer look at the affluent and retail segments to assess any untapped
potential for the bank.6
Another study by BCG reveals significant weaknesses in the affluent
segment, in particular, with regard to the range of services currently
offered. At the same time, the study ascribes the greatest potential for
future growth to this segment: For example, investable assets in the global
affluent client segment—defined here as those with assets ranging from
USD 250,000 to USD 1 million and comprising 76 million people—are
expected to grow at an above-average CAGR of 6.2% over the next five
years (Zakrzewski et al. 2019, pp. 12–13). This segment has a great market
potential. Although retail banks, private banks, discount brokers, insurance
companies and fintechs have developed a broad range of products, the
affluent segment continues to be underserved (Zakrzewski et al. 2019,
l.c.). It is dominated by expensive products from well-known brands or
simplified offerings from retail banks that do not meet the needs of affluent
clients (Zakrzewski et al. 2019, l.c.). In order to successfully serve this
client segment and develop the associated market potential as a bank, it is
necessary to offer products that meet the specific needs and preferences of
affluent sub-segments (Zakrzewski et al. 2019, l.c.). In specific terms, this
means expanding services and products to include the right technology.
A bank must be seen as a hybrid business model that combines digital
and human engagement to provide clients with a more personalized,
convenient and at the same time successful asset solution and performance
(Zakrzewski et al. 2019, l.c.).
BCG has closely examined the weaknesses of existing providers in the
affluent segment (Zakrzewski et al. 2019, l.c.). Although competition for
this segment is increasing, most banks and asset managers have not yet
found an answer to their clients’ needs. Private banks with a focus on
wealth management tend to overserve their affluent clients at the beginning
of the relationship, offering them their own relationship manager and
processes designed primarily for wealth management clients. But when
the relationship fails to live up to its perceived potential, the bank may be
tempted to reduce its engagement. From a cost perspective, this is of course
understandable, but may leave clients feeling neglected. Private banks with
a focus on affluent clients have in some cases barely done anything to tailor
6 Data for this paragraph taken from Nicolaisen (2018).

their product range (e.g. broader product range) and service offering (e.g.
digital access) to client needs and are thus hardly in a position to offer
sophisticated financial or investment advice to a wide range of affluent
clients.
Retail banks are still more focused on the sales and reach of their brands
and products than on the specific needs of this client group. Although
fintechs or robo-advisors have the advantage over their rivals in innovation,
they do not have the market access, reputation or the trusted advisor status
enjoyed by private banks. Insurance companies are also trying to tap into
this promising segment. However, they lack the resources and the necessary
expertise to develop individual asset-management solutions. Some asset
managers have again tried to expand their range of services by buying
and deploying robo-advisors, with mediocre results so far. That said, some
providers have regrouped after their first efforts and have begun to work on
digital additions to existing investment models that should ultimately allow
advisors to serve this segment better and more cost-effectively. Brokerage
firms have largely underserved this segment from a product perspective,
focusing too much on short-term trading and too little on long-term
wealth creation. Also here, there are some signs of improvement.
Overall, however, this market remains poorly served and thus offers
great earnings and market potential to an established traditional brand
such as H&A. The last step was therefore to take a closer look at the
current market for digital offers in the field of asset management and the
client groups addressed. It is also noticeable that even established private
banks, but also new robo-advisors, are focusing on the retail rather than
the affluent segment. The positionings of current market participants tend
to follow two paths: either they emphasize the level of digitization, which
in the case of robo-advisory is usually tied to a smart algorithm, or they
emphasize the proximity to the bank and the brand behind it. This also
depends on whether the companies are fintechs or robo-advisors (e.g.
Scalable, ginmon, or Liquid) and thus mostly relatively young companies
or a bank spin-off or the digital asset management of an already established
bank (e.g. Solidvest, Warburg Navigator or Castell Insight). There are
also offshoots of banks that, like fintechs, consciously position themselves
through automated risk profiling in conjunction with semi-automated
rebalancing.7 The competitors relevant to H&A can thus be grouped into

three clusters (Fig. 10.3), which, from a positioning perspective, range
between “smart algorithm” and “proximity to a parent company” (see
Fig. 10.3). For H&A’s private banking, the question was how to leverage
its strengths in positioning its digital offering and how it can differentiate
itself from other providers. Following in-depth studies of client needs and
an analysis of the market environment, it has become clear that there
is a great potential for H&A in the affluent segment, which has so far
received only rudimentary service from other providers. As this is a group
of affluent clients who have been denied access to the bank’s products and
services because of their assets, the reduction in the minimum investment
requirement for the bank is a logical consequence of the new technological
possibilities. H&A not only has the necessary expertise in the field of
investment but also the necessary reputation to harness the potential of
the affluent segment.
10.2 THE IMPLEMENTATION AND INTEGRATION

PROCESS
The following key findings can be formulated based on the market and
customer analysis:
• The following three features are crucial for positioning in the private
banking market:
1. convenience and technology
2. service quality
3. brand reputation
• The demand for technology and innovation is increasing among
clients, whether it is today’s average client or the inherited wealth
client, and is even more pronounced among the under-40 age group.
• This is where robo-advisors come in, which are expected to grow faster
than traditional private banking providers because they offer a better
user experience, greater transparency and more attractive terms.
7 Rebalancing is defined as the regular adjustment of a portfolio in line with the defined
investment strategy.
Fig. 10.3 Products and services offered by existing market providers in the robo-
advisory segment
• The analysis of the client segments shows that, although the private
banking market is expected to grow faster, the percentage share of the
affluent and retail segments is much bigger.
• While most financial investment providers concentrate on the retail
segment, the affluent segment is underserved by established financial
service providers and robo-advisors.
• Existing robo-advisors have so far focused on their positioning on the
level of digitization or the proximity to the parent company/parent
brand.
These findings then flow into the development of growth scenarios and
possible next steps for H&A’s private banking, to answer the overarching
question at the heart of the case study. Robo-advisory services are intro-
duced on five8 levels: strategy, product, marketing and sales, regulatory
and the technical platform.
8 See Accenture (2019).

Fig. 10.4 Strategic objective of an integrated omni-channel
10.2.1 Strategy
The omni-channel9 emerged as a result of the key findings of the market
and client analysis, long-standing expertise in working with private clients
and the analysis of strengths of H&A in a seamless transition between a new
digital sales channel to be established for affluent clients with a minimum
investment of EUR 50,000 and a local sales network for conventional
private banking customers with a minimum investment of EUR 500,000,
through personal advice offered in the branches and supported by modern
applications. H&A’s strategic approach is explained in more detail below
(Fig. 10.4).
The market and client analysis showed that the robo-advisory business
model offered to date by fintechs is less tailored to the needs of private
banking and wealth management clients serviced by private banks such as
9 In this context, omni-channel means a cross-channel business model, e.g. the combina-
tion of high-street and digital sales.
H&A in the traditional sense and it thus does not directly compete in the
core group segment. At the same time, the technology-based approach
represents a smart way of making H&A’s asset-management solution and
associated long-term investment management expertise available to a new
client group below the previous minimum investment amount, thereby
unlocking new revenue streams. While such a reduction in the minimum
investment amount was previously out of the question for cost reasons, it
is now viable thanks to new technology. However, the growth forecasts for
the robo-advisory market mentioned at the beginning, as well as local client
contacts, show that there is also a clear demand for the asset-management
product in this segment. There was particularly strong interest among the
new generation, technology-savvy clients and high-net-worth clients who
would like to try out the bank first by investing a small amount.
While most robo-advisors with minimum investment amounts of EUR
10,000 or even less concentrate on the retail segment and the offering of
local private banks, similar to H&A, starts on average at EUR 500,000,
there is a clear market niche between EUR 50,000 and EUR 500,000
with very few targeted offers to date (see Sect. 10.1.3). At the same
time, the asset situation in the affluent segment is associated with other
needs and investment opportunities. In this area, it is already possible to
invest in individual securities (in contrast to pure ETFs or funds) without
jeopardizing adequate diversification. As an asset manager for high-net-
worth clients, H&A has been well-aware of this need.
A comparison of the company’s strengths with the market potential and
the competitive situation revealed that H&A’s new digital sales channel
should target the affluent segment, as the segment with the greatest
potential. In addition, internal analyses have shown that this segment plays
a key role for H&A when it comes to attracting the next generation
of private banking and wealth management clients. H&A also realized
that its brand and the expertise of its client advisors count as particular
strengths, potentially giving the bank a competitive edge when acquiring
new clients in this segment (see Sect. 10.1.2). This resulted in the omni-
channel strategy (see details below) with a digital distribution channel
supplemented by a hybrid advisory approach and a seamless transition
to the local distribution network with increasing complexity of financial
circumstances.
Even though the analyses showed that the pure robo-advisory busi-
ness model had a lesser effect on traditional private banking or wealth
management client segments, in-house client surveys carried out at the
same time have shown that the technological demands of clients have risen
considerably in recent years as a result of digitization (with respect to online
banking, among other things).
The first phase of the project therefore focused on modernizing the set-
up of the local sales channel and the needs of the existing client group—for
example, by implementing a new, modern online banking system with real-
time access to asset portfolio (including detailed analysis of asset structure,
e.g. by asset class, country or industry). This included measures such
as the introduction of a WhatsApp news channel and a virtual tour of
the branch offices on the bank’s website. The measures were based on
client and market analysis results as well as a client survey carried out in
advance in local sales. The project prioritization reflected the need for
a modern online banking system for the digital sales channel and the
central requirement of addressing the needs of existing client groups.
In line with the methods of agile project management, H&A relied on
MVP for the introduction of the new online banking, that is, a product
that meets the minimum requirements of its customers. In contrast to
conventional development methods, this approach allowed clients to see
the first technological improvements after just a few months. This basic
version was then equipped with additional features in subsequent releases,
and it is being continuously developed. Thanks to the agile approach, it was
possible to obtain feedback from clients based on MVP and to incorporate
client feedback with respect to ease of use and new requirements into the
prioritization for the next project phases.
The second step consisted in developing a completely new concept for
digital asset management, Zeedin.10 This included all aspects of a market
launch—starting from the product and service range, including pricing
conditions, putting together a team responsible for this and creating the
interface with existing private banking and other departments, developing
the underlying operational processes and the technical platform through
to targeted marketing and communications.
The following three strategic questions played a central role for H&A
in determining the range of services:
• What would an intelligent interaction between a human and a

machine look like?
10 The reasons for naming the service "Zeedin" are discussed in detail in the Sect. 10.2.3.
• What options does H&A have to ensure that its product range meets
the increasing demand for personalization?
• How can H&A provide its customers with personal access in the
digital channel despite the cost pressures?
As a result, the bank opted for a hybrid approach, that is, an intelligent
interaction between human beings and technology. Before we get into
further details, it is important to understand the basic functions and
features of Zeedin: the starting point is Zeedin’s website, which offers
prospective clients a comprehensive overview of the services Zeedin offers.
This is followed by an introduction to the range of investments offered
by Zeedin and information about the related investment process, which is
divided into three steps:
• In the first step, the prospective client is required to complete an

online questionnaire to determine their personal investment profile,
taking into account their investment objectives, risk appetite and
existing financial knowledge.
• On this basis, Zeedin selects an optimal investment proposal for the
prospective client in the second step. Where required, the investor can
further personalize the investment proposal, for example, by deselect-
ing individual asset classes or choosing the investment universe. In
line with the high level of information in local sales, Zeedin’s digital
investment proposal is also very detailed and includes, in addition to
the strategic asset class allocation, a simulation of future performance,
historical performance information, regional and sector allocation and
top portfolio picks.
• In the third step, the client can choose between a fully digital
process using the latest video authentication technology and a semi-
digital variant with authentication in a post office. In both cases, the
necessary personal data is collected and the contract documents based
on this data are made available through Zeedin’s digital application
process.
Zeedin’s product range stands out thanks to the following exclusive

features, state-of-the-art technology and expertise: through Zeedin, clients
who have EUR 50,000 or more to invest can now access H&A’s multi-
Fig. 10.5 Intelligent interaction between people and technology
award-winning11 asset management product, developed over many years,

which until recently was only available to high-net-worth clients with
EUR 500,000 or more to invest. In line with H&A’s mission statement
of “Achieving a return through active management across all liquid asset
classes” (Hauck & Aufhaeuser 2020b), Zeedin’s investment management
approach also involves a combination of strategic asset class allocation
specified by the client online and a tactical asset allocation depending on
the market phase. The tactical asset allocation and the selection of the
securities in which to invest is then carried out by H&A’s investment
management experts based on a structured process. H&A also relies on
Zeedin’s combination of human and computer intelligence with respect
to the ongoing monitoring and adjustment of the portfolio (Fig. 10.5).
In this way, any portfolio adjustments are transparent and can be retraced
by the client at any time, and can also be explained based on fundamental
valuations carried out by our investment managers.
11 Received the Golden Bull—“Asset Management of the Year 2019” award from the
magazines e uro, e uro am Sonntag and Boerse Online, and the transparency in asset
management award—“R&P VV Ausweis” from Roedl & Partner.
In addition, Zeedin stands out through its ability to adapt to the various
requirements of the target client group. This begins with the product
selection: Zeedin offers clients the option to choose initially between three
types of discretionary portfolio management solutions (see Sect. 10.2.2)
depending on the required investment universe. Overall, clients can pick
from a total of over 100 individual investment strategies based on further
configuration options (such as the deselection of individual asset classes or
specification of the equity allocation) in conjunction with the investment
and risk profile derived from the questionnaire. The service range can also
be customized through optional additional services such as a portfolio
review in the first year, provision of monthly macroeconomic analysis or
participation in six-monthly macroeconomic webinars.
In addition to the individual configurability of the range of products
and services and the investment management track record supported by
human expertise and technological leadership, another key advantage of
Zeedin is its hybrid approach to investment advice. Based on offline
sales experience12 and the findings from Sect. 10.1.2, it was clear that
personal contact should not be underestimated with respect to digital
sales channels. Zeedin’s clients and prospective clients have access to a
dedicated customer service team—via email, chat or telephone. In addition,
prospective Zeedin clients have the opportunity to discuss their investment
approach with a client advisor from local private banking before signing a
contract. Here H&A is again relying on modern applications such as a
digital appointment calendar for arranging meetings and co-browsing for
joint, location-independent discussion of the investment approach.
With the resulting positioning and the omni-channel approach involving
Zeedin and local sales, the bank is pursuing several strategic goals:
• a hybrid and transparent approach—a prospective client can, for

example, obtain comprehensive information via the website and draw
up an individual investment proposal using the Zeedin application
route without obligation.
• a customer-specific approach with a range of products and services
tailored to the individual needs and financial circumstances of the
client,
12 Based on interviews with client advisors and analysis of the conventional customer
journey in private banking at H&A.
• a seamless transition between the two sales channels, including an

integrated overview of the company’s website, the design of the
service range and the platform (both channels use the same online
banking).
In principle, the following also applies: the higher the investment volume,
the more individual and personal the range of services. The range of
services offered by H&A’s traditional private banking—from an investment
volume of EUR 500,000—includes additional key elements such as inheri-
tance planning, further personalization of asset management (e.g. through
the option of deselecting individual industries or considering regular
distributions) and investment advice as an alternative to discretionary
portfolio management. In addition, clients with an investment volume of
EUR 500,000 or more are assigned a permanent personal advisor.
10.2.2 Product
As part of the implementation of the digital asset management platform
Zeedin, it was important to reconcile several goals from the strategic
perspective. The main focus, however, was on unlocking the potential of a
new client group in the affluent segment, designed to complement H&A’s
focus on the private banking segment.
Furthermore, it should be based on the proven investment approach
of the portfolio management of the bank, that is, the bank would seek to
develop a complementary digital sales channel rather than an algorithm-
driven selection process. The product range should be transparent in
relation to all costs incurred and provide clients with convenient access
from any location. As H&A wants to achieve a positive alpha factor
compared to the benchmark for its clients through targeted stock pick-
ing, Zeedin should also offer a variant with individual stock selection.
In addition, the bank can look back at a long tradition of ethical and
sustainable investment: in 1995 H&A launched the first and thus the
oldest ethical fund in continental Europe. This expertise should also be
reflected in Zeedin’s product range (Schraner 2011). In order to offer
a financial investment without the relatively high minimum investment
amounts required for the individual security variants from as little as EUR
50,000, the offering was expanded to include an ETF-based solution
(Fig. 10.6). These unit-linked asset management and individual security
solutions should allow clients to deselect mandatory components such as

gold and risk-adjusted investments (certificates with a negative correlation
effect as a portfolio supplement). As already mentioned, transparent and
regulation-compliant pricing was an important aspect (under MiFID II, in
particular, banks are required to provide a much greater level of detail from
2018 and illustrate the resulting reduction in returns). From the client’s
perspective, all relevant factors must be clearly visible and comprehensible.
If possible, this should not give rise to any additional costs. For this reason,
a pricing model has been developed for the client which includes all third-
party expenses, administrative costs, transaction and depositary costs of the
bank as well as any currency conversion expenses. VAT at the applicable rate
is also taken into account in accordance with the requirements of the Price
Indication Ordinance. In addition, some financial instruments used incur
product-related costs (e.g. running costs related to funds and certificates).
These tend to be specified by nearly all market participants, but the level of
detail of this information varies considerably among providers. In Zeedin’s
case, for example, the bank not only specifies the running costs and the
depositary bank fee for funds (these are usually listed in the prospectus
as a total expense ratio, but do not take into account the transaction
costs incurred within a fund). Even though transaction costs for ETFs
are typically low, active funds can still add substantial positions that are
not always disclosed to the investor.13 These are therefore calculated and
disclosed under costs in EUR and as percentage of the investment amount
in accordance with the requirements of MiFID II.
10.2.3 Marketing and Sales

Once the product had been defined at the strategic level and the develop-
ment started at the operational level, the bank had to plan the customer
journey14 and thus the marketing concept, in order to position Zeedin’s
13 Certificates may incur product-related costs when purchased through the issuer, where
the fair value of the product differs from the subscription price. We have also selected and
specified these costs.
14 In this context, the customer journey refers to the entire customer experience with the
product—from the first contact through marketing through examination of the product to
conclusion of the contract and subsequent use of the online banking platform.
Fig. 10.6 Zeedin product range (H&A webpage)

Fig. 10.7 Customer journey with Zeedin
USP on the market.15 The customer journey is divided into six process
steps: marketing, website, investment process/proposal, retargeting,16
onboarding/securities account opening, online banking and other ser-
vices. The online banking platform had already been modernized.17 This
sub-project also included the development of the onboarding process.
Marketing for Zeedin therefore focused on marketing the product in order
to lead potential customers to their own website, which in turn leads them
to the investment section and finally to the creation of an investment
proposal. The next step was to define a lead management process that
would enable systematic retargeting (Fig. 10.7).
The development of the marketing concept began with the detailed
analysis of target groups. Although it is largely determined by the volume
of initial investment, characteristics that are just as independent of the
15 The entire marketing strategy for Zeedin was developed in collaboration with the
Munich-based communications agency Saint Elmo’s; see Annual Report 2018 Hauck &
Aufhaeuser (2018).
16 In this context, retargeting describes marketing measures used to approach a potential
customer again.
17 Zeedin’s onboarding process is technically linked to H&A’s online banking to provide
customers with a consistent product experience.
Fig. 10.8 Analysis of customer needs
assets such as occupation, education or hobbies, play a role when it

comes to a targeted and individual approach. In the preparation of the
so-called personae, the bank used empirical values and studies of the
needs of potential clients. This process has generated four personae, whose
individual lifestyles were outlined: Markus, Laura, Simon and Christiane.18
The personae were then superimposed with one another. Although the
trigger varies from person to person (digital access, convenience, “private
bank” status, attractive terms and “active risk management”), a common
feature quickly became apparent: the needs of the target group were based
on large amounts of unused cash (Hauck & Aufhaeuser 2019a) (Fig. 10.8).
18 These names were chosen at random and represent the personae used internally.
Fig. 10.9 Zeedin logo
The overlap of the target group consists of the fact that the assets remain
unused for a longer period of time and are actually reduced by negative
interest rates and inflation. H&A’s new product should offer an innovative
solution to activate the untapped potential of cash assets to invest based on
individual ideas. This insight should then be reflected in the entire history
of the brand. Firstly, the bank needed to find a suitable name. At a strategic
level, the name had to reflect the bank’s multifaceted approach and the
combination of different values and perspectives. Therefore, the team has
to ensure that the name did not sound too technical, but also human, and
thus reflected the connection of the best of both worlds as the core of a
hybrid identity. A number of expectations were thus attached to the new
product name Zeedin (Fig. 10.9).
The word Zeedin [si:d In] is intentionally based on the English word
“seed in/seeding” and should refer to the planting/growing/sowing
aspect. The term seed or zeed stands for growth and profit, future, spirit
of the time and flexibility, while in stands for intelligence, innovation and
individuality. On this basis, the logo was developed, and the bank added
the subtitle “Investment Intelligence by Hauck & Aufhäuser” to reflect
the client demand for a well-known, exclusive brand, proximity to the
bank and the bank’s own private banking business. This is also reflected
in Zeedin’s corporate design (CD), for example, in the investment process
(Fig. 10.10).
Based on the fundamental elements of the CD, it was necessary to
develop a campaign that would not only draw the attention of the target
clients to the product on launch, but also beyond that, and at the same time
pick them up at their point of pain, which is the unused cash. The messages
“Money is not stupid” and “If you invest intelligently” were developed for
this purpose. Each personae (Markus, Laura, Simon and Christiane) was
assigned an individual motif that transfers the message of the campaign
into their individual environment (Fig. 10.11).
A further focus of Zeedin’s marketing concept was the design and
implementation of the website (Hauck & Aufhaeuser 2020d). In contrast
Fig. 10.10 Layout of Zeedin’s investment process
Fig. 10.11 Visualizations—personae and messages

to other H&A services, this is a sales site that not only enhances the image
of the bank but also aims to motivate potential clients to go through the
investment section and prepare an investment proposal. At this point, the
bank had to decide whether to create a separate website or to integrate it
into the H&A corporate website. For this purpose, all competitors with a
close proximity to a banking house were examined.19 It was noticeable that
none of them had an integrated website, although all providers advertised
their new digital offerings on their respective corporate websites. This was
likely due to organizational reasons, like technical hosting, or strategic
reasons.
For Zeedin, an integrated approach was chosen after weighing all the
advantages (e.g. positive impact of the parent brand H&A on digital
asset management and thus greater trust on the part of clients, potential
linking of marketing activities, integration into the existing website is
associated with lower technical costs) and disadvantages (e.g. proximity to
the parent brand must always be maintained; the marketing presence must
not counteract other messages, reputation risk if digital asset management
does ends in failure), not least in order to highlight the proximity to offline
private banking and the hybrid advisory approach. Today, when website
visitors click on private banking, they can choose between conventional
and digital asset management. The website itself is divided into a total of six
sections: Home, Investment strategy, Expertise, Products & Conditions,
Security and Frequently asked questions (FAQ). An additional “Login”
button allows returning visitors to re-enter the investment process by
entering their login details at the point where they left when they last visited
the website.20,21 The implementation of the Zeedin section on the H&A
website also focused on transparency and information. The costs and terms
& conditions are always just a few clicks away. The individual subpages
19 This refers to banks which, like H&A, have expanded their offline presence by adding a
digital sales channel.
20 Before the prospective client can view their investment proposal, they are required by
law to enter their personal data. Next, the prospective client receives a password from the
investment section, which they can use to login again at any time and change their data.
21 As soon as the prospective client enters their personal data into the Zeedin investment
section, they become part of the lead management process established and developed
specifically for this purpose. This then allows a personalized approach to the prospective
client/client (depending on the status) at any point in the customer lifecycle.
have their own FAQ area, which provides answers to the most important
questions visitors may have (Hauck & Aufhaeuser 2020d).
In addition to the website as the central sales platform, the bank also had
to develop an adequate communications mix for Zeedin. The creation of
high-quality content, that is, content marketing, was the first priority. This
involved the creation of various articles about strategy and the functions
and features of Zeedin, together with a high-quality video.22 Due to
the digital affinity of the target clients, the marketing was focused at the
beginning on targeted online marketing measures: Google, social media,
Amazon, display ads and cross-channel retargeting. These measures were
complemented by a print campaign and earned media (e.g. interviews).
The advantages of the integrated marketing approach quickly became
apparent: The launch of the product on October 26, 2018, was spread
across all H&A’s communication channels, enabling Zeedin to benefit
from the bank’s existing followers.
A further focal point for publicity was Zeedin’s presence at trade fairs
and selected events. The aim here was to find out which trend themes were
of interest to the different personae, for example, “women and finance”. In
addition to the conventional marketing approaches, the corporate network
of the investor Fosun was also involved from the outset.23 For example,
Zeedin’s first customers were offered a discount on a Club Med trip
(Aufhaeuser 2018). Later, this cooperation was also extended to Club Med
customers. Depending on their status, club members of the travel group
have the opportunity to benefit from various discounts when concluding an
asset management contract with Zeedin.24 The integration of Zeedin into
the ecosystem of the investor of H&A Fosun25 thus increases synergies
on both sides, which ultimately benefit the clients of all the companies
involved.
10.2.4 Regulatory Requirements

For the project team at H&A, it was important from the outset to consider
all regulatory requirements and where applicable, ensure that they are
22 Available at (Hauck & Aufhaeuser 2020d).

23 Fosun has also invested in Club Med since 2015: (BBC 2015).
24 See Hauck & Aufhaeuser (2019b).
25 Since 2016 H&A has been majority owned by the international investment group Fosun.
fully complied with. As a financial services institution with a full banking

license, H&A is subject to supervision by the German Federal Financial
Supervisory Authority (BaFin). In this context, the bank is required to
undergo a comprehensive audit yearly conducted by an external and
independent auditing firm. The results of this audit are presented in a
consolidated report and made available to the supervisory authority. The
concept therefore took into account the very broad scope of this audit in
order to meet all documentation requirements.
A digital service not only has to pay attention to national regulatory
requirements but also to the regulatory requirements in the countries of
potential clients. When selling products or services to persons residing in
the USA, for example, the bank would have to comply with the far-reaching
US consumer protection law, which in some cases deviates significantly
from European legal standards. Therefore, in the first step, the bank had
to narrow down the potential target client group. To keep the complexity
of the implementation in the first phase (initial product development)
manageable, the bank decided to focus initially only on clients residing
in Germany. To avoid complex tax structures in the initial phase, the bank
also decided to open only individual client portfolios for the time being.
The bank planned from the outset to make this service available to clients
residing in the European Union in the next phase.
The bank is, in particular, required to comply with MiFID II and the
General Data Protection Regulation (GDPR) of the European Union.26
While the provisions of the MiFID II Directive are not directly binding,
as they are transposed into the respective national laws, GDPR as an EU
regulation is directly binding. In early 2018, the far-reaching changes to
MiFID II came into effect, and they are thus relevant to the project.
Compared to other EU countries, Germany has already anticipated a
number of national requirements in previous years, so that we can speak
here of an even stronger, established, regulated market.
While the analysis in this case study focuses on the investment rec-
ommendation aspect of a given implementation form, it disregards the
far-reaching requirements for actual portfolio management (as there have
not been many innovations that would be particularly worth mentioning
compared to non-digital distribution channels). The requirements for the
26 Regulation (EU) 2016/679 of the European Parliament and of the council of 27 April
2016.
client acceptance process or further topics such as the prevention of money

laundering will not be discussed here.
While the implicit meaning of the term robo-advisor is automated
advice, the definitions still vary. However, from a legal perspective, this
does not constitute investment advice, as this investment service under
the German Banking Act (KWG) explicitly and exclusively targets the
recommendation of a particular financial instrument, such as a security
or a fund (Bundesanstalt für Finanzdienstleistungsaufsicht & Deutsche
Bundesbank 2019, p. 1). However, the recommendation for a financial
service such as portfolio management—as in this case—is not subsumed
under the KWG. Although this does not constitute investment advice
under the relevant rules and regulations, the MiFID II requirements set out
in the Securities Trading Act (WpHG) stipulate that financial institutions
must nevertheless ensure that the recommendation is suitable for the
client. This means that knowledge and experience in securities transactions
must be available in a relevant form or must be conveyed as part of
the advice/recommendation and the investment must be suitable for the
client, taking into account the client’s financial circumstances, investment
objectives and individual risk tolerance (see Article 64 (3) WpHG). As
already mentioned, the bank is regularly checked for compliance with these
requirements as part of the so-called WpHG audit (this strict annual audit
does not exist in other European countries or only to a limited extent),27
so the bank is required to produce the relevant documentation.
According to initial hints from BaFin, it can be assumed that the
recommendations from robo-advisors could in the future fall within the
scope of investment advice (Bundesanstalt für Finanzdienstleistungsauf-
sicht 2017). From the client’s perspective, robo-advisors also provide
a service comparable to investment advice by reviewing very personal
information and offering specific recommendations. At the same time,
clients do not distinguish between the various regulatory requirements.
The Zeedin project team has therefore decided to take a proactive
step and to apply the same high regulatory standards to this new digital
sales channel and the recommendation of a specifically designed asset
management solution as the bank would apply to personal discussions with
discerning private clients at any of the bank’s branches . For this reason,
27 Exceptions include e.g. Austria, where financial institutions must submit to regular
independent audits.
the client questionnaire and investment objectives are also recorded in

the relevant suitability report in accordance with MiFID II, as would be
required when providing investment advice to private clients with respect
to individual financial instruments.
The complete digitalization and automation of this process was only
made possible by the introduction of MiFID II. By the end of 2017,
financial institutions in Germany had to prepare an advisory protocol for
investment advice. This document was intended to provide the client with
a tool to reflect the investment advisory situation and the main arguments
before closing the transaction and, in case of doubt, enable reconstruction
of the conversation afterwards. This process involved recording of every
individual information about the conversation. In addition, precise formal
requirements did not allow for a purely technical recording. For example,
the advisory protocol always had to be signed by hand by the relevant
investment advisor.
With the implementation of MiFID II, this instrument has been
removed from the German WpHG and replaced by the suitability report.
This has much lower formal requirements (e.g. no signature) and rather
than being a protocol of a conversation, it gives substantive reasons for
a recommendation. As a result, it is now also possible to reproduce this
document completely by technical means.
These considerations had a direct influence on the necessary design of
the digital sales channel. In order to be able to ensure that the suitability
report can also be submitted on time, it is mandatory to provide an e-
mail address before the recommendation is displayed to the client. This
means that the document can always be verifiably transmitted as soon as a
recommendation has been made, even if the potential client simply aborts
the process or closes the browser. Since Zeedin offers a very individual
product range, it was also necessary to ensure that any changes to client
information resulted in a new, updated suitability report. Clients can,
for example, log into an existing consultation again and change their
details if necessary, which would result in a modified recommendation.
In this context, it was necessary to define which personal data should
be collected during customer evaluation. The data obtained in this way
can then also be used to address clients correctly and to transfer them
to the onboarding process. Information such as education level and field
of activity of the client, which is collected and taken into account for
investment advice, was designed in such a way that this information can also
constitute mandatory information when the client makes an investment.
All the GDPR requirements had to be taken into account, which meant
that we had to limit ourselves in the aforementioned information to what
is absolutely necessary to make a recommendation. The relevant data
erasure/retention periods were also implemented and the information
about our data protection measures was adjusted accordingly to reflect
information about the new sales channel.
10.2.5 Technical Platform

The development of Zeedin and the online banking backend infrastructure
could not be started with a greenfield approach, but had to be based on
existing and central banking applications. Zeedin’s front-end application,
on the other hand, could be developed and designed independently and
according to the latest insights to enhance the customer experience. The
technical challenge here was the seamless integration of state-of-the-art
technologies and specialized banking service providers or fintechs into the
established banking systems and processes, in particular the core banking
system.28
A key challenge here was to enable external access to the real-time
information of the core banking system (master data, account and securities
transactions). The previously closed core banking system, which supplied
the original online banking system exclusively with data from batch-based
overnight processing, had to be opened up for web access via a secure
infrastructure. This was the only way to meet the requirement for an
automated and direct new client onboarding process when opening a
Zeedin securities account, that is, direct exchange during interaction with
a new customer.
The core banking system, which was a legacy system from the techno-
logical standpoint, was extended to include a modern web services layer, in
this case a REST API. As a result, all relevant business transactions, from
new customer investments to online banking processes such as foreign
transfers, were available to third-party systems in real time. In addition,
this architectural approach allowed sensitive data to be kept strictly within
the bank while integrating robotic and online functionalities from external
service providers.
28 Client accounts and securities accounts are managed in the core banking system, among
other things.
The integration of web services into the core banking system was new
territory for the bank. In addition to the technical security requirements
for setting up the network and firewall infrastructure, the technical aspects
of implementing a sophisticated, multidimensional authorization concept
for online clients also had to be taken into account.
The latest security standards were implemented as part of the infrastruc-
ture development, the effectiveness of which had to be demonstrated by
various penetration tests. Communication between the servers was secured
through the integration of server certificates. In addition, users authen-
ticate themselves with client certificates, which also enable encrypted
data transmission (using TSL).29 As is customary in the industry, the
transactions of bank customers are protected by a two-factor authentication
system, which is being further expanded under PSD2.
To develop the digital onboarding with direct entry of all client mas-
ter data including accounts, securities accounts and online access to
the core banking system, numerous processes had to be automated as
required by law. The KYC process, which is usually run asynchronously and
semi-automatically in conventional and stationary processes, is particularly
worthy of mention here. For the first time, the KYC check had to be
carried out without any disruption with minimal latency as a regulatory
prerequisite for opening an account. To do this, the relevant compliance
application was connected via REST API and following video legitimation,
the checking process was automatically triggered and its results evaluated.
Here too, real-time access via web services was implemented for the first
in the relevant application using real-time API. As a result, not only has
the technology been upgraded but the business process flow of the KYC
check, which is based on it, has also been fundamentally changed.
For the first time, the chosen technical architecture allows online
customers to access H&A’s document management system via SOAP web
services.30 Current and historical client records such as account statements
can be made available in an electronic mailbox at any time without having
to create memory-intensive copies of the documents for online access and
29 TSL stands for transport layer security, and it is a protocol designed to provide
communications security over computer networks.
30 SOAP stands for “Simple Object Access Protocol”, which can be used to exchange data
between systems or to call a remote procedure.
redundant data storage. The documents remain stored at their source at all
times in a manner that ensures data integrity.
In addition, both the Zeedin application and online banking from H&A
offer modern technologies for communication between end customers and
the bank. An external partner provides both a cloud-based co-browsing
solution31 and live chat functionality, which are seamlessly integrated into
the website. As a result, the client communication is “web-enabled”, which
means that investment products requiring explanation can be discussed
with the client on a split screen.
Through the consistent implementation of real-time interfaces and
the elimination of batch processes,32 it is possible to complete client
onboarding processes without any disruption or manual interactions. This
process integrates industry-leading specialist vendors with features such
as KYC33 compliance checks or videoident procedures. For initial asset-
allocation purposes, the bank integrated a service that calculates the
minimum variance portfolio depending on the risk class and based on the
Markowitz (1952) method.
The complex infrastructural development and redesign of Zeedin’s
business processes as well as the upstream introduction of modern online
banking as a digital hub show how to move successfully into a new, strategic
business area through an optimal combination of customer focus, design
thinking,34 agile implementation methods and the targeted use of smart
technologies.
10.3 THE REAL QUESTION IS: “HOW CAN THE BANK

USE THE LATEST TECHNOLOGY TO EXPAND ITS
RANGE OF SERVICES IN THE INTEREST OF THE
CLIENT?”
Many aspects had to be taken into account when answering this ques-
tion, resulting in a multifaceted approach. Although the question of
31 With co-browsing, two or more people can navigate the same website at the same time
and can mark certain places or use visual aids such as arrows to show them to the client.
32 These are process chains that are queued together to form a batch.
33 KYC stands for “Know your customer”, which is a criminal background check in
accordance with Article 8 of the 3rd EU Anti-Money Laundering Directive.
34 Design thinking is a method for developing projects and innovations.
the appropriate technical implementation (especially the “make or buy”

decision35 ) and the integration into an existing bank played a key role and
the implementation required a dedicated project team, the task went far
beyond the introduction of a technical application. Rather, the question
of strategic positioning and the needs of the target client group ran as a
thread through the entire project.
During implementation, it was also necessary to parallelize task packages
and coordinate numerous internal and external experts from IT, marketing,
sales, compliance, and so on. For instance, the application process was com-
pleted in several two-week sprints36 developed in parallel to the account
opening process and the website. In this context, it was important not to
lose sight of the integrated customer journey (see Sect. 10.2.3; Fig. 10.7)
to avoid any process-related breaking points between different applications.
This demanded high levels of communication and teamwork as well as
quick and straightforward decision-making processes. The entrepreneurial
spirit often associated with small and medium-sized businesses, combined
with a high level of personal engagement, a sense of responsibility and a
certain willingness to take risks—that is, a team committing to an idea as if
it was their business—played a central role in the implementation process.
However, this only works if the team is fully trusted by the management
and given the appropriate decision-making authority.
It was also important to see things consistently from the client’s
perspective: Which pain point drives our target client group? Where does
the client expect the latest technology to be deployed to help him manage
his finances? In which areas does the client consider personal interaction
to be essential? Which factors can strengthen trust, in particular, on the
digital customer journey? These and other questions took up a lot of space
alongside the regulatory requirements and technical feasibility. With the
omni-channel strategy and the digital asset management platform Zeedin,
H&A has opted for an approach that gives customers a choice in key
areas. Based on the results of client and market analysis (see Sect. 10.1.2
Client needs and 10.1.3 Market analysis), the bank has given priority to
those parts of the customer journey where clients require information
(e.g. information about the range of services and terms offered by H&A
35 This refers to the decision as to whether a company prefers to produce something itself
(make) or to buy something from a third party (buy).
36 A sprint is one of several cycles within an agile project.
provided on a transparent website, non-binding and advisor-independent

preparation of an investment proposal using Zeedin, daily, location- and
time-independent monitoring of the development of the portfolio using
modern online banking across all end devices). Zeedin offers clients the
option to discuss the investment strategy with a client before signing
a contract, which is the key decision point in the investment process.
Accordingly, the experienced investment team from H&A is responsible
for picking the securities for the portfolio and its day-to-day management,
assisted by state-of-the-art portfolio management systems rather than
automated algorithms. Zeedin clients can thus benefit from the existing
track record of investment experts managing the funds of high-net-worth
clients. Daily client interactions show that historical performance of real
client portfolios is a central element in building trust and thus also a key
factor in the decision-making process.
After market launch, it quickly became clear that any strategy is only
as good as the team that lives by it every day. There were many soft
factors and personal incentives that played a decisive role in the consistent
implementation of the omni-channel approach. In addition, close personal
interaction between the teams was essential for success.
REFERENCES
Accenture. 2019. Omnikanal Vertrieb im Banking. Last Accessed 24 March
2019. https://bankingblog.accenture.com/deutsch/omnikanal-vertrieb-im-
banking?lang=de_DE.
BBC. 2015. China’s Fosun Wins Bidding for Club Med After Two Years. Last
Accessed 20 March 2020. https://www.bbc.com/news/business-31432322.
Buess, Robert, Daniel B. Hildebrand, Dominik Löber, and Jeanette Okwu. 2018.
Rewiring Asset and Wealth Management, P. 8. Munich: Roland Berger GMBH.
Bundesanstalt für Finanzdienstleistungsaufsicht. 2017. Robo-Advice –
Automatisierte Anlageberatung in der Aufsichtspraxis. BaFin Journal. 18–22.
https://www.bafin.de/SharedDocs/Veroeffentlichungen/DE/Fachartikel/
2017/fa_bj_1708_RoboAdvice.html.
Bundesanstalt für Finanzdienstleistungsaufsicht, and Deutsche Bundesbank. 2019.
Gemeinsames Informationsblatt der Bundesanstalt für Finanzdienstleistungsauf-
sicht und der Deutschen Bundesbank zum Tatbestand der Anlageberatung. Last
Accessed 20 March 2020. http://www.bafin.de/SharedDocs/Downloads/
DE/Merkblatt/dl_mb_110513_anlageberatung_neu.pdf .
Deloitte. 2015. Wealth Management and Private Banking – Connecting with
Clients and Reinventing the Proposition Value. https://www2.deloitte.com/
content/dam/Deloitte/ch/Documents/financial-services/ch-en-financial-
services-efma-wealth-management-private-banking-032015.pdf .
Ernst & Young. 2019. EY Banken-barometer 2019–Zeichen der Zeit. Tech. rep.
2019. https://www.ey.com/Publication/vwLUAssets/ey-bankenbarometer-
2019-broschuere/\protect\LY1\textdollarFILE/ey-bankenbarometer-2019-
broschuere.pdf .
European Bank for Financial Services GmbH. 2019. Robo Advice in Deutschland –
Deutschland – Status quo und Entwicklungsperspektiven. ebase. P. 6.
Hauck & Aufhaeuser. 2016. Hauck & Aufhauser Acquires Luxembourg-Based
Sal. Oppenheim Asset Servicing Business from Deutsche Bank. Last Accessed
20 March 2020. https://www.hauck-aufhaeuser.com/en/hauck-aufhaeuser-
acquires-luxembourg-based-sal-oppenheim-asset-servicing-business-from-
deutsche-bank.
Hauck & Aufhaeuser. 2017. Annual Report 2017. Tech. rep. Last Accessed 24
March 2019. https://www.hauck-aufhaeuser.com/fileadmin/Publikationen/
Geschaeftsbericht/180606_HA-Jahresbericht-2017_DE.pdf .
Hauck & Aufhaeuser. 2018. Mit Zeedin 20% Rabatt auf Ihre Club-Med-Reise. Last
Accessed 20 March 2020. https://www.hauckaufhaeuser.com/mit-zeedin-20-
rabatt-auf-ihre-club-med-reise.
Hauck & Aufhaeuser. 2019. Annual Report 2019. Tech. rep. Last Accessed 1
August 2020. https://www.hauck-aufhaeuser.com/fileadmin/Publikationen/
Geschaeftsbericht/GB_ENG_BANK.pdf .
Hauck & Aufhaeuser. 2019. Club Med. Last Accessed 24 March 24 2019. https://
www.hauck-aufhaeuser.com/ad/club-med.
Hauck & Aufhaeuser. 2020. About Us. Last Accessed 20 March 2020. https://
www.hauck-aufhaeuser.com/en/about-us.
Hauck & Aufhaeuser. 2020. Anlagekonzept und Investmentansatz. Last Accessed
20 March 2020. https://www.hauck.aufhaeuser.com/zeedin/anlagekonzept.
Hauck & Aufhaeuser. 2020. Hauck & Aufhaeuser–History. Last Accessed 20
March 2020. https://www.hauck-aufhaeuser.com/en/about-us/company-
profile/history.
Hauck & Aufhaeuser. 2017. Hauck & Aufhauser Successfully Completes Acquisition
of the Luxembourg-based Sal. Oppenheim Companies. Last Accessed 20
March 2020. https://www.hauck-aufhaeuser.com/en/hauck-aufhaeuser-
successfully-completes-acquisition-of-the-luxembourg-based-sal-oppenheim-
companies.
Hauck & Aufhaeuser. 2018. Produktangebot im Bereich Private Banking
um eine Digitale Vermögensverwaltung. Last Accessed 20 March 2020.
https://www.hauck.aufhaeuser.com/hauck-aufhaeuser-erweitert-mit-
zeedin-sein-produktangebot-im-bereich-private-banking-um-eine-digitale-
vermoegensverwaltung.
Hauck & Aufhaeuser. 2020. Zeedin kümmert sich um Ihr Portfolio. Last Accessed
20 March 2020. https://www.hauck-aufhaeuser.com/zeedin.
Linden, Erik, and Andreas Wittmer. 2018. Zukunft Mobilität: Gigatrend Dig-
italisierung und Megatrends der Mobilität. Universität St. Gallen. P. 1,
7, 13, 16. https://www.alexandria.unisg.ch/253291/2/Zukunft%20Mobilit
%C3%A4t%20%20Gigatrend%20Digitalisierung_A5_final.pdf .
Markowitz, Harry M. 1952. “Portfolio Selection”. Journal of Finance 7(1): 77–91.
Mussler, Hanno. 2016. Chinesen Dürfen das Bankhaus Hauck & Aufhäuser
Übernehmen. Last Accessed 20 March 2020. https://www.faz.net/aktuell/
wirtschaft/unternehmen/fosun-darf-das-bankhaus-hauck-aufhaeuser-
uebernehmen-14383183.html.
Nicolaisen, Christian. 2018. Margen im Private Banking Sinken auf Kritisches
Niveau. Last Accessed 20 March 2020. https://www.private-banking-
magazin.de/zeb-studie-2018-margen-im-private-banking-sinken-auf-
kritisches-niveau-1.
Scalable Capital. 2018. Scalable Capital Durchbricht Milliardengrenze. Last
Accessed 20 March 2020. https://de.scalable.capital/presse/scalable-capital-
durchbricht-milliardengrenze.
Schraner, Remy. 2011. Es Fehlt eine Echte Bereinigung. Last Accessed 20
March 2020. https://www.n-tv.de/deleteMe/migration/Es-fehlt-eine-echte-
Bereinigung-article4475046.html.
Statista. 2019. Aktuell genutzte Geldanlagemöglichkeiten. Last Accessed 24 March
2019. https://de.statista.com/statistik/daten/studie/838/umfrage/aktuell-
genutzte-geldanlagemoeglichkeiten/.
Statista. 2019. Entwicklung der Sparquote privater Haushalte seit 1991. Last
Accessed 24 March 2019. https://de.statista.com/statistik/daten/studie/
2699/umfrage/entwicklung-der-sparquote-privater-haushalte-seit-1991.
Statista. 2019. Geldvermögen der Privathaushalte. Statista. Last Accessed
24 March, 2019. https://de.statista.com/statistik/daten/studie/37880/
umfrage/geldvermoegen-der-privathaushalte-in-deutschland/.
Statista. 2019. Robo-Advisors: Deutschland. Last Accessed 20 March 2020.
https://de.statista.com/outlook/337/137/robo-advisors/deutschland.
Zakrzewski, Anna, Tjun Tang, Galina Appell, Renaud Fages, Andrew Hardie,
Nicole Hildebrandt, Michael Kahlich, Martin Mende, Federico Muxi, and
Andre Xavier. 2019. Global Wealth 2019 Reigniting Radical Growth. Tech.
rep. Last Accessed 30 June 2019. Boston Consulting Group. http://image-
src.bcg.com/Images/BCG-Reigniting-Radical-Growth-June-2019_tcm9-
222638.pdf .
PART IV
The Future of Robo-Advisory

CHAPTER 11
The Role of Artificial Intelligence

in Robo-Advisory
Alexander D. Beck
Contents
11.1 Introduction to Artificial Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
11.1.1 Historical Development of AI in the Investment
Management Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
11.1.2 What Can be Expected from AI Today? . . . . . . . . . . . . . . . . . . . . . . . . . . 230
11.2 Value Potential for AI Along the Core Processes of Robo-Advisory . . . . . 231
11.2.1 Customer Data and Collection Strategies . . . . . . . . . . . . . . . . . . . . . . . . . 231
11.2.2 Marketing, Sales, and Service. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
11.2.3 Matching Customers to Portfolios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
11.3 AI Supporting Portfolio Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
11.4 The Way Forward .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
A. D. Beck ()
Quoniam Asset Management, Frankfurt am Main, Germany

228 A. D. BECK
11.1 INTRODUCTION TO ARTIFICIAL INTELLIGENCE

Owing to the highly digital and scalable nature of web-based asset manage-
ment, artificial intelligence (AI) and robo-advisory have a lot of potential
to go hand in hand in manifold ways. However, even though AI is a strong
and powerful technology, it is not by default the right choice for any
emerging business challenge. As fields of application need to be suitable in
certain aspects, they need to be deliberately selected. The following chapter
describes areas in robo-advisory where AI can be used with good chances
of success, but it also points to impediments and limitations.
From the perspective of information processing, AI describes a technol-
ogy that automatically detects patterns in all sorts of data. These patterns
are correlated to the likelihood of certain (future) events to occur. Based on
these likelihoods, the AI system will make decisions and take actions, such
as suggesting a product and selling a stock. The list of examples where AI is
utilized today is long and diverse: It ranges from the detection of icebergs
from satellite images to the prediction of the next best movie to watch.
Among the most well-known applications of AI are Amazon’s product
recommendation service or Apple’s Siri. Hidden from the perception of
retail customers, AI is used for optimizing logistics processes or marketing,
sales, and service events. Regarding robo-advisory, the playing ground for
AI appears quite large, ranging from customer-related topics to portfolio
management. Predicting the stock market with AI is an exciting but chal-
lenging field of application: Historic price information and other market
and non-market data are utilized to predict future price developments
(Domingos 2012; Hastie et al. 2009).
From a bird’s-eye perspective, AI emerged from the field of statistical
and machine learning. It adds some human-like behavior to “pure”
machine learning (which usually takes place on structured, tabular data
only). As such, AI is often referred to when humanoid cognitive functions
are imitated by a computer, such as machine vision and machine-based text
processing. Examples include estimating the meaning of a written text or
determining its sentiment.
With respect to the machine learning part of AI, many different mod-
eling approaches are utilized and continuously further explored. Two of
the most prominent are artificial neural networks and random forests, just
THE ROLE OF ARTIFICIAL INTELLIGENCE IN ROBO-ADVISORY 229
to name a few (Breiman 2001).1 These modeling techniques are powerful

methods for describing real world non-linear and complex phenomena,
and they usually outperform classical modeling techniques such as linear
regression in terms of prediction power. Non-linear phenomena need to
be well-considered in modeling because they are impactful and all around
us. Black swan events and panic reactions at the financial markets are
prominent examples.
From a practitioner’s perspective, AI falls into two categories—general
(or strong) AI and applied (or weak) AI. General AI denotes a technology
which can perform any intellectual task the human brain is capable of. This
type of AI still lies in the future and is today not yet existent. On the other
hand, applied AI denotes a technology that is used to perform specific
tasks. One example of applied AI is the Statoil iceberg challenge (see
kaggle.com), where an AI algorithm is utilized to distinguish icebergs from
ships based on satellite imagery. As it turns out, what is not possible for
the human eye is feasible for a well-trained AI algorithm. As such, applied
AI tends to outperform humans when it comes to such highly specialized
problems. In the course of this chapter, only applied AI is considered.
11.1.1 Historical Development of AI in the Investment Management

Industry
The philosophical roots of AI date back hundreds of years and even
philosophers like Aristotle had early ideas of “mechanical thinking” (Giles
2016). A conference at Dartmouth College in 1956 laid the foundation of
today’s state of AI. AI was a much-hyped field of research in its early years,
and machines on a human intelligence level were expected within a few
tens of years. Obviously, these estimates drastically underestimated the real
academic effort and necessary computational power to achieve this goal. AI
research suffered from two funding crises (also known as “AI—Winters”)
in the late 20th century due to unmet expectations in the development
progress. Today, AI has reached a productive state where its application is
a natural part of many software projects. However, a human-like intelligent
machine is still more science fiction than reality.
1 Cf. also Hinton, who won the Turing Award (Association for Computing Machinery
2018).
230 A. D. BECK
When it comes to financial markets, machine learning, a subfield of AI,

and especially neural networks have quite a history in the field of predicting
the financial markets. Even though it appears very tempting and promising
to throw financial market data at some strong AI model, the reality of
predicting financial markets looks somewhat different: Financial markets
are dominated by randomness to a large extent, and it is notoriously
difficult to squeeze out the predictable part of market movements. After
some hype around applying expert systems (a subfield of AI) to the
financial markets applications, practitioners were rather disappointed by the
performance of this approach. The reasons behind this frustration could be
either of the following: (1) highly complicated systems or (2) the unmet
expectation that AI would render human intellectual input into the system
needless (Durkin 2002). However, two lessons from these days are still
valid today: First, “crap in crap out”, i.e. (financial) data needs to be at its
best quality and deliberately selected and second, overfitting is always just
one tiny step away (overfitting happens when a model is so flexible that
it adjusts to random fluctuations on the training data and cannot discover
the actual correlations any more). However, in retrospect, today’s even
stronger models would have significantly outperformed markets in the past
(Fischer and Krauss 2018). This might lead to the assumption that financial
markets are constantly balanced at the level of what forecasting technology
allows at each time point.
Driven by the latest technological advancements in data availability and
model performance, AI models are again in vogue in the financial industry.
11.1.2 What Can be Expected from AI Today?

Today, the field of AI has strongly evolved, and additional aspects of
modeling markets as well as predicting customer behavior come into
play. On the one hand, the increase in available computational power
at a comparatively low cost allows the development and utilization of
more sophisticated AI models. On the other hand, data is collected more
extensively than ever before and is now widely available, many times even
for free.
Moreover, the application perspective has changed: During the past
decade, the sophisticated scientific component of utilizing AI increasingly
disappeared behind easy-to-use technology frameworks. Today, it is even
possible to assemble AI systems by simply dragging together modules on
a graphical user interface. Due to this development, AI has become easily

accessible to a wide range of users, comparable to a car driver who does
not need to have working knowledge of the car engine in order to drive.
Naturally, this encourages a broader use of AI especially in areas where
returns on investment need to be achieved after short development times.
In the highly digital field of robo-advisory, this leads to a fruitful playing
field for AI. Especially in digital business models and in situations where
decisions are made repeatedly in a high frequency, the game-changing
character of AI will become increasingly visible.
11.2 VALUE POTENTIAL FOR AI ALONG THE CORE

PROCESSES OF ROBO-ADVISORY
In a strongly digital business model such as robo-advisory, process automa-
tion and decision optimization are obvious levers behind business success.
Especially, customer-related processes that repeat themselves frequently are
perfectly suited for applying AI-driven decision support. Such processes are
found along the customer journey in manifold situations. Due to the digital
nature of robo-advisory, each customer journey creates an electronic foot-
print. This footprint is very valuable from a business analytics perspective,
and its proper recording and data management is the foundation of a data-
driven and AI-optimized customer experience.
11.2.1 Customer Data and Collection Strategies

Customer data form the foundation for applying AI to improve customer
processes. They are collected across touchpoints of the customer journey,
and typical customer data sources include the following:
• Master data (annual income, gender, age, pension plans, assets like
real estate, family situation, etc.)
• Marketing interaction data (interaction with past marketing cam-
paigns, reaction to outbound calls, participation in surveys)
• Website/app interaction data (after login, which parts of the website
or the app are especially important to the customer? Where do
customers spend a lot of time?)
• Service data (inbound calls, service requests, type of service requests)
• Product purchases (transactions)
232 A. D. BECK
For companies with data spread across different (legacy) systems, it is an

expensive challenge to map the above-mentioned data sources properly to a
single customer. However, this investment pays out once customer-related
data is used to improve the customer journey. With respect to the technical
aspect of the data collection process, it is important to choose only tools
and services that allow easy access to the stored customer information on
a single-user granularity. Additionally, data quality is of high importance,
and the right quality-assurance measures and futureproof data designs
should be in place as early as possible in the robo-advisor’s life. Another
important aspect of customer data is data privacy, which should be in line
with regulation.
Due to its digital and scalable nature, a robo-advisor is not limited to
a maximum number of customers, and AI-driven customer actions play a
crucial role for the robo-advisor’s success. Obviously, when the number of
AI use cases grows across the marketing, sales, and service landscape, and
the customer base grows as well, increasing computing power is needed.
The technological foundation should not be a limiting factor on this
growth path. Consequently, it is important to build all data and analytics
systems as scalable designs and on futureproof infrastructure (typically
cloud infrastructure).
Generating a data-driven customer view requires an appropriate data-
collection strategy throughout the whole customer journey. For some
business models, this is a serious challenge: For example, a typical problem
of insurance companies is that they do not hear much from their customers
after closing a contract until certain payout events occur. This is a poor
data basis for managing customer processes by AI. Robo-advisory service
providers, on the other hand, have much better possibilities of interacting
with their customers on a regular basis. Easy accessibility, for example, by
providing a smartphone app and offering interesting ways of interacting
with the portfolio management service, creates points of attraction and
makes the service sticky for its users. This paves the way for receiving
important information from clients on a frequent basis. For example, it
is a key task for a robo-advisor to gauge its clients’ risk appetite regularly.
This figure will change over time, and this information enables the robo-
advisory service to adjust individual portfolios accordingly. Thus, the
customer is always provided with an appropriate investment plan and has a
feeling of safety.
To make the customer journey even more attractive, the approach
of gamification can be used to induce enjoyable and playful interaction
elements. This increases the willingness of customers to use the service

and spend some time on things like customer surveys (Seaborn and Fels
2015). For example, clients’ reactions to political events might be used to
gauge their risk sensitivity and awareness in a playful way.
11.2.2 Marketing, Sales, and Service

Marketing, sales, and service are typically the three pillars of customer
relationship management. All three have great potential to be drastically
improved by applying automated decision-making with the help of AI.
In the digital world, and from the robo-advisor’s perspective, these three
pillars benefit from each other and should be strongly intertwined. Con-
sidering the client’s perspective, interconnections lead to a smooth and
enjoyable customer journey. Imagine, for example, a service request that
leads to an offer for a better product experience without reaching out to the
sales department separately. From a technological viewpoint, the intercon-
nection between the three pillars takes place naturally. Think, for example,
of a product recommendation engine which not only suggests products
and generates sales but also offers service touchpoints by suggesting what
previous customers found helpful when purchasing a certain investment
product.
Especially in marketing, AI has a huge potential for increasing the return
on marketing expenditure. Here is an example: Assume a cross-selling
campaign where an incentive is paid to those customers who add a risk
management module to their robo-advisory service. How can AI help in
this scenario? At first sight, this problem appears to be a plain and simple
propensity-scoring problem; however, it is more complicated than that.
Four types of customers need to be distinguished:
1. Customers who would have added the risk-management module

either way because they are risk-averse and seek ways to minimize
risk.
2. Customers who are being convinced by the campaign and purchase
the risk management module only because of the incentive.
3. Customers who receive the campaign and see it as a reminder to
cancel their portfolio.
4. Customers who will not enter the deal no matter what.
234 A. D. BECK
The modeling approach to tackle this challenge is called uplift modeling.

Only those customers who are triggered to buy through the incentive are
selected by the model and all others are left untouched. Typically, returns
on investment in uplift-modeling campaigns turn out to be significantly
larger than with other approaches, because the target is so precisely
defined and in line with monetary goals. Such an uplift model can be
constructed with classifying neural networks or other statistical learning
models. Most important is the data-generation part, on which such a model
is constructed. The data needs to be generated on a pilot campaign carried
out on a randomly drawn subset of the customer base. Only by assuring
randomness in the customer-selection process, a bias-free training data set
can be achieved.
The role of AI in this scenario is to correlate the available customer data
to the uplift. One advantage of this method is that customer data needs to
be explicitly collected in pilot studies for this approach, which means there
is no need to depend on existing customer data. A sufficiently large number
of customers (typically > 1000) needs to be available for this approach.
11.2.2.1 Chatbots
Chatbots are another example of how the customer journey can be
supported with the help of AI. Chatbots have emerged over the past
years (e.g. Microsoft’s Chatbot Service “LUIS” was launched in late 2017
(Hernandez 2017)) and provide an interesting way to utilize written
speech recognition to support customer interactions with AI.
Human-based customer service is typically quite expensive and is diffi-
cult to scale. Web-based services like robo-advisory that target a digitally
affine group of people should consider utilizing digital service channels
and employing their human service reps for qualified cases only, that is,
where an answer can only be found or prepared by a human. Consequently,
chatbots offer a way to scale customer services quite effectively. Chatbots
provide various ways to support customer interactions (Cava 2016):
1. Directly communicating with a customer and answering straight-

forward questions. Example: “Where can I find my YTD portfolio
performance?”
2. Providing links for self-help in case of less well-defined questions and
handing over to a service rep if the answer is unsatisfactory for the
client. Example: “How can I find out which portfolio is right for
me?” -> “Website: Portfolio Finder”
3. Preparing a conversation with a service or sales rep: Example: “Please

provide your name and email address, and you will be directly handed
over to the next free assistant.”
4. Being supportive to a service representative. While the service rep-
resentative talks to a client, she can use a bot service simultaneously,
for example, to quickly find reference cases. Example: “Tax issues for
a German robo-advisor when the customer is a US tax-payer.”
5. Guiding through interactions with multiple levels, for example,
guiding through a process of signing up to a service. Example:
Purchasing an investment product where certain investor criteria
need to be fulfilled.
Chatbots mix the three pillars of marketing, sales, and service because they
are designed to assist the customer in the most beneficial way. During a
chatbot conversation, producing a service ticket might be as equally likely
as suggesting a product or service to satisfy the customer’s need. In the
basic working principle of chatbots, the intent of the customer request is
determined by a specially designed AI algorithm. Based on the discovered
intents, the chatbot dialogue is managed and processed. Usually, chatbots
also possess socializing small talk and response delay modules to make the
conversation more of a real-world experience.
11.2.2.2 Product Recommendations

Product recommendations are an especially exciting field within robo-
advisory because they have a direct impact on the topline of the robo-
advisory service. In a provision-based business model such as portfolio
management services, regular customer actions are an important revenue
driver. At the same time, customers need to be taken care of and be ensured
on a regular basis that their investment is in good hands. This provides a
good starting position for event-based marketing actions and production
recommendations.
Customers who show a high affinity for the happenings in the financial
market are interesting from a marketing perspective. Scheduled financial
events can be used to promote products, for example, to offer a hedging
product in a risk-mitigation scenario.
For example: A customer holds stocks of company XYZ worth USD
10,000 in her portfolio. Quarterly financial results are going to be disclosed
for this company in 10 days, and recent quarterly financial reports regularly
236 A. D. BECK
led to a price jump of 5% the day after the reporting day. Based on these
historic findings, USD 500 are potentially at risk. Potential robo-advisory
action: A put option is being offered to the customer in order to protect
against possible price jumps.
The role of AI in this scenario is manifold. First, customers who show
an actual desire to be approached by event-driven marketing actions
need to be identified. Approaching all customers would quickly lead to
overcommunication, especially when a high number of market events is
considered. Consequently, the brute-force way is typically not an option.
Second, a product-event match must be found. AI plays a crucial role in this
matching process. Typically, the upcoming trigger events’ characteristics
need to be understood correctly, and well-matching products (e.g. a
product for risk mitigation) need to be determined automatically.
11.2.3 Matching Customers to Portfolios

The process of matching customers to portfolios is an important task in
robo-advisory, and AI is a supportive factor here. Portfolios of financial
assets have varying risk and return levels, and they distinguish themselves
in some important characteristics, such as asset class, country/currency
exposures, and in the special case of equity investments, industry sectors
or other stock attributes on which customers might look for investment
opportunities. Income, wealth, and the age of the investor, hence the
remaining investment time until retirement, are other critical factors that
need to be considered for a proper individual portfolio construction.
Consider two examples:
Example 1: A woman in her mid-20s wants to start putting money
in an investment vehicle. She is a supporter of actions against climate
change and has an average income but no significant savings. She lives
in Europe and does not support Asian environmental politics. For her,
a cost-averaging investment approach in a stock portfolio consisting of
green US and European companies might be a smart way to (a) be in line
with her investment preferences and (b) use the higher long-term return
expectations from a stock portfolio. Cost averaging matches her financial
situation well.
Example 2: A woman in her late 50s inherits a significant sum of money.
She mainly wants to protect this money against market losses and inflation,
but she does not want to be exposed to the risk of market fluctuations. To
her, the situation looks much different. A multi-asset portfolio consisting

of (inflation-protected) high-grade bonds to secure the amount of money
mixed with little volatile blue-chip stocks for protecting against inflation
might be an answer.
It is a challenging task for human investment advisors to both provide
the right ways of investing and accompany investors on a regular basis
with their changing investment needs. AI can pre-select a range of well-
matching investment opportunities based on the investment needs and
preferences of each individual customer. Such an AI system must be
skilled in basic investment theory and develop an understanding of which
investment opportunities resonate well with certain types of customers.
Through this, investment suggestions are generated that are a) economi-
cally reasonable and b) appealing to individual customers. One common
approach to learn about a customer’s investment preferences and economic
situation is to guide the person through a set of initial questions. This
can help to determine risk appetite, monetary liquidity, and investment
goals. Such a service typically falls into the category of investment advice
and is a regulated financial service in most jurisdictions. Depending on the
jurisdiction, AI might play a role in this matching process or not.
11.3 AI SUPPORTING PORTFOLIO MANAGEMENT

Predicting where financial markets are developing has kept researchers and
financial markets participants busy for decades. As it turns out, financial
markets tend to show some predictability, especially when today’s strong
analytics capabilities are applied to financial data from a decade ago (Fischer
and Krauss 2018). It is, however, notoriously difficult to tap these sources
of predictability, even for experienced researches.
The roots of quantitative finance date back to the beginning of the
twentieth century, to Louis Bachelier, a French mathematician, who
applied the Brownian motion theory in his PhD thesis to evaluate stock
options. Investigating the random-walk behavior of financial products was
the starting point to develop a mathematical view of financial processes that
led to achievements like the Black–Scholes formula for options pricing,
or the Markowitz and later the Capital Asset Pricing Model (CAPM) to
manage portfolios in a risk-return adjusted manner. By utilizing stochastic
calculus, quantitative finance is used to calculate the fair value of assets
and to estimate future risk and return expectations. One famous approach
238 A. D. BECK
for managing a portfolio of stocks is the three-factor model by Fama and

French (1993). In addition to market risk, as discussed in the CAPM,
market capitalization and book-to-market ratios are considered to produce
buy and sell decisions for each single stock in a portfolio. This approach
received a further modification in 2015 by Fama and French, and a five-
factor model was proposed. This shows that the quantitative approach to
producing profitable financial portfolios is still an actively researched field.
Technical chart analysis is another way how buy and sell signals are
generated with quantitative methods. But this approach should not be
confused with quantitative finance as described above. In a technical chart
analysis, trading rules and trading signals are deducted with the help of
technical indicators. An example is the “death cross”, denoting a situation
when the 50-day moving average price line drops below the 200-day
moving average price line. Situations like these only occur quite rarely (e.g.
48 times since 1928 for the S&P 500 Index). From a statistician’s view-
point, quantitative trading rules, such as that mentioned above, typically
miss statistical significance. This means that these rules do not possess a
sufficiently high degree of reliable forecasting power.
The quantitative finance approach and AI share the idea of predicting
future asset prices with mathematical modeling techniques. The application
of machine learning is in many ways a natural advancement of the ideas
already existing in quantitative finance. Consequently, knowledge about
the underpinnings of quantitative finance like the random-walk theory,
stationarity requirements of time series, stylized facts of financial time
series, to name just a few, are valuable to AI researchers and AI portfolio
managers as well. In contrast to classical quantitative approaches, AI and
machine learning approaches facilitate a much broader field of explaining
variables, such as news comments and sentiments on stocks. Additionally,
they rely on more advanced statistical models (such as neural networks in
contrast to linear regression) to produce price forecasts.
Today, in the era of big data and advanced analytics, it is even more
worthy to investigate this field by applying novel modeling approaches
and tapping new data sources. However, new challenges occur and making
money by simply throwing powerful algorithms at data has turned out to
be not a successful approach in the development of investment strategies.
Besides generating alpha (alpha is a common metric for determining the
profitability of an investment scheme compared to a benchmark, e.g. a
stock market index) with the application of AI, there are other ways in
which AI supports the investment process. Especially, when AI is compared

to a human investor, some points are worth considering.
• Humans are not free from cognitive biases and are unconscientiously
driven by emotions. Behavioral economist and Nobel laureate Daniel
Kahneman has discovered and investigated many of these biases,
showing how irrationally and illogically people behave in certain
situations (Kahneman 2011).
• Decisions driven by “gut feeling” are excluded from the investment
process when AI is involved.
• AI typically stimulates process automation and leads to a cleaner and
better structured information flow, which in turn increases trans-
parency and traceability.
However, AI is by no means the ideal version of an investor and new issues

arise, such as:
• Information may be biased or censored, and an AI algorithm will

consequently only learn what it “sees”. For example, US stock mar-
ket prices experienced an extended period of growth (bull market)
between 2008 and 2017. An AI that is designed on this period has no
understanding of a declining (bear) market.
• An AI algorithm cannot detect correlations in such a way that the
induced causational relation really applies. One example is an image-
classification AI that is trained to classify images showing wolves
and huskies. As it turns out, wolves are often photographed with
a snowy background. Consequently, the AI will classify an animal
as “wolf” if the background of the image is white. This inference
is of course nonsense and in the field of investing money highly
dangerous. However, humans are by no means protected from making
this mistake but at least have the possibility to question their own
decisions on a critical-thinking level.
• Spurious correlations: The strength of an AI algorithm is its capability
to digest huge amounts of data, which also becomes its weakness:
Correlations may appear at random, and this effect becomes stronger
with a growing number of information variables.
240 A. D. BECK
A cooperative working model between the human investor and an AI-

driven suggestion engine looks very reasonable and combines the strengths
of these two approaches. This means, however, that an AI needs to explain
itself and pure black-box methodologies do not fall into this category.
There are well-known approaches today that “open” the black boxes of
AI algorithms and explain in visual terms the ingredients of the decision-
making process (Ribeiro et al. 2016). There is, however, one financial
scenario where black-box applications are acceptable: When assets are
bought and sold on short time scales (e.g. an average asset-holding period
of 5 minutes), a sufficiently high number of profit and loss events are
collected quickly. It is then easy to determine with statistical significance
if an algorithm works in a profitable manner or not. But this scenario is
rather rare for retail-oriented investment vehicles, where rebalancing fre-
quencies are much lower. Even a weekly rebalancing scheme only produces
approximately 52 profit and loss events per year, and it is hard to determine
the statistically significant correctness of the investment decisions. Hence,
pure black-box AI models are not acceptable in this regime.
When developing investment strategies with a multi-parameter AI sys-
tem, brute-force backtesting (changing parameters until a backtest looks
satisfying) is an often applied but certain way to fail. Given the manifold
ways for modeling data and additionally considering the many different
data sources, some of these combinations will just by coincidence show
high profit numbers in a backtest. An ex-post justification of why the best-
performing algorithm is also economically reasonable is typically driven by
the desire to find supportive explanations and consequently not a reliable
way for discovering profitable investment strategies. Hence, some rules of
thumb are helpful to prevent this from happening:
1. Start with an investment story that sounds appealing and economi-

cally reasonable.
2. Choose zero-parameter systems. These are systems that find their
optimal settings by themselves, for example, by parameter optimiza-
tion. This is helpful because it prevents the strategy developer from
playing with the parameters until a good-looking backtest strategy is
found.
3. Choose your data wisely. Have an idea why you add a variable to
the system and be prepared to explain its behavior from an economic
perspective.
4. Always put aside some of the data before the modeling process starts,
and these are only unboxed when a satisfying investment model is
found.
Apart from finding profitable investment strategies with the help of AI, it
might also be used to mitigate risk or reach other valuable goals, such as
constructing green portfolios or portfolios that are free from child labor.
One key element for achieving these other goals is by tapping alternative
data sources, such as text data. Relevant texts can be professional news, as
well as information from Twitter or trader forums that have a less curated
character.
One stock market example is to blacklist or even short-sell stocks of
companies with a bad news sentiment and double down on those with
an extraordinary good news sentiment. In a first step, one or more news
sources need to be tapped and relevant news articles are extracted. A text
sentiment algorithm is used to process these texts, and a sentiment time
series is constructed for each stock. Based on the idea that a bad sentiment
correlates with a bad company outlook, stocks are either blacklisted or
purchased.
Such a sentiment-driven approach can be combined with purchasing
decisions from an AI model that generates trading signals based on pure
market data. In such a scenario, it would play the role of a risk-mitigating
overlay, where stocks with a bad sentiment are removed from the portfolio
of stocks. A wide range of data providers promise to deliver such signals in
a ready-to-digest format.
Forecasting the stock market is an ongoing field of research for both
academia and financial services companies. These research efforts are driven
by a constant stream of technological advancements and the desire to find
new ways to outperform other market participants.
11.4 THE WAY FORWARD

AI systems are developing rapidly with no end in sight. These innovations
are strongly pushed by billion-dollar research investments from software
companies like Microsoft, Google, SAP, and more. This is especially true
for systems that have strong automation and process digitalization aspects,
such as chatbots.
242 A. D. BECK
Disruptive innovation through AI will be mostly noticeable on the cus-

tomer end. Here it supports customer interactions, onboarding processes,
and service communications to run smoothly, quickly, and with a clear
customer focus. AI will be a key driver in scaling customer processes and
services and will have a dominating impact on the top and bottom line of
a robo-advisory service.
Even though highly interesting from an intellectual perspective, invest-
ment portfolios run by AI will be more of a cost versus benefit discussion
and less of an alpha-generation question. Hence, some investment deci-
sions and processes may well be entirely replaced by AI systems, given that
the benefit of introducing these systems outweigh the significant cost of
installing them. However, one question that has been around for decades
is how safe consumers feel when their money is managed by algorithms
instead of humans. This is certainly a question of time, as people feel safe in
aircrafts without questioning the auto pilot. AI systems will most certainly
not replace skilled professionals but rather aid in their work.
For the highly digital business model of robo-advisory, having AI at the
core of the strategic development is a key success ingredient. This is true
not only from a customer satisfaction perspective but also from a revenue
and operational cost perspective. Especially for early-stage robo-advisory
companies, a structured and scalable data strategy is a key component for
a later profitable utilization of AI.
The bottom line is that robo-advisory is a highly attractive playing field
for AI systems. The strongest impact of AI systems will be felt on the
customer-management side and much less on the asset-management side.
REFERENCES
Association for Computing Machinery. 2018. Fathers of the Deep Learning Revolu-
tion Receive ACM A.M. Turing Award. Last Accessed 20 March 2020. https://
awards.acm.org/about/2018-turing.
Breiman, Leo. 2001. “Random Forests”. Machine Learning 45(1): 5–32. https://
doi.org/10.1023/A:1010933404324.
della Cava, Marco. 2016. Microsoft CEO Nadella: ‘Bots are the new apps’. Last
Accessed 13 September 2019. https://eu.usatoday.com/story/tech/news/
2016/3/3/microsof-ceo-nadella-bots-new-apps/82431672/.
Domingos, Pedro. 2012. “A Few Useful Things to Know About Machine Learn-
ing”. Communication of the ACM 55(10): 78–87. https://doi.org/10.1145/
2347736.2347755.
Durkin, John. 2002. “History and Applications”. In Expert Systems, ed. Cornelius
T. Leondes, 1–22. New York: Academic Press. https://doi.org/10.1016/
B978-012443880-4/50045-4.
Fama, Eugene F., and Kenneth R. French. 1993. “Common Risk Factors in the
Returns on Stocks and Bonds”. In Journal of Financial Economics 33(1): 3–56.
https://doi.org/10.1016/0304-405X(93)90023-5.
Fischer, Thomas, and Christopher Krauss. “Deep Learning with Long Short-
Term Memory Networks for Financial Market Predictions”. European Journal
of Operational Research 270(2): 654–669. https://doi.org/10.1016/j.ejor.
2017.11.054.
Giles, Timothy D. 2016. “Aristotle Writing Science: An Application of His
Theory”. Journal of Technical Writing and Communication 46(1): 83– 104.
https://doi.org/10.1177/0047281615600633
Hastie, Trevor, Robert Tibshirani, and Jerome Friedman. The Elements of Statistical
Learning: Data Mining, Inference and Prediction. Springer Series in Statistics.
New York: Springer. https://doi.org/10.1007/978-0-387-84858-7.
Hernandez, Pedro. 2017. Microsoft Officially Releases Azure Chatbot Service.
Last Accessed 13 September 2019. https://www.eweek.com/enterprise-apps/
microsoft-officially-launches-ai-bot-creation-services-for-developers.
Kahneman, Daniel. 2011. Thinking, Fast and Slow. New York: Farrar Straus and
Giroux.
Ribeiro, Marco T., Sameer Singh, and Carlos Guestrin. 2016. “ “Why Should I
Trust You?”: Explaining the Predictions of Any Classifier”. Machine Learning
97–101. https://doi.org/10.18653/v1/N16-3020.
Seaborn, Katie, and Deborah I. Fels. 2015. “Gamification in Theory and Action:
A Survey”. International Journal of Human-Computer Studies 74: 14–31.
https://doi.org/10.1016/j.ijhcs.2014.09.006.
CHAPTER 12
What Role Do Social Media Play

for Robo-Advisors?
Ana-Maria Climescu
Contents
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
12.2 Social Media vs. Robo-Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
12.3 Robo-Advisor Feat. Social Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
12.4 Social Advisory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
12.4.1 Social Media as Marketing and Sales Support . . . . . . . . . . . . . . . . . . . . 252
12.4.2 Social Media as an Alternative to Robo-Advisory . . . . . . . . . . . . . . . . 253
12.4.3 Social Media as Potential Cooperation Partner for
Robo-Advisors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
A.-M. Climescu ()


246 A.-M. CLIMESCU
12.1 INTRODUCTION
Looking at the current number of users, it becomes clear that social media1
today play an important role both privately and professionally. Their rise
is mainly attributable to the increasing use of the internet in general
(Statista 2019a) and smartphones and other mobile devices in particular
(Statista 2019b). In January 2019 Facebook (FB) had approximately 2.27
billion active users per month worldwide, making it the largest social
network in the world (Statista 2019c). Second place goes to YouTube with
1.9 billion users, which is owned by Google (Statista 2019c). WhatsApp
follows in third place with 1.5 billion active users per month (Statista
2019c). The overall distribution of worldwide use of social media channels
is summarized in Fig. 12.1: This presentation of worldwide user figures
shows the relevance of social media in people’s everyday lives. Nevertheless,
there are of course large differences between individual countries. In the
USA, for example, Facebook (68% usage) and YouTube (73%) are the
leading social media in 2018 (Sandmeier 2018). In addition, Instagram is
growing rapidly in the USA, while Facebook user numbers are stagnating
(Sandmeier 2018). In Asia, on the other hand, WeChat and QQ dominate
the market (Mao 2018). In Germany, Facebook (31%) Instagram (15%),
and Snapchat (9%) are again in the lead in 2018 (ARD & ZDF 2018). The
age structure is also interesting here. For example, Instagram and Snapchat
are preferred in the 14–19 age group, while Facebook is increasingly used
in the 20–29 age group (ARD & ZDF 2018).
This high number of users and frequency has led to social media already
being seen as an established marketing instrument—not least because the
business models are largely dependent on advertising revenues. And so
almost every company today is expected to have social media channels at its
disposal in order to win new customers and retain those already acquired
(Kontor4 2019). In the context of robo-advisory, three relevant questions
can be derived for this section:
1. Have social media favored the emergence of robo-advisors?

2. What role do social media in general play for robo-advisors?
3. How can social media support robo-advisors?
1 Social media in this context are understood as communication platforms such as Face-
book, Instagram, Twitter, YouTube, LinkedIn, Xing, and Pinterest, which enable users to
share content digitally and interact with other users via an online application/app. In addition
there are messenger services like WhatsApp, Snapchat, FB Messenger, Skype, and iMessage.
This definition is based on the lexicon of Gruenderszene, but has been extended to include
those channels that are relevant to this section (Gründerszene 2019).
WHAT ROLE DO SOCIAL MEDIA PLAY FOR ROBO-ADVISORS? 247
Fig. 12.1 Active users of social media worldwide (Own illustration based on
Kontor4 2019)
12.2 SOCIAL MEDIA VS . ROBO-ADVISOR

To answer the first question, we will first take a look at the evolution of
social media in comparison to robo-advisors to find out whether there
has been temporal parallelism or interdependence. Although there were
already various platforms in the 1990s that sought to connect people and
enable the direct exchange of messages, it was LinkedIn that laid the
foundation for today’s social and digital media landscape in 2003 (Kroker
2018). One year later, in 2004, Facebook came onto the market (Kroker
2018). This was followed by YouTube in 2005 and MySpace and Twitter
in 2006 (Kroker 2018). In 2008 FB launched its Messenger. Instagram
did not follow until 2010 (Kroker 2018). Since then, the number of users
has been rising constantly, as already mentioned (Kontor4 2019). The
comparison with the evolution of robo-advisory is intriguing.
The chronological development of robo-advisors has been traced in a
study by Becchi et al. (2018, p. 3). Accordingly, their emergence can be
traced back to 2008, with the first services from Wealthfront and Better-
ment, among others. In 2010, Bank of America and Merrill Lynch Merrill
248 A.-M. CLIMESCU
Edge introduced a robo-advisor to optimize investments for clients. Two

years later, Australia and New Zealand Banking Group Limited (ANZ)
began using artificial intelligence (AI) technology to better understand
the customer. In 2015, Blackrock acquired a digital advisory platform to
improve investment decisions. Meanwhile, in 2017, Betterment was one
of the first robo-advisors to announce that it would develop its digital
business through client advisors in the direction of a hybrid business model.
Robo-advisors have therefore experienced a massive growth in the period
from 2008 to today, which is also reflected in searches for this term
on the net.2 The first robo-advisor followed about five years after the
first social medium. There is therefore no immediate chronological par-
allelism. But the close chronological succession in which the emergence of
robo-advisory follows social media suggests an interdependence. Indeed,
between 2005 and 2010 in general, there was a major technological
advance that affected many industries. The first iPhone (2007) is one of
many technological developments that originated in the same time span
and have digitized the lives of a large number of people. With smartphones
and social media, people’s user behavior began to change step by step. This
was accompanied by an increased desire for digital, location-independent
services and access to products. The new technological possibilities have
had an extreme impact on user behavior and thus on customer wishes;
and have opened the door to new digital business models (Ernst & Young
2018, p. 42).
In the same period (2008), however, there was another significant
change for banking services that has also shaped the behavior of bank
customers worldwide: the financial crisis. In addition to heavy economic
losses, it was above all the loss of confidence in existing institutions,
accompanied by increasing cost pressure on financial service providers,
that paved the way for the emergence of new alternatives (Ernst & Young
2018, p. 42). Against this background, banks were forced to address on
the one hand the options for automating their own processes and on
the other hand the greater skepticism on the part of customers, coupled
with new customer requirements. This put existing business models to the
test, while some existing financial service providers, together with some
new ones, seized the opportunity offered by the latest technologies (Vogel
2 An analysis of the terms fintech (as a general umbrella term) and robo-advisor using Google
Trends shows an increase in public interest from 2014 (in the US and Germany).
2016). The emergence of robo-advisors and the digitalization of existing

banking processes, such as creating an investment proposal, is therefore
due to three factors: the financial crisis, technological opportunities and the
digital affinity of customers. The last of these factors was promoted by social
media, which, along with e-commerce, have the greatest direct influence on
day-to-day user behavior. In addition, they were used by the robo-advisory
market as a new digital marketing platform that can accompany users along
their entire customer journey—from acquisition to after-sales service.
12.3 ROBO-ADVISOR FEAT. SOCIAL MEDIA

Social media is much more than a simple extension of marketing funnels
and has a significant influence on the business models of many companies.
Particularly after the financial crisis, social media offered a place for those
affected and interested users to interact, which created a whole new level of
transparency (Planung & Analyse 2011). This has opened up new opportu-
nities for digital business models such as robo-advisors. In concrete terms,
four major areas of influence of social media can be identified for robo-
advisors: Marketer/enabler (Becchi et al. 2018, p. 4), communication
channel, data source and (on a broader level) even strategy. According
to a study by EY, social media platforms are considered “enablers”, that
is, pioneers, of robo-advisors which firstly serve to cultivate customer
relationships and secondly increase the reach of the brand (Becchi et al.
2018, p. 4). They provide a space in which corporate content can be made
publicly available. In addition, they function as a platform on which an open
exchange of experience can take place and a relevant community can be
established.3 Industry-relevant influencers can help strengthen brand trust
and increase reach within the community (Mindruta 2016). Furthermore,
providers can communicate directly with users, for example, via chat, which
promotes customer acquisition and retention (Becchi et al. 2018, p. 5).
Social media thus also function as a communication channel and offer
direct access to the people behind the digital service, which is particularly
valuable for digital investment models without constant advice. Direct
contact with users and customers is also elementary from the company’s
3 The term “fintech marketing” was coined in this context, which describes the marketing
of digital financial service providers and banks with the help of social media. The goal: mutual
interaction through contributions and thus support in building a community (Finleap 2017).
250 A.-M. CLIMESCU
Fig. 12.2 Areas of influence of social media on robo-advisors (authors’ own

image)
point of view: feedback and evaluations from users can reveal valuable
improvement potential in their own products (Fig. 12.2).
Finally, social media offers opportunities for effective (social media
enables accurate tracking) and personalized marketing (user behavior on
social media allows conclusions to be drawn about other needs of a
(potential) customer) and thus to attract customers and increase revenue
(Becchi et al. 2018, p. 4). Advertising created on the basis of personal
characteristics increases the relevance of the offer for the customer and
reduces wastage on the part of the provider.4 If you look beyond robo-
4 Thus, FB, Instagram, YouTube and LinkedIn, for example, enable personalized adver-
tising to be displayed that takes factors such as demographics, purchasing preferences and
friends into account, so that advertising banners are shown only to those users who might
be interested in the respective company or product/service according to their personal
characteristics.
advisors, you can already find impressive examples of this. For instance,
American Express has given its customers the opportunity to link their
Facebook or Foursquare accounts to their Amex cards so that they can be
provided with personal offers based on data from these accounts (Hamilton
2016). As a result, social media no longer simply offer the possibility of
distributing content, but can also act as an interface with the producer
through the use of AI in order to influence the buying behavior of
customers (Finleap 2017). The primary focus is on big data evaluations:
social media are a rich source for this, as FB (in possession of a European
banking license) has already proven impressively. Today users voluntarily
provide companies like FB with a great deal of personal information that
can be used for targeted marketing/personalized ads if analyzed correctly.
However, the use of AI is particularly promising in this context, as it
could help robo-advisors, among others, to exploit the full potential of
social media beyond their traditional marketing function. What this means
in concrete terms will be explained briefly in the following, using robo-
advisors as an example. Until now, robo-advisors have exclusively used
the existing technology and functions that social media offer them, for
example, as a publication platform, an appointment agreement tool or a
chatbot. The turning point in terms of efficiency and customer satisfaction
would now come if a provider were able to use AI to collect and process the
personal data of users of its own social media channels and forward it in
real time to the product platform, which then adapts itself automatically
to customer needs. For example, market trends for certain investment
products would lead to an immediate adjustment of the product portfolio.
Decisions about costs or marketing presence would then no longer be made
by the company itself during ongoing operations, but would be handled
by an integrated AI platform. Through the use of AI, social media can thus
point the way for the entire strategy of robo-advisors in future (FinTech
Futures 2018).
12.4 SOCIAL ADVISORY

In general, some interesting approaches for the current, but above all for
the future significance of social media for robo-advisory can now be derived
from the previous discussion. In relation to the question as to whether
social media have favored the emergence of robo-advisors and what their
252 A.-M. CLIMESCU
overall role is for the robo-advisory market, the following conclusions can
be drawn:
1. The use of social media continues to grow worldwide, with regional

and continental differences in relation to preferred media. While
Facebook, YouTube and Instagram dominate the market in the USA,
Facebook, Instagram and Snapchat lead the market in terms of user
numbers in Germany, for example. For the robo-advisor market, this
means that media must be evaluated on a regional basis with regard
to their use. The age of the target group for the respective product
also plays a crucial role.
2. Looking at the historical development of social media in comparison
to that of robo-advisors, it is also evident that social media are
about five years ahead of the first robo-advisor. This goes hand in
hand with increasing digital affinity, with digital consumption and
overall attraction to digital technology now increasingly taking hold
in people’s lives. This different world is increasingly reflected in
related customer demand and digital expectations of companies in
general and financial services in particular.
3. In relation to the role of social media for robo-advisors, this results
in a total of four central areas of application: marketer/enabler,
communication channel, data source and strategy. Social media
enable robo-advisor providers to maintain their customer relation-
ships through direct contact and increase the reach of their brand. As
a communication channel, social media can also enable personalized
offers and targeted advertising to be provided on the basis of the
data collected. This not only provides valuable insights for the
strategy (products in demand and pricing), but also potentially allows
automated adjustments to the product platform via AI.
Against this background, the following points can be considered going

forward with regard to the scope for the use of social media to support
robo-advisory models:
12.4.1 Social Media as Marketing and Sales Support

Social media can serve as an interface between the end consumer and the
digital financial services provider. In concrete terms, this means that they
are used as a platform to communicate (one-to-one-to-many)—that is, via

chat, telephone, in public or in private. For example, FB already offers
a scheduling function that companies can use to arrange meetings with
potential customers. For many e-commerce providers, initiating business
via social media has already become part of everyday life.
12.4.2 Social Media as an Alternative to Robo-Advisory

And this potential is now leading to a consideration of why social media
should not offer more financial services. Not only do they have the
necessary data, such as place of residence, education or profession, to
present tailor-made offers to customers, but they also have the necessary
resources and know-how to evaluate and use these data volumes. Since
2016 Facebook has had a license for money services with the Irish Central
Bank (Degel 2016). It was not until mid-2019 that the group announced
its plans for its new payment service libra. The new platform, in which
PayPal and Visa, among others, would like to participate, will be based on
blockchain technology (Schuler 2019). It has also been reported on occa-
sion that Snapchat is working on a robo-advisor (Wolff-Mann 2016). The
reasons for the failure to exploit this potential to date probably lie in the
regulations for financial service providers and in the complexity of banking
products. WeChat, on the other hand, has already implemented a payment
function (Taylor 2019). However, more complex bank transactions such
as financial investments are not possible in China either.
12.4.3 Social Media as Potential Cooperation Partner for

Robo-Advisors
Nevertheless, these tendencies and attempts on the part of social media
to enter the banking business lead to the conclusion that a cooperation
between social media and robo-advisors, that is, banking services in the
field of investment, could certainly be a strategic consideration. The
technical course for this has already largely been set by the social media
and modern robo-advisors, in the form of social-advisors. For both business
models, the next step is to evaluate the potential of such a cooperation from
a business and strategic point of view using concrete case studies.
254 A.-M. CLIMESCU
REFERENCES
ARD & ZDF. 2018. Onlinecommunities – Nutzung von Onlinecommunitys
2018. Last Accessed: 7 September 2019. http://www.ard-zdf-onlinestudie.de/
whatsapponlinecommunities/.
Becchi, Scott M., Ugur Hamaloglu, Taroon Aggarwal, and Samit Panchal. 2018.
The Evolution of Robo-advisors and Advisor 2.0 Model: The Future of Investment
Management and Financial Advisory. Ernst & Young Global Limited.
Degel, Joachim. 2016. Facebook Hat Jetzt eine E-Bank-Lizenz für die EU. Last
Accessed 17 March 2019. http://www.bgk-p.de/de/blog/posts/Facebook-
hat-jetzt-eine-E-Bank-Lizenz-fuer-die-EU.php.
Ernst & Young. 2018. EY Banken-barometer 2018 – 10 Jahre nach der Finanzkrise
– Aufbruchstimmung? Tech. rep. https://www.de.ey.com/Publication/
vwLUAssets/ey-bankenbarometer-mai-2018/\protect\LY1\textdollarFILE/
ey-bankenbarometer-2018.pdf .
Finleap. 2017. Fintech Trends: Social Media. Last Accessed 17 March 2019.
https://www.finleap.com/insights/fintech-trends-social-media/.
FinTech Futures. 2018. Social Media and Fintech – What Kind of Relationship? Last
Accessed 17 March 2019. https://www.bankingtech.com/2018/05/social-
media-and-fintech-what-kind-of-relationship/.
Gründerszene. 2019. Social Network. Last Accessed 17 March 2019. https://www.
gruenderszene.de/lexikon/begriffe/social-network?interstitial.
Hamilton, Josch. 2016. How Social Media is Changing Fintech and Other Startups.
Last Accessed 17 March 2019. https://cloudtweaks.com/2016/12/social-
media-changing-fintech-startups/.
Kontor4. 2019. Social Media 2019: Aktuelle Nutzerzahlen. Last Accessed 17 March
2019. https://www.kontor4.de/beitrag/aktuelle-social-media-nutzerzahlen.
html.
Kroker, Michael. 2018. Die Social-Media-Zeitleiste: Von der LinkedIn-Gründung
2002 bis zum IGTVStart 2018. Last Accessed 17 March 2019. https://
blog.wiwo.de/look-at-it/2018/07/16/die-social-media-zeitleiste-von-der-
linkedin-gruendung-2002-bis-zum-igtv-start-2018/.
Mao, Bianca R. 2018. Asian Social Media: 5 Networks You Should Know About. Last
Accessed 7 September 2019. https://www.synthesio.com/blog/asian-social-
media/.
Mindruta, Ruxandra. 2016. Was ist das Geheimnis? Eine Einführung in das Influ-
encer Marketing – Markenreichweite erhöhen. Glaubwürdigkeit verstärken. Viele
Euros einsparen. Dies sind nur einige Vorteile von effizienten Influencer Marketing
Strategien. Last Accessed 17 March 2019. https://www.brandwatch.com/de/
blog/was-ist-das-geheimnis-eine-einfuehrung-in-das-influencer-marketing/.
Planung & Analyse. 2011. Social Media thematisiert die Finanzkrise. https://
www.horizont.net/planung-analyse/nachrichten/Social-Media-thematisiert-
die-Finanzkrise-148096.
Sandmeier, Michelle. 2018. Pew Research Center: Social Media in den USA 2018.
Last Accessed 7 September 2019. https://bernet.ch/blog/2018/03/08/
pew-research-center-social-media-in-den-usa-2018/.
Schuler, Marcus. 2019. Holt Facebook die Kryptowährungen aus der Nische?
Last Accessed 17 March 2019. https://www.deutschlandfunk.de/neues-
digitalgeld-libra-holt-facebook-die-kryptowaehrungen.684.de.html?dram:
article_id=452038.
Statista. 2019. Anzahl der Internetnutzer weltweit in den Jahren 2005 bis 2017
sowie eine Schätzung für das Jahr 2018. Last Accessed 17 March 2019.
https://de.statista.com/statistik/daten/studie/805920/umfrage/anzahl-
der-internetnutzer-weltweit/.
Statista. 2019. Prognose zum Anteil der Smartphone-Nutzer an den
Mobiltelefonnutzern weltweit von 2014 bis 2021. Last Accessed 17 March
2019. https://de.statista.com/statistik/daten/studie/556616/umfrage/
prognose-zum-anteil-der-smartphone-nutzer-an-den-mobiltelefonnutzern-
weltweit/.
Statista. 2019. Ranking der größten sozialen Netzwerke und Messenger nach der
Anzahl der monatlich aktiven Nutzer (MAU) im Januar 2019. Last Accessed
17 March 2019. https://de.statista.com/statistik/daten/studie/181086/
umfrage/die-weltweit-groessten-social-networks-nach-anzahl-der-user/.
Taylor, Twain. 2019. Alipay and Wechat Pay: The Quest for a Cashless China. Last
Accessed 17 March 2019. http://techgenix.com/alipay-and-wechat-pay/.
Vogel, Michaela. 2016. Social Media-Analyse zur FinTech-Branche - In unserer
Social-Media-Analyse warfen wir einen Blick auf die Online-Gespräche rund um
FinTechThemen und Unternehmen. Last Accessed 17 March 2019. https://
www.brandwatch.com/de/blog/social-media-analyse-zur-fintech-branche/.
Wolff-Mann, Etan. 2016. Snapchat Wants to Be Your Next Financial Adviser. Last
Accessed 7 September 2019. https://money.com/snapchat-robo-financial-
adviser-investing/.
CHAPTER 13
Success Factors for Robo-Advisory: Now

and Then
Madeleine Sander
Contents
13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
13.2 Growth Factors Identified . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
13.2.1 No. 1: Overall Market Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
13.2.2 No. 2: Replacing Existing Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
13.2.3 No. 3: Attracting Previously “Uninvested” People to the
Financial Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
13.3 Success Factors for Achieving the Identified Potentials and
Increasing the Market Penetration of Robo-Advisors . . . . . . . . . . . . . . . . . . . . 265
13.3.1 Extending the Reach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
13.3.2 Improving Financial Literacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
13.3.3 Greater Transparency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
13.3.4 Targeting Women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
13.3.5 Hybrid Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
13.3.6 Expanding and Improving the Product Range . . . . . . . . . . . . . . . . . . . 267
13.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
M. Sander ()

258 M. SANDER
13.1 INTRODUCTION
After reading all the implementation details, facts, and best practice in the
field of robo-advisory—what are the key takeaways to make robos a success
story? What are the hurdles that robo-advisory will need to overcome in
order to succeed? And what are the crucial benchmarks? To answer these
questions, it is worth looking first at the US and thus at the roots of the
robo-advisory market. The first robo-advisor, the fintech Betterment, was
launched in the US in 2008 at the climax of the financial crisis. The robo-
advisory market in the US not only has the longest history but is also
currently the global leader in terms of assets under management (AuM); for
details on market data see Chap. 1.4. Although the robo-advisory market is
still largely in its infancy, the amount of empirical data it has at its disposal
is unique on a global scale.
In 2018, the top five robo-advisors accounted for an estimated 70% of
assets under management in the US (Friedberg 2018). While Betterment
ranked third in terms of assets under management (AUM) in 2018, asset
managers and online brokers such as Vanguard and Charles Schwab topped
the list with “Vanguard Personal Advisor Services” on first place and
“Schwab Intelligent Portfolios” on second place (Friedberg 2018). The
involvement of these established financial institutions has made a significant
contribution to growth. This is mainly due to the high brand recognition,
the trust placed in the brand, and leveraging the cross-selling potential of
existing clients. These are the two key factors that give traditional asset
managers a competitive edge over fintechs, which are the actual inventors
of robo-advisory.
Similar patterns are emerging on the German market. In 2018, the
fintech Scalable, in partnership with the direct bank ING Diba, was a clear
market leader with a market share of approx. 25% (for more details on
market data please refer to Chap. 1.4). Furthermore, the current trend
of mergers between established providers and the new robo-advisors will
continue; the latest examples include Deutsche Bank with Robin, as well as
Hauck & Aufhaeuser Privatbankiers with Zeedin. At the same time, the first
signs of market consolidation started to emerge in 2018 with Werthstein
discontinuing business operations (Robo-Advisor Portal 2018) and the
end of cooperation between Haspa and the fintech Investify (Brummer
2018), as well as the takeover of Vaamo by Moneyfarm (Robo-Advisor
Portal 2018). On first glance, the underlying reasons seemed to be quite
diverse. In the case of Werthstein, sales fell short of expectations (l.c.),
SUCCESS FACTORS FOR ROBO-ADVISORY: NOW AND THEN 259
while Moneyfarm and Vaamo intend to use the merger to set up an online
asset manager operating throughout Europe (IT Finanzmagazin 2018). In
contrast, the Finanz-Szene portal reported that “Haspa’s robo plans failed,
not least because of internal resistance” (Dohms 2018). Nevertheless, there
are two distinct patterns emerging from the cases described. Given the
considerable marketing costs for new customer acquisition and the low
cross-selling potential of robo-advisory, the robo-advisory market is a scale
market. Market participants who fail to generate reach within the first two
to three years (e.g. by moving up to the top 10 in their home market), or to
leverage synergies with existing business areas and to optimize the growth-
related KPIs, will sooner or later be swept up by the consolidation wave.
On the other hand, parallel interests may play a central role in the successful
implementation of cooperative ventures, as well as the introduction of
new products at established institutions, supported by appropriate internal
incentives. This can also mean that higher-margin products are sacrificed
in favor of lower-margin, progressive products in the interest of sustainable
success. This is because the entire global robo-advisory market is still
expected to experience immense growth. The consulting firm Roland
Berger expects the CAGR of robo-advisors’ assets under management to
be between 20 and 40% by 2022 (Buess et al. 2018, p. 8).
When considering these growth forecasts, however, the question
arises—as with any product—where will this growth come from.
Essentially, there are two options: either it will be driven by general growth
in the respective market (i.e. the investment volume, for example, through
an average increase in private assets) or by pushing out other products
(i.e. investment opportunities, such as a self-directed securities account or
a direct investment in ETFs or actively managed funds). Another factor
to keep in mind when it comes to investing is the fact that people still
have large amounts of money and thus potential investment volume sitting
in their checking accounts—especially in Germany. Here, assets parked
in demand deposits, and cash amounted to EUR 1.6 trillion in the last
quarter of 2018, representing an increase of 188% since 1999.1 Releasing
this potential could be the third factor, although experience has shown that
it is not that easy. According to a study by JP Morgan Asset Management
in cooperation with GfK, only 5% of Germans surveyed switched to more
profitable investment products after a decade in a low-interest environment
1 See TagesgeldVergleich (2019).

260 M. SANDER
(Bradtmöller and Düchting 2018). At the same time, 81% of the women
and men surveyed are dissatisfied with the performance of their savings
products (Bradtmöller and Düchting 2018).
13.2 GROWTH FACTORS IDENTIFIED
13.2.1 No. 1: Overall Market Growth

While the global markets—measured in AuM or as assets per capita—are
expected to grow by 6 to 7% p.a. up to 2022, the fastest growing markets
in terms of assets per capita are expected to be in Eastern Europe with 11%
p.a. and Central Asia with 10% p.a. until 2022 (Beardsley et al. 2018, p.
11). This has the potential to be an important piece of the puzzle from
which all investment products can benefit. The simultaneously expected
increase in e-commerce—approx. 12% p.a. up to 2022 in retail online sales
in the US and Western Europe versus a growth rate of 3% p.a. in the retail
sales as a whole (Bhave et al. 2018, p. 7)—should also favor digital financial
solutions such as robo-advisors.
13.2.2 No. 2: Replacing Existing Products

With respect to the second factor, the relative customer benefits of robo-
advisors compared to conventional investment opportunities for private
clients, whether in the retail or affluent segment—such as buying an
actively managed fund or investing in ETFs—point toward a significant
potential. At around 0.3% p.a., for example, the ETF fees are relatively
attractive and about 1.3% p.a. lower than the costs of active funds.2 On
the other hand, with the conventional ETF, the client participates one-
to-one, both in the value gains and in the value declines of the replicated
index. ETFs do not offer services such as risk management in the form
of ongoing rebalancing in line with the previously determined investor’s
investment and risk profile or other methods typically used by robo-
advisors. Furthermore, it is also debatable how many and which ETFs
to pick to achieve a differentiated portfolio. Private clients attempting
to invest into ETFs beyond the equities asset class will need specialist
knowledge—for example, on the functioning of bonds and correlations
2 Fonds und ETFs: Aktiv versus passiv (Focus Money No. 28/2017, p. 53).
between asset classes. However, the growth in popularity of ETFs over

the last few years has highlighted the importance of transparency and low
costs. With a conventional robo-advisor, the prospective client receives
an investment proposal for a diversified portfolio with several financial
securities such as ETFs, actively managed funds through to individual
stocks or certificates in line with the previously determined investment
profile. The cheapest robo-advisors in Germany cost 0.6% of the invested
amount per year (Test 2018). The growth potential of robo-advisors
by replacing existing investment opportunities or in line with existing
investment products—ETFs are currently robo-advisors’ main investment
product—is, therefore, quite substantial.
Nevertheless, some transparency and education is required to come
close to the reach of established investment opportunities. For example,
the global Google search requests for robo-advisors in 2018 were just
under 400,000 vs. 4.4 million for ETFs, at the same time, according
to an IPSOS 2018 (Antoniotti et al. 2018, p. 5) study in the US, only
23% of respondents said they were very familiar with robo-advisors, and
5.3% said they were already using robo-advisors (Fig. 13.1). However,
compared to 2015, the figures of 11 and 3.8% have increased significantly.
A similar survey was conducted by Ebase in early 2019 for the German
market. Roughly, only every fifth person knew what a “robo-advisor” was
(Nicolaisen 2019). After all, the figure was just over 28% for men and
46% for individuals with a net income of at least EUR 4000 (Fig. 13.2).
At the same time, the projections for the future are quite positive. For
example, across all respondent groups, 42% rate the likelihood of using a
robo-advisor to invest over the next 12 months as medium to high. At 46%,
women are still ahead of men with 40%. The respondents considered the
factors of transparency, low fees, and traceability of the investment strategy
to be most important (Fig. 13.3).
13.2.3 No. 3: Attracting Previously “Uninvested” People to the

Financial Markets
The third lever is certainly the most challenging, although there is plenty of
potential behind it, and rising inflation coupled with continued low interest
rates in Germany in 2018, for example, has amplified the pain point for
private customers who park their cash in a checking account. The question,
therefore, is what would it take for these largely inexperienced investors to
choose a robo-advisor?
262 M. SANDER
Fig. 13.1 IAI Q1 2018 USA Barometer on awareness of robo-advisor based on

Antoniotti et al. (2018)
Fig. 13.2 Ebase survey Q1 2019 on the awareness of the term robo-advisor based
on Nicolaisen (2019)
Fig. 13.3 Ebase survey Q1 2019 on the likelihood of using a robo-advisor to

invest in the next 12 months Nicolaisen (2019)
According to a representative Forsa survey in 2018, 43% of respondents

in Germany stated that they parked their money in a savings account. On
the other hand, according to the Forsa survey, 28% of those surveyed
indicated that they invest in investment funds and 24% in equities. The
results are even lower for women: 60% even rule out ever investing their
money in stocks or funds (Wenig 2018). One of the main reasons appears
to be their lack of knowledge about investing (l.c.). For example, 48%
admitted that they were not well informed about the possibilities of
investing money, and “(...) one in four respondents (24%) admitted[,] they
had little or no knowledge about investment opportunities,” (l.c.) while in
a survey conducted by the Association of German Banks (Bundesverband
deutscher Banken), only 44% stated that they knew what an investment
fund was (Verbrauchermagazin-Redaktion 2017).
Furthermore, a study from Investopedia in cooperation with Chirp
Research shows that the reservation against stock market investments
264 M. SANDER
Fig. 13.4 Investopedia affluent millennials study shows the wealth distribution
amongst the different investor generations (Gobell 2019)
is surprisingly increasing amongst the group of younger generations.

Figure 13.4 shows that affluent millennials have a higher probability to
put their money in low return investments than the generation X. Only
34% of the millennials in the study invest in stocks compared to 47% of the
generation X (Gobell 2019). Interestingly, the confidence of the millennials
regarding investing is increasing, the earlier they came in contact with the
topic of money and wealth. Fifty nine percentage of the millennials, who
had the first contact with investing after the age of 21, feel that they do
have not enough knowledge in this field. If the millennials in the study had
contact with investing before the age of 15, only 12% have the impression
of being under-educated in the subject (Gobell 2019). A targeted approach
that addresses the individual needs of this target group can, in principle,
have a positive effect on the growth of all existing financial market products.
At the same time, however, the answers cited above suggest that investment
opportunities with an edge are those which are easy to understand and/or
make the process as straightforward as possible for the investor. The first
step when investing with the help of a robo-advisor is to produce a personal
risk profile for the investor based on their individual goals, risk appetite,
and experience. With robo-advisors, investors do not have to manage their
portfolio themselves as robo-advisors pick the stocks for them and do the
ongoing rebalancing as is the case with conventional, analog discretionary
portfolio management, which has so far only been typically available to
wealthy clients with at least EUR 0.5 million to invest. These factors, as
well as the points mentioned under growth factor 2, point to a significant
potential for robo-advisors. To unlock this potential, a targeted expansion
of the existing offering is required.
13.3 SUCCESS FACTORS FOR ACHIEVING THE

IDENTIFIED POTENTIALS AND INCREASING THE
MARKET PENETRATION OF ROBO-ADVISORS
The potential behind the highlighted factors is quite immense and supports
the growth forecasts of some well-known management consultants and
statistics portals. What accompanying activities could help to unlock this
potential? What can be done to address the cited arguments such as the
lack of reach, transparency, financial market knowledge, and the perceived
complexity? Six possible approaches are outlined below:
13.3.1 Extending the Reach

Cooperation with established partners and brands can make a significant
contribution here. These can involve cooperation within the financial
sector (such as the cooperation between Scalable and ING Diba) as well
as innovative, cross-sector ventures (for example, Ikea partnering with
Sonos,3 the men’s fashion designer Virgil Abhol of Louis Vuitton (Samann
2019) or Adidas (Baur 2018)). In this respect, it is key to seek win-
win partnerships offering broadly equal benefits to both parties. This is
especially the case if both partners offer something that the other party
lacks. In addition to the much-discussed financial industry collaboration
between fintechs (a robo-advisor fintech, in this instance) and established
financial market participants with existing reach in the relevant target cus-
tomer groups, robo-advisor providers should also consider more creative,
cross-sector approaches (think Ikea), especially in the era of open banking.
To realize the identified cooperation potentials, it is also very helpful if
the cooperation partners have compatible corporate cultures and the key
individuals from both sides involved in the project get on well, even on a
personal level. Buzz campaigns are another important element, illustrated
by Red Bull with its event marketing campaigns featuring extreme sports
(e.g. “Red Bull Stratos—Felix Baumgartner’s Space Jump” Sampiero 2013
or “Red Bull Crashed Ice” Red Bull 2019). In addition, even in the age
of digitalization, and maybe because of it, personal networks can play a
pivotal role in the promotion of a product (especially in the form of word-
of-mouth advertising).
3 See Computer Bild (2019).

266 M. SANDER
13.3.2 Improving Financial Literacy

This is an extremely important and at the same time fundamental issue for
all investment products. Today, it is still not uncommon to go through
school and even further education, including universities, without really
coming across the topic of financial education. Where is the financial
knowledge supposed to come from if it is not acquired privately? The same
applies to industries such as law or medicine. While experts are usually
trusted here, trust in the financial industry has suffered greatly in recent
years. One effective remedy here is to offer help with self-learning to build
up some basic knowledge. This calls for the involvement of various parties.
For example, robo-advisory providers can certainly make a contribution
with webinars, podcasts, a financial encyclopedia, or a macroeconomic
magazine. Even better, of course, are neutral bodies such as independent
platforms or schools, taking on this task with the support of policymakers or
the private sector. As people live longer, financial investment can and must
play an important role in pension planning, while improvement of financial
literacy can also become a key factor supporting economic growth.
13.3.3 Greater Transparency

Offering a complete and easy-to-understand overview of costs and related
services is certainly a very important factor. However, this also includes
information about the investment management approach presented in an
accessible way. In addition to illustrations and images, videos, or local or
virtual open-house events featuring the investment management team can
play an important role.
13.3.4 Targeting Women

Figure 13.3 shows that 24.3% of women, but only 15.1% of men, rate
the likelihood of using a robo-advisor to invest in the next 12 months
as high to very high. At the same time, Figure 13.2 shows that 12.7%
women vs. 28.5% men are familiar with the term robo-advisor. To an
extent, this may be because the marketing campaigns and online presence
of many robo-advisors is too male-focused. Individual events, information
materials, and studies aimed at women can help to increase the identified
potential. Parallel efforts, such as increasing the proportion of women
on supervisory boards and in senior management in companies, should
help women to accumulate wealth over time, which in turn is likely to

play an increasingly important role in boosting the growth potential going
forward.
13.3.5 Hybrid Access

In addition to the measures mentioned under the previous points for
attracting previously “uninvested” participants to investment and pushing
out other existing investment products, a hybrid model can play a key role
in building trust. This is where robo-advisors and client advisors act in
something of a tandem. For example, a potential client can first create a
personal investment proposal online without obligation using the robo-
advisor’s investment onboarding process. The potential client can then
contact a client advisor to clarify any questions regarding content or to
further personalize the investment strategy. Charles Schwab and the private
bank, Hauck & Aufhaeuser have implemented this model with Schwab
Intelligent Portfolios PremiumTM in the US (Charles Schwab & Co. Inc.
2020) and Zeedin in Germany.
13.3.6 Expanding and Improving the Product Range

Besides the continuous improvement of the user interface (on mobile
devices in particular) and the intuitiveness of the application based
on customer feedback and innovative methods such as solution-based
thinking, the robo-advisory industry has to ask itself what can be done to
truly revolutionize the market. In view of the points mentioned above,
progressive technologies such as voice-controlled applications could prove
to be promising here. However, this also includes applications that
introduce the user to the topic of investment depending on their particular
circumstances and related interests (e.g. saving for a college fund for their
children, buying a property, pensions). Playful elements that can help
clients identify potential needs easily and intuitively can also be useful here.
13.4 SUMMARY
All these measures or activities can be leveraged to raise awareness and
achieve the critical mass required for robo-advisors to be profitable. As
clients pay a recurring, annual fee for the robo-advisor services they use,
268 M. SANDER
long-term customer loyalty and the continued acquisition of new clients

will be a key factor for the profitability of robo-advisors in the medium-
term. Robo-advisors become profitable when the total fee received for
both the volume of existing business and new business in the respective
year exceeds the costs for the associated new customer acquisition (the
marketing costs are typically quite considerable) and the costs for ongo-
ing operations (e.g. personnel costs for portfolio management, customer
service, middle and back office areas as well as material costs such as IT
costs).
REFERENCES
Antoniotti, Joe, Donna DenBraven, and Greg DePalma. 2018. Ipsos Affluent
Intelligence: Fintech: The Technology and the Early Adopters Disrupting Finan-
cial Services. Tech. rep. Ipsos. https://www.ipsos.com/sites/default/files/ct/
publication/documents/2018-05/iai-fintech-whitepaper-5-2018.pdf .
Baur, Susanne. 2018. Einmal Ausatmen Bitte: Ikea Bringt eine Kollektion mit
Adidas Raus. Last Accessed 20 March 2019. https://www.instyle.de/fashion/
ikea-adidas-kollektion.
Beardsley, Brend, Daniel Kessler, Martin Mende, Federico Muxi, Matthias Nau-
mann, Jürgen Rogg, Tjun Tang, Tyler Woulfe, André Xavier, and Anna
Zakrzewski. Global Wealth 2018: Seizing the Analytics Advantage. Tech. rep.
P. 11. Boston Consulting Group.
Bhave, Ashwin, Chris Biggs, Peter Burggraaff, Brad Loftus, and Shishir Pathak.
2018. Accelerating Digital Innovation in Retail. Tech. rep. P. 7. Boston
Consulting Group.
Bradtmöller, Pia, and Annabella Düchting. 2018. J.P. Morgan Asset Management
– Income-Barometer 2018. Last Accessed 24 March 2019. http://www.
jpmorganassetmanagement.de/CMSMediaStorage/RADEditorDocuments/
2018-07-05_JPMAM%20Income-Barometer%20Deutschland%202018.pdf .
Brummer, Thomas. 2018. Sebastian Hasenack (Investify): “Testphase mit Haspa
beendet”. Last Accessed 20 March 2019. https://de.extraetf.com/news/
finanznews/sebastian-hasenack-investify-testphase-mit-haspa-beendet.
Buess, Robert, Daniel B. Hildebrand, Dominik Löber, and Jeanette Okwu. 2018.
Rewiring Asset and Wealth Management, P. 8. Munich: Roland Berger GMBH.
Charles Schwab & Co. Inc. 2020. Intelligent Advisory. Last Accessed 20
March 2019. https://www.schwab.com/public/schwab/investment_advice/
intelligent_advisory.
Computer Bild. 2019. IKEA und Sonos: Symfonisk RegalWiFi-Speaker im Test. Last
Accessed 23 March 2019. https://www.computerbild.de/artikel/cb-News-
Vernetztes-Wohnen-Smarthome-IKEA-Sonos-Symfonisk-19465133.html.
Dohms, Heinz-Roger. Immer Mehr Deutsche Banken Begraben ihre Robo-Projekte.

Last Accessed 20 June 2019. https://finanz-szene.de/digital-banking/immer-
mehr-deutsche-banken-begraben-ihre-robo-projekte/.
Friedberg, Barbara. 2018. Robo-Advisors with the Most Assets Under Management
(AuM). Last Accessed 10 March 2019. https://www.roboadvisorpros.com/
robo-advisors-with-most-aum-assets-under-management/.
Gobell, Joetta. 2019. The Affluent Millennial Investing Survey. Investope-
dia. Last Accessed 1 December 2019. https://www.investopedia.com/the-
investopedia-affluent-millennials-study-4769751.
IT Finanzmagazin. 2018. Sebastian Hasenack (Investify): “Testphase mit Haspa
beendet”. Last Accessed 20 June 2019. https://www.it-finanzmagazin.de/
online-vermoegensverwalter-vaamo-und-moneyfarm-vor-zusammenschluss-
81232/.
Nicolaisen, Christian. 2019. Studie von Ebase: Nur jeder fünfte Deutsche kann etwas
mit Robo-Advice anfangen. Last Accessed 9 March 2019. https://www.private-
banking-magazin.de/ebase-umfrage-nur-jeder-fuenfte-deutsche-kann-etwas-
mit-robo-advice-anfangen/.
Red Bull. 2019. Red Bull Crashed Ice Boston 2019. Last Accessed 20 March 2019.
https://www.redbull.com/se-en/events/crashed-ice-boston.
Robo-Advisor Portal. 2018. Schluss, Aus, Vorbei: Werthstein schließt die Pforten! Last
Accessed 20 March 2020. https://www.roboadvisor-portal.com/schluss-aus-
vorbei-werthstein-schliesst-die-pforten/.
Samann, Lena. 2019. Warum Sie Jetzt ein Design von Virgil Abloh bei IKEA
Kaufen Können. Last Accessed 20 March 2020. https://www.vogue.de/
mode/artikel/virgil-abloh-ikea.
Sampiero, Josh. 2013. Flieg mit Baumgartner: Red Bull Stratos-Video. Last
Accessed 20 March 2019. https://www.redbull.com/de-de/flieg-mit-
baumgartner-red-bull-stratos-pov-video.
TagesgeldVergleich. 2019. Geldvermögen Privater Haushalte. Last Accessed 31
March 2020. https://www.tagesgeldvergleich.net/statistiken/geldvermoegen.
html#entwicklung.
Test. 2018. Was die Automatisierte Vermögensverwaltung Taugt. Last Accessed
20 March 2019. https://www.test.de/Robo-Advisor-Was-die-automatisierte-
Vermoegensverwaltung-taugt-5107535-0/.
Verbrauchermagazin-Redaktion. 2017. Umfrage: Die Deutschen überschätzen ihr
Finanzwissen. Last Accessed 20 March 2019. https://bankenverband.de/
blog/umfrage-die-deutschen-uberschatzen-ihr-finanzwissen/.
Wenig, Mirko. 2018. Deutsche Parken Geld am Liebsten auf Sparkonto. Last
Accessed 20 March 2019. https://www.versicherungsbote.de/id/4862540/
Deutsche-parken-Geld-auf-Sparkonto/.
GLOSSARY
Active management Whereas most robo-advisors follow a passive

approach, that is, to follow indices as closely as possible, some robos
also apply active management, that is, they are trying to outperform a
given benchmark.
Application programming interface (API) Intermediary interface
between multiple parts of software in order to facilitate ease of
implementation, execution, and maintenance of interrelated programs.
Artificial intelligence Artificial Intelligence (AI) is a broadly used term
to describe the capability of a computer system, which resembles cog-
nitive functionality, especially in decision making. General AI denotes a
type of artificial intelligence that is not bound to a certain task and mim-
ics human behavior. It is, however, not yet existent. Applied AI, on the
other hand, is focused on solving specific real world problems, mostly
in a data-driven way and makes use of machine learning algorithms.
Asset allocation The distribution of assets in a portfolio to diversify the
wealth of the investor. Typical asset classes comprise stocks, bonds, and
cash. Further asset classes for robo-advisory may include real estate,
commodities (mostly gold), hedge funds, and alternative investments.
Behavioral finance A theory, which describes the impact from emo-
tional biases on financial decision-making. By and large, the findings
from social psychology are adapted to similar situations in finance to
derive conclusion about the behavior of investors. Financial models

Technology, https://doi.org/10.1007/978-3-030-40818-3
272 GLOSSARY
often assume rational agents on financial markets, the concept of homo

economicus.
Black-Litterman model The Black–Litterman model is a mathematical
model for portfolio allocation by Fischer Black and Robert Litterman
published in 1992. It takes into account uncertainties in risk and (most
important) return estimators and gives therefore more robust results
than the older Markowitz model (simple mean-variance optimization).
Chatbot A computer system that deals with digitalized written or spoken
language and follows a certain conversation protocol, based on the
provided textual or spoken content. The goal is to provide a human-
machine interface to solve specific problems, such as product-related
questions. Hence, for example, chatbots interact with people in sales or
support scenarios, but manifold other areas of interaction are possible.
Compound annual growth rate (CAGR) The CAGR is a measure for
the average growth over a multitude of periods.
Core banking system The core banking system is a software, which
stores and processes the core data of a bank.
Cross-selling campaign Cross-selling denotes a process in the
customer-vendor interaction, in which the customer is recommended
additional items to buy, typically following a pre-defined sales strategy.
For example, after purchasing a pair of shoes, material for shoe care is
offered to the customer.
Customer journey The entire sequence of touch points a customer
experiences when interacting with a company’s offering, from awareness
to purchase to advocacy.
Customer-salesperson interaction technologies Technologies that
enable customers to inform about product and services completely
on its own.
Diversification breakdown In “normal” times different risky assets will
behave differently. This is the reason why the total risk of a portfolio
which consists of different asset classes lies normally well below the
weighted sum of the individual risks. During crisis, however, investors
tend to get out of their risky positions globally and simultaneously and
therefore this diversification effect is vanishing in times when it is needed
most.
Diversification ratio The diversification ratio is defined as the ratio of
the weighted average of volatilities divided by the portfolio volatility. In
other words, it is the portfolio risk without taking diversification into
account divided by the portfolio risk with diversification.
GLOSSARY 273
Exchange traded fund (ETF) An ETF, very often also called index
fund, is an investment fund, which tries to minimize the tracking error
compared to a given benchmark. It is hence used for passive investing
and is part in many robo-advisory portfolios.
FinTech Companies which provide financial services based on modern
technologies are denoted as FinTechs. Very often, they are start-ups that
offer just one specific service and not a whole range of services. The idea
of FinTech is rather old because there has always been a tendency to use
innovations for financial purposes to gain advantage over other market
participant, for example, the telegraph for transmitting stock market
information.
General data protection regulation (GDPR) The GDPR is an EU
law, which regulates data protection and privacy in most of the European
countries.
Home bias Describes an emotional bias in decision-making. Following
academic research, investors have a tendency to overweight domestic
assets and to underweight foreign ones. This effect can be found with
private as well as professional investors and creates portfolios with
limited diversification benefits.
Hybrid model Robo-Advisors that integrate a human advisor as well
in their process are called hybrid because they are link two different
approaches: the strength of the robot by implementing an automatized
investment process, as well as a human for personal contact and empathy
if necessary, for example, during market downturns.
Irrational behavior Decisions and choices which do not maximize util-
ity, against the classical economic theory assumption of rationality.
Mental accounting Describes the tendency to separate different invest-
ments into different internal accounts. In some cases, this can be quite
useful, for example, if different investment goals are pursued. But if the
economic situation is the same, sometimes still different decisions are
made—which would be inconsistent. Mental accounting links to the
Prospect Theory of Kahneman and Tversky.
Markets in financial instruments directive (MiFID) Established in
2007, MiFID is a European regulation to improve transparency
and efficiency of the European financial markets. It also determines
standards for regulatory disclosures. Meanwhile, there is an updated
regulation in act, which is called MiFID 2.
Minimum viable product The minimum viable product denotes an
early version of a product, which typically includes core features only.
274 GLOSSARY
The idea is that customers like early adopters provide enough feedback
for successful enhancements of the product.
Omni-channel Denotes a sales strategy, which includes the use of several
distribution channels. For the improvement of customer experience, the
user decides which channel to use for a certain service and can switch
between the different channels seamlessly.
Overconfidence Describes the observation that people tend to overes-
timate their knowledge or abilities. Overconfidence is a sub-form of
control illusion. Investors who are suffering from overconfidence may
estimate the investment’s risk incorrectly and may create portfolios with
poor diversification.
Passive management In contrast to active portfolio management, pas-
sive management tries to track a given benchmark as closely as possible.
The aim is to recreate, for example, an index return as accurate as
possible and to avoid any deviations. Exchange Traded Funds (ETFs)
are a popular instrument for passively managed portfolios.
Payment services directive 2 (PSD 2) European directive aiming at
furthering the integration of the market for electronic payments within
the EU in order to increase competition and attain better prices for
consumers.
Perceived risk Consumer’s subjective belief of suffering a loss in the
pursuit of desired outcomes.
Perceived use Individual’s perception that the use of a new technology
will enhance or improve his or her performance.
Psychological reactance Threat of perceived freedom by elimination.
Questionnaire Set of questions, which is usually applied to measure the
risk profile of investors in order to give adequate investment advice.
Rationality In financial models, it is very often assumed that fully
rational investors are decision-makers. This comprises the assumption of
no emotional biases, complete information, and a stable utility function.
Risk In contrast to return, risk is the opponent in the story of investing.
Risk can be measured in different ways: it can be the deviation from
a given benchmark—no matter in which direction; or it can be the
potential loss an investor might bear.
Risk capacity Amount of risk an investor could bear based on his or her
wealth or investment requirements. For example, investors with a longer
investment horizon could accept higher equity levels than short-term
investors.
GLOSSARY 275
Risk parity This investment approach seeks to optimize the balance of

risk within a portfolio by assigning each asset the same risk budget.
Risk profile In the risk profile, the risk tolerance and risk capacity of an
investor is measured, typically by a questionnaire.
Risk tolerance Amount of risk an investor could bear based on his or her
risk appetite: from risk-averse to risk-loving, depending on individual
preferences. However, people who love to take risks do not necessarily
prefer risk in every field.
Robo-advice Digital investment advice tools that match customers on
the base of their personal preference to financial products.
Robo-advisor A financial service based on two components: investment
advice and automatization through robots. The term robo-advice may
comprise different levels of automatization: whereas the onboarding and
risk profiling is typically provided by algorithms, the investment process
can either be completely rule-based or driven by committees.
SEC The US Securities and Exchange Commission (SEC) is a federal
agency. It is responsible for the regulation of markets and exchanges.
Self-profiling A process through which financial advisory clients create
a risk-tolerance, risk-budget profile of themselves, aided by automated
computer systems.
Social media Describes platforms, which allow internet users to com-
municate and hence to share user-generated content like knowledge,
opinions, evaluations, impressions and so on.
Trust Confidence in the exchange partner’s reliability and integrity.
White-label Financial service or product, which is offered without the
brand of the originating bank.
INDEX
Agile project management, 201 Chatbots, 234, 241, 251

Algorithm, 5, 46, 54, 63, 113, 196 China, 9, 189, 192, 253
Ambiguity aversion, 55 Clients
Anti-fraud provisions, 108 affluent, 194, 195
API economy, 27, 184, 193 private banking, 194, 200
Appropriateness assessment, 146 retail, 194
Artificial intelligence, 16, 183, 228, wealth management, 194, 200
248, 251 Cloud services, 158
Asset allocation, 6, 24, 63, 64, 170 Compliance, 117
Assets under management, 8 Consumer behavior, 22, 192, 248,
251
Core banking system, 217
BaFin, 135, 214 Cross-selling, 259
Behavioral economics, 54 Customer journey, 206, 249
Behavioral finance, 54, 172 Customer needs, 11, 181, 190, 201,
Bias, 25, 54, 64, 65 251
Black-Litterman model, 82 Customer–salesperson interaction
Blockchain, 183, 253 technologies, 94, 95
Brand reputation, 191, 197 Customization, 89
B2B market, 176
Data protection, 173

CAGR, 259 Design thinking, 219
Capital preservation, 84 Digitalization, 10, 22, 167, 168, 189,
Cash lock, 85 191, 216, 249, 265

Technology, https://doi.org/10.1007/978-3-030-40818-3
278 INDEX
Digital revolution, 4 rule-based, 29, 56, 64, 72

Diversification, 55, 62, 200 Investment
breakdown, 79 advice, 138, 216
ratio, 77 automation, 24
Due diligence, 154 broking, 140
committees, 6, 56, 64
process, 4, 54, 202
Early adaptors, 7, 26 universe, 88, 89
Emotion, 29, 54 Investment Advisers Act, 106
Employee Retirement Income Security Investment Company Act, 122
Act, 125
ETF, 6, 8, 63, 65, 89, 120, 136, 165,
172, 206, 260, 261 Know-your-client, 40, 172, 218
Execution only, 146 KWG, 135, 215
Fees, 11, 24, 113, 119, 120, 175, 212, Level playing field, 135
261 Liability umbrella, 143
Fiduciaries, 125
Fiduciary duties, 109
Financial crisis, 7, 134, 248 Market
Financial instrument, 135 access, 196
FinTech, 4, 11, 22, 27, 134, 142, 144, volume, 8
153, 164, 178, 192, 193, 196, Marketing, 198, 206, 231, 233, 235
217, 248, 258, 265 Maximum diversification, 79
Maximum loss, 84, 85
Mean-variance optimization, 81
General Data Protection Regulation, Mental accounting, 58
214 MiFID 2, 155, 206, 214
General regression neural network, 83 Millennials, 10, 24, 264
Minimum investment amount, 200
Minimum viable product, 193
Home bias, 55 Moral hazard, 7
Homo economicus, 55
Human advisor, 6, 46, 169
Hybrid model, 7, 15, 24, 65, 195, Onboarding, 149, 153, 172, 216, 218
204, 248, 267 Open banking, 265
Overconfidence, 56, 61
Overweighting, 56
Illusion of control, 62
Information processing, 54
Investing Passive management, 24, 63
goal-based, 29, 60, 114, 119 Payment Services Directive 2, 27, 218
INDEX 279
Perceived risk, 94, 96, 97, 99, 100 Risk-bearing capacity, 86

Perceived use, 94, 96, 98, 99, 101 Robo-advice, 96, 99–101, 188
Performance, 12, 191 Robo-advisor, 4, 21, 23, 45, 54, 73,
Plug-and-play, 193 135, 164, 246
Portfolio evaluation criteria, 178
management, 142, 232, 235, 237 registration of, 106
minimum variance, 219 Robo-advisory, 258
optimization, 63 Robo economicus, 55
Predictive banking, 16
Product recommendation, 228, 233,
235 Sales, 233, 235
Prospect theory, 42, 59 Sales channel, 201
Psychological, 38 Salesperson–customer interaction
Psychological reactance, 94, 96, 97, technologies, 101
99, 100 SEC
Psychometry, 41, 43 Investor Alert, 118
Staff Guidance, 112
Securities Exchange Act, 121
Quantitative strategies, 73, 75, 91 Service quality, 191, 197
Quant models, 91 Simplification, 47
Questionnaire, 43, 114, 116, 202, Smartphone, 248
216 Social advisory, 251
Social media, 190, 213, 246, 247,
249, 253
Rationality, 54 Supervision, 121, 134
Rebalancing, 120, 152
Recommendation, 119, 138, 148
suitability of, 115 Tax, 39, 214
Regret aversion, 56 advice, 114
Regulation, 22, 36, 40, 134, 172, status, 110
198, 213 Test
Retail clients, 7 reliability, 40
Return targets, 86 validity, 40
Risk, 96 Trust, 94, 96, 99, 100, 172
budget, 6
capacity, 37
minimization, 80 UCITS funds, 136
parity, 79
preference, 8, 36
profile, 4, 36 Volatility, 120
profiling, 25, 45, 172, 173
targets, 84, 90
tolerance, 6, 26, 60, 119 White-label, 165, 168, 173

Robo-Advisory - Investing in The Digital Age (2021 Peter Scholz)

Uploaded by

Copyright:

Available Formats

Robo-Advisory - Investing in The Digital Age (2021 Peter Scholz)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Robo-Advisory - Investing in The Digital Age (2021 Peter Scholz)

Uploaded by

Copyright:

Available Formats

PALGRAVE STUDIES IN

FINANCIAL SERVICES TECHNOLOGY

More information about this series at

ISSN 2662-5083 ISSN 2662-5091 (electronic)

way, globally diversified and secured 24/7, with a state-of-the-art risk

Munich, Germany Jochen Werne

Digitalization—it is a megatrend that has accompanied our economy for

Hamburg, Germany Peter Scholz

A large project like this book on robo-advisory, where so many perspectives

Furthermore, I owe a big thank you to my student assistant Laura

Part I Status Quo of Robo-Advisory 1

1 Robo-Advisory: The Rise of the Investment Machines 3

Part II Implementation of Robo-Advisory 33

3 Risk Preferences of Investors 35

4 Robo Economicus? The Impact of Behavioral Biases on

5 Quant Models for Robo-Advisors 71

5.3 What Quantitative Approaches Does the Robo-Advisory

6 Analysis of the Use of Robo-Advisors as a Replacement

7 Regulation of Robo-Advisers in the United States 105

8 Regulation of Robo-Advisory in Europe and Germany 133

8.4 Conclusion of Contract 155

Part III Case Studies of Robo-Advisory 161

9 (Re-)Launching a Robo-Advisor as a Bank 163

10 How Can Robo-Advisory be Implemented and

10.1.3 Market Analysis 193

Part IV The Future of Robo-Advisory 225

11 The Role of Artificial Intelligence in Robo-Advisory 227

12 What Role Do Social Media Play for Robo-Advisors? 245

12.4.2 Social Media as an Alternative to

13 Success Factors for Robo-Advisory: Now and Then 257

Alexander Beck is an experienced technology expert in the field of

Ana-Maria Climescu is Senior Associate at Hauck & Aufhäuser and is

Melanie L. Fein is an attorney who advises financial institutions on

Joachim Goldberg has analyzed the interaction between investors and

Goetz Greve is Professor of marketing and sales at the HSBA Hamburg

David Grossmann earned his Ph.D. from the Andrássy University

Christian Hammer worked as a savings bank manager for private and

his home as an Executive Partner at NFS. By writing Basic knowledge for

Christian von Keitz is Senior Project Manager in Hauck & Aufhäuser’s

Sinan Krueckeberg has more than 10 years of experience in conventional

Frederike Meyer lives in Hamburg and works in a wealth management

business administration. She is pursuing her master’s degree in business

Monika Mueller founded FCM Finanz Coaching in 1999 and works as an

Yvonne Quint is a partner in the segment Banking & Capital Markets

Paul Resnik is a founder of the world’s first psychometric financial risk

Thorsten Ruehl is partner and head of research and portfolio manage-

Ruehl has two decades of experience in quantitative portfolio manage-

Madeleine Sander is Managing Director and Chief Financial Officer at

Craig Saunders is a communications and marketing consultant who

Theodor Schabicki has worked in financial services for 24 years, of which

and Artificial Intelligence (AI), Business Process Outsourcing (BPO) and

Peter Scholz is Professor of Banking and Financial Markets at Hamburg

Soeren Schroeder is a manager at BearingPoint with 10 years of experi-

Michael Tertilt holds a bachelor’s degree in Technology and Business and

bank management. He has been following the developments in the Fintech

Jochen Werne is a German manager with a proven track record in business

FINRA Financial Industry Regulatory Authority

VaR Value at Risk

Fig. 1.1 Wealth management process adapted from Evensky et al.