See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/237622306

The World of Funds of Funds

Article  in  Investment Management and Financial Innovations · January 2008

CITATIONS READS
4 1,282

3 authors:

Michael Stein Svetlozar Rachev

University of Duisburg-Essen Texas Tech University
29 PUBLICATIONS   86 CITATIONS    277 PUBLICATIONS   4,369 CITATIONS   

SEE PROFILE SEE PROFILE

Edward W. Sun
Kedge Business School, Bordeaux
60 PUBLICATIONS   782 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Risk assessment for Lithuanian pension funds View project

Big Data Mining; P&C policy pricing; Energy Forecasting; Optimal Trading View project

All content following this page was uploaded by Edward W. Sun on 24 August 2014.

The user has requested enhancement of the downloaded file.

 The World of Funds of Funds

Michael Stein, University of Karlsruhe & Credit Suisse Asset Management*

Svetlozar T. Rachev, University of Karlsruhe and KIT& University of Santa Barbara**
Wei Sun, University of Karlsruhe***

* Michael Stein: Institute for Statistics and Mathematical Economics, University of

Karlsruhe, Germany & Credit Suisse Asset Management, Frankfurt, Germany. //
[email protected]

** Prof. Svetlozar T. Rachev (Corresponding Author): Institute for Statistics and

Mathematical Economics, University of Karlsruhe, Germany & Department of Statistics and
Applied Probability, University of California, Santa Barbara, USA. // Chair of Econometrics,
Statistics and Mathematical Finance, School of Economics and Business Engineering,
University of Karlsruhe, Kollegium am Schloss, Bau II, 20.12, R210, Postfach 6980, D-
76128, Karlsruhe, Germany, [email protected], Chief-Scientist,
FinAnalytica, http://www.finanalytica.com

*** Dr. Wei Sun: Institute for Statistics and Mathematical Economics, University of
Karlsruhe, Germany. // [email protected]

1
 Abstract

In this paper we provide a review of developments in the fund of funds (FoF)

industry as well as related academic research work. Since the genesis of this sort
of financial product in the 1960s, huge capital inflows have generated a total of
US $1,200 billion of worldwide FoF assets under management as of year-end
2007. Being portfolios built out of other portfolios and assets, FoFs are a special
type of investments and therefore have distinct advantages and disadvantages in
comparison with other investment vehicles. When reviewing both the academic
research related to the unique, or at least distinct, features of FoFs and discussing
market developments, we address a good number of interesting questions and
challenges. Specifically, we find that classical methods of research and
assessment may be insufficient to address these questions in both theoretical and
quantitative aspects. The need for a distinct treatment of FoFs is even more
evident when focusing on the dynamic nature of financial market developments
and increasing global competitiveness. As past research work on FoFs is scarce,
especially when being compared to the amount of research work on funds, we
conclude that further investigation of this sort of financial asset is necessary.

Keywords: Fund of Funds, FoF, Hedge Funds, Multi Manager Funds, Fund
Portfolios

JEL Classifications:

2
 1. Introduction

The fund of funds (FoF) concept has its origin in the 1960s, with the industry steadily
growing since then. A FoF is a fund which invests in other funds and is sometimes referred to
as a multi-manager fund. There are many different types of FoF. They include fund of hedge
funds, fund of private equity funds, fund of mutual funds, and fund of real estate funds. With
FoF as of year-end 2007 having about US $1,200 of assets under management, a closer look
at the FoF industry is needed and we do so in Section 2 of this paper. In 2007, funds of hedge
funds received about US $60 billion of net new assets under management, increasing the
amount of global capital invested in this type of FoF to around US $800 billion.
Investments in FoF can be advantageous for both retail and institutional investors due
to the distinct features of this kind of financial product. However, as with any other
investment product, disadvantages and sources of possible dissatisfaction exist as well.
One out of several striking advantages of FoF concepts is the possibility for retail investors to
get access to financial products in which they could not directly invest. Many funds ─ and
especially hedge funds ─ are not accessible for most private, retail investors due to high
minimum investments, prohibitive high transaction costs, lack of information or simply
missing distribution channels.
With FoFs, retail investors are able to get exposure to sectors, asset classes, markets,
and products which otherwise would not have been included in their portfolios. Such
structural aspects, albeit largely differing between countries, markets, and sectors, stem from
the fact that business ties and related costs are crucial in determining the investment product
universe. With most FoFs pooling money from large and diverse investor bases, they are able
to invest in assets that demand high minimum investments or that offer discounts for
management fees, costs or loads when investing amounts above specified marks. Investing in
special share classes of funds, which are generally open for all kinds of investors with pre-
defined minimum investments and lower management fees, is another path to cost reduction.
Because banks and fund management companies generally have their own trading
infrastructure, accounting, and clearing offices and desks as well as special agreements with
other market players and counterparties, absolute and relative operational costs also can be
reduced significantly.

3
 However, on the con side, the double cost structure of FoFs caused debates in the past
and is still subject to discussions both in the academic world and among practitioners.
In addition to these organizational economies of scale, direct contracting between
financial institutions may impose another beneficial factor when it comes to market access.
With direct contracts between financial institutions, banks, endowments, management
companies, and/or advisors, discounts to fund load fees for example may be agreed on, or the
institution may be able to trade without paying issuance fees. Structural aspects and the
effects of business ties in the fund industry have been the subject of numerous studies, which
will be discussed in the fund (of fund) industry review in Section 2.
Besides constraints at the cost-of-investment side or barriers to entry, retail investors
face another problem when building portfolios out of a large variety of assets and financial
products: The problem of information and overview. As the fund industry is offering a huge
range of products, it is difficult for retail investors to get an overview concerning funds in
which they are interested. Performing the task of market screening may be both time-
consuming and inefficient. Furthermore, even when having found a pool of investment
possibilities, selecting the ones which suit the investor’s needs and preferences is a
challenging task, sometimes even for experienced investors.
As is the task to define an investment universe, the evaluation task is challenging,
because based on the respective needs each single investor has, information building with
respect to the quality of target funds is crucial. This stems from two interrelated facts. First,
retail investors generally do not have access to sophisticated data systems or information
systems. Second, even if such sources are at their disposal, retail investors may find it
difficult to use such information properly.
Being exposed to some kind of informational blinkers, the only way to remedy may
be delegating investment decisions. This can happen in various ways, for example with
investment advisors or wealth managers. Investing in pension funds or insurance plans could
be a solution, too. However, none of the mentioned forms of investment decision delegation
is free from shortcomings or disadvantages. Costs have to be incurred in any case, and one is
always exposed to the classical problems of moral hazard, divergence of interests,
uncertainty, and, once again, insufficient information. For FoFs, the same holds true of
course, and one may argue that indirectly paying a FoF manager via management fees may
result in the same problems as paying directly for investment consultancy or wealth
management.

4
 However, the emergence of the industry in recent years and the steady path of growth
that the branch has found, suggests another view. Seemingly, the FoF industry delivers
products and investment possibilities that attract retail investors all over the world. If it would
not pay in the most direct sense of the word, why should people put their money into FoFs? Is
it the inexistence of better solutions, advertisement effects, or do FoFs really suit retail
investors that well? These questions have yet to be answered, where attention should be
drawn on the double cost structure imposed by FoFs and their management fees. This often
emphasized double fee structure of FoFs is subject to the studies of (Brown et. al., 2004) and
(Reddy, 2007).
Many of the problems that retail investors face when making investment decisions do
not arise for institutional investors in the same manner. As mentioned above, information
flows are completely different for institutional investors such as pension plans, asset
managers, wealth managers, endowments, or state-owned investment funds than they are for
private, retail investors. The same holds true for different cost burdens, resulting from the
structures discussed above. Reconsidering the decision to choose between types of delegating
investment decisions, questions concerning the value added by market professionals have to
be answered.
Naturally related to management fees and advisor compensation is the question of
how well the services provided suit the investors. When deciding on the sector, asset class or
country to invest in, the problem is not only to separate the ones which one wants to be
exposed to, but to decide on how this can be achieved. Investing in index or basket
certificates or exchange traded funds (ETFs), for example, are ways to gain exposure to
specific markets, sectors, countries or strategies. Most index products are very transparent
when it comes to underlying constituents, have very low management fees, and offer the
ability to participate directly in the movements of the underlying index. If exposure is gained
through index or asset tracking products, the investor receives a return profile with zero alpha
(no excess returns over the benchmark or index) and a beta of one (the returns are exactly
proportional to the underlying benchmark or index).
Passively managed ─ or at least benchmark oriented ─ funds are another way to
participate (almost) one-by-one, although some funds exist, which are marketed as actively
managed ones, but are merely tracking their benchmark. In contrast to investing in index
profiles, both retail and institutional investors are demanding excess returns from their
investments, that is, they expect the managers to outperform their benchmark.

5
 Finding fund or portfolio managers which seem to possess superior ability to
outperform the market and thereby keeping track of the imposed costs is also a strain of
research of its own. Questions that arise when searching for alpha include ─ among many
others ─ the following: Is past performance due to pure luck or ability? Are the returns
achieved driven by timing, selectivity, superior strategies or are unobservable factors
responsible? May the investors expect the past performance to persist over time? As even
market professionals and academics may struggle to identify winners and losers, the quest for
alpha is understood as being one of the most challenging. Selection problems, performance
analyses, and efforts to identify winners and losers are the subject of Section 3. These
problems arise on both sides of the FoF ─ investors are interested in selecting the best FoFs
and in turn FoF managers are seeking to invest in the best funds.
After discussing selectivity and identification problems that arise when deciding on
investments, we lay out a problem that is very much a special FoF problem. FoF managers
may choose among a large variety of funds depending on the branch they are investing in.
Building portfolios out of funds may result in multiple exposures to one and the same asset or
risk factor. For example, when investing in European real estate equity funds like the
Henderson Horizon Pan-European Property Equity Fund or the Morgan Stanley European
Property Fund, one has significant exposure to the shares of Unibail Rodamco, a real estate
company that makes up more than 10% of the European Public Real Estate Association
(EPRA) Europe Index.
One should be careful when selecting related funds in order to avoid the trap of
ending up with a market-representing portfolio of top-holdings, while at the same time
incurring higher costs than when investing in the related indices. From this, it should be clear
that limits to diversification arise not only from the structure of the underlying assets but
from the paralleling of holdings. Benefits of including additional funds therefore need to be
weighed against the disadvantageous increased monitoring burden and the diversification
drain. Of course, this problem is not limited in dimensions, as FoF-Squared (fund of fund of
funds) structures exist as well, for example when institutions decide between building
portfolios out of funds or investing in FoFs. We will cover FoF specific portfolio construction
problems in Section 4.
2. The Fund (of Funds) Industry

As noted earlier, many different FoF types exist. Although the variety is large and growing,
hedge FoFs (HFoFs) have attracted the majority of capital invested in FoFs. Hedge Fund
6
Research (HFR) reports net new assets of US $9.5 billion, $49.7 billion, and $59.2 billion
invested in the HFoF industry in 2005, 2006, and 2007, respectively. According to HFR, by
year-end 2007 assets under management of HFoFs amounted to US $798.6 billion
worldwide, with the expectation of further growth. Compared with the industry-wide total of
US $ 1,208 billion of assets under management of FoF as reported by Barclay Group, HFoFs
represent about two thirds of total FoF global assets under management.
The large fraction of HFoFs in the industry has a straightforward structural
interpretation, as one crucial benefit from investing in this type of FoF is the possibility of
investing in hedge funds at all. Generally, hedge funds are not accessible for most non-
institutional investors except high-net-worth individuals, and by pooling investors’ money the
HFoFs open the door to this asset class for nearly everybody. Of course, minimum
investments exist for HFoFs too, but especially when accessing investible hedge fund indices
those are found to be lower.
Diversification benefits are another source of attractiveness of all FoFs, especially
when multiple hedge fund strategies such as Event Driven, Convertible Arbitrage, Distressed
Securities or Global Macro for example are included in the HFoFs. Large and growing, the
US $800 billion of worldwide hedge fund assets under management show a wide range of
investment possibilities. Due to the fact that hedge funds are way less transparent than mutual
funds and do not have the strong and strict reporting obligations that are imposed on mutual
funds, the task of selection and identification in the investment process is especially tough
when building portfolios that consist of or at least contain hedge funds. In this respect, HFoFs
deliver a precious service to investors by screening the hedge fund market, performing due
diligence processes, and selecting the most prospective investment possibilities.
As with any other asset class, the layout of the investment process is crucial to the
success of the investments made. Following the due diligence process and the manager
selection, the HFoF asset allocation (bottom-up or top-down approach, diversification
considerations, expectation building among others) is done, followed by continuously
monitoring the risks and returns of the investments made. Investment processes’ setup and
quality are the determining factors for the success or failure of HFoFs. For example, Standard
& Poor’s defines fund rating criteria that are underlying their decisions such as investment
culture, due diligence approach, portfolio monitoring systems and controls, operational risk
assessment, experience of fund management teams, selected managers’ experience, and
performance success.

7
 Private equity FoFs (PEFoFs) parallel many features and advantages of HFoFs
outlined above. These invest in leveraged buyout (LBO) or venture capital (VC) funds and by
doing so serve as investment channels to otherwise not accessible investment possibilities.
PiperJaffray, who offers a variety of PEFoFs, published a special report – hereafter referred
to as (PiperJaffray, 2003) – on this type of FoFs, which describes the distinct features of this
sort of investment, some of which are explained in the following.
LBO or VC funds invest directly in companies that are not traded publicly on stock
exchanges and are not listed. While LBO funds make use of leverage after purchasing part of
a target company, VC funds typically make serial equity investments without taking debt. Of
course, both sub-types of private equity try to identify companies that seem to be the most
promising concerning actual and future returns. Due diligence and subsequent close
monitoring enhance the possibility of high prospective returns on capital invested. Especially
LBO funds when taking over whole firms are directing the path of the companies in which
they are investing. The difference between VC and LBO funds can be roughly seen in the
maturity of their target companies, with the former commonly investing in young immature
companies and the latter targeting more mature firms with more or less stable cash flows.
High capital amounts are demanded to perform this kind of business, and the pooling of
money by PEFoFs serves as an appropriate way of raising those.
In addition to the HFoFs and the PEFoFs, which make up most of the industry, many
other FoF types have emerged in recent years. For example, FoFs that consist of stock funds
and bond funds provide high diversification benefits due to the opposing movements that the
fixed income and equity markets naturally take. Investors do not need to shift between bonds
and stocks; the adjustments are made by the FoF managers, whose timing on the markets is
crucial to the performance of this type of investment vehicle.
Sector specific or industry mutual FoFs exist as well, being portfolios that comprise
investments in a certain sector, country or class of investments. For example, some Real
Estate FoFs invest in both open-ended real estate funds (which are directly investing property
funds with a bond-like risk and return profile) and real estate equity funds. Depending on
their market expectations, the fund management teams can quickly increase their real estate
equity exposure or stick to the safe-haven directly investing real estate funds.
Not all FoF are limited to invest solely in other funds. Some have the possibility to
invest certain fractions of the fund volume in shares of companies, corporate or government
bonds, certificates or derivatives. While increasing the flexibility and enlarging the
investment universe of these FoFs, these additional investment possibilities represent both
8
opportunities and threats. Consider a fund manager who has a strong bullish view on one
single company, to which he wants to get more exposure than is possible through the
underlying fund holdings. By buying ordinary shares or derivatives on that company, the
fund manager may tweak the exposure to that company to the desired level. Another strategy
example would be to use derivatives or reverse index trackers to isolate underlying fund
performances or alpha, or to reduce exposure to certain parts of the underlying funds while
maintaining the remaining structure. Non-fund investments may therefore be a tool to
sophisticated FoF strategies, with hedge fund-like strategies then being accessible by
managers of long-only funds. However, if FoF managers are able to discretionally invest in
non-fund assets, the FoF concept may loose its stability or the structure that was expected by
investors. Put another way, the abilities of FoF managers need to be high enough to reap the
benefits of non-fund investment possibilities.
Therefore, the skills of the management team are once again the crucial determinant
of the success of investments. When it comes to performance measuring and attribution, a
variety of questions and problems arises, such as comparability, factor selection, statistical or
technical problems, measurement decisions and many more. To address these issues, the next
section will be devoted to an overview concerning performance analysis and identification
problems in the fund and fund of fund world.

3. Performance Analysis and Identification Problems in

Fund (of Funds) Management

This Section highlights the problems of performance analysis, the search for alpha and
identification problems inherent in FoF business. In doing so, we turn the focus on several
problems which especially apply to FoF investments.
When building FoFs, the product management and portfolio management teams are
confronted with a large set of questions. First, one has to choose how the investment universe
should be defined. Generally, FoFs are set up as products that focus on a certain industry, a
country, a sector, or an asset class of financial products. Several possible types of FoFs have
been discussed in section 1. After the “topic” of the FoF is selected, the next step is whether
to constrain the investment universe further. For example, if a FoF is bond oriented, the
question is whether the FoF should be able to invest in bond funds of any kind, or whether
certain profiles or countries may be excluded or limited.

9
 In addition, some FoFs are allowed to allocate a certain fraction of their assets under
management to non-fund investments, such as single stock shares, bonds, derivatives or
others. As mentioned in the introductory section, this may lead to two opposing outcomes.
On the one hand, the profile of the FoF could be greatly improved. With FoF managers
having the ability to (partly) hedge fractions of their investments, to gain or tweak exposures
to preferred sectors or companies which may be underrepresented in the fund holdings, or to
circumvent structural and institutional constraints, able managers may perform better than
they would when being limited to fund-only investment schemes.
A very actual example is that of real estate company Unibail-Rodamco. The EPRA
(European Public Real Estate Association) Europe Index, which serves as the benchmark for
most European real estate equity mutual funds, consists currently (February 2008) of about
14% of Unibail-Rodamco. As UCITS (Undertakings for the Collective Investment of
Transferable Securities) regulation limits the single allocation of mutual funds in one
company to 10% of the fund volume, this has led to all funds underweighting Unibail-
Rodamco relative to the benchmark. If the FoF management team is bullish on Unibail-
Rodamco, they may heal the expected underperformance of their regulated fund holdings by
investing directly in Unibail-Rodamco shares or derivatives. Another example would be if the
FoF managers want to pursue a strategy of picking small companies for which they have
promising information, but which are only small fractions in the target fund holdings due to
their small role in the benchmark index.
On the other hand, non-fund investment allowances for FoF managers may lead to
undesirable effects. If managers take the wrong steps and have a large amount of
discretionary freedom, they may dis-stabilize the FoF and introduce performance flaws. Put
another way, the possibility of non-fund investments is increasing both risk and uncertainty
concerning future performance from the perspective of FoF shareholders. As the investment
universe and therefore the allocation possibilities may be exploding due to non-fund
investments, the investor holding a FoF is confronted with increased problems concerning
expectation building. Therefore, fund manager ability is the crucial factor dividing pro and
con of non-fund investment possibilities for FoFs.
When it comes to ability and performance attribution as well as the identification of
“better” funds and FoFs, i.e. investments that deliver “alpha”, we are in the favourable
position of having a huge research work body concerning performance measurement at our
disposal. While the several studies are differing largely in their very nature, the aim of the
most is to conduct an analysis that may be useful for selecting funds, i.e. managers. Before
10
the various approaches will be discussed and put in relation to the FoF world, some preceding
arguments are due.
One important aim of performance analyses and identification in the search for alpha
should be comparability, that is, when trying to analyse various mangers’ skills and fund
performances, the study needs to focus on the right factors and benchmarks. The classical
model of portfolio selection and the single-index model by (Markowitz, 1952 and 1959) as
well as the Capital Asset Pricing Model (CAPM) that has been developed by (Sharpe, 1964)
and (Lintner, 1965), use simple linear ordinary least squares regressions (OLS). The
regression is run on the excess return of an asset on its benchmark’s (the market portfolio)
excess return, with the excess return generally being defined as over a risk-free rate. In these
models, the higher the intercept that represents the alpha, the higher the risk adjusted return,
while risk is measured as beta, relative to the benchmark.
(Fama and French, 1992) in their seminal study augment the analysis with additional
factors. They introduce two factors in addition to the benchmark or market portfolio, the
excess return of small capitalisation stocks over large capitalisation stocks (small-minus-big,
SMB) and the excess return of stocks with high ratios of book-to-market-value over ones
with low book-to-market-value (high-minus-low, HML). Not representing the end of the
factor model developments, the (Fama and French, 1992 and 1993) model had an invention
by (Carhart, 1997), who introduced the momentum of one-year stock returns as an additional
characteristic component, after (Jegadeesh and Titman, 1993) having proposed the
momentum factor. The resulting four-factor model has been used extensively in the past and
builds the baseline for many studies on performance analysis. The initial work on portfolio
theory and benchmark-oriented performance measurement has triggered a lot of following
research work, such as the arbitrage pricing theory by (Ross, 1976), an alternative to the
CAPM.
Before discussing the nature of performance analyses for selection processes, a few
technical facts concerning the use of alpha and beta as a measure of superior fund (manager)
quality are noteworthy. As alpha is simply the intercept of an OLS regression, it tells the
analyst about the (excess) return that a fund would achieve if all the explaining factors (for
example the SMB excess return) were set to zero. It is understood that the intercept is
somehow a bin for all non-random effects not caught up by the explaining factors and
therefore may be the result of a large variety of effects, not only representing the superior
ability of a funds’ manager. When it comes to beta, used as a measure of fund exposure to the
explaining factors, the use of linear regression analyses may be inappropriate especially when
11
analysing funds that show highly non-linear dependencies, for example hedge funds and
other vehicles that are subject to option-like payoff structures. However, the non-linear
effects may be included even in linear regression analyses when using respective explaining
factors.
Not criticizing the use of the four-factor model or related setups, we stress that the
four-factor model is not suitable for all kinds of performance analyses, especially when
constructing FoF portfolios. For example, if the universe of the funds under review is not
restricted very much, that is, if fund managers may have invested in a large variety of stocks
(for example a country oriented mutual fund), the four-factor model lets conclude about the
source of performance. Nevertheless, the alpha, the regressions’ intercept, is not measuring
the ability of the fund manager.
An intuitive example: A fraction of fund managers in a study sample overweighs
small caps against large caps stronger than other fund managers. In a year where small caps
subsequently perform better than their blue chip counterparts, the higher returns of the small
cap biased funds are a result of their superior ability to forecast the small cap outperformance.
In the four-factor model, this ability is not identified as ability, but is “soaked up” by the
SML term in the regression, leaving an alpha that neglects the good decision made by the
fund managers overweighing small caps. If the aim of the study is to identify where the
performance comes from, this is a favourable effect, if the study was to compare fund
managers in a selection process, it is not. This problem is crucial to any performance analysis
in the fund universe and calls for sensible selection of the factors in relation to which
information is to be obtained, thereby carefully interpreting the results obtained.
Especially when analysing funds in a FoF portfolio building process, the caveats of
simply picking high Carhart-alpha funds are clear-cut. As FoFs should be well diversified
portfolios build out of the most promising target funds, the misleading effects discussed
above may introduce biases that lead to significant deviations from this goal. In the
mentioned example, one would be underweighting funds that successfully chose the right
strategy, possibly harming future performance. Therefore, it is key to use factor models such
as the (Carhart, 1997) model in the right way. If the FoF management team is aiming at
identifying the strategy or sector relations of target funds, the factor models may suit them
well, for example when aiming at including a heterogeneous set of target funds. If they want
to identify the ones that made the right investment decisions, they should change the
respective view.

12
 In the fund selection process for a FoF, we propose a multi-step use of the (Carhart,
1997) model or similar factor models. First, the model should be used to identify by which of
the observable and identifiable factors or characteristics a fund’s performance was driven, on
a very aggregated level, for example indeed with the four factors proposed by the (Carhart,
1997) model. Second, from the initial analysis, separate classes are built for which the
analysis is re-run, yielding a more reliable picture of the underlying funds’ quality. In the re-
run(s), the factors may be adjusted in relation to the class characteristics. For example after
separating the small cap-benefited funds, one could introduce further more dis-aggregated
sector benchmarks such as the S&P Technology or the Wilshire Micro Caps. Obviously, the
use of the multi-step procedure may better suit FoF selection processes due to the possibility
of both finding different characteristic classes of funds and finding the ones which are the
best performers in the respective classes. How deep the analyses are conducted, and in which
order the analyses are performed, depends on the respective needs and the structures of the
target funds under consideration.
The arguments proposing a multi-step approach to fund selection are broadly in line
with (Daniel et. al., 1997), who favour characteristic benchmark portfolio models over the
four- factor model by (Carhart, 1997). However, even when using the proposed multi-step
analysis or benchmark portfolio building processes, the analyst may struggle to identify the
funds which steadily perform in the way that is found in the analysis.
As for any other investor, the search for performance persistence and the
interpretation of past fund returns (and the projection of those into the future) is an important
task in the FoF portfolio building process. Besides the problems discussed above, one has to
take the analysis from the cross-section to the intertemporal dimension. Work on this subject
goes back to (Jensen, 1969) and (Beebower and Bergstrom, 1977), and it were (Grinblatt and
Titman, 1989a, 1989b, 1992), (Brown and Goetzmann, 1995), (Hendricks et.al., 1993),
(Malkiel, 1995), (Elton et. al., 1996), (Daniel et. al., 1997) and (Carhart, 1997) heavily
influencing the work on performance persistence. The issue of survivorship bias in
performance persistence studies is an often discussed problem, as are the problems of short-
history samples and non-normally distributed alphas across the funds. The latter two
problems have led to the use of Bayesian and bootstrap methods, see (Pastor and Stambaugh,
2002a and 2002b) and (Kosowski et. al., 2006 and 2007) among others.
Concerning FoF performance, (Rachlin and Castro, 2007) discuss hedge fund
performance measuring for FoF managers and on the FoF layer, (Chiang et. al., 2008)
investigate the performance of real estate mutual FoFs.
13
 As all of the studies above are related to analysing the performance of funds and/or
assets, the question being addressed is how to identify winners and losers, and to identify
which of them tend to be of the same type in the future. Following the performance analysis
and identification problem, the process of fund of fund portfolio building necessitates an
appropriate selection process in the task of picking the respective funds to include in the
portfolio. This leads to a discussion of the problems concerning portfolio optimization and is
covered in the next section.
4. Building Fund of Funds

When constructing portfolios of funds, it is critical to consider both the nature of any fund, as
well as the common factors driving them. Diversification benefits stemming from low or
even negative relationships among assets are important for the expected risk and return
structure of the resulting portfolio. Since the seminal work of (Markowitz, 1952 and 1959),
this topic has been among the most researched and discussed by both academics and
practitioners, see (Steinbach, 2001) for an overview on mean-variance optimization.
(Black and Litterman, 1990 and 1992) have developed a framework that lets the
investor include his subjective views, a setup being more robust to estimation errors.
Extensions to the Black-Litterman approach have been made by (Giacometti et. al., 2007)
who use stable distributions and therefore propose models that do not suffer from the
shortcomings caused by the normality assumption of the classical models. The use of stable
Paretian distributions in financial and portfolio modelling has been studied in detail by
(Mittnik and Rachev, 1993 and Rachev and Mittnik , 2000), (Samorodnitsky and Taqqu,
1994), (Rachev and Han, 2000), and (Ortobelli et. al., 2002 and 2003).
When it comes to FoFs, a few words concerning the distinctiveness of FoF portfolio
optimization are due. In the FoF building process, one often has to choose one or several
funds out of a family of funds with very similar exposures and/or strategies. This introduces
the possibility of very high correlations among underlying funds, stemming from the fact that
those may be invested in the same companies, assets, sectors or markets. It is therefore
crucial to identify the holding structures or risk factors of target funds, as well as common
factors that are influential to the funds’ performances. Only by doing so does the FoF
management team avoid the risk of unnecessary and inefficient double or multiple exposures
to the same companies, assets, sectors or markets. This may be done by either running factor
analyses on the funds’ data or by investigating the reports of funds and/or taking into account
information available on them. If the unique and common features of the respective funds are
14
found and a set of funds in which the FoF management wants to invest is defined, the
question remains how the FoF portfolio will be build. Thereby, it is not possible to build a
FoF by viewing any target fund as a single asset.
Using an appropriate risk measure is crucial for FoF portfolio building, with
(Goodworth and Jones, 2007) focussing on non-parametric risk measurement for hedge funds
and FoFs, and (Christie, 2007) using downside leverage and event risk measures in FoFs. For
a general discussion of risk measures, see (Rachev et. al., 2008).
As is the choice of the appropriate risk measure far from trivial, so is its interpretation,
especially when being applied for optimizing a FoF portfolio. When viewing any target fund
as a single asset, one ignores the possibility that the risk that is included in one fund may also
be included in other funds.
Related to this problem is the issue of choosing not only which funds or what kinds of
funds to include in a FoF, but also how many. The question is whether including additional
funds really helps in diversifying the portfolio and thereby not averaging or counter-investing
away the characteristics of the target funds. Among others, (Connelly, 1997), (O’Neal, 1997),
(Park and Staum, 1998), (Saraoglu and Detzler, 2002), (Brands and Gallagher, 2005),
(Louton and Saraoglu, 2006), (Amo et. al., 2007) and (Kooli, 2007) discuss the problem of
FoF portfolio building. Especially the (Connelly, 1997) paper hits the point with the
discussion surrounding so-called unintended indexing, which means that by choosing too
many funds, one could end up with a costly index type investment portfolio. This does not
only come from the fact that especially many mutual funds label themselves as being actively
managed and thereby only slightly over- or underweight their holdings relative to their
benchmark. (Connelly, 1997), in citing a speech of William E. Jacques at the Institute for
International Research sponsored conference of Active vs. Passive Investment Management
argues that mixing for example a growth fund with a value fund counters the investment
strategies, thereby increasing the portfolio holdings deadweight.
With all the problems introduced in the preceding discussion, it is clear that FoF
portfolio building is by no means a trivial or at least easy task. While only a few studies on
fund portfolios exist, the literature did not yet provide a concluding answer on the questions
raised, leaving open the door for further investigations and insights.

5. Conclusion and Outlook

15
Being a large and growing part of worldwide financial investment possibilities, funds of
funds are increasingly in the focus of both practitioners and academic researchers. We lined
out some specific features, (dis) advantages, developments, questions and problems that are
special to FoFs. By doing so, we reviewed past research work in both the FoF world as well
as in the field of funds as the natural FoF underlyings.

Among the most crucial questions and challenges that we found are the following;
FoFs are very diverse according to their investment universe and need to be treated
accordingly; the fees charged by FoFs have initiated the discussion of the double-cost fee
structure; fund portfolios have to be build by taking into account that the target funds may not
be seen a single assets.

With these problems and many more still being unresolved, we stress the importance
that FoF research needs to be done with tools that are sensible in light of the special nature of
fund portfolios. Identifying the nature, risk factors and exposures of the target funds, thereby
assessing their similarities and differences, needs sophisticated and sensible approaches and
techniques.

Especially when it comes to non-linear (inter)dependencies and relationships,

classical measures and methods may not be sufficient to perform the needed analyses.
Copulas, simulation models and other inventions may be needed to identify the factors that
are crucial in FoF management. Of course, the obtained results need to be used with other
information from the due diligence and compliance assessment processes.

When building portfolios out of funds, it is clear that one cannot rely on the classical
models with an assumed normal distribution. As both the most target funds as well as their
investments exhibit non-normally distributed returns, we propose the use of stable
distributions in the portfolio building process for FoFs. Furthermore, the fact that several
target funds may be influenced by the same factors calls for methods that detect multiple
exposures or holdings.

Concluding, we stress that it is especially the need for measuring (inter) dependencies
and diversification possibilities between target funds as well as the use of modelling the
target funds returns in the appropriate way that should drive future fund and FoF research.

16
References

Amo, A.-V., H. Harasty, P. Hillion. Diversification Benefits of Funds of Hedge Funds:

Identifying the Optimal Number of Hedge Funds // Journal of Alternative Investments,
2007. - №10. - pp. 10-22.
Beebower, G.L., G.L. Bergstrom. A Performance Analysis of Pension and Profit-Sharing
Portfolios: 1966-1975 // Financial Analysts Journal, 1977. - №33. - pp. 31-42.
Black, F., R. Litterman. Asset Allocation: Combining Investor Views With Market
Equilibrium - Goldman Sachs Fixed Income Research, 1990.
Black, F., R. Litterman. Global Portfolio Optimization // Financial Analyst Journal, 1992. -
№48. -, pp. 28-43.
Brands, S., D.R. Gallagher. Portfolio Selection, Diversification and Fund-of-Funds: A Note //
Accounting & Finance, 2005. - №45. -, pp. 185-197.
Brown, S., W. Goetzmann. Performance Persistence // Journal of Finance, 1995. - №50. - pp.
679-698.
Brown, S.J., W.N. Goetzmann, B. Liang. Fees on Fees in Funds of Funds // Journal of
Investment Management, 2004. - №2. - pp. 39-56.
Carhart, M.M. On Persistence in Mutual Fund Performance // Journal of Finance, 1997. -
№52. - pp. 57-82.
Chiang, K.C.H., K. Kozhevnikov, M.-L. Lee, C.H. Wisen. Further Evidence on the
Performance of Funds of Funds: The Case of Real Estate Mutual Funds // Real Estate
Economics, 2008. - №36. - pp. 47-61.
Christie, S. Downside Leverage and Event Risk in Fund of Funds Portfolios // Journal of
Alternative Investments, 2007. - №10. - pp. 68-75.
Connelly, T.J. Multi-Fund Diversification Issues // Journal of Financial Planning, 1997. -
№10. - pp. 34-37.
Daniel, K., M. Grinblatt, S. Titman, R. Wermers. Measuring Mutual Fund Performance with
Characteristic-Based Benchmarks // Journal of Finance, 1997. - №52. - pp. 1035–1058.
Elton, E.J., M.J. Gruber, C.R. Blake. The Persistence of Risk-Adjusted Mutual Fund
Performance // Journal of Business, 1996. - №69. - pp. 133–157.
Fama, E.F., K.R. French. The Cross-Section of Expected Stock Returns // Journal of Finance,
1992. - №47. - pp. 427-465.
Fama, E.F., K.R. French. Common Risk Factors in the Returns on Stocks and Bonds //
Journal of Financial Economics, 1993. - №33. - pp. 3-56.
17
Giacometti, R., M. Bertocchi, S.T. Rachev, F.J. Fabozzi. Stable Distributions in the Black-
Litterman Approach to Asset Allocation // Quantitative Finance, 2007. - №7. - pp. 423-
433.
Goodworth, T.R.J., C.M. Jones. Factor-Based, Non-Parametric Risk Measurement
Framework for Hedge Funds and Fund-of-Funds // European Journal of Finance, 2007.
- №13. - pp. 645-655.
Gregoriou, G.N., R. Christopherson. Corporate Governance of Hedge Funds // Derivatives
Use, Trading & Regulation, 2005. - №10. - pp. 331-337.
Grinblatt, M., S. Titman. a). Mutual Fund Performance: An Analysis of Quarterly Portfolio
Holdings // Journal of Business, 1989, - №62. - pp. 393-416.
Grinblatt, M., S. Titman. b). Portfolio Performance Evaluation: Old Issues and New Insights
// Review of Financial Studies, 1989. - №2. - pp. 393-421.
Grinblatt, M., S. Titman. The Persistence of Mutual Fund Performance // Journal of Finance,
1992. - №47. - pp. 1977-1984.
Hendricks, D., J. Patel, R. Zeckhauser. Hot Hands in Mutual Funds: Short-Run Persistence of
Relative Performance 1974-1988 // Journal of Finance, 1993. - №48. - pp. 93-130.
Jegadeesh, N, S. Titman. Returns to Buying Winners and Selling Losers: Implications for
Stock Market Efficiency // Journal of Finance, 1993. - №48. - pp. 65-91.
Jensen, M. Risk, the Pricing of Capital Assets, and the Evaluation of Investment Portfolios //
Journal of Business, 1969. - №42. - pp. 167-247.
Kooli, M. The Diversification Benefits of Hedge Funds and Funds of Hedge Funds //
Derivatives Use, Trading & Regulation, 2007. - №12. - pp. 290-300.
Kosowski, R., A. Timmermann, R. Wermers, H. White. Can Mutual Fund Stars Really Pick
Stocks? New Evidence from a Bootstrap Analysis // Journal of Finance, 2006. - №61. -
pp. 2551-2596.
Kosowski, R., N.Y. Naik, M. Teo. Do Hedge Funds Deliver Alpha? A Bayesian and
Bootstrap Analysis // Journal of Financial Economics, 2006. - №84. - pp. 229-264.
Lintner, J. The Valuation of Risk Assets and the Selection of Risky Investments in Stock
Portfolios and Capital Budgets // Review of Economics and Statistics, 1965. - №47. -
pp. 13-37.
Louton, D., H. Saraoglu. Performance Implications of Holding Multiple Mutual Funds with
the Same Investment Objective // Journal of Investing, 2006. - №15 - pp. 62-78.
Malkiel, B.G. Returns from Investing in Equity Mutual Funds 1971 to 1991 // Journal of
Finance, 1995. - №50. - pp. 549-572.
18
Markowitz, H.M. Portfolio Selection // Journal of Finance, 1952. - №7. - pp. 77-91.
Markowitz, H.M. Efficient Diversification of Investments - New York: Wiley, 1959.
Mittnik S., S.T. Rachev. Modelling Asset Returns with Alternative Stable Distribution //
Economic Review, 1993. - №12. - pp. 261-330.
O'Neal, E.S. How Many Mutual Funds Constitute a Diversified Mutual Fund Portfolio? //
Financial Analysts Journal, 1997. - March/April - pp. 37-46.
Ortobelli, S., I. Huber, E. Schwartz. Portfolio Selection with Stable Distributed Returns //
Mathematical Methods of Operations Research, 2002. - №55 - pp. 265-300.
Ortobelli, S. I. Huber, S.T. Rachev, E. Schwartz. Portfolio Choice Theory with Non-Gaussian
Distributed Returns - Handbook of Heavy-Tailed Distributions in Finance, ed. Rachev,
S.T. - Amsterdam: North Holland Handbooks of Finance, 2003.
Park, J., J. Staum. Fund of Funds Diversification: How Much is Enough? // Journal of
Alternative Investments, 1998. - №1. - pp. 39-42.
Pastor, L., R. Stambaugh. a). Mutual Fund Performance and Seemingly Unrelated Assets //
Journal of Financial Economics, 2002. - №63. - pp. 315–349.
Pastor, L., R. Stambaugh. b). Investing in Equity Mutual Funds // Journal of Financial
Economics, 2002. - № 63. - pp. 351–380.
PiperJaffray. Alternative Assets- Private Equity Fund of Funds // PiperJaffray Special Report,
2003. - August -
Rachev, S.T., S. Han. Portfolio Management with Stable Distributions // Mathematical
Methods of Operations Research, 2000. - №51. - pp. 341-353.
Rachev, S., S. Ortobelli, S. Stoyanov, F.J. Fabozzi, and A. Biglova. Desirable Properties of
an Ideal Risk Measure in Portfolio Theory // International Journal of Theoretical &
Applied Finance, 2008. - №11. - pp. 19-54.
Rachev, S.T., Mittnik, S. Stable Models in Finance - New York: Wiley, 2000.
Rachlin, E., M. Castro. Hedge Funds' Delta Sharpe Score for the Fund of Funds Portfolio
Manager // Journal of Investing, 2007. - №16. - pp. 83-88.
Reddy, G., P. Brady, K. Patel. Are Funds of Funds Simply Multi-Strategy Managers with
Extra Fees? // Journal of Alternative Investments, 2007. - №10. - pp. 49-61.
Ross, S.A. The Arbitrage Theory of Capital Asset Pricing // Journal of Economic Theory,
1976. - №13- pp. 341–360.
Samorodnitsky G., M.S. Taqqu. Stable Non-Gaussian Random Variables - New York:
Chapman and Hall, 1994.

19
 Saraoglu, H., M.L. Detzler. A Sensible Mutual Fund Selection Model // Financial Analysts
Journal, 2002. - №58. - pp. 60-73.
Sharpe, W.F. Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of
Risk // Journal of Finance, 1964. - №19.- pp. 425-442.
Steinbach, M.C. Markowitz Revisited: Mean-Variance Models in Financial Portfolio
Analysis // SIAM Review, 2001. - №43. - pp. 31-87.

20

View publication stats