Accenture Risk Analytics Network

Credit Risk Analytics

Accenture, as a leader in risk analytics,
works closely with banks and other
financial institutions in developing the
building blocks of credit risk analytics.

1
Lending, the core business of banks, is their Probability of Default –
daily activity for profit generation, while
risk management serves as the controlling Loss Given Default –
function for lending. Through innovation Exposure at Default Modeling
in the financial services industry over the
last decade, regulators have gradually Beginning in 2004, Basel II imposed a
encouraged banks to develop their own risk standard methodology for credit risk
management tools and enhance their risk management and introduced more
management framework. flexible regulatory supervision. This in
turn led banks to move towards the
For managing credit risk, many banks
still use expert judgment models without development and implementation of
the benefit of an accurate or integrated accurate modeling methodologies on an
framework to support their often Internal Ratings-Based (IRB) approach,
complicated risk management needs in and the quantitative-based measurement
a changing and evolving environment. of credit risk factors - Probability of
A bank’s exposure to risks which have Default (PD), Loss Given Default (LGD)
not been fully measured might lead and Exposure at Default (EAD). (See
to unexpected default rates and high Figure 1.)
write-offs, which would influence
their profit and capital requirement.

Figure 1: Banks can help reduce their capital charge by using an advanced IRB (Internal Ratings-Based) approach

Pillar 1: Banks who move up the ladder are rewarded by a reduced capital charge
Minimum Capital
Requirements

Banks use internal estimations

Advanced Internal of PD, loss given default (LGD)
Ratings-Based and exposure at default (EAD)
Approach to calculate risk weights for
exposure classes.
Increased Sophistication

Banks use internal estimations

Foundation Internal of probability of default (PD) to
Ratings-Based calculate risk weights for exposure
Approach classes. Other risk components are
standardized.

Risk weights are based on

Standardized
assessments by external credit
Approach
assessment institutions.

Potential Reduced Capital Requirements

Source: Accenture

2
Benefits of Adopting Basel
Accord Compliance
With all the relevant models in place Improved credit processing efficiency due to:
within a Basel II framework, banks can
• Streamlined and/or automated
enjoy a broad range of potential benefits,
credit processing
including:
• Improved collections management
Improved Credit Risk Return profile due to:
Reduced operational losses due to:
• Improved credit rating and monitoring
• Improved allocation of capital
• Enhanced risk-based pricing
• Reduced non-performing loans and This can also help banks potentially
bad debts improve their credit rating and
thereby provide a competitive edge over
• Optimized credit portfolio structure competitors.
Reduced Economic capital requirement
due to:
• Lower risk-weighted assets through
the adoption of an internal ratings-based
approach
• Improved capital allocation

3
Our Approach The Accenture Risk Analytics Network Probability of Default – Loss
consists of experienced members who
Gradual implementation of the complete are all dedicated to developing and Given Default and Model
credit risk management framework – implementing accurate and robust PD, Validation
using PD/LGD/EAD models as the basic LGD and EAD models for corporate,
building blocks – can help banks realize small and medium enterprises (SME) and The critical role played by internal
these benefits. Accenture typically uses a the consumer sectors. This experience models, industry leading practices
six-step credit risk management process includes providing credit scoring such and regulatory requirements dictate
consisting of: as application scorecard, behavior that financial institutions implement
scorecard and collection scorecard for an independent model validation
1. Risk identification all consumer lending products, such process to assess the quality and
as credit cards, installment loans and accuracy of their internal models.
2. Risk measurement mortgages. With many successful
assignments providing IRB approach Independent validation of internal IRB
3. Approval and control models is in increasing demand under
and risk scoring to international and
regional banks, we have deep industry Basel II.
4. Reporting and monitoring
insight and a broad array of industry
Banks worldwide need to invest and
5. Provision and capital benchmarks to support such initiatives.
implement a strong mechanism via
6. Portfolio management and capital systems to authenticate the precision and
allocation reliability of rating systems, processes,
and the appraisal of all relevant risk
These steps take into account the components.
organization’s data, transaction and
portfolio levels while aligning key In addition, a bank must also demonstrate
components such as governance, to regulators the completeness of its
policies and processes, and information internal model validation process.
technology enablement to set appropriate
As required by the Basel Accords, IRB
transaction and portfolio limits.
model validation is necessary to meet
external and internal compliance. While
the various aspects of model quality
can be assessed with complicated
quantitative procedures, qualitative
judgment is essential to guarantee
Figure 2: Validation is an integral part of the Model Life Cycle that the financial institutions are
using the correct model. As a
consequence, the efforts involve a
combination of in-depth knowledge
in analytical validation techniques as
Validate well as banking industry practices.

Financial institutions are also expected to

have frameworks in place to enable:

Design & • Initial model validation-- review of the

Monitor
Develop model development, the processes and
Model the execution of the model
Life Cycle
• Ongoing model validation-- ongoing
validation of rank-order performance
using industry wide standard metrics.

Use Implement

Source: Accenture

4
Benefits of Model Validation
Process
Accenture can add value by helping • In-depth model enhancements based on
clients implement a model validation real-world applications during validation
process. The prospective benefits include: process
• Access to experience and know-how • Prompt reporting capabilities including
gained from implementing robust practice “from issue to outcome” analysis
methodologies and processes, and model
• Top-down evaluation process for
validation efforts during previous client
the design and implementation of risk
assignments
mitigation controls
• Access to implementation benchmarks
• Support through the Accenture
that can be used by clients as part of
Analytical Network
their assessment effort
• Knowledge transfer program
• Expertise gained from working
with regional regulators on IRB
model reviews

Figure 3: Advantages of outsourcing independent validation

Basel II Regulatory Perspective Risk Management Unit

• A key requirement for IRB compliance Independent

• A key element of board and senior Department 1 Department 2 Validation Team
management reports
• A well-defined, actionable process around
ongoing reporting on model quality with Team 1 Team 2 Team 3 Team 4 Team 5
clearly defined responsibilities, metrics
and thresholds for acceptable quality
In order to comply with Basel II regulation, most banks should consider establishing internal
• Independent of model development independent validation teams to meet the requirements.
• The responsibilities of banks, Consideration should also be given to having the independent validation teams focus on
not supervisors providing effective feedback and recommendations for strengthening the models.

5
 Our Approach Our validation services include: The primary objective of the
countercyclical capital buffer is to
The goals of model validation are to: • Independent model validation: achieve the broader macro prudential
reviewing model methodologies, goal of protecting the banking sector
• Improve Basel compliance through an assumptions, data inputs, intermediate from periods of excess aggregate credit
efficient approach to risk assessment, adjustments, expert judgment and growth that have often been associated
confirming the model is operational as outcomes with the build-up of system-wide risk.
expected
• Program management: identifying As such, the common reference point
• Improve model development and program managed capabilities and end- put forward by the BCBS for taking
best-in-class validation, by identifying to-end delivery countercyclical buffer decisions is the
model inadequacies and determining credit to GDP guide.1 This buffer is up
the situations where the model is • Model and process matching:
to an additional 2.5 percent of risk-
inappropriate conducting a complete process map using
weighted assets (RWAs). The direct
template or interview inputs
implication is that the minimum capital
• Find the tradeoff between analytical requirement will increase by 30 percent,
foundation and risk judgment, on behalf • Outsourcing of independent validation,
regular model monitoring and reporting from eight percent to 10.5 percent at a
of a standard process incorporating these maximum and at the national supervisor’s
essential needs (See Figure 3.) function
discretion, depending on the different
level of Credit to GDP ratio.
In order to fulfill regulatory requirements Countercyclical Capital Buffer
prior to the Basel II compliance
application, banks had to perform an in Basel Regulation The Credit to GDP2 ratio for Hong Kong
from 1995 to 2010 and its long-run
independent validation through an Pro-cyclicality in the banking industry trend are shown in Figure 3. Using the
external vendor (a common industry was said to exacerbate the impact of the countercyclical capital buffer rule (buffer
practice). Today there are some practical banking crisis. While this is inherent to derived from the Credit to GDP ratio for
challenges to performing these “external” the industry and cannot be completely Hong Kong), we can observe the capital
independent validations. eliminated, the Basel Committee on buffer that would have been required in
Accenture can work with clients to Banking Supervision (BCBS) introduced the years 1990, 1996, 1997, 1998, 1999,
perform such validations, as well as the a framework for countercyclical capital 2009 and 2010.
regular monitoring and reporting of buffers beyond the minimum capital
requirement set in the Basel II framework. During the Asian Financial Crisis (1998
internal models to external parties. (See to 1999), it is likely that the high Credit
Figure 4.) to GDP ratios were due to very low GDP
figures rather than significant systematic
Figure 4: Accenture’s approach to performing an independent validation credit risk, and therefore the additional
capital buffer might be misleading.

Optimized Internal Resources Usage In order to understand the benefits, an

example is provided to illustrate the
Greater emphasis placed on the validation of newly implemented models. Stable impact of countercyclical capital buffer.
validation team creates value

and robust models and proper monitoring can help reduce the validation effort.
Using a common external

Re-allocation of internal risk analytics resources can help create more value:
Resources can be released for new risk management research, such as LR or
counterparty risk.

Compliance Purpose
Standardized independent validation program/Approval: Market consistency,
Acceptable thresholds.
Cost Efficiency: Economy of scale.
Consultants with adequate experience can provide advice during tests.
Provide industry benchmarks: Peer to peer comparison for all participants.

1
It is important to note that the BCBS has caveats with respect to its use, not the least of which is that the common reference point could give misleading signals if used as
a standalone measure. The BCBS proposed supervisory judgment is also exercised when countercyclical buffer decisions are made. The key role given to judgment by relevant
national authorities, and the designation of which will be left to each jurisdiction, could result in an unleveled playing field.

2
The precise methodology can be found in the Countercyclical capital buffer proposal, Issued for comment on September 10, 2010. More recent information on this topic can be
found at: http://www.bis.org/publ/bcbs187.htm and http://www.bis.org/publ/bcbs189.htm.

6
 Benefits of Countercyclical
Capital Buffer
We constructed a hypothetical3 corporate The graph in Figure 6 plots the internal For the impact analysis, we constructed
segment portfolio of HK$ 150 billion rating distribution for different years the rating distribution for the internal
from 1000 obligors in Hong Kong (HK) and Figure 7 presents the assumed model based upon Point-in-Time (PIT)
for the years 1995 to 2006 and used HK average LGD per rating grade, which and Through-the-Cycle (TTC) models.
Credit to GDP ratio to derive the timing is a downward trending curve where By comparing the impact of the
and magnitude of the countercyclical greater collateral coverage requirements internal model capital requirements,
capital buffer. (See Figure 5.) were imposed to higher default the PIT and TTC models, we can see
risk customers. With the exposure the differences attributable to the
amount fixed at HK$ 150 billion, the countercyclical capital buffer.
distribution of the exposure across
rating grades is merely determined
by customer rating distribution.

Figure 5: Countercyclical capital buffer distribution

2.5 2.5 2.5 2.5

200 2.5
190
2.0
180 2.0

170
Credit to GDP Ratio

160 1.5

150
140 1.0
0.8
130
120 0.5
0.3
110
100 0.0
1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

Capital Buffer (RHS) Credit to GDP (LHS) HP Trend (LHS)

Source: Accenture

3
This hypothetical portfolio is built from Accenture’s corporate experience with the rating migration of an internal
rating model, average LGD and EAD across rating grades.

7
 Figure 6: Internal rating distribution of a hypothetical portfolio

20%
18%
16%
Percentage of Obligors

14%
12%
10%
8%
6%
4%
2%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Rating Grades
1995 1996 1997 1998 1999 2000
2001 2002 2003 2004 2005 2006

Source: Hypothetical example created by Accenture

Figure 7: LGD distribution by rating grades

40%

38%
Percentage of Obligors

35%

33%

30%

28%

25%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Rating Grades

Source: Hypothetical example created by Accenture

8
 The chart that follows (Figure 8) The differences in impact of the For the years 1998 and 1999, banks
shows the capital requirement of our countercyclical buffer can be clearly using a TTC model will benefit from
hypothetical portfolio. We can observe observed in Figure 9. In the years 1996 a lower capital requirement of about
that for a perfect TTC model, the and 1997, the capital requirements of one percent compared to those using
capital requirement is stable across the using a PIT model are 3.6 percent and 2.8 a typical internal model, while, for
economic cycle, while the PIT model is percent lower than those associated with those using a PIT model, their capital
more volatile and highly affected by a typical internal model. These numbers requirement will be about one percent
the economic cycle. The fluctuation of are more than enough to cover the higher due to more significant rating
the capital requirement for the internal additional countercyclical capital buffer downgrades incurred during the
model is somewhere in between. The of 2 and 2.5 percent in 1996 and 1997. crisis.4 Nevertheless, as mentioned
countercyclical capital buffers in the Banks using a TTC model are different above, imposing a countercyclical
years 1996 to 1999 were imposed from those using a typical internal model. capital buffer may not have been
as an additional capital requirement appropriate during 1998 and 1999.
for the bank using the internal
From the impact analysis results, it is
model as shown in the bar chart.
clear that banks using a PIT internal
model would have been less affected
by countercyclical capital buffer due
to capital savings from the rating
upgrades of their internal models
during the credit expansion period.

Figure 8: Comparison of capital requirements for TTC (Through-the-Cycle) & PIT (Point-in-Time) models

7000
6000
Capital Buffer

5000
4000
3000
2000
1000
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Capital Requirement

Model + Buffer Model PIT TTC

Source: Hypothetical example created by Accenture

4
Increase in capital requirement will occur one year after the crisis.
9
 Figure 9: Capital requirement differences as a percentage of internal model’s RWAs

3.0%
% Difference of Capital Requirement

2.0%

1.0%

0.0%

-1.0%

-2.0%

-3.0%

-4.0%

-5.0%
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Capital Buffer PIT TTC

Source: Accenture

10
Our Approach The impact differences of PIT-focused We have designed a comprehensive
internal models will not only affect the approach to help clients undertake a
The BCBS as well as banking supervisors return on equity of the banks, but it model enhancement project, including
in many countries do not explicitly will also impact the dividend payout internal model enhancements to reduce
prohibit the use of different types of ratio, share buybacks and discretionary the impact of countercyclical buffer. (See
rating approaches. In fact, based on a bonus payments due to minimum capital Figure 11.)
survey conducted by the BCBS5, most of conservatism ratio.
the banks find it difficult to use a TTC
rating method. However, the level of As shown in Figure 10, during the
PIT focus of the internal models used by countercyclical capital buffer periods
banks may differ depending on the rating banks are required to conserve a higher
model design and the risk factors chosen. percentage of earnings for
the same level of capital ratio
(Tier 1 capital).

Figure 10: Comparison of capital requirements during countercyclical buffer period and normal period

120%
Percentage of Earnings

100%
80%

60%
40%
20%
0%
4.675%

4.875%

5.075%

5.275%

5.475%

5.675%

5.875%

6.075%

6.275%

6.475%

6.675%

6.875%

7.075%

7.275%

7.475%

7.675%

7.875%

8.075%

8.275%

8.475%

8.675%

8.875%

9.075%

9.275%

9.475%

9.675%

9.875%
Common Equity Tier 1 Ratio (including other fully loss absorbing capital)
Capital Conservatism Ratios Capital Conservatism Ratios with Countercyclical Buffer
Source: Accenture

Figure 11: The Accenture approach to enhancing internal ratings

Model Design Pilot Run and

Current State Analysis Model Enhancement
Review/Diagnostic Impact Analysis
• Review credit portfolio • Review high level • Build/enhance model • Rating migration
• Obtain model model design building blocks, such as analysis
inventories • Review each of the industry rating and • Analysis impact of
components of the model overlays rating changes on RWA
• Review segmentation
model building block • Analyze predictiveness • Pilot run
• Define scope of model of the risk factors
enhancement • Analyze the “PIT-ness” • Feedback analysis
of the models • Correlation analysis
• Documentation
• Analyze the source of • Multifactor analysis
the “PIT-ness” or • Calibration
“TTC-ness” of the model
• Master scale
• Define areas of enhancement
enhancement
Source: Accenture

5
The Internal Ratings-Based Approach, http://www.bis.org/publ/bcbsca05.pdf
11
Why Accenture

There are many reasons why Accenture Accenture’s approach focuses on

is the right partner for risk analytics collaboration, prioritizes what needs to
initiatives. Accenture has a Risk be undertaken early in an assignment
Analytics Network with experienced and aligns itself with a client’s needs and
professionals from local Asia Pacific comfort level. This delivers ‘quick wins’ to
countries. Our people have extensive stimulate organizational confidence, buy-
in-market experience, with broad and in and create focus and momentum.
diversified modeling skills. They also
bring broad industry insights, knowledge, Accenture’s solutions can accommodate
and familiarity with industry specific an adaptable business strategy, operating
benchmarking standards to each client model and solution architecture.
assignment. These solutions also provide the
necessary flexibility, built on scalable
Accenture’s broad corporate knowledge platforms to meet future needs and
acquired through the years by working growth opportunities and respond to
with leading firms developing end-to-end evolving environmental challenges,
solutions allows us to support clients in including new regulatory requirements
important transformation projects and to seize a competitive advantage.
initiatives. Our risk analytics services give
clients access to a mature quantitative
methodology, qualitative assessment
capabilities in addition to a systematic
approach, and proprietary assets to assist
them in their risk analytics capability
development and implementation.

12
Asia Pacific Risk Tokyo
Analytics Network Shingo Yamamoto
[email protected]
Global Lead Risk Analytics Direct: +81 3 3588 3820
Phillip Straley Mobile: +090 8812 1373
[email protected] India
Direct: +852 2249 2939
Mobile: +852 9186 2929 Sanjay Ojha
[email protected]
Singapore Direct: +91 124 467 2191
Christopher Loh Mobile: +91 995 369 0574
[email protected]
Direct: +65 6410 6450
Mobile: +65 9069 3860
Beijing
Kent Tianshi Xu
[email protected]
Direct: +86 10 5870 5881

13
14
About Accenture About Accenture
Management Consulting Accenture is a global management
consulting, technology services
Accenture is a leading provider of
and outsourcing company, with
management consulting services
approximately 266,000 people serving
worldwide. Drawing on the extensive
clients in more than 120 countries.
experience of its 16,000 management
Combining unparalleled experience,
consultants globally, Accenture
comprehensive capabilities across
Management Consulting works with
all industries and business functions,
companies and governments to achieve
and extensive research on the world’s
high performance by combining broad
most successful companies, Accenture
and deep industry knowledge with
collaborates with clients to help them
functional capabilities to provide
become high-performance businesses and
services in Strategy, Analytics, Customer
governments. The company generated
Relationship Management, Finance &
net revenues of US$27.9 billion for the
Enterprise Performance, Operations, Risk
fiscal year ended Aug. 31, 2012. Its
Management, Sustainability, and Talent
home page is www.accenture.com.
and Organization.

About Accenture Risk

Management
Accenture Risk Management consulting
services works with clients to create
and implement integrated risk
management capabilities designed
to gain higher economic returns,
improve shareholder value and
increase stakeholder confidence.
For more information about
Accenture Risk Management please visit
www.accenture.com/riskmanagement

Copyright © 2013 Accenture

All rights reserved.
13-2558
Accenture, its logo, and
High Performance Delivered
are trademarks of Accenture.