International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

Credit Risk Modelling: A wheel of Risk Management

Dr. Gupta Shilpi1

Abstract

Banking institutions encounter two broad types of risks in their everyday business – credit risk
and market risk. Credit risk may be defined as the risk that borrowers might default on their
obligations, whereas market risk reflects the variability in the value of their financial position
due to changes in interest rates, exchange rates, etc. Over the last decade, rapid strides have been
made in developing Value at Risk (VaR) models for managing market risks in a portfolio
context. Such models have also been recognized for regulatory capital setting for market risks.
However, a similar approach to measure credit risk in a portfolio context was found difficult on
account of certain crucial differences between credit risk and market risk.
In the last few years, credit risk models, which attempt to measure risk in a “Portfolio” context,
and compute VaR due to credit, have emerged in the market. While significant hurdles,
especially relating to data limitation and model validation, still need to be addressed before a
VaR type model for credit risk can be accepted as an alternative to the standardized approach to
the measurement of capital, such modeling techniques have caught considerable attention
amongst the community of bankers and banking supervisors. The present study explores ball-by
ball description of credit risk modeling in reference to the Indian public sector bank
Keywords:

Credit risk, Risk Management, Credit Risk Models, Probability Density Function, Portfolio

1
Assistant Professor
Central University of Rajasthan
9, State Bank Colony Tonk Phatak Jaipur-302015
[email protected]
[email protected]

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 985
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

Introduction Credit Risk Models: Definition

Banking institutions encounter two broad and Advantages

types of risks in their everyday business –
credit risk and market risk. Credit risk may
be defined as the risk that borrowers might Credit risk models attempt to measure and
default on their obligations, whereas market manage credit risk, taking into account the
risk reflects the variability in the value of correlations in credit quality between
their financial position due to changes in difference borrowers by virtue of the fact
interest rates, exchange rates, etc. Over the that they may operate in the same industries
last decade, rapid strides have been made in and/or countries, and be influenced by the
developing Value at Risk (VaR) models for same economic forces.
managing market risks in a portfolio The primary objective of credit risk models
context. Such models have also been is to treat credit risk on a “Portfolio” a basis
recognized for regulatory capital setting for to address issues, such as, qualifying
market risks. However, a similar approach to aggregate credit risk, identifying
measure credit risk in a portfolio context concentration risk, quantifying marginal
was found difficult on account of certain risk, i.e., the effect on portfolio risk on
crucial differences between credit risk and account of the addition of a single asset,
market risk. setting risk limits, and last but not least,
While market rates change from one second quantifying economic and regulatory capital.
to the next, credit events are rare, and hence, The traditional techniques for managing
the amount of credit data available is much credit risk, the use of limits. The common
smaller. Also, whereas the historic data limits used by banks are individual/group
necessary to calculate market rate borrower limits, which seek to control the
correlations are readily available, size of exposure, concentration limits, which
correlations in credit quality cannot be seek to control concentration within
readily observed and may have to be industry, instrument type, country, tenor
inferred from other sources like equity limits, which seek to control the maximum
prices. maturity of exposures to borrowers, etc.
In the last few years, credit risk models, While the limit system takes care of the
which attempt to measure risk in a various factors, which contribute to the
“Portfolio” context, and compute VaR due magnitude of credit risk, viz., size of
to credit, have emerged in the market. While exposure, concentration risk of the
significant hurdles, especially relating to borrowers, the maturity of the exposure, etc.,
data limitation and model validation, still on a “stand-alone” basis, it does not provide
need to be addressed before a VaR type a satisfactory measure of the “concentration
model for credit risk can be accepted as an risk” and “diversification benefits” of a
alternative to the standardized approach to portfolio of exposures.
the measurement of capital, such modeling Concentration risk refers to additional
techniques have caught considerable portfolio risk resulting from increased
attention amongst the community of bankers exposure to one borrower or groups of
and banking supervisors. correlated borrowers. The common method

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 986
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

used to control concentration risk is among different borrowers. The

exposure based credit limits methodology also ignores the effect of
(individual/group limits). However, such portfolio diversification on credit risk. In
limits tend to be arbitrary in nature. Credit addition the current risk-based capital
risk models have the potential to address standards have provided incentives to banks
concentration risk in a more systematic to indulge in regulatory capital arbitrage – a
manner as they provide risk estimates, prime example is the use of asset
which give an idea of the ‘relative riskiness’ securitization by banks in the United States
of the various exposures in a portfolio. to achieve significant reduction in capital
Further, a portfolio view of credit risk requirements without materially reducing
facilitates a rational assessment of portfolio the credit risk in their books (although this is
diversification. For example, the decision to not relevant in the Indian context as the
take an ever higher exposure to a borrower securitization market is yet to take off).
will result in higher marginal risk, which Credit risk models facilitate computing a
will increase exponentially with increasing measure of economic capital reflecting more
exposure to the borrower. On the other hand, closely the perceived riskiness of the
a similar additional exposure to another underlying assets of an institution.
borrower, although having a higher absolute
risk, offers a relatively small marginal Types of Credit Risk Models
contribution to the overall portfolio risk due
to diversification benefits. Credit risk
models, therefore, help in quantifying Essentially, credit risk models can be
marginal contribution to portfolio risk on classified into two types based on the
account of addition/deletion of exposures, definition of credit loss. First, Default Mode
which in turn aid in quantifying portfolio (DM) models, also called as “two-state”
diversification benefits. models, recognize credit loss only if a
borrower defaults within the planning
Perhaps, the most significant objective,
horizon, i.e., in such models only two
which the output of a credit risk model can
outcomes are relevant – non – default and
address is in the estimation of the amount of
default. Such models are useful in situations,
capital needed to support a bank’s credit
where secondary loan markets are not
risk, termed as ‘economic capital’ it is now a
sufficiently developed to support a full
well-recognized fact, that the current risk-
mark-to-market approach. Second, Mark-to-
based capital standards for banks established
market (MTM) models, also called “multi-
by the 1988 Basle Accord have significant
state” models, recognize that ‘default’ is the
shortcomings inasmuch as the quantum of
only one of the several possible credit rating
capital arrived at under the standard may not
grades to which the instrument could
be a true measure of the riskiness of a
migrate over the planning horizon.
bank’s business. A notable weakness under
Therefore, a credit loss under the MTM
the current regime is that the risk-weighted
paradigm is defined as an unexpected
assets in the denominator of the capital ratio
reduction in portfolio value over the
may not represent the true risk, as all
planning horizon due to either deterioration
commercial credits are assigned 100 percent
in credit ratings on the underlying loans or a
risk weight, and therefore, the methodology
widening of credit risk spreads in financial
ignores the crucial difference in credit risk
markets.

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 987
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

Credit Risk Model: Building planning horizon is favoured by banks on

account of the fact that (a) accounting
statements are prepared on a yearly basis,
Blocks Type of Model (b) credits are normally reviewed on a
yearly basis, and (c) one year is a reasonable
time over which new capital could be raised
As described above, there are two types of and / or other loss mitigating action could be
models to choose from DM models and taken to eliminate future risk from the
MTM models, and the approach to the portfolio.
measurement of credit loss depends on the
type of model chosen. While the MTM Internal Credit Rating and
modeling technique is conceptually
attractive, as it recognizes the fact that
changes in an asset’s creditworthiness and Transition Matrices
its potential impact on a bank’s financial
position can occur not only on account of A reliable internal credit rating system is a
defaults but also due to events short of key component needed to implement a credit
defaults (rating downgrade), its adoption in risk model, as the probability of a credit
countries, which do not have a well- facility defaulting within the planning
developed secondary market for loans (to horizon is determined solely on the basis of
support a full MTM approach), is difficult. its internal rating.
Choice of Planning Horizon Another component required is a ‘rating
transition matrix’, which indicates the
probability of a customer migrating from the
Along with deciding on the conceptual current rating category to any other category
definition of credit loss, a bank has to within the time horizon. A sample one-year
choose a time horizon over which to transition matrix showing the credit rating
measure this loss. Generally, a constant time one year in the future is shown in the Table
horizon, such as, one-year or a hold-to- below:
maturity time horizon under which each
facility is assessed according to its maturity
is chosen. Normally, a uniform one-year

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 988
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

Table 1
Sample credit rating transition matrix
(Probability of migrating to another rating within one year as a percentage)
Credit rating one-year in the future
Transition Matrix
Initial Rating at year end (%)
Rating
AAA AA A BBB BB B CCC Default
AAA 85.50 13.10 0.40 0.60 0.15 0.14 0.10 0.01

AA 0.70 88.50 7.43 2.05 1.10 0.11 0.10 0.01

A 0.25 1.40 89.40 7.28 1.46 0.12 0.07 0.02

BBB 1.60 1.76 4.55 85.04 6.60 0.27 0.13 0.05

BB 0.05 2.85 3.43 5.04 74.50 8.81 4.04 1.28

B 0.20 0.39 9.17 8.30 2.28 64.34 10.04 5.28

CCC 0.08 0.27 1.87 2.07 12.35 25.87 40.01 17.48

Note: The credit rating transition matrix is based on the historical migration frequencies of publicly rated
corporate bonds.
Proprietary rating transition matrices of the not default, the future value would be the
above type are prepared by external rating book value at the end of the planning
agencies like the Standard & Poor and the horizon, after adding back the interest and
Moody’s based on historical migration principal payments received during the
frequencies of publicly rated corporate planning horizon. The future value of a
bonds. Above the transition matrix indicates, defaulting loan would be the recovery rate
for example, that the likelihood of an AA- measured as the loan’s book value
rated loan migrating to single A within one multiplied by (1-its loss rate given default).
year would be 7.43%. For DM models, only Computation of loss rate give defaults
the last column would be relevant, as such (LGDs) is a difficult task. Banks compute
models recognize only two states, viz., LGDs from a variety of sources which
default and non-default. include: a) internal data on the bank’s own
LGD, wherever available, b) loss rate from
Loan Valuation external reports like publicly available
regulatory reports, c ) intuitive judgments of
experienced lending officers, etc.
The current and future values of each credit
instrument at the beginning and end of the In respect of many types of credit
planning horizon have to be computed under instruments, a bank’s exposure is not known
both DM and MTM models. In the DM with certainty, but will depend on the
model, the current value of a loan is its book occurrence of future random events. In
value and the future value depends on respect of a committed line of credit, for
whether or not the borrower defaults during example, the customer’s drawdown rate
the planning horizon. If the borrower does would tend to increase as his credit quality

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 989
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

deteriorates, reflecting the higher costs of that the fortunes of the tyre industry are
alternative sources of funds. linked to that of the automobile industry.
Therefore, a rating downgrade of an
Credit risk models treat such ‘credit related
exposure in the automobile sector is very
optionality’ associated with a line of credit
likely to trigger off a similar downgrade
by treating the draw-dawn rate as a known
amongst borrowers in the tyre industry.
function of the customer’s end-of-period
However, while bankers are well aware of
credit rating. For example, consider a one-
such correlation, its quantification is
year line of credit that is initially undrawn.
difficult in practice. Quantification of such
Then, depending on the customer’s credit
correlations is the most challenging and the
grade at the end of the planning horizon,
least evolved area in credit risk modeling.
assumed end-of-period draw-down rate
Under the current generation of credit risk
would be based on the average historical
models, correlation between details/rating
draw-down experience of customers having
migrations and LGDs, between LGDs and
that future grade. In the DM framework, the
term structure of credit spreads and between
undrawn facility is converted into a loan
rating migrations and changes in credit
equivalent exposure (LEE) to make it
spreads are assumed to be zero. The only
comparable to a term loan. The LEE is
correlation effects considered are the
calculated as the expected draw-down under
correlation between defaults/rating
the line in the event the customer becomes
migrations of different customers.
insolvent by the end of the period (if the
customer remains solvent, the size of the One of the methods used to estimate the
draw-down is irrelevant in a DM model, as correlation among defaults/rating migrations
credit losses would be Zero). of different customers is to represent each
customer’s credit migration at the end of the
Credit Events Correlation planning horizon in terms of a future
realization of a migration risk factor (e.g.,
customer asset value or net worth). Thus, a
After determining the current and future
customer might be assumed to default if the
values of each credit instrument, the next
underlying value of its assets falls below
step is to consider the correlation among the
some threshold, such as, the level of his
factors determining credit-related losses.
liabilities. For MTM models, the change in
According to modern portfolio theory,
the value of a customer’s assets in relation to
portfolio credit risk is not just the
the various thresholds is often assumed to
summation of the credit risk of the
determine the change in its risk rating over
individual credit instruments comprising the
the planning horizon. It is the correlation
portfolio, there is also an element of system
between these migration risk factors, which
risk on account of joint movements in loan
determines, implicitly, the correlation
values arising from their dependence on
among borrowers’ defaults/rating
common influences. Across customers,
migrations.
correlations exist among (a) rating
migrations/default events (b) loss rate given Probability Density Function
defaults (LGDs) and (c) term structure of
credit spreads and LGDs. (Under DM
models, of course, credit spreads are Once all the parameters are specified as
irrelevant). For example, it is well known described in the above paragraphs, the credit

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 990
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

risk model can be used to quantify credit provisions are made to take care of expected
risk through a concept called ‘probability losses (which represent the amount of losses
density function (PDF)’ of credit losses over a bank expects to incur in the normal
the chosen time horizon. PDF is computed course), economic capital covers unexpected
essentially using either of the two methods credit losses (which is the amount by which
(a) Monte Carlo Simulation, or (b) actual losses exceed the expected losses).
Approximation using a mean/standard The amount of economic capital depends on
deviation approach. the target credit loss quintile chosen. Due to
the long-tailed nature of distribution of
The concept of PDF and the process of
credit losses, a target quintile in the range of
setting economic capital using the same is
99.0-99.8% interval is chosen when
explained with the help of the graph below:
compared to 95.0-99.0% interval chosen in
market models. Thus, if the confidence
interval is set at 99.97% (which corresponds
to a target insolvency rate of 0.03%), it
means that there is only a 0.03% estimated
Graph 1: PDF and Economic Capital probability that the unexpected credit losses
would exceed the amount of economic
Probability Density capital set aside by a bank corresponding to
Allocated economic capital
Function of Losses the chosen insolvency rate.
Under the mean/standard deviation approach
(which is mainly used in the context of DM
models), the PDF is assumed to take the
shape of a beta or normal distribution and
the economic capital allocation process
generally simplifies to setting capital at
some multiple of the estimated standard
deviation of the portfolio’s credit losses. The
overall portfolio risk under the method is
Expected Losses summarized as follows:
X The portfolio expected credit loss (u) over
Losses
the chosen time horizon equals the sum of
the expected losses for the individual credit
While a standard shape of PDF is yet to
facilities:
emerge unlike in the case of market risk
models (where the normal distribution has
evolved as the standard), observed portfolio
credit loss distributions are typically skewed
towards large losses as shown in the graph –
the PDF of a risky portfolio has a relatively Where, for the i facility,
long and fat tail. An important property of EDF i is the facility’s expected probability
the PDF is that the probability of credit of default,
losses exceeding a given amount X (as
shown in the graph) is equal to the shared LEE i is the bank’s expected credit
area under the PDF to the right of X. while exposure, and

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 991
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

LGD i is the expected loss rate given portfolio. For each scenario, the portfolio is
default. revaluated to reflect the new credit ratings.
Thus, a large number of possible future
The portfolios expected standard deviation
portfolio values are generated, with which
of credit losses (o) can be expressed as:
the distribution of portfolio values is
estimated. Thereafter, a target insolvency
rate is chosen and the corresponding
quantum of economic capital is computed.
Where, σ i is the stand-alone standard Conclusion
deviation of credit losses for the i facility,
and p i denotes the correlation between
credit losses on the i facility and those on While credit risk models are not a substitute
the overall portfolio. for sound credit appraisal systems, it is now
Further, the stand-alone standard deviation generally accepted that such models have
of credit losses for the i facility can be the potential to contribute significantly to
expressed as: enhancing the internal risk management
systems in banks. The future scenario in this
regard is articulated in the consultative paper
issued by the Basel Committee on Banking
Where VOL is the standard deviation of the Supervision on the new capital adequacy
facility’s LGD. framework. The Committee intends to
explore ways in which such models could
The mean/standard deviation method play an explicit part in the regulatory capital
attempts to approximate the PDF setting process. The Indian banks, especially
analytically and does not take much time for those which are internationally active, would
execution. Monte Carlo simulation do well to critically study the credit risk
technique, on the other hand, characterizes models available in the market (the two
the full distribution of portfolio losses. prominent models for which extensive
However, it is computationally burdensome technical documentation is available are
and can take several days for execution. The Credit Metrics by J.P. Morgan and Credit
technique, which is used in MTM models, Risk’ by Credit Suisse Financial Products),
involves generating scenarios, with each and prepare themselves for a model-based
scenario corresponding to a possible credit approach to the measurement of risk capital
rating of each of the obligors in the in the future.
References:

1. Jose A. Lopez, Marc R. Saidenberg (1999) “Evaluating Credit Risk Models”, Economic
Research Department Federal Reserve Bank of San Francisco
2. Munich and Frankfurt, Christian Bluhm, Ludger Overbeck, Christoph Wagner (2002),
“An Introduction to credit Risk Modeling”, Chapman & Hall/CRC press USA
3. Michel Crouhy a,*, Dan Galai b, Robert Mark (2000) “A comparative analysis of current
credit risk models” Journal of Banking & Finance 59-117

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 992
 International Journal of Research (IJR) Vol-1, Issue-4, May 2014 ISSN 2348-6848

4. Carty, L.V., Lieberman, D. (1996). “Defaulted bank loan recoveries Global Credit
Research”, Special report. Moody’s Investors Service.
5. Chen, N., Roll, R., Ross, S. (1986). “Economic forces and the stock market” Journal of
Business 59, 368-403.
6. Credit Metrics, (1997). Technical Document. JP Morgan.
7. Credit Suisse, (1997). “Credit Risk+: A Credit Risk Management Framework”. Credit
Suisse Financial Products
8. Lucas, D.J. (1995). Default correlation and credit analysis. Journal of Fixed Income, 76-
87
9. Altman, E.I. and Saunders, A. (1997). “Credit Risk Measurement: Developments over the
Last Twenty Years,” Journal of Banking and Finance, 21, 1721-1742.
10. Basle Committee on Banking Supervision, (1999). “Credit Risk Modelling: Current
Practices and Applications,” Basle Committee on Banking Supervision, Basle.
(http://www.bis.org/press/index.htm)
11. Berkowitz, J. (1999). “Evaluating the Forecasts of Risk Models,” Manuscript, Trading
Risk Analysis Group, Federal Reserve Board of Governors.
12. Carey, M., (1998). “Credit Risk in Private Debt Portfolios,” Journal of Finance, 53, 1363-
1388.
13. Diebold, F.X., Gunther, T.A. and Tay, A.S., (1997). “Evaluating Density Forecasts with
Applications to Financial Risk Management,” International Economic Review, 39, 863-
883.
14. Federal Reserve System Task Force on Internal Credit Risk Models, (1998). “Credit Risk
Models at Major U.S. Banking Institutions: Current State of the Art and Implications for
Assessments of Capital Adequacy.” Manuscript, Board of Governors of the Federal
Reserve System. (http://www.federalreserve.gov:80/boarddocs/press/General/1998/
19980529/study.pdf)
15. Nickell, P., Perraudin, W., and Varotto, S., (1998). “Ratings- Versus Equity-Based Credit
Risk Modelling: An Empirical Analysis.” Manuscript, Conference on Credit Risk
Modelling and Regulatory Implications.
16. Treacy W.F. and Carey, M. (1998). “Credit Risk Rating at Large U.S. Banks,” Federal
Reserve Bulletin, 897-921.

Credit Risk Modelling: A wheel of Risk Management Dr. Gupta Shilpi Page 993