Credit Risk

Types of Loans
Standalone financing: When a FI is provided
funds to a single client is known as stand along
financing.
Syndicated financing: When a group of FI is
provided funds to a single client is known as
syndicated financing.
Secured loan: A loan that is backed by a first claim
on certain assets (collateral) of the borrower if
default occurs.
Unsecured loan: A loan that only has a general
claim to the assets of the borrower if default occurs.

5 Cs of Credit
Character
Capital
Capacity
Collateral
Conditions

Calculating the return on a loan

The interest rate on the loan (BR);

Any fees relating to the loan (f);
The credit risk premium on the loan (m);
The collateral backing of the loan; and
Other nonprice terms [especially compensating
balances (b) and reserve requirements(R)]
1 + k = 1 + f + (BR + m)
1+[b(1-R)]

Example Calculation of ROA on a Loan

Suppose a bank does the following:
Sets the loan rate on a prospective loan at 14 percent (where BR
12% and m 2%). Charges a 1/8 percent (or 0.125 percent) loan
origination fee to the borrower. Imposes a 10 percent
compensating balance requirement to be held as non-interestbearing demand deposits. Sets aside reserves, at a rate of 10
percent of deposits, held at the Federal Reserve (i.e., the Feds
cash-to-deposit reserve ratio is 10 percent).
1 + k = 1 + 0.00125 + (0.12 + 0.02)
1 [(0.10)(0.9)]
k = 15.52%

N U M E R I C A L
Metrobank offers one-year loans with a 9% stated
or base rate, charges a 0.25% loan origination
fee, imposes a 10% compensating balance
requirement, and must pay a 6% reserve
requirement to the Federal Reserve. The loans
typically are repaid at maturity.
If the risk premium for a given customer is 2.5%, what
is the simple promised interest return on the loan?
What is the contractually promised gross return on the
loan per dollar lent?
Which of the fee items has the greatest impact on the
gross return?

The Expected Return on a Loan

The promised return on the loan (1 + k ) that the
borrower and lender contractually agree on
includes both the loan interest rate and non
interest rate features such as fees.
The promised return on the loan, however, may
well differ from the expected and, indeed, actual
return on a loan because of default risk.
Default risk is the risk that the borrower is
unable or unwilling to fulfill the terms promised
under the loan contract.

The Expected Return on a Loan

Default risk is usually present to some degree in all loans. Thus, at the time the
loan is made, the expected return [ E ( r )] per dollar lent is related to the promised
return as follows:
1 + E(r) = p(1+k) + (1-p)0
where p is the probability of complete repayment of the loan (such that the FI
receives the principal and interest as promised) and (1 - p) is the probability of
default (in which the FI receives nothing, i.e., 0). Rearranging this equation, we get:
E(r) = p(1 +k) - 1
To the extent that p is less than 1, default risk is present. This means the FI
manager must (1) set the risk premium ( m ) sufficiently high to compensate for this
risk and (2) recognize that setting high risk premiums as well as high fees and
base rates may actually reduce the probability of repayment (p).
That is, k and p are not independent. Indeed, over some range, as fees and loan
rates increase, the probability that the borrower pays the promised return may
decrease (i.e., k and p may be negatively related).

The Expected Return on a Loan

FIs usually have to control for credit risk along two
dimensions: the price or promised return dimension
(1+k) and the quantity or credit availability dimension.
Further, even after adjusting the loan rate (by increasing
the risk premium on the loan) for the default risk of the
borrower, there is no guarantee that the FI will actually
receive the promised payments.
The actual payment or default on a loan once it is issued
may vary from the probability expected.

??
Why could a lenders expected return be lower when
the risk premium is increased on a loan? In addition to
the risk premium, how can a lender increase the
expected return on a wholesale loan? A retail loan?
An increase in risk premiums indicates a riskier pool of
clients who are more likely to default by taking on riskier
projects. This reduces the repayment probability and
lowers the expected return to the lender. In both cases
the lender often is able to charge fees that increase the
return on the loan. However, in both cases also, the fees
may become sufficiently high as to increase the risk of
nonpayment of default on the loan.

??
Why are most retail borrowers charged the
same rate of interest, implying the same risk
premium or class? What is credit rationing?
How is it used to control credit risks with
respect to retail and wholesale loans?

??
Why are most retail borrowers charged the same rate of interest, implying the
same risk premium or class?

Most retail loans are small in size relative to the overall investment portfolio of an FI,
and the cost of collecting information on household borrowers is high. As a result, most
retail borrowers are charged the same rate of interest that implies the same level of risk.

What is credit rationing? How is it used to control credit risks with respect to retail and
wholesale loans?

Credit rationing involves restricting the amount of loans that are available to individual
borrowers. On the retail side, the amount of loans provided to borrowers may be
determined solely by the proportion of loans desired in this category rather than price or
interest rate differences, thus the actual credit quality of the individual borrowers. On
the wholesale side, the FI may use both credit quantity and interest rates to control
credit risk. Typically more risky borrowers are charged a higher risk premium to control
credit risk. However, the expected returns from increasingly higher interest rates that
reflect higher credit risk at some point will be offset by higher default rates. Thus
rationing credit through quantity limits will occur at some interest rate level even though
positive loan demand exists at even higher risk premiums.

??
Why is the degree of collateral as specified in
the loan agreement of importance to the
lender? If the book value of the collateral is
greater than or equal to the amount of the
loan, is the credit risk of the lender fully
covered? Why, or why not?
Collateral provides the lender with some assets
that can be used against the amount of the loan
in the case of default. However, collateral has
value only to the extent of its market value, and
thus a loan fully collateralized at book value may
not be fully collateralized at market value.
Further, errors in the recording of collateralized
positions may limit or severely reduce the

Default Risk Models

Qualitative models
Borrower-specific factors:

Reputation: Based on the lending history of the borrower; better

reputation implies a lower risk premium.
Leverage: A measure of the existing debt of the borrower; the larger
the debt, the higher the risk premium.
Volatility of earnings: The more stable the earnings, the lower the
risk premium.
Collateral: If collateral is offered, the risk premium is lower.

Market-specific factors include:

Business cycle: Lenders are less likely to lend if a recession is

forecasted.
Level of interest rates: A higher level of interest rates may lead to
higher default rates, so lenders are more reluctant to lend under
such conditions.

Credit Scoring Model

Credit scoring models are quantitative models that use
observed borrower characteristics either to calculate a
score representing the applicants probability of default
or to sort borrowers into different default risk classes. By
selecting and combining different economic and financial
borrower characteristics, an FI manager may be able to:
Numerically establish which factors are important in explaining
default risk.
Evaluate the relative degree or importance of these factors.
Improve the pricing of default risk.
Be better able to screen out bad loan applicants.
Be in a better position to calculate any reserves needed to meet
expected future loan losses.

Credit Scoring Model

The primary benefit from credit scoring is that credit lenders can more accurately
predict a borrowers performance without having to use more resources.
With commercial loan, credit scoring models taking into account all necessary
regulatory parameters and posting an 85% accuracy rate on average, according to
credit scoring experts, 25 using these models means fewer defaults and write-offs
for commercial loan lenders.
To use credit scoring models, the manager must identify objective economic and
financial measures of risk for any particular class of borrower.
For consumer debt, the objective characteristics in a credit scoring model might
include income, assets, age, occupation, and location. For commercial debt, cash
flow information and financial ratios such as the debtequity ratio are usually key
factors. After data are identified, a statistical technique quantifies, or scores, the
default risk probability or default risk classification.
Credit scoring models include these three broad types: (1) linear probability
models, (2) logit models, and (3) linear discriminant analysis.

Linear Probability and Logit Model

The linear probability model uses past data, such as financial ratios, as inputs
into a model to explain repayment experience on old loans. The relative
importance of the factors used in explaining past repayment performance than
forecasts repayment probabilities on new loans. That is, factors explaining past
repayment performance can be used for assessing p, the probability of
repayment and the probability of default (PD).
Briefly, we divide old loans (i) into two observational groups: those that
defaulted (PDi = 1) and those that did not default (PDi =0). Then we relate
these observations by linear regression to a set of j causal variables ( Xij) that
reflect quantitative information about the ithborrower, such as leverage or
earnings. We estimate the model by linear regression of this form:
PD = + error
where j is the estimated importance of the jth variable (e.g., leverage) in explaining past repayment
experience. If we then take these estimated j s and multiply them by the observed Xijfor a prospective
borrower, we can derive an expected value of PD ifor the prospective borrower. That value can be interpreted
as the probability of default for the borrower: E(PDi) = (1 - pi) = expected probability of default, where pi is
the probability of repayment on the loan.

Example 11-2

Estimating the Probability of Repayment on a Loan Using Linear Probability Credit Scoring Models

Suppose there were two factors influencing the past

default behavior of borrowers: the leverage or debt
equity ratio (D/E) and the salesasset ratio (S/A).

Based on past default (repayment) experience, the linear

probability model is estimated as:
PDi = 0.5(D/Ei) + 0.1(S/Ai)

Assume a prospective borrower has a D/E .3 and an S/A

2.0. Its expected probability of default (PDi) can then be
estimated as:
PDi = 0.5(.3) + 0.1(2.0) = 0.35

Linear Probability and Logit Model

While this technique is straightforward as long as
current information on the Xij is available for the
borrower, its major weakness is that the
estimated probabilities of default can often lie
outside the interval 0 to 1.
The logit model overcomes this weakness by
restricting the estimated range of default
probabilities from the linear regression model to
lie between 0 and 1. Essentially this is done by
plugging the estimated value of PDi from the
linear probability model (in our example, PDi = .
35) into the following formula:
F(PDi) =
1___
1 + e-PDi

Linear Discriminant Models

While linear probability and logit models project a
value for the expected probability of default if a
loan is made, discriminant models divide
borrowers into high or low default risk classes
contingent on their observed characteristics (Xj).
Similar to these models, however, linear
discriminant models use past data as inputs into
a model to explain repayment experience on old
loans.
The relative importance of the factors used in
explaining past repayment performance then
forecasts whether the loan falls into the high or
low default class.

Linear Discriminant Models

Altmans discriminant function (credit-classification model)
takes the form:
Z = 1.2X1 + 1.4X2 + 3.3X3 + 0.6X4 + 1.0X5
where
X1 = Working capital/total assets ratio
X2 = Retained earnings/total assets ratio
X3 = Earnings before interest and taxes/total assets ratio
X4 = Market value of equity/book value of long-term debt ratio
X5 = Sales/total assets ratio
According to Altmans credit scoring model, any firm with a Z
score of less than 1.81 should be considered a high default risk
firm; between 1.81 and 2.99, an indeterminant default risk firm;
and greater than 2.99, a low default risk firm.

Example 11-3
Calculations of Altmans Z-Score

Suppose that the financial ratios of a potential

borrowing firm took the following values:
X1 = 0.2, X2 = 0, X3 = 0.20, X4 = 0.10, X5 = 2.0

The ratio X2 is zero and X3 is negative, indicating that the

firm has had negative earnings or losses in recent periods.
Also, X4 indicates that the borrower is highly leveraged.
However, the working capital ratio (X1) and the sales/assets
ratio (X5) indicate that the firm is reasonably liquid and is
maintaining its sales volume. The Z score provides an
overall score or indicator of the borrowers credit risk since
it combines and weights these five factors according to
their past importance in explaining borrower default. For
the borrower in question:
Z = 1.2(.2) + 1.4(0) + 3.3(.20) + 0.6(.10) + 1.0(2.0) = 1.64

With a Z score less than 1.81 (i.e., in the high default risk

??
What are the purposes of credit scoring models? How do these
models assist an FI manager in better administering credit?
Credit scoring models are used to calculate the probability of
default or to sort borrowers into different default risk classes.
The primary benefit is to improve the accuracy of predicting
borrowers performance without using additional resources. This
benefit results in fewer defaults and charge offs to the FI.
The models use data on observed economic and financial
borrower characteristics to assist an FI manager in (a) identifying
factors of importance in explaining default risk, (b) evaluating the
relative degree of importance of these factors, (c) improving the
pricing of default risk, (d) screening bad loan applicants, and (e)
more efficiently calculating the necessary reserves to protect
against future loan losses.

??
Suppose the estimated linear probability model is
PD = .3X1 + .2X2 0.5X3 + error, where X1 =
0.75 is the borrowers debt/equity ratio, X2 =
0.25 is the volatility of borrower earnings, and X3
= 0.10 is the borrowers profit ratio.
What is the projected probability of default for the
borrower?
What is the projected probability of repayment if
the debtequity ratio is 2.5?
What is a major weakness of the linear probability
model?

??
Describe how a linear discriminant analysis model
works. Identify and discuss the criticisms which have
been made regarding the use of this type of model to
make credit risk evaluations.

Linear discriminant models divide borrowers into high or low

default classes contingent on their observed characteristics. The
overall measure of default risk classification (Z) depends on the
values of various financial ratios and the weighted importance of
these ratios based on the past or observed experience. These
weights are derived from a discriminant analysis model.

Several criticisms have been levied against these types of models.

First, the models identify only two extreme categories of risk,
default or no default. The real world considers several categories
of default severity. Second, The relative weights of the variables
may change over time. Further, the actual variables to be
included in the model may change over time. Third, hard to
define, but potentially important, qualitative variables are omitted
from the analysis. Fourth, the real-world database of defaulted
loans is very incomplete. Finally, the model is very sensitive to

??

MNO, Inc., a publicly traded manufacturing firm in the United States,

has provided the following financial information in its application for a
loan. All numbers
are in thousands of dollars.
Assets
Liabilities
Cash

$20 Accounts payable

Accounts
receivables

90 Notes payable

90

Inventory

90 Accruals

30

Long-term debt

$30

150

Plant and
500 Equity
400
equipment
Also assume
sales $500, cost of goods sold $360, taxes $56, interest
payments
net income $44,
dividend
ratio is$700
50 percent,
Total $40,
assets
$700the Total
liab.payout
+ equity
and the market value of equity is equal to the book value.
What is the Altman discriminant function value for MNO, Inc.?
Should you approve MNO, Inc.s, application to your bank for a $500 capital
expansion loan?
If sales for MNO were $300, the market value of equity was only half of
book value, and the cost of goods sold and interest were unchanged, what
would be the net income for MNO? Assume the tax credit can be used to
offset other tax liabilities incurred by other divisions of the firm. Would your
credit decision change?
Would the discriminant function change for firms in different industries?

??
Consider the coefficients of Altmans Z score.
Can you tell by the size of the coefficients
which ratio appears most important in
assessing creditworthiness of a loan
applicant? Explain.

Although X3, or EBIT/total assets has the highest coefficient

(3.3), it is not necessarily the most important variable.
Since the value of X3 is likely to be small, the product of 3.3
and X3 may be quite small. For some firms, particularly
those in the retail business, the asset turnover ratio, X5 may
be quite large and the product of the X5 coefficient (1.0)
and X5 may be substantially larger than the corresponding
number for X3. Generally, the factor that adds most to the
Z score varies from firm to firm and industry to industry.

??
If the rate of one-year T-Bills currently is 6 percent, what is
the repayment probability for each of the following two
securities? Assume that if the loan is defaulted, no
payments are expected. What is the market-determined
risk premium for the corresponding probability of default for
each security?
a. One-year AA rated bond yielding 9.5 percent?
b. One-year BB rated bond yielding 13.5 percent?

??
A bank has made a loan charging a base lending
rate of 10 percent. It expects a probability of
default of 5 percent. If the loan is defaulted, it
expects to recover 50 percent of its money through
the sale of its collateral. What is the expected
return on this loan?

??
Assume that a one-year T-bill is currently yielding 5.5
percent and an AAA rated discount bond with similar
maturity is yielding 8.5 percent.
If the expected recovery from collateral in the event of
default is 50 percent of principal and interest, what is the
probability of repayment of the AAA rated bond? What is the
probability of default?
What is the probability of repayment of the AAA-rated bond
if the expected recovery from collateral in the case of
default is 94.47 percent of principal and interest? What is
the probability of default?
What is the relationship between the probability of default
and the proportion of principal and interest that may be
recovered in case of default on the loan?

RAROC Models
An increasingly popular model used to evaluate
(and price) credit risk based on market data is the
RAROC model.
The RAROC (risk-adjusted return on capital) was
pioneered by Bankers Trust (acquired by
Deutsche Bank in 1998) and has now been
adopted by virtually all the large banks.
The essential idea behind RAROC is that rather
than evaluating the actual or contractually
promised annual ROA on a loan, the lending
officer balances expected interest and fee income
less the cost of funds against the loans expected
risk.

RAROC Models
Further, rather than dividing annual loan income by assets lent, it is divided
by some measure of asset (loan) risk or what is often called capital at risk,
since (unexpected) loan losses have to be written off against an FIs
capital:
RAROC

= One year net income on a loan

Loan (asset) risk or capital at risk

A loan is approved only if RAROC is sufficiently high relative to a

benchmark return on capital (ROE) for the FI, where ROE measures the
return stockholders require on their equity investment in the FI.

RAROC Models
The idea here is that a loan should be made only if the risk-adjusted return
on the loan adds to the FIs equity value as measured by the ROE required
by the FIs stockholders. Thus, for example, if an FIs ROE is 15 percent, a
loan should be made only if the estimated RAROC is higher than the 15
percent required by the FIs stockholders as a reward for their investment in
the FI. Alternatively, if the RAROC on an existing loan falls below an FIs
RAROC benchmark, the lending officer should seek to adjust the loans
terms to make it profitable again. Therefore, RAROC serves as both a
credit risk measure and a loan pricing tool for the FI manager.
The numerator of the RAROC equation is relatively straightforward to
estimate. Specifically,
One year net income on loan = (Spread + Fees) x Dollar value of the loan
outstanding

??
A bank is planning to make a loan of $5,000,000 to a firm in the steel
industry. It expects to charge a servicing fee of 50 basis points. The loan has
a maturity of 8 years with a duration of 7.5 years. The cost of funds (the
RAROC benchmark) for the bank is 10 percent. Assume the bank has
estimated the maximum change in the risk premium on the steel
manufacturing sector to be approximately 4.2 percent, based on two years of
historical data. The current market interest rate for loans in this sector is 12
percent.
Using the RAROC model, determine whether the bank should make the
loan.
What should be the duration in order for this loan to be approved?
Assuming that the duration cannot be changed, how much additional
interest and fee income will be necessary to make the loan acceptable?
Given the proposed income stream and the negotiated duration, what
adjustment in the loan rate would be necessary to make the loan
acceptable?

Internal Credit Risk Rating System

Obligor Rating
A. Financial Condition including:
a. Economic and financial situation
b. Leverage
c. Profitability
d. Cash flows
B. Management and ownership structure
a. Ownership structure
b. Management and quality of internal
controls
c. Promptness/ assessment of the
willingness to pay
d. Strength of Sponsors

Facility Rating
A. Facility
a. Nature and purpose of loan
b. Loan structure
c. Product type
d. Priority of rights in case of bankruptcy
e. Degree of collateralization
f. Composition of collateral

Internal Credit Risk Rating System

Obligor Rating
C. Qualitative factors:
a. CIB report
b. Sector of business
c. Industry properties and its future
prospects
D. Others:
a. Country risk
b. Comparison to external ratings.
c. Credit information from other sources

Facility Rating
B. Collateral
a. Nature
b. Quality
c. Liquidity
d. Market value
e. Exposure of the collateral to different
risks
f. Quality of the charge

How to measure efficiency of banks

Gross profit / Loss as percentage of total assets

Net profit/ Loss as percentage of total assets
Spread as percentage of total assets
Business per employee
Return on assets
Net profit per employee percent
Non-performing assets