®

CreditMetrics Monitor page 46

Third Quarter 1998

A one-parameter representation of credit risk and transition matrices

Daniel H. Wagner Associates This paper presents a one-parameter representation of credit risk and transition matrices. We start
Dr. Barry Belkin with the CreditMetrics view that ratings transition matrices result from the “binning” of a standard
(1-610) 644-3400 normal random variable X that measures changes in creditworthiness. We further assume that X splits
[email protected] into two parts: (1) an idiosyncratic component Y, unique to a borrower, and (2) a systematic compo-
Dr. Stephan Suchower nent Z, shared by all borrowers. Broadly speaking, Z measures the “credit cycle,” meaning the values
(1-610) 644-3400 of default rates and of end-of-period risk ratings not predicted, using historical average transition
[email protected] rates, by the initial mix of credit grades. In good years Z will be positive, implying for each initial
credit rating, a lower than average default rate and a higher than average ratio of upgrades to down-
grades. In bad years, the reverse will be true. We describe a way of estimating Z from the separate
KPMG Peat Marwick LLP
transition matrices tabulated each year by Standard & Poor’s (S&P) and Moody’s. Conversely, we
Dr. Lawrence R. Forest, Jr. describe a method of calculating transition matrices conditional on an assumed value for Z.
(1-202) 739-8771
[email protected]
The historical pattern of Z depicts past credit conditions. For example, Z remains negative for most
of 1981–89. This mirrors the general decline in credit ratings over that period. In 1990–91, Z drops
well below zero as the U.S. suffers through one of its worst credit crises since the Great Depression.
The relatively high proportion of lower grade credits inherited from the 1980s together with the
1990–91 slump (Z < 0) accounts for the high number of defaults. Over 1992–97, Z has stayed positive
and credit conditions have remained benign. The movements of Z over the past 10 years correlate
closely with loan pricing.

Our focus is on how Z affects credit rating migration probabilities. However, one can also model the
effect of Z on the probability distribution of loss in the event of default (LIED), on credit par spreads,
and ultimately on the value of a commercial loan, bond, or other instrument subject to credit risk. By
parametrically varying Z, one can perform stress testing to assess the impact of changing credit con-
ditions on the value of an individual credit instrument or an entire credit portfolio to changing credit
conditions. One can also quantify how volatility in Z translates into transaction and portfolio value
volatility.

1. Defining Z risk
Following the CreditMetrics approach described by Gupton, Finger, and Bhatia (1997), we assume
that ratings transitions reflect an underlying, continuous credit-change indicator X. We further as-
sume that X has a standard normal distribution. Then, conditional on an initial credit rating G at the
G G
beginning of a year, we partition the X values into a set of disjoint bins (x g , x g + 1 ] .1 To simplify
references, we use the indices G and g to represent sequences of integers rather than letters or other
symbols. We then define the bins such that the probability of X falling within a given interval equals
the corresponding historical average transition rate (see Chart 1).

1 We observe an inconsistency among the bins for different initial ratings. For an initial borrower rating of G0, consider suc-
cessive yearly values for X of x1 and x2, in which x1 implies a rating change to G1 and x2 a change to G2. We will not find,
in general, that an X value of x1+x2 in the first year implies a rating change from G0 to G2.
 ®
CreditMetrics Monitor page 47
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

Chart 1
Relationship between continuous credit index X and rating transitions
Historical average transition rates determine bin thresholds
Probability density for X
BBB
0.40

0.35
XBBB = 0
0.30

0.25

Bin thresholds
0.20 BBB
(88.0%)

0.15

BB A
0.10 (5.0%)
(5.0%)
B
0.05 CCC (1.0%) AA
D (0.2%) (0.6%)
(0.1%)
AAA
(0.1%)
0.00 X
BBB

We write the conditions defining the bins as follows:

[1] P ( G, g ) = Φ ( x gG+ 1 ) – Φ ( x gG )

in which P(G, g) denotes the historical average G-to-g transition probability and Φ(.) represents the
standard normal cumulative distribution function. The default bin D has a lower threshold of −∞. The
AAA bin has an upper threshold of +∞. The remaining thresholds are fit to the observed transition
probabilities.

Suppose there are N ratings categories, including default. Then there are N − 1 initial grades, which
represent all the ratings, excluding default. For each of those initial grades, we observe N − 1 histor-
ical average transition rates. The Nth value results from the condition that the probabilities sum to 1.
We must determine N − 1 threshold values defining the bins. Thus, we can solve for all of the bin
boundaries.

We illustrate the process below. The starting point is the smoothed version of the 1981–97 historical
average transition matrix tabulated by S&P for 8 grades, including default (see Table 1). The corre-
sponding bins are computed using Eq. [1]. 2

2
The smoothing applied to the matrix enforces default rate monotonicity, row and column monotonicity and several of the
other regularity conditions listed in the CreditMetrics™—Technical Document. Default rate monotonicity means that
default rates rise as credit ratings go down. Row and column monotonicity means that transition rates fall as one moves
away from the main diagonal along either a row or a column. We note one exception to this rule. Default is a trapping
state. Thus, the default rate may rise above the probability of transition to neighboring non-default states.
 ®
CreditMetrics Monitor page 48
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

Consider transitions from BBB. We observe a 15 bp default rate. Using the inverse probability func-
tion for a standard normal distribution, we compute a value of about −2.97 for the upper threshold
for the default bin. Next consider the CCC bin. We get a value of about 25 bp for the sum of transition
rates to CCC or to default. Again applying the inverse probability function, we get an upper threshold
value for CCC of about −2.81. Now consider B. We compute a probability of about 1.3 percent for
transitions to B or to lower grades. Once again applying the inverse probability function, we get an
upper threshold value of −2.23. Continuing in this way for each terminal and each initial grade, we
derive all of the bin values.

Table 1
Smoothed historical average transition matrix and associated bins
Initial End-of-year credit rating
rating AAA AA A BBB BB B CCC D
Smoothed historical AAA 91.13% 8.00% 0.70% 0.10% 0.05% 0.01% 0.01% 0.01%
average transition ma- AA 0.70% 91.03% 7.47% 0.60% 0.10% 0.07% 0.02% 0.01%
trix A 0.10% 2.34% 91.54% 5.08% 0.61% 0.26% 0.01% 0.05%
BBB 0.02% 0.30% 5.65% 87.98% 4.75% 1.05% 0.10% 0.15%
BB 0.01% 0.11% 0.55% 7.77% 81.77% 7.95% 0.85% 1.00%
B 0.00% 0.05% 0.25% 0.45% 7.00% 83.50% 3.75% 5.00%
CCC 0.00% 0.01% 0.10% 0.30% 2.59% 12.00% 65.00% 20.00%
Bins corresponding to AAA (∞, −1.35) [−1.35, −2.38) [−2.38, −2.93) [−2.93, −3.19) [−3.19, −3.54) [−3.54, −3.72] [−3.72, −3.89) [−3.89, −∞)
smoothed historical AA (∞, 2.46) [2.46, −1.39) [−1.39, −2.41) [−2.41, −2.88) [−2.88, −3.09) [−3.09, −3.43) [−3.43, −3.72) [−3.72, −∞)
average transition ma- A [1.97, −1.55) [−1.55, −2.35) [−2.35, −2.73) [−2.73, −3.24) [−3.24, −3.29) [−3.29, −∞)
(∞, 3.10) [3.10, 1.97)
trix
BBB (∞, 3.50) [3.50, 2.73) [2.73, 1.56) [1.56, −1.55) [−1.55, −2.23) [−2.23, −2.81) [−2.81, −2.97) [−2.97, −∞)
BB (∞, 3.89) [3.89, 3.05) [3.05, 2.48) [2.48, 1.38) [1.38, −1.29) [−1.29, −2.09) [−2.09, −2.33) [−2.33, −∞)
B (∞, 4.11) [4.11, 3.29) [3.29, 2.75) [2.75, 2.43) [2.43, 1.42) [1.42, −1.36) [−1.36, −1.64) [−1.64, −∞)
CCC (∞, 4.27) [4.27, 3.72) [3.72, 3.06) [3.06, 2.64) [2.64, 1.88) [1.88, 1.04) [1.04, −0.84) [−0.84, −∞)

As in Belkin, Suchower, and Forest (1998), we decompose X into two parts: (1) a (scaled) idiosyn-
cratic component Y, unique to a borrower, and (2) a (scaled) systematic component Z, shared by all
borrowers. Thus, we write

[2] X = 1 – ρ Y + ρZ

We assume that Y and Z are unit normal random variables and mutually independent.3 The parameter
ρ (assumed positive) drives the correlation between Z and X; Z explains a fraction ρ of the variance
of X.

In any year, the observed transition rates will deviate from the norm (Z = 0). We can then find a value
of Z such that the probabilities associated with the bins defined above best approximate the given
year’s observed transition rates (see Chart 2).

3 The variate Z actually changes from year to year, and is modeled as following a stochastic process; therefore, it is more
proper to denote it by Zt. A reasonable stochastic model is the Ornstein-Uhlenbeck (O-U) process
dZ t = – βZ t dt + σdW t with parameters β (reciprocal of time constant) and σ (volatility); Wt is a standard Wiener pro-
cess. The O-U process is mean reverting (capturing the analogous property of the business cycle) and has a limiting sta-
tionary Gaussian distribution. The condition σ2/2β = 1 is imposed to insure that the stationary distribution has unit
variance. See Arnold (1974) for a discussion of the O-U process.
 ®
CreditMetrics Monitor page 49
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

Chart 2
Illustration of the Z value for a particular initial rating in a given year
This y ear’s ρ 1/2Z
f(x)
0.40
Distribution best Distribution best approxim ating
approxim ating this transitions in a norm al (Z = 0)
0.35 y ear’s transitions credit y ear

0.30

0.25

0.20

0.15

0.10

0.05

0.00 X
D CCC BB B BBB A AA AAA

We label that value of Z for year t, Zt.4 We determine Zt so as to minimize the weighted, mean-
squared discrepancies between the model transition probabilities and the observed transition proba-
bilities.

For this we define

 x gG+ 1 – ρZ t   x gG – ρZ t 
[3] ∆( x gG+ 1 , x gG, Z t ) = Φ  ------------------------------  – Φ  -------------------------  .
 1–ρ   1–ρ 

This is the fitted value for the G-to-g transition rate in year t. Then for a fixed ρ and a fixed t, the
least-squares problem takes the form

2
n t, g [ P t ( G, g ) – ∆( x gG+ 1 , x gG, Z t ) ]
[4]
min ∑ ∑ -----------------------------------------------------------------------------------------
∆( x gG+ 1 , x gG, Z t ) [ 1 – ∆( x gG+ 1 , x gG, Z t ) ]
- ,
Zt G g

4 It can be shown that one recovers the historical average transition matrix by integrating the transition matrices conditioned
on Zt = z with respect to the stationary unit normal distribution for z. Thus, the historical average matrix is the expectation
of the conditioned matrices over all possible values of Z.
 ®
CreditMetrics Monitor page 50
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

where Pt (G, g) represents the G-to-g transition rate observed in year t and n t, G is the number of tran-
sitions from initial grade G observed in that year. In this formula, we weight observations by the in-
verses of the approximate sample variances of Pt (G, g).5

Since we do not know the value of ρ a priori, we estimate it as follows. We apply the minimization
in Eq. [4] for 1981–97 using an assumed value of ρ. We then obtain a time series for Zt conditional
on ρ and compute the mean and variance of this series. We repeat this process for many values of ρ,
and use a numerical search procedure to find the particular ρ value for which the Zt time series has
variance of one.

We illustrate this process of solving for Zt at a single time t. We start with the S&P transition matrix
observed for 1982 (see Table 2). We hold fixed the bins determined from the historical average ma-
trix (Table 1) and fix ρ at the value determined by the search process, i.e., .0163. The indicated value
for Zt of –0.89 provides the best fit to the observed 1982 transition rates.

Table 2
S&P transition matrix for 1982 and calculations leading to Z estimate
Initial End-of-year credit rating
Statistic rating # Obs. AAA AA A BBB BB B CCC D
Observed transition matrix AAA 85 92.94% 4.71% 2.35% 0.00% 0.00% 0.00% 0.00% 0.00%
AA 220 0.46% 92.52% 6.08% 0.47% 0.47% 0.00% 0.00% 0.00%
A 480 0.00% 4.45% 84.95% 9.54% 0.64% 0.00% 0.00% 0.42%
BBB 298 0.37% 0.37% 3.26% 85.52% 9.78% 0.37% 0.00% 0.34%
BB 168 0.00% 0.68% 0.00% 2.68% 82.42% 10.05% 0.00% 4.17%
B 161 0.00% 0.00% 0.72% 0.72% 2.89% 87.50% 5.06% 3.11%
CCC 16 0.00% 0.00% 0.00% 0.00% 0.00% 7.39% 73.86% 18.75%
Fitted transition matrix AAA 85 89.34% 9.54% 0.89% 0.13% 0.07% 0.01% 0.01% 0.01%
AA 220 0.48% 89.56% 8.93% 0.77% 0.13% 0.09% 0.03% 0.01%
A 480 0.06% 1.72% 90.88% 6.14% 0.78% 0.34% 0.01% 0.07%
BBB 298 0.01% 0.20% 4.39% 88.03% 5.72% 1.33% 0.13% 0.20%
BB 168 0.00% 0.07% 0.38% 6.19% 81.63% 9.39% 1.06% 1.29%
B 161 0.00% 0.03% 0.17% 0.32% 5.56% 83.41% 4.38% 6.14%
CCC 16 0.00% 0.01% 0.06% 0.20% 1.94% 10.09% 64.53% 23.16%
Z value −0.89

Broadly speaking, Zt measures the “credit cycle,” meaning the values of default rates and of end-of-
period risk ratings not predicted (using historical average transition rates) by the initial mix of credit
grades. In good years Zt will be positive, implying for each initial credit rating a lower than average
default rate and a higher than average ratio of upgrades to downgrades. In bad years, the reverse will
be true.

5
∆( x gG+ 1 , x gG, Z t ) [ 1 – ∆( x gG+ 1 , x gG, Z t ) ]
In Eq. [4], we normalize each squared deviation by the factor ---------------------------------------------------------------------------------------------------- . This weighting fac-
n t, G
tor represents the sample variance for the G-to-g transition rate under a binomial sampling approximation such that “suc-

cess” is the occurrence of a G-to-g transition and “failure” is any other transition. A full multinomial treatment would

account for the constraint that the sample transition rates across a row must sum to one.
 ®
CreditMetrics Monitor page 51
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

2. Zt’s historical patterns

Zt’s historical movements describe past credit conditions not evident in the initial mix of ratings (see
Chart 3). Zt’s history is erratic, more so than the term “credit cycle” suggests. In particular, the fluc-
tuations do not show a stable sinusoidal pattern.

Chart 3
Zt as estimated from S&P annual transition matrices

Z
t
2.00

Turbulent 80s 90 - 91 Crisis Benign 90s

1.50

1.00

0.50

0.00

-0.50

-1.00

-1.50

-2.00
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

Zt is mostly negative over 1981–89. Credit ratings generally declined over that period as many cor-
porations increased leverage. In 1990–91, Zt drops below zero, as the U.S. suffers through one of its
worst credit crises since the Great Depression. The relatively high proportion of lower grade credits
inherited from the 1980s together with the 1990–91 credit slump (Zt < 0) accounts for a high number
of defaults. Over the period 1992–97, Zt has stayed positive and credit conditions have remained be-
nign.

Loan prices over the past 10 years correlate quite closely with the credit indicator Zt (see Chart 4).
One observes that loan spreads have generally lagged abrupt changes in credit conditions.
 ®
CreditMetrics Monitor page 52
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

Chart 4
Zt index and BB spreads

BB par spreads* Z
t
140 -2

130 -1.5

120 -1

110 -0.5

100 0

90 0.5

80 1

70 1.5

60 2
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

* Source: Loan Pricing Corporation.

Loan spreads in North America and Europe over the past 2–3 years have remained near record lows.
This suggests that the past 6 years of favorable credit conditions have made many lenders optimistic
about the future. One might ask whether the past patterns exhibited by Zt justify this optimism or any
other forecast of credit conditions.

Applying the weighted least-squares scheme for estimating the Zt, we get a ρ value of 0.0163. Thus,
systematic credit migration risk accounts for only about 1.6% of total credit migration risk over the
period 1981–97. This contrasts with equity price data, which suggests that systematic risk accounts
for about 25 percent of the total variance in an average company’s stock price. Still, the seemingly
small estimated variations in Zt translate into substantial swings in default and downgrade rates (see
the discussion following Table 3 on page 54).

The discrete time counterpart to the O-U model is a first-order autoregressive process. We fitted such
a model for Zt to the data for the 1982–97 period and obtained the following:

[5] Z t – Z t – 1 = – 0.54Z t – 1 + 1.04ε t .

Here εt is a standardized white noise sequence. The sample mean of the Zt values is −0.16, which is
statistically consistent with the assumption that the Zt process has zero mean. From Eq. [5] we obtain
the sample estimates β = .54 yr −1 and σ = 1.04 yr −1/2 for the O-U process parameters. Thus, the Zt
process has an estimated mean relaxation time of about 2 years and an estimated annual volatility of
about 1.

We performed several statistical tests on Eq. [5] for model goodness of fit. The sample estimate of
the mean of the Zt process mean is −0.16, with a standard error of 0.25. As a result, there is no sta-
 ®
CreditMetrics Monitor page 53
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

tistical basis to reject the hypothesis that the Zt process has zero mean. Based on a t-statistic value of
2.18, the hypothesis that there is no mean reversion (i.e., that β = 0) can be rejected at the .025 sig-
nificance level. The Kolmogorov-Smirnov test statistic for the model residuals has a value of
d = .17, indicating that the residuals are statistically indistinguishable from a white noise sequence
(α = .71).

The calculated R2 for the model in Eq. [5] is .24, indicating that a first-order autoregressive model
for the Zt has modest predictive power.6 However, the utility of the model is not in predicting future
values of Zt. Rather, it is to quantify how the variability in Zt that is predictable and the variability
in Zt that is not predictable each influence credit risk and the pricing of that risk.

3. Determining transition matrices as functions of Zt

We have already described a way of imputing the Zt variable from observed transition matrices. By
inverting this process, we can determine transition matrices from values of Zt.

We again use the bin values xG for each initial grade G and end-of-year grade g. Now, conditional on
Zt, we compute the probability of a G-to-g transition as

 x gG+ 1 – ρZ t   x gG – ρZ t 
[6] P t ( G, g ) = Φ  ------------------------------ – Φ  ------------------------- 
 1–ρ   1–ρ 

Table 3 shows matrices for a good year (Zt = 1), an average year (Zt = 0), and a bad year (Zt = −1).
Note that an absolute value of 1 for Zt represents a 1-standard deviation variation from “normal”
credit conditions.

6 The R2 for a second-order autoregressive model is only .33, so going to a higher order model adds little in the way of pre-
dictive power. The simply reality is that the Zt process, at least over the 17-year historical period analyzed, is quite vola-
tile.
 ®
CreditMetrics Monitor page 54
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

Table 3
Transition matrices computed using Z parameterization
Initial End-of-year credit rating
Statistic rating AAA AA A BBB BB B CCC D
Calculated transition matrix for good AAA 93.17% 6.25% 0.47% 0.06% 0.03% 0.01% 0.00% 0.00%
year (Zt = 1) AA 0.95% 92.72% 5.81% 0.41% 0.06% 0.04% 0.01% 0.01%
A 0.14% 3.02% 92.33% 3.88% 0.42% 0.17% 0.01% 0.03%
BBB 0.03% 0.41% 7.03% 88.00% 3.65% 0.73% 0.06% 0.09%
BB 0.01% 0.15% 0.73% 9.50% 82.00% 6.32% 0.61% 0.67%
B 0.00% 0.07% 0.34% 0.59% 8.58% 83.68% 3.03% 3.70%
CCC 0.00% 0.01% 0.14% 0.40% 3.30% 14.12% 65.60% 16.42%
Calculated transition matrix for AAA 91.31% 7.87% 0.67% 0.09% 0.05% 0.01% 0.00% 0.00%
average year (Zt = 0) AA 0.66% 91.24% 7.34% 0.57% 0.09% 0.06% 0.02% 0.01%
A 0.09% 2.26% 91.79% 4.98% 0.58% 0.24% 0.01% 0.05%
BBB 0.02% 0.28% 5.52% 88.28% 4.66% 1.01% 0.09% 0.14%
BB 0.00% 0.10% 0.52% 7.63% 82.13% 7.84% 0.82% 0.95%
B 0.00% 0.04% 0.23% 0.43% 6.87% 83.85% 3.71% 4.86%
CCC 0.00% 0.01% 0.09% 0.28% 2.51% 11.91% 65.39% 19.81%
Calculated transition matrix for bad AAA 89.09% 9.75% 0.92% 0.14% 0.07% 0.01% 0.01% 0.01%
year (Zt = −1) AA 0.46% 89.34% 9.13% 0.79% 0.14% 0.10% 0.03% 0.01%
A 0.06% 1.66% 90.75% 6.28% 0.80% 0.35% 0.01% 0.07%
BBB 0.01% 0.19% 4.27% 87.96% 5.85% 1.37% 0.14% 0.21%
BB 0.00% 0.07% 0.37% 6.03% 81.53% 9.58% 1.09% 1.33%
B 0.00% 0.03% 0.16% 0.31% 5.42% 83.32% 4.47% 6.30%
CCC 0.00% 0.00% 0.06% 0.20% 1.88% 9.88% 64.39% 23.58%

One observes here significant variation in the migration probabilities between the good, average, and
bad years, particularly off the main diagonal. For example, the default probability starting in grade
B is .0486 in an average year but increases to .063 in a bad year and drops to .037 in a good year. In
relative terms these are about 30% variations. Consequently, the effect of systematic risk on migra-
tion probabilities is significant.

4. Applications
The Z variable and its related formulas provide a simple one-factor description of credit portfolio risk
and credit pricing. On the pricing side, changes in credit spreads for a given grade reflect shifting
expectations regarding expected and unexpected loss. By “unexpected loss,” we mean the premium
(over expected loss) that a loan must pay to compensate for its contribution to volatility in a well-
diversified portfolio.

We can explain changes in credit spreads using Z. Suppose that the expected value of Z increases.
Then the anticipated transition rates to default and to near default go down. The probability distribu-
tion for LIED can shift downward (and change shape) as well. The effect is that both expected and
unexpected losses fall, lowering credit spreads. Suppose, alternatively, that the expected value of Z
decreases. Expected and unexpected losses go up, raising credit spreads. Thus, we can relate spread
volatility to changes in expected credit conditions.

Given a stochastic specification for Z, we can incorporate spread volatility into loan pricing models.
We are currently modifying KPMG’s Loan Analysis SystemSM by incorporating Z risk along with its
 ®
CreditMetrics Monitor page 55
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

effect on rating migration probabilities, on the distribution of LIED, and on par credit spreads. In ad-
dition, we are including Z risk as one of the factors in a multifactor model for the interest rate term
structure.

The Z variable provides a simple way of running credit scenarios. For example, one might want to
simulate the value of a credit portfolio under conditions similar to those in 1990–91. To accomplish
this, one would run a two-year simulation, setting Z equal to its 1990 value in year 1 and to its 1991
value in year 2. One would compute the associated transition matrices and use those matrices in cal-
culating credit value-at-risk.

Alternatively, one could run a large number of simulations drawing Z from a time series model, such
as the O-U process. This would provide valuable insight into how volatility in Z in response to chang-
ing credit conditions induces volatility in the mark-to-market value of a credit portfolio.

In closing, we note that Z offers only a one-factor explanation of credit risk. International data sug-
gest that one needs several factors to describe credit risk globally. The assumption that a single factor
can satisfactorily represent all systematic risk in valuing a credit portfolio needs to be tested by com-
paring model predictions of mark-to-market prices with observed market prices.

5. Summary
We have described a one-parameter representation of credit risk and transition matrices in the form
of a single systematic credit factor Z. The historical record of Z provides a succinct description of
past credit conditions. We have described a stochastic process model for Z and a way of estimating
Z from past ratings transition matrices and applied the method to rating migration data for the period
1983–97.

Our results indicate that specific risk dominates systematic risk in terms of explaining the variance
of X, the continuous variate that governs credit migration under the CreditMetrics model. Nonethe-
less, Z has a significant effect on migration probabilities, and the framework that we described can
be used to stress test a credit portfolio, i.e., to quantify the impact of changing credit conditions on
individual transaction value and portfolio value.

The Z variate can be incorporated into models for stochastic LIED and stochastic par credit spreads.
It also provides a basis for modeling the correlation between credit migration and interest rates, for-
eign currency exchange rates, and other market variables subject to systematic risk.

The information provided here is of a general nature and is not intended to address the specific cir-
cumstances of any individual or entity. In specific circumstances, the services of a professional
should be sought. The views and opinions are those of the authors and do not necessarily represent
the views and opinions of KPMG Peat Marwick LLP.
 ®
CreditMetrics Monitor page 56
Third Quarter 1998

A one-parameter representation of credit risk and transition matrices (continued)

References
Arnold, Ludwig. Stochastic Differential Equations: Theory and Applications. John Wiley & Sons,
1974.

Belkin, Barry, Stephan J. Suchower, and Lawrence R. Forest, Jr. “The effect of systematic credit risk
on loan portfolio value-at-risk and loan pricing.” CreditMetrics Monitor, 1st quarter 1998.

Gupton, Greg, M., Christopher C. Finger, and Mickey Bhatia. CreditMetrics™—Technical Docu-
ment. New York: Morgan Guaranty Trust Co., 1997.
 ®
CreditMetrics Monitor page 57
Third Quarter 1998
 ®
CreditMetrics Monitor page 58
Third Quarter 1998