Special Comment

November 2006

Contact* Phone
New York
David T. Hamilton 1.212.553.1653
Richard Cantor

Measuring Corporate Default Rates

Summary
Measurement of the probability of default for a corporate exposure over a given investment horizon is often the first
step in credit risk modeling, management, and pricing. Many market practitioners base their parameter estimates on
results reported in rating agency default studies. Although the comparability of default rates reported by the agencies
has increased in recent years, many differences in default rate calculation methodologies remain and care should be
taken to understand how these differences may limit their comparability.
One important and poorly understood methodological difference is whether default rate estimates are statistically
adjusted for issuer rating withdrawals, which occur when borrowers shift from rated public to unrated private debt
finance or when all their debts are extinguished outright. Unadjusted default rates report the share of rated issuers that
were observed to have experienced a default over a particular measurement horizon. Withdrawal-adjusted default
rates, however, are estimates of the share of rated issuers expected to default under the assumption that withdrawn
issuers would have faced the same risk of default as other similarly rated issuers if they had stayed in the data sample.
Both methods have legitimate uses under appropriate assumptions, but each method makes a different statement about
default risk for a given historical data set.
In addition to their being statements of historical fact, unadjusted default rates may be useful benchmarks for the
expected likelihood of default for obligations that have fixed maximum potential tenors and expected rating withdrawal
rates similar to those exhibited by issuers in the empirical sample on which default rates were estimated. These
requirements severely limit their applicability as proxies for expected default rates in practice, however. Expected
rating withdrawal patterns for specific credit exposures are unlikely to be closely related to the historical average rating
withdrawal pattern for corporate issuer rating histories. Furthermore, unadjusted default rate estimates are likely to be
downwardly biased because rating agencies have incomplete knowledge of subsequent defaults once firms are no
longer rated.
In contrast, withdrawal-adjusted default rates are the appropriate estimates of expected default rates for
obligations with specific expected realized tenors. Adjusted default rates provide common yardsticks for default risk for
credit exposures across all sectors, regardless of differences in rating withdrawal rates. Moreover, marginal default rates
calculated using the withdrawal-adjusted method may be interpreted as default intensities, which are critical inputs to
most credit pricing models. Moody's typically reports its default rates on a withdrawal-adjusted basis, although it also
provides unadjusted default statistics as well.
In this Special Comment we review the mechanics and rationale behind Moodys corporate default rate calculation
methodology. We discuss the relative merits of adjusting or not adjusting for rating withdrawals and the importance of
the assumption that firms whose ratings are withdrawn would have faced similar default risk as firms that did not
withdraw if had they remained in the data sample. We demonstrate that the available evidence suggests this is a
reasonable assumption.

* The authors would like to thank Edward Altman, Lea Carty, Jerry Fons, Martin Fridson, Gus Harris, David Lando, and Til Schuermann for helpful discussions and com-
ments on earlier drafts of this Special Comment. The views expressed herein are solely those of the authors and Moody's Investors Service.
 Table of Content
1. Introduction ................................................................................................................................ 3
2. Cumulative Default Rate Methodology ........................................................................................ 4
3. Adjusting For Rating Withdrawals ............................................................................................... 7
4. Assessing The Neutrality Of Issuer Rating Withdrawals ............................................................. 10
Appendix A. 20-year Cumulative Default Rate Tables .................................................................... 13
Appendix B. Corporate Bond Pricing With Rating Withdrawals....................................................... 14
References ................................................................................................................................... 15

Moodys Special Comment 3

1. Introduction
The measurement of the probability of default for a corporate exposure is often the first step in credit risk modeling,
management, and pricing. Rating agency default studies are widely-used sources for estimates of these important
parameter values. The default statistics reported in rating agency studies are based on rich source data sets, containing
a large number of corporate rating histories and credit events.1 It is frequently assumed that the default statistics
reported by the rating agencies are calculated using more or less the same methodology and may, therefore, be used
interchangeably, compared, and interpreted more or less consistently. Furthermore, it is often taken for granted that
the default statistics reported by the rating agencies are equally appropriate measures of risk for a given purpose.
In the past decade there has indeed been a convergence in the methodologies used by the agencies to calculate
cumulative default rates, and their methodologies currently share many similarities. Most rating agencies emphasize
issuer-based default statistics rather than dollar-volume based statistics; average default rate estimates for an historical
time period are calculated using a cohort-based approach; and, long-term multi-year default rates are derived using a
discrete-time hazard rate method.2 Despite these similarities, the default rates for corresponding rating categories
reported by the rating agencies often differ significantly. While variations in default rates by rating category across
agencies are to be expected due to differences in rating methodologies, discongruities in the distributions of the
underlying rated populations, variations in the agencies' definitions of default, the historical time periods under study
and the periodicity of observation, an additional reason to expect differences is that each rating agency's default rate
calculation methodology differs in its statistical treatment of issuer rating withdrawals.3
Default rate calculation methodologies generally take one of two approaches to dealing with rating withdrawals
when calculating default rates: ignore them and make no adjustment; or adjust for rating withdrawals by treating them
as randomly censored data. Under the no adjustment for withdrawals method, issuers whose ratings are withdrawn are
treated as if they remained in the data sample over the entire measurement horizon. An attempt is made to monitor
their subsequent default status. If no default is observed, the firm is assumed not to have defaulted. Hence, the no
adjustment method takes a relatively simple view of the evolution of credit risk in that there are only two possible
outcomes, default or non-default. Under the withdrawal-adjusted method, issuers whose ratings are withdrawn are
treated as randomly censored data, meaning that it is assumed that firms whose ratings are withdrawn would have
faced the same risk of default as other similarly rated issuers if they had stayed in the sample. The adjusted-for-
withdrawals method recognizes that there are three possible end-of-period outcomes: default, survival, and rating
withdrawal. Rating withdrawals represent losses from the data sample before the final outcome of interest (default or
survival) is observed. Moody's default statistics are most often reported using the adjusted-for-withdrawals method,
although Moody's also reports unadjusted default rates.
Both calculation methods have legitimate uses under the appropriate assumptions, but each method makes a
different statement about default risk. As they are derived from historical corporate rating histories and default data,
the default rate estimates generated by each method represent a view of the "actual" default experience of a given data
sample. However, empirical default rates are frequently used as proxies for expected default probabilities, and it is for
this purpose that the treatment of rating withdrawals becomes an important concern. Unadjusted default rates may be
useful benchmarks for the expected likelihood of default for obligations that have fixed maximum potential tenors and
expected rating withdrawal rates similar to those exhibited by issuers in the empirical sample on which default rates
were estimated. In contrast, withdrawal-adjusted default rates are the appropriate estimates of expected default rates
for obligations with specific expected realized tenors. Withdrawal-adjusted default rates therefore provide common
yardsticks for comparing default risk for credit exposures across all sectors, regardless of difference in rating
withdrawal rates.
In many respects, the issue is similar to that studied in Altman (1989). Altman (1989) maintained that prevailing
methods for calculating multi-year bond default rates were unsuitable as estimates of expected default risk because they
failed to account for maturities, calls, and other early redemptions that occur prior to the end of a given measurement
horizon. Altman's mortality rate estimator recognized that calculating default rates based on the surviving population
was the relevant measure of expected default risk. Coming to a similar conclusion, Asquith, et. al. (1989) showed that
default rates estimates are materially affected by early bond redemptions, as nearly two-thirds of high yield bonds in
their data sample had been called, defaulted, or exchanged within 10 years of issuance.

1. Moody's database records the rating histories and defaults of over 19,000 Moody's-rated corporate and sovereign bond issuers since 1919. See Hamilton and Varma
(2006).
2. Moodys (and other rating agencies) also reports default rates derived by calculating multi-period rating transition matrices. Although we do not discuss this method in
this Special Comment, transition matrix-derived default rates which generally report rating withdrawals as a distinct state are very close to those derived using the
unadjusted method discussed later in this study.
3. Differences in default rate calculation methods aside, it is important to keep in mind that disparities in default rates across agency rating scales are likely to result from
differences across agencies fundamental rating practices. Moreover, Moodys ratings, for example, are relative rankings, and Moodys does not attempt to hit partic-
ular default rate targets when assigning corporate ratings. See Fons, Cantor, and Mahoney (2002).

4 Moodys Special Comment

 The adjustments advocated by Altman (1989) and Asquith et. al. (1989) therefore amounted to adjusting for
survival bias. But because rating agency default rates are typically issuer (or corporate family) based, adjusting for
withdrawals depends critically on the assumption of random censoring. An issuers rating may be withdrawn for a
variety of reasons. One common reason is that a company has extinguished all of its rated public debt due to scheduled
maturities, company-initiated calls, investor-initiated puts, or mergers and acquisitions. In many cases, the issuer is no
longer at risk of default after a rating withdrawal because the withdrawal event corresponds to the extinguishment all
of its debt obligations. However, in many other cases an issuer remains at risk of default after its rating has been
withdrawn because it has replaced all of its public, rated debt with unrated, typically private, debt. The relevant
question is whether issuer rating withdrawals are uninformative events or are correlated with changes in credit quality.
The remainder of this Special Comment is organized into four sections. In the first section we review the general
cumulative default rate calculation methodology followed by Moody's and other rating agencies. We also identify certain
features of Moody's default rates that distinguish them from other approaches. In Section 3 we explain the mechanics of
Moody's adjustment for rating withdrawals and discuss the rationale underlying the unadjusted and withdrawal-adjusted
methods. Following a long line of academic research, we argue that withdrawal-adjusted default rates have the most general
use for applications requiring estimates of expected future default risk for a stated investment horizon. In Section 4 we
analyze the hypothesis of the neutrality of issuer rating withdrawals. We demonstrate that the available evidence suggests the
assumption of random censoring is reasonable.

2. Cumulative Default Rate Methodology

The cumulative default rate calculation methodology used by Moody's (and other agencies) is a discrete-time
approximation of the nonparametric continuous-time hazard rate approach.4 A pool of issuers, called a cohort, is
formed on the basis of the rating held on a given calendar date (or set of dates), and the default/survival status of the
members of the cohort is tracked over some stated time horizon. The time horizon T for which we desire to measure a
default rate is divided into evenly spaced time intervals (e.g. months, years) of length t. Hence, the data is discrete in
that the time to default is not measured continuously. In each time interval, some fraction of the cohort that has
survived up to that time may default. The marginal default rate is the probability that an issuer that has survived in the
cohort up to the beginning of a particular interval t will default by the end of the time interval. The T-horizon
cumulative default rate is defined as the probability of default from the time of cohort formation up to and including
time horizon T.
Cohorts of issuers can be formed on the basis of their original ratings or on the ratings held as of the cohort
formation date. The original rating method, studied by Altman (1989), groups issuers into pools based on the first
rating that was assigned to the issuer (or one of its obligations); such pools consist only of first-time issuers that were
rated as of the cohort formation date(s).5 In contrast, the cohort rating method on which Moody's and other agencies
corporate default studies often rely are based on pools of issuers holding a given rating on the cohort date regardless of
original rating or time since issuance. Because Moody's long-term corporate ratings address the likelihood of default
over multiple time horizons, regardless of age or time to maturity, Moody's usually reports corporate default rates
based on the rating held on the cohort date rather than on original ratings.6
Mathematically, the marginal default rate in time interval t, d(t), for a cohort of issuers formed on date y holding
rating z is defined as the number of defaults x(t) from the cohort that occur in the time interval t divided by the
effective size of the cohort, n(t), at the start of time t:

(2.1) x yz (t )
d (t ) =
z
y
nyz (t )
Initially, n(t) is equal to the number of issuers in the pool holding rating z on the cohort formation date. As time
from the initial cohort date passes the size of the denominator falls because some issuers in the cohort fail to survive to
the next time interval. As we discuss in detail in the next section, differences in the default rates reported by the rating
agencies arise to a large extent because each rating agency models the default/survival process differently.

4. The method is essentially that of Cutler and Ederer (1958). This approach is sometimes referred to as the life-table or actuarial method.
5. The original rating method captures the impact of the now well-known aging or seasoning effect (i.e. the term structure of default risk for a given issuance year and rat-
ing category). Marginal default (hazard) rates exhibit more pronounced "humps" relative to the cohort rating method.
6. Bond level ratings are statements about expected loss severity, which incorporates loss-given-default as well as default probability. The ratings referenced in Moody's
default studies are senior unsecured (or estimated senior unsecured) issuer-level ratings, which control for loss severity (see Hamilton (2005)).

Moodys Special Comment 5

 Cumulative default rates for investment horizons of length T, denoted D(T), are built up from the marginal default
rates, and are found by subtracting the product of the fraction of surviving cohort members in each of the t time
intervals from unity:
T
Dyz (T ) = 1 [1 d yz ( t )]
(2.2)

t =1

Or, expanding equation 2.2 (and dropping indices for brevity):

T 1
(2.3) D (T ) = d (1) + d ( 2)[1 d (1)] + d (3)[(1 d (1))(1 d (2))] + ... + d (T )( [1 d (t )])
t =1

Equation 2.3 highlights the fact that a cumulative default rate is a conditional probability. In the first time period,
a fraction of the credit exposures in the cohort either defaults or survives. The credit exposures that survive period one
may then go on to default or survive in period two; those that survive period two may go on to default or survive in
period three, etc. Because the time periods are non-overlapping and the probability of default in each period is
assumed to be independent, the T-period cumulative default rate is defined as one minus the product of the T marginal
survival rates.
Issuer-based default rates receive particular emphasis in the rating process because the expected likelihood of
default of a bond issuer holding a given rating is expected be the same regardless of differences in the nominal sizes of
the exposures.7 For example, the expected likelihood of default for a B-rated corporate issuer should be the same
whether the size of the exposure is $200 million or $2 billion, everything else equal. Issuer-based default rates give
equal weight to all issuers in the default rate calculation. Dollar volume based default rates, which weight each
exposure by the total face (or market) value of its outstanding bonds, are useful statistics for portfolio benchmarking,
but they are less useful for forming expectations about future default probabilities.8
The frequency with which cohorts are formed also impacts the accuracy of the average default probability
estimates for a given rating category. The higher the sampling frequency equivalently, the shorter the time interval
between cohorts the more accurate the estimates of expected default rates for a given rating category become. Closer
cohort spacing captures rating changes and default events that occur in small time intervals, an important
consideration when an issuer's rating is undergoing rapid change. The effect of cohort spacing on default rate
estimates becomes clear in the following example. Consider the senior unsecured rating history for LTV Steel
Company up to its default on July 17, 1986:

Table 2.1 LTV Steel Company Rating History

Rating Date Rating Event
11/18/1970 A First rating assigned
4/26/1982 A3 Alphanumeric rating assigned
5/5/1982 Baa2 Downgraded
10/18/1982 Baa3 Downgraded
11/18/1983 Ba1 Downgraded
3/20/1985 Ba3 Downgraded
8/9/1985 B3 Downgraded
7/17/1986 Caa Defaulted

Using annual cohort spacing, LTV Steel Company's default is recorded for the A-rated cohorts from 1971-1982,
the Baa3-rated 1983 cohort, the Ba1-rated cohorts in 1984 and 1985, and the B3 1986 cohort. If one instead formed
cohorts at monthly intervals, the default event gets captured at the appropriate time horizon for every rating in its
rating history, including its A3, Baa2, Ba3 and Caa ratings that are ignored under annual cohort spacing. Moody's has
traditionally reported its average cumulative default rates calculated using annual cohort spacing (cohorts of issuers
formed on January 1 of each year). In Moody's 2005 default study, Moody's moved to monthly cohort spacing in
calculating its average cumulative default rates. Moody's believes that monthly cohort spacing strikes a reasonable
balance between the competing goals of informational efficiency and tractability.9

7. When a firm defaults on one bond it usually defaults on all its bonds due to cross-default clauses in bond indentures. Additionally, in some bankruptcy codes (e.g. U.S.
Chapter 11 and France's "sauvegarde" procedure) an automatic stay provision triggered upon a bankruptcy filing creates perfect cross default, causing all debt to
default at the same time (unless the bankruptcy judge grants a waiver). This approach is also consistent with the structural view of credit risk (e.g. Merton (1974))
which regards default as an issuer-level phenomenon that is primarily a function of firm-level characteristics, such as its operating performance and liability structure.
8. Fridson (1991) is an interesting discussion of the many different ways to measure default rates that addresses this and other topics.

6 Moodys Special Comment

 While investors may be interested in the cumulative default experience of a particular cohort, averages taken over
many cohort periods (which capture the effects of several macroeconomic and credit cycle peaks and troughs) are
required to estimate expected cumulative default probabilities. The average cumulative default rate for a given historical
time period is calculated by first averaging the period t marginal default rates across all available cohort dates y in the
historical data set Y, then calculating the cumulative rates using equation 2.2 or 2.3. Moody's average cumulative
default rates are weighted averages, where each period's marginal default rate is weighted by the relative size of the
cohort (proportion of issuers) in each time interval t.10
z
Equation 2.4 shows that the calculation of the average cumulative default rate for rating class z, D (T ) ,
z
is derived from the weighted average marginal default rates, d (t ) , calculated from all the available cohort
marginal default rates in the historical data set Y:

T
(2.4)
D (T ) = 1 [1 d z (t )]
z

t =1

where

(2.5) x z
y (t )
yY
d (t ) =
z

n
yY
z
y (t )

If, for example, one were calculating the average three-year cumulative default rate for the 2003-2005 time period
(with annual cohort spacing), one would first take the weighted average of the d(1) from each of the three cohort years.
The second year's average marginal default rate would consist of the weighted average of the two cohort years (2003
and 2004) with two years of exposure available, d(2). The third year average marginal default rate would simply consist
of the 2003 cohort's third year marginal default rate since it is the only cohort with three years of history available (i.e.
it would receive 100% weight). The average cumulative default rate would then simply be calculated according to
equation 2.4 using the weighted average marginal default rates derived using equation 2.5.
Note that this procedure for calculating average cumulative default rates maximizes the existing historical information
by using all the available rating and marginal default rate data, not just issuers with rating histories that endure for a period of
at least length T. While the third year's marginal default rate is calculated from just one cohort year (the 2005 cohort), the
three-year cumulative default rate reflects information on conditional default/survival derived from all three cohort years.
What may seem like the simplest method deriving average cumulative default rates directly from the cohort cumulative
default rates limits the estimated average to the set of cohorts with at least T full periods of data. For long-horizon default
rate averages, this requirement throws away much useful data (as well as raising the noise of the estimate). More importantly,
however, deriving average cumulative default rates directly from cohort cumulative default rates may result in seriously
biased and inconsistent estimates of expected cumulative default risk. For example, average cumulative default rates could
possibly be decreasing if the historical data sample consists of default rates that have been very high in recent cohorts but
very low in past cohorts.

9. There is a tradeoff between informational efficiency and tractability when calculating default rates using duration methods. Default/survival times are precisely mea-
sured using continuous time methods, but the resulting output may be quite unwieldy. Making default event times discrete by arbitrarily choosing the width of the mar-
ginal time intervals the distance between cohort formation dates results in some loss of information, but greatly facilitates the presentation and interpretation of cohort
cumulative default rates. For example, investors are often interested in default rates for certain discrete time horizons (e.g. one, five, ten years). As the time interval t is
allowed to shrink so as to be so small that at most one default occurs within an interval, the derived default rates approach the continuous time estimate (Kaplan and
Meier (1958)).
10. Weighted averages place greater weight on more recent cohorts as both the number and total dollar volume of bond issuance has experienced secular growth over
time. This is appealing from a statistical sampling point of view, but also because defaults tend to be correlated with periods of active bond issuance. Simple averaging
may be appropriate in some circumstances; e.g. the impact of macroeconomic fluctuations on multi-year default rates.

Moodys Special Comment 7

3. Adjusting For Rating Withdrawals
The calculation methodologies for cohort and average cumulative default rates described in the previous section are
generally followed by all the major rating agencies (again, with minor variations). Default rate calculation
methodologies diverge, however, on their assumptions about the default/survival process. Whereas most rating
agencies marginal default rates are incrementally adjusted for defaults that occurred in the past, Moody's default rates
also account for rating withdrawals that occur prior to the end of the measurement period. Rating withdrawals
complicate the calculation of default rates because there will be some issuers initially included in a cohort that will be
lost from the data sample before the final outcome of interest (default or survival) is observed. In practice, a rating
agency's approach to modeling the survival process is reflected in its calculation of the effective cohort size in each
time interval; i.e. the denominator of equation 2.1.
Under the unadjusted method, the effective size of a cohort of issuers rated z formed on date y in time interval t is
equal to the initial size of the cohort less the total number of issuers that have defaulted prior to the current time
interval (and therefore cannot default in the future).11 The denominator of the cohort marginal default rate (equation
2.1) for the unadjusted method is therefore calculated:
t 1
(3.1)
n (t ) = n (0 ) x zy (i )
z
y
z
y
i= 2

In contrast, the withdrawal-adjusted method recognizes that there are three possible end-of-period outcomes:
default, survival, and rating withdrawal.12 The cohort size at time t is calculated as in equation 3.1, but with an
additional adjustment for the number of issuers that have had their ratings withdrawn in periods prior to the current
time interval. Additionally, a small adjustment is made for rating withdrawals that occur within the current time
interval. Withdrawn ratings that occur within an interval are treated as if they were censored at the midpoint of the
interval; i.e. were at risk for half the time.13 Equation 3.2 shows the calculation of the denominator for the adjusted for
withdrawals method.
t 1 t 1
(3.2)
n yz ( t ) = n yz ( 0 ) x yz (i ) w zy ( i) 12 w yz ( t )
i =2 i =2

Unadjusted default rates are highly intuitive. They report the share of issuers that were observed to have
experienced a default over a particular time horizon. Unadjusted default rates are clearly useful benchmarks for the
likelihood of default for obligations that have fixed maximum tenors and expected rating withdrawal patterns similar to
those of the empirical sample from which the default rate estimate was derived. 14 In contrast, withdrawal-adjusted
default rates are more complex in both calculation and interpretation. Withdrawal-adjusted default rates are based in
part on hypothetical data whose accuracy depends on the assumption that issuers whose ratings are withdrawn would
have defaulted at the same average rates as other similarly-rated issuers. One might reasonably ask, therefore, why
bother adjusting default rates for rating withdrawals? Unadjusted default rates, it turns out, have three shortcomings
not shared by withdrawal-adjusted default rates that limit their usefulness as measures of expected default risk for
similarly rated obligations.
Firstly, unadjusted cumulative default rates are downwardly biased measures of default risk because one cannot
observe all defaults experienced by issuers after their ratings are withdrawn.15 Figure 3.1 gives an indication of the
magnitude of the problem. Of the 1,201 corporate bond issuers since 1980 that have defaulted and had their ratings
withdrawn, the percentage of defaults observed after the rating withdrawal date (5%) is relatively low compared to that
before the withdrawal date (95%). Many rating withdrawals occur when issuers retire their public debt with proceeds
raised through private bank borrowings. These firms remain at risk of default but rating agencies cannot easily track

11. The unadjusted method is often referred to as the "static pool" method. However, the term has been subject to some confusion. Sometimes, the term static pool is
meant to imply no adjustment for withdrawals. Other times, the term static pool has been used to refer to what we have defined (in Section 2) as the cohort approach
to calculating multi-year default rates. Using our terminology, the static pool method can be defined as a cohort-based method that does not adjust for rating withdraw-
als.
12. Moody's considers the three possible end-of-period outcomes mutually exclusive. Issuers that default and have their rating withdrawn in the same time interval are
categorized as defaults, not withdrawals.
13. The within-period adjustment is valid only if one is confident that defaults are observable after a rating withdrawal within the current time interval. The time interval
must, therefore, be reasonably short (such as one year or less). Of course, for small enough time intervals the effect of the within period adjustment is immaterial.
14. A prime example is static synthetic corporate CDOs, which reference the debt obligations of a large number of corporations over a common and fixed maturity. In the
event that all of the public debts and syndicated loans of a corporation are paid off, the risk in the CDO associated with that entity disappears. In such a structure, the
historical average rating withdrawal pattern of the typical corporate issuer may be very relevant. For cash CDOs, on the other hand, the pattern of issuer rating with-
drawals is less relevant, since it is unlikely to be closely related to the realized maturity patterns loans and bonds that comprise the structures collateral pool.

8 Moodys Special Comment

their subsequent default experience. Hence, without adjustments for rating withdrawals, measured default rates are
likely to understate true long-term issuer default rates. Under the withdrawal-adjusted method, rating withdrawals are
viewed as events that randomly censor rating histories and events of default that occur after rating withdrawals need
not be tracked. Rather, it is assumed that issuers whose ratings are withdrawn, had they remained in the sample, would
have experienced the same rate of default as that experienced by similarly-rated issuers that did remain in the sample.

Figure 3.1 Distribution of Distance between Observed Withdrawal and Default Times

100%

90% N = 1,201
C umulative Percent of Issuers

80%

70%

60%

50%

40%

30%

20%

10%

0%
<-10 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 >10

Y ears between Withdrawal and Default Date

Secondly, the relevance of unadjusted default statistics as guides of future expected default experience is limited to
sets of issuers with similar rating withdrawal patterns. While all issuers that carry the same rating can reasonably be
expected to have roughly the same withdrawal-adjusted default rates, their unadjusted default rates are likely to vary
significantly. Borrowers from different industries within the non-financial corporate sector and across different sectors
altogether (non-financial corporates, financial corporates, and sovereigns) have very different rating withdrawal
patterns and should therefore be expected to have markedly different unadjusted default rates by rating category.
Moreover, structured finance securities have rating withdrawal patterns that differ considerably from one another and
from those of the broad universe of corporate issuers.16 As a result, it is hard to imagine how one can realistically use
unadjusted corporate issuer default rates to benchmark and compare the risks of such diverse obligations. Withdrawal-
adjusted default rates, in contrast, facilitate comparisons of default rates across asset classes with markedly different
rating withdrawal patterns.
Lastly, credit spreads are unlikely to be closely related with unadjusted default rates. For a risk neutral investor, the
appropriate discount rate for debt obligations subject to default risk is, in theory, equivalent to the risk free rate r(t)
plus a spread to account for expected loss in default.17 Hence, R(t) = r(t) + d(t)L, where d(t) is the expected marginal
default rate and L is the expected (and assumed constant) rate of loss given default. If we assume that investors recovery
nothing in the event of default (i.e. L=1), then the required spread s(t) for a risk-neutral investor is simply the expected
marginal default rate: s(t) = R(t)-r(t) = d(t). The appropriate measure of d(t) is the withdrawal-adjusted marginal default
rate, not the unadjusted marginal default rate. If an investors expected investment horizon were, say, 10 years, then
(s)he would only require compensation for default risk only on exposures expected to survive for at least 10 years. In

15. It has been asserted (e.g. DeRosa-Farag, et. al. (1999)) that withdrawal-adjusted cumulative default rates are "biased" too high due to the correction for data censor-
ing. However, this perspective confuses concerns about sample size and statistical significance with issues of bias. Consider the following oft-used hypothetical
example. Suppose that there were 10 bond issuers in a cohort, nine of which had their ratings withdrawn over a 10-year time span for benign reasons such as merg-
ers or retirement of debt. If, in the 10th year, the one company that was still rated were to default, the withdrawal-adjusted marginal default rate would be 100%. This
sample statistic is not biased. However, because only one issuer was at risk of default in its tenth year, the statistical reliability of the 100% point estimate is virtually nil.
In order for the unadjusted default rate to reach 100%, all nine of the censored issuers would need to default after their ratings were withdrawn. The unadjusted
method assumes that there were ten issuers at risk of default in the tenth year, lowering the empirical marginal default rate estimate to 10%. The small sample prob-
lem does not go away by changing the definition of the default rate.
16. See Hu (2004).
17. In addition to compensation for expected credit losses, credit spreads may also be influenced by tax effects, interest rate risk premia, and a variety of other potential
sources of risk premia. The academic literature in this area is large. See Duffie and Singleton (2003) for an overview.

Moodys Special Comment 9

fact, as long as withdrawal-adjusted marginal default rates are used in pricing, no further adjustments for realized
rating withdrawals are required. In Appendix B we demonstrate this argument with an example.
As proxies for expected default probabilities, the advantages of withdrawal-adjusted default rates are consequently
threefold: they avoid the downward bias that can arise from incomplete knowledge of defaults for firms whose ratings
are withdrawn; they provide a common yardstick for measuring default risk for issuers and obligations across different
sectors regardless of differences in rating withdrawal rates and, thus, associate a single time profile of default rates for
each rating category for all types of credit exposures. Lastly, they provide useful and relevant data for pricing a wide
variety of debt obligations.
Adjusting cumulative default rates for rating withdrawals, however, changes the meaning of the reported default
rates from the simple concept of a share of a pool of issuers that were observed to default over a specific horizon.
Rather, withdrawal-adjusted default rates report the historical frequency of default over a particular horizon
conditional on having a rating outstanding for that length of time. For example, the withdrawal-adjusted 10-year
cumulative default rate provides the answer to the following question: if one expects to have exposure to an issuer for T
years, what is its probability of default? As long as rating withdrawals are randomly censored, then withdrawal-
adjusted rates are appropriate estimates. The unadjusted default rate provides the answer to a simpler question: if one
has a maximum potential maturity of T years, what is the expected probability of default? The answer provided,
however, is potentially downwardly biased (as discussed above) and is only accurate if the issuer withdrawal experience
of the underlying sample matches the expected rating withdrawal experience of the obligation in question.
An example should make the differences between the two methods clear. Table 3.1 shows detailed calculation of the
1- through 10-year cumulative default rates for the January 1, 1996 cohort of B-rated corporate bond issuers. The table
shows the number of defaults, x(t) and rating withdrawals, w(t), in each year after the cohort formation date, as well as the
effective size of the denominator in each time interval, n(t), for each of the two methods. The table also shows the
marginal default rates and the resulting cumulative default rates for each method (calculated using equation 2.2).

Table 3.1 10-Year Cumulative Default Rates: Adjusted vs. Unadjusted Methods
January 1, 1996 Cohort of B-Rated Corporate Issuers
Withdrawal-Adjusted Method Unadjusted Method
t x( t ) w( t ) n( t ) d( t ) D( t ) n( t ) d( t ) D( t )
0 0 0 519 0.00% 0.00% 519 0.00% 0.00%
1 7 55 491.5 1.42% 1.42% 519 1.35% 1.35%
2 13 51 431.5 3.01% 4.39% 512 2.54% 3.85%
3 19 61 362.5 5.24% 9.41% 499 3.81% 7.51%
4 12 42 292 4.11% 13.13% 480 2.50% 9.83%
5 17 23 247.5 6.87% 19.10% 468 3.63% 13.10%
6 21 12 213 9.86% 27.07% 451 4.66% 17.15%
7 19 28 172 11.05% 35.13% 430 4.42% 20.81%
8 8 22 128 6.25% 39.18% 411 1.95% 22.35%
9 4 14 102 3.92% 41.57% 403 0.99% 23.12%
10 1 14 84 1.19% 42.26% 399 0.25% 23.31%

Table 3.1 shows that, at any given measurement horizon, the withdrawal-adjusted method results in higher default
rate estimates than the unadjusted method, with the difference growing larger as the time horizon lengthens. The
unadjusted 10-year cumulative default rate shows that 23.31% of issuers originally in the cohort defaulted by the tenth
year. The unadjusted method is calculated as if the 322 issuers whose ratings were withdrawn had remained in the
cohort and did not default over the entire 10 year measurement period. In contrast, the withdrawal-adjusted 10-year
cumulative default rate method yields an estimate of 42.26%. Default rates calculated using the withdrawal-adjusted
method are based on the number of issuers that remain at risk (i.e. have not previous defaulted nor had their ratings
withdrawn) of default in each time interval. For example, the marginal default rate in the tenth year is 25 basis points
under the unadjusted method; under the withdrawal-adjusted method, it is nearly five times higher, 1.19%. The
withdrawal-adjusted approach recognizes that at the start of the tenth year only 84 issuers actually remained at risk of
default. Appendix A shows average cumulative unadjusted and withdrawal-adjusted default rates for a 20 year time
horizon.
In addition to the three advantages of the withdrawal-adjusted method discussed above, the method also generates
default probability estimates with intuitive and appealing statistical characteristics. The data in Table 3.1 illustrates that
cumulative default rates calculated using the withdrawal-adjusted method will, at sufficiently long time horizons,
approach 100% much more quickly than unadjusted default rates. This has a natural statistical interpretation:

10 Moodys Special Comment

conditional on survival, all firms will likely eventually default. Cumulative default rates for a given cohort calculated
using the unadjusted method, on the other hand, may never approach 100% over any measurement horizon. In order
for the cumulative default rate to approach 100%, all the issuers whose ratings were withdrawn would need to be
observed to ultimately default.
As we have mentioned several times in the preceding sections, the accuracy of the withdrawal-adjusted default rate
measure depends critically upon the validity of the assumption of random data censoring. We analyze the validity of
this assumption in the next section.

4. Assessing The Neutrality Of Issuer Rating Withdrawals

An issuer rating withdrawal indicates that Moody's has ceased to rate all the publicly rated bonds of an issuer. At the bond
level, rating withdrawals are overwhelmingly correlated with scheduled or anticipated redemptions. Table 5.1 shows the
reasons for rating withdrawals organized into five categories. The table shows that of the 137, 414 bond18 rating
withdrawals between 1980 and 2005, 97% were associated with maturity, calls, puts, conversions, or mergers. The
business reasons category includes instances where the issuer chose to stop paying for a rating or the size of the bond issue
was increased or decreased (and re-rated), or Moody's removed the rating because of lack of information from the issuer.
The defaulted category includes cases where the bond rating was withdrawn on or shortly after the date of default.

Table 5.1 Reasons for Rating Withdrawals, 1980-2005

Reason %Bonds %Issuers
Matured 68.84% 32.00%
Called, put, converted, etc. 27.85% 43.97%
Reason unknown 3.15% 21.31%
Business reasons 0.11% 2.02%
Defaulted 0.05% 0.70%
N 137,414 3,275

Table 5.1 also shows that 76% of issuer rating withdrawals corresponded to the final maturity, call, etc. of its bonds.
However, even if it were known that all the bonds of an issuer were withdrawn due to, say, maturity (therefore making
default on those particular bonds impossible), the rationale behind the firm's decision exit the rated bond market
might reveal information about its default risk. Bond rating withdrawals, which are closely associated with maturities
and redemptions of specific bonds, and issuer rating withdrawals, which are related to a firm's decision to exit or issue
in the rated public bond market, reflect different corporate finance choices of a firm. It is impossible to know with
complete certainty whether an issuer defaults (or would have defaulted) after all its ratings are withdrawn. It is in this
sense that issuer rating withdrawals are censoring events.
An issuer rating withdrawal might signal heightened future credit risk. For example, a bond issuer experiencing
financial distress may be forced into the private or short-term debt market. Contrarily, an issuer rating withdrawal
might be negatively correlated with default risk if issuers experiencing improving credit quality choose to pay off their
rated debt obligations or replace debt with equity. As a matter of rating policy, the rating outstanding immediately
prior to a withdrawal is intended to reflect Moody's view of the credit at the time of the withdrawal.19 Hence, rating
withdrawals are supposed to be neutral events that are not systematically correlated with changes in default risk.
The neutrality of issuer rating withdrawals is, ultimately, an empirical question, yet there is almost no published
research on the subject. Carty (1997) is the only study that has attempted to assess whether treating rating withdrawals
as randomly censored data is justified. Carty's analysis using Moody's data was, however, indirect because he analyzed
the reasons for bond rating withdrawals and their correlation with issuer rating withdrawals. In this section we attempt
to asses whether the assumption that firms whose ratings are withdrawn would have faced similar default risk as firms
that did not withdraw if had they remained in the data sample is valid. We seek to answer two specific questions.
Firstly, does default risk increase or decrease leading up to or shortly following the rating withdrawal date? Secondly, is
the level of default risk correlated with rating withdrawal events? Establishing the neutrality of issuer rating
withdrawals is not straightforward as there is no direct statistical test for random versus informative censoring. We
therefore attempt to infer the neutrality of withdrawals by examining several indications of default risk near rating

18. Includes coupon, discount, and convertible bonds.

19. See Moody's (2004).

Moodys Special Comment 11

withdrawal dates. Collectively, the results we present in this section provide evidence that the assumption of random
censoring is reasonable.
A large body of literature20has shown that default rates vary with past rating actions: default rates are relatively
higher conditional on a past downgrade and relatively lower conditional on a past upgrade. If rating withdrawal rates
exhibit the same dependence on past rating actions, then we might have cause to doubt the random censoring
hypothesis. For example, if rating withdrawal rates are higher conditional on a past downgrade, then rating
withdrawals might represent "hidden" defaults. Table 4.1 shows average one-year issuer rating withdrawal rates
conditional on rating upgrades, downgrades, and no changes in the prior year based on monthly cohorts of corporate
issuers between 1983 and 2005. (A rating action represents a change of one alphanumeric rating notch or more; e.g.
Ba2 to Ba3.) For investment-grade rated issuers in the aggregate, issuer rating withdrawal rates do not appear to
exhibit strong systematic differences when conditioned on rating changes in the year prior to the rating withdrawal.
Among speculative-grade rated issuers, those whose ratings were unchanged in the past year exhibit somewhat higher
rates of rating withdrawal. Overall, the likelihood of a rating withdrawal does not appear to be influenced by rating
changes in the past year.
Similar results are evident when issuer rating withdrawal frequencies are conditioned on outlook and review
(Watchlist) status. Moody's rating outlooks and reviews provide indications of the likely direction and timing of future
credit rating changes. Cantor and Hamilton (2004) and Cantor and Hamilton (2005) showed that default rates for
similarly rated issuers differ when conditioned on rating outlook and Watchlist status. Similar to the analysis of rating
actions, a positive correlation between outlooks/review and rating withdrawal rates would cast a doubt on the validity
of the assumption of uninformative rating withdrawals. Table 4.2 shows average one-year issuer rating withdrawal
rates between 1995 and 2005 conditional on outlook and Watchlist status held at the start of each monthly period.21
Issuers with outlooks other than stable have exhibited higher rating withdrawal rates, but there does not appear to be a
systematic bias in one direction: one-year rating withdrawal rates are roughly symmetric for positive and negative
outlooks and Watchlist assignments.

Table 5.2 One-Year Rating Withdrawal Rates Conditional on Prior Rating Action, 1983-2005
Rating Action in Prior Year
Upgraded Unchanged Downgraded
Investment-Grade 5.39% 4.20% 4.30%
Speculative-Grade 9.62% 12.64% 9.88%
All Rated 6.79% 6.76% 7.02%

Table 5.3 One-Year WR Rates Conditional on Outlook Status, 1995-2005

Outlook on Cohort Date
Watch Up Positive Stable Negative Watch Down
Investment-Grade 6.25% 4.79% 3.70% 3.62% 5.78%
Speculative-Grade 7.04% 7.99% 6.56% 7.51% 5.54%
All Rated 6.54% 6.41% 4.92% 5.51% 5.70%

Moody'sKMV EDFs offer a way to analyze changes in default risk both before and after an issuer rating
withdrawal occurs. We analyzed monthly Moody'sKMV EDFs in the 12 months before and the 6 months after
withdrawal dates for 5,577 issuers between 1998 and 2005. The data was divided into two samples: 206 issuers with
EDF data available that actually experienced a rating withdrawal, as well as a control group consisting of 5,371 issuers
that did not experience a rating withdrawal. For the non-withdrawn subset, we measured EDFs around the same
rating withdrawal dates as for the withdrawn subset. Because EDFs can vary between 0 and 0.20, and because we are
interested in changes in EDFs around withdrawal dates, we normalized the EDFs so that they are equal to 1 in the
twelfth month prior to the rating withdrawal date. Figure 4.1 shows the average EDF indices around rating withdrawal
dates for the two groups of issuers. The graph shows that the changes in the average EDFs of the two subsets exhibit
very similar patterns in the months leading up to and after the rating withdrawal date. Both series exhibit a slight
upward trend leading up to the rating withdrawal date, but the magnitude of the change is relatively small.

20. See, for example, Altman (1991), Carty (1997), and Cantor and Hamilton (2004).
21. Moody's introduced the Watchlist in 1991 and rating outlooks in 1995.

12 Moodys Special Comment

 Figure 5.1 Average EDF Index around Withdrawal Dates
1.6
1.5
1.4
1.3

EDF Index
1.2

1.1
1.0
0.9
0.8
0.7
0.6
-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6

T ime to R ating W I thdrawal (Months)

Not W ithdrawn W ithdrawn

As a final assessment of the neutrality hypothesis we compared the weighted average levels of credit ratings for
issuers that experienced a rating withdrawal to those that did not. We formed annual cohorts for 4,833 corporate bond
issuers over the 1983 to 2005 time period, and recorded their estimated senior unsecured ratings from the cohort year
to cohort year plus 5 years. To calculate the average rating level at each point in time, we weighted ratings using
Moody's CDO rating factors.22 To the extent that rating levels are correlated with future default risk (as clearly
demonstrated by Moodys default studies), any significant differences observed in the rating levels in the stratified
samples would cast doubt on the neutrality hypothesis.
Table 4.3 presents the results of this exercise. The table shows the number of notches of difference between the
ratings of the withdrawn sub-sample and the non-withdrawn sub-sample. By construction, the average ratings for the
two sub-samples on the cohort dates are the same; as time elapses rating changes may occur. Negative values in the
table indicate the number of rating notches lower the withdrawn sub-sample was relative to the non-withdrawn sub-
sample. The relevant question is whether the sub-sample that experienced a rating withdrawal exhibited significant
rating improvement or degradation relative to the sub-sample that did not experience a rating withdrawal. The data
shows that 47% of the time the average ratings of the sub-samples is the same at year 5; 42% of the time the average
ratings of the withdrawn sub-sample is lower than the not withdrawn sample by one rating notch at year 5. Hence, it
appears that the average ratings of a sub-set of issuers whose ratings are ultimately withdrawn exhibit little difference
compared to the average ratings of issuers whose ratings are not withdrawn over the same time period.

Table 5.4 Withdrawn vs. Non-Withdrawn Factor-Weighted Rating Notch Differences

Cohort Rating +1 year +2 years +3 years +4 years +5 years
Aaa 0 0 0 -1 0
Aa1 0 0 -1 -1 0
Aa2 0 0 0 -1 0
Aa3 -1 -1 0 0 -1
A1 -1 -1 0 -1 -1
A2 0 0 0 -1 -1
A3 -1 -1 -1 -1 -1
Baa1 0 0 -1 -1 -1
Baa2 0 0 1 0 0
Baa3 0 -1 -1 -1 -2
Ba1 0 0 -1 -1 -1
Ba2 0 0 0 0 0
Ba3 0 0 0 -1 -1
B1 0 0 0 -1 -1
B2 0 0 0 0 0
B3 0 0 0 0 0
Caa1 0 0 -1 -1 -1
Caa2 0 0 0 0 0
Caa3 -1 0 0 0 0
Sample period: 1983-2005

22. Moody's CDO rating factors are described in Yoshizawa and Witt (2003).

Moodys Special Comment 13

 Appendix A. 20-Year cumulative default rate tables
Table A1 Average Cumulative Default Rates by Whole Letter Rating, Unadjusted vs. Withdrawal-Adjusted
Unadjusted
Years After Cohort Formation Date
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Aaa 0.00 0.00 0.00 0.02 0.08 0.14 0.19 0.25 0.30 0.36 0.41 0.47 0.52 0.55 0.57 0.60 0.63 0.64 0.64 0.64
Aa 0.01 0.02 0.04 0.09 0.15 0.20 0.25 0.29 0.32 0.34 0.37 0.40 0.45 0.49 0.51 0.54 0.57 0.60 0.62 0.65
A 0.02 0.10 0.21 0.31 0.41 0.51 0.60 0.70 0.79 0.87 0.95 1.02 1.08 1.13 1.21 1.28 1.36 1.43 1.50 1.55
Baa 0.18 0.49 0.86 1.25 1.60 1.91 2.18 2.41 2.64 2.87 3.08 3.28 3.47 3.64 3.80 3.92 4.03 4.12 4.20 4.27
Ba 1.18 2.98 4.85 6.52 7.86 8.90 9.68 10.34 10.90 11.40 11.84 12.25 12.59 12.89 13.12 13.34 13.52 13.64 13.75 13.84
B 5.41 11.04 15.41 18.46 20.66 22.17 23.21 23.85 24.29 24.59 24.78 24.91 25.00 25.08 25.14 25.18 25.19 25.20 25.20 25.20
Caa-C 19.90 29.37 34.86 37.81 39.32 40.16 40.58 40.88 41.06 41.18 41.23 41.23 41.23 41.23 41.23 41.23 41.23 41.23 41.23 41.23

Withdrawal-Adjusted
Years After Cohort Formation Date
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Aaa 0.00 0.00 0.00 0.03 0.11 0.18 0.27 0.36 0.45 0.56 0.66 0.78 0.90 0.97 1.04 1.12 1.22 1.25 1.25 1.25
Aa 0.01 0.02 0.05 0.12 0.19 0.29 0.38 0.46 0.51 0.58 0.65 0.76 0.92 1.06 1.16 1.29 1.44 1.58 1.74 1.91
A 0.02 0.10 0.24 0.37 0.51 0.67 0.83 1.01 1.22 1.42 1.63 1.82 2.02 2.21 2.49 2.81 3.17 3.53 3.88 4.20
Baa 0.18 0.53 0.98 1.52 2.06 2.60 3.13 3.65 4.23 4.89 5.50 6.17 6.85 7.56 8.24 8.84 9.41 9.97 10.44 10.91
Ba 1.23 3.31 5.75 8.26 10.57 12.65 14.48 16.28 18.05 19.86 21.62 23.41 25.15 26.82 28.29 29.78 31.14 32.17 33.15 33.97
B 5.65 12.35 18.65 24.09 29.06 33.50 37.47 40.71 43.59 46.12 47.56 48.77 49.65 50.51 51.26 51.77 51.96 52.12 52.12 52.12
Caa-C 21.12 33.53 43.47 51.01 56.52 61.05 64.58 68.50 71.98 74.72 75.16 75.16 75.16 75.16 75.16 75.16 75.16 75.16 75.16 75.16
Sample Period: 1970-2005, monthly cohort spacing
Moodys Special Comment
14
Appendix B. Corporate bond pricing with rating withdrawals
Using a reduced form model similar to that of Fons (1994), we show that the credit spread required by a risk neutral
investor is independent of the rate withdrawal; i.e., the frequency that a bond prepays prior to maturity. The equation
below represents the price a risk neutral investor would pay at time zero for a risky bond with face value of 1 payable at
maturity date T. The bond pays an annual coupon c(t). The bond is subject to three possible end-of-year outcomes: it
may default, it may withdraw, or it may survive. In each year, the issuer may default with probability d(t); if the issuer
does not default, it may then pay off the bond early and withdraw with probability w(t). For simplicity, we assume that
loss-given-default for the bond is 100%. r(t) denotes the risk-free rate.

T
T
T
c (i 1) 1 + c (i )
P(0, T ) = [1 d (i )][1 w(i 1)]w(i ) T
+
i =1 i =1 i =1 T
(1 + r (i 1)) (1 + r (i ))
t =1 t =1

T
T
T
c (i 1)
+ [1 d ( i 1)][1 w(i 1)]d (i ) T

i =1 i =1 i =1
(1 + r (i 1))
t =1

T

+ [1 d (i )][1 w( i )] T +
T
c (i ) 1
i =1 i =1 T
(1 + r (i )) (1 + r ( i ))
t =1 t =1
The three terms on the right-hand side of the equation are the probability weighted payoffs in the event of
withdrawal, default, and survival to maturity, respectively. In the event of a withdrawal prior to maturity, an investor
receives all the coupons and principal up to the withdrawal date. If default occurs in year t, an investor only receives
coupons paid prior to t and loses the entire face value upon default. The last term in the equation is the payoff if the
bond survives i.e. does not withdraw and does not default to maturity.
Suppose the coupon paid in each year is equal to the sum of the risk free rate and the marginal default probability
adjusted for survival,

r (t ) + d (t )
c(t ) =
1 d (t )
Upon substituting this value of c(t) into the formula, it can easily be seen that the price of the bond at time 0
should be 1 and all the terms involving w(t) drop out. That is, the bond will price at par if the coupon rate is set equal
to the risk free rate plus an adjustment for default risk; however, the withdrawal rate has no impact on the required
spread. Since credit spreads are compensation for bearing risk, only marginal default rates matter for pricing, not
marginal withdrawal rates.

Moodys Special Comment 15

References
Altman, E. I., 1989. "Measuring Corporate Bond Mortality and Performance," Journal of Finance 44: 909-22.
Asquith, P., Mullins, D. W., Wolff, E. D, 1989. "Original Issue High Yield Bonds: Aging Analyses of Defaults,
Exchanges, and Calls," Journal of Finance 44: 923-52.
Carty, L. V., 1997. "Measuring Changes in Corporate Credit Quality," Moody's Global Credit Research, July.
Cutler, J. S., Ederer, F., 1958, "Maximum Utilization of the Life Table Method in Analyzing Survival," J. Chronic
Dis. 8:699-712.
DeRosa-Farag, S., Blau, J., Matousek, P., Chandra, I., 1999. "Default Rates in the High Yield Market," Journal of Fixed
Income, June: 7-31.
Duffie, D. and K. Singleton, 2003, Credit Risk: Pricing, Measurement, and Management. Princeton University Press.
Fons, J., 1994. "Using Default Rates to Model the Term Structure of Credit Risk," Financial Analysts Journal,
September/October: 25-32.
Fons, J., Cantor, R., and Mahoney, C., 2002. "Understanding Moodys Corporate Bond Ratings and Rating Process,"
Moody's Global Credit Research, June.
Fridson, M. S., 1991. "Everything You Ever Wanted to Know about Default Rates," Extra Credit, Merrill Lynch High
Yield Strategy, July/August: 4-14.
Hamilton, D. T., 2005. "Moody's Senior Ratings Algorithm and Estimated Senior Ratings," Moody's Global Credit
Research, July.
Hamilton, D. T., Varma, P., 2005. "Default and Recovery Rates of Corporate Bond Issuers," Moody's Global Credit
Research, January.
Hu, J., 2004. Structured Finance Rating Transitions: 1983-2003, Moody's Global Credit Research, February.
Kaplan, E. L. and P. Meier, 1958, "Non-Parametric Estimation from Incomplete Observations," Journal of the
American Statistical Association, 53: 437-481.
Merton, R. C., 1974. "On the Pricing of Corporate Debt: The Risk Structure of Interest Rates," Journal of Finance,
29: 449-470.
Moody's Investors Service, 2004. "Moody's Guidelines for the Withdrawal of Ratings," Moody's Global Credit Research,
January.
Yoshizawa, Y. and G. Witt, 2003. "Moody's Approach to Rating Synthetic CDOs," Moody's Structured Finance Rating
Methodology, July.

16 Moodys Special Comment

To order reprints of this report (100 copies minimum), please call 1.212.553.1658.
Report Number: 100779

Authors Production Associate Senior Production Associate

David T. Hamilton Tara Cheparev Wing Chan
Richard Cantor

Copyright 2006, Moodys Investors Service, Inc. and/or its licensors and affiliates including Moodys Assurance Company, Inc. (together, "MOODYS"). All rights reserved. ALL
INFORMATION CONTAINED HEREIN IS PROTECTED BY COPYRIGHT LAW AND NONE OF SUCH INFORMATION MAY BE COPIED OR OTHERWISE REPRODUCED, REPACKAGED,
FURTHER TRANSMITTED, TRANSFERRED, DISSEMINATED, REDISTRIBUTED OR RESOLD, OR STORED FOR SUBSEQUENT USE FOR ANY SUCH PURPOSE, IN WHOLE OR IN PART, IN
ANY FORM OR MANNER OR BY ANY MEANS WHATSOEVER, BY ANY PERSON WITHOUT MOODYS PRIOR WRITTEN CONSENT. All information contained herein is obtained by
MOODYS from sources believed by it to be accurate and reliable. Because of the possibility of human or mechanical error as well as other factors, however, such information is provided as
is without warranty of any kind and MOODYS, in particular, makes no representation or warranty, express or implied, as to the accuracy, timeliness, completeness, merchantability or fitness
for any particular purpose of any such information. Under no circumstances shall MOODYS have any liability to any person or entity for (a) any loss or damage in whole or in part caused by,
resulting from, or relating to, any error (negligent or otherwise) or other circumstance or contingency within or outside the control of MOODYS or any of its directors, officers, employees or
agents in connection with the procurement, collection, compilation, analysis, interpretation, communication, publication or delivery of any such information, or (b) any direct, indirect,
special, consequential, compensatory or incidental damages whatsoever (including without limitation, lost profits), even if MOODYS is advised in advance of the possibility of such
damages, resulting from the use of or inability to use, any such information. The credit ratings and financial reporting analysis observations, if any, constituting part of the information
contained herein are, and must be construed solely as, statements of opinion and not statements of fact or recommendations to purchase, sell or hold any securities. NO WARRANTY,
EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OF ANY SUCH RATING OR OTHER
OPINION OR INFORMATION IS GIVEN OR MADE BY MOODYS IN ANY FORM OR MANNER WHATSOEVER. Each rating or other opinion must be weighed solely as one factor in any
investment decision made by or on behalf of any user of the information contained herein, and each such user must accordingly make its own study and evaluation of each security and of
each issuer and guarantor of, and each provider of credit support for, each security that it may consider purchasing, holding or selling.
MOODYS hereby discloses that most issuers of debt securities (including corporate and municipal bonds, debentures, notes and commercial paper) and preferred stock rated by
MOODYS have, prior to assignment of any rating, agreed to pay to MOODYS for appraisal and rating services rendered by it fees ranging from $1,500 to $2,400,000. Moodys Corporation
(MCO) and its wholly-owned credit rating agency subsidiary, Moodys Investors Service (MIS), also maintain policies and procedures to address the independence of MISs ratings and rating
processes. Information regarding certain affiliations that may exist between directors of MCO and rated entities, and between entities who hold ratings from MIS and have also publicly
reported to the SEC an ownership interest in MCO of more than 5%, is posted annually on Moodys website at www.moodys.com under the heading Shareholder Relations Corporate
Governance Director and Shareholder Affiliation Policy.
This credit rating opinion has been prepared without taking into account any of your objectives, financial situation or needs. You should, before acting on the opinion, consider the
appropriateness of the opinion having regard to your own objectives, financial situation and needs.

17 Moodys Special Comment