November 1996

Phone

New York
Lea V. Carty
Dana Lieberman

(212) 553-1653

Defaulted Bank Loan Recoveries

Summary
In 1995, Moodys began assigning credit ratings to bank loans in the belief that access to
analysis and information relating to the credit aspects of bank loans is critical to the continued development of the secondary market for those instruments. For a firm with both rated
bank debt and bonds, Moodys typically rates the bank debt at least as high as the bonds.
Variation in the ratings spread between bank loans and bonds of the same company are,
in part, attributable to Moodys opinion of the relative value of these instruments in default.
This research examines both secondary market loan pricing and actual payments to
defaulted loan holders to arrive at estimates of the value and recovery rate of defaulted bank
loans. Briefly, the study finds that:
Secondary market pricing for defaulted U.S. senior secured syndicated bank loans indicates an average recovery rate of 71% and a median recovery rate of 77%. The standard
deviation of these recovery rate estimates is 21%, and they range from a low of 15% to a
high of 98% of par. The tremendous dispersion indicates that the likelihood that an individual loans recovery rate will be much more or much less than average is high.
A second set of recovery estimates was derived using a distinct methodology and a dataset
consisting of the post-default cash flows to a sample of senior secured loans to mostly small
and mid-size U.S. firms. The average of these recovery rate estimates is 79% and the median
is 92%. The standard deviation of these recovery rate estimates is also large, 29%, and loan
holders within the sample recovered from as little as 1% to as much as 110% of par.
Bank loans are typically worth more upon default than unsecured bonds, a fact that is
reflected in Moodys bank loan ratings. Defaulted bank loans tend to be priced 13% of
par higher than the defaulted senior unsecured bonds and 37% of par higher than the
defaulted subordinated bonds of the same company.

Defaulted Bank Loan Recoveries

Contact

Average Defaulted Bank Loan and Bond Prices

One Month After Default
$80
$70

$71
$57

$60
$50

$46

$40

$34

$30
$20
$10
$0
Bank Loans

Sr. Secured Bonds

Sr. Unsecured Bonds Subordinated Bonds

continued on page 3

Special Report

(from September 1989 to September 1996)

 Copyright 1996 by Moody's Investors Service, Inc., 99 Church Street, New York, New York 10007. All rights reserved.
ALL INFORMATION CONTAINED HEREIN IS COPYRIGHTED IN THE NAME OF MOODY'S INVESTORS SERVICE, INC. ("MOODY'S"), 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 MOODY'S PRIOR WRITTEN CONSENT. All information contained herein is obtained by MOODY'S 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 MOODY'S, 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 MOODY'S 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 MOODY'S 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 MOODY'S 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, 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 MOODY'S 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. Pursuant to Section 17(b) of the Securities Act of 1933, MOODY'S hereby discloses that most issuers of debt securities
(including corporate and municipal bonds, debentures, notes and commercial paper) and preferred stock rated by MOODY'S have, prior to assignment of
any rating, agreed to pay to MOODY'S for appraisal and rating services rendered by it fees ranging from $1,000 to $350,000.
PRINTED IN U.S.A.

Moodys Special Report

Introduction
In the 1990s, the secondary market for bank loans has grown dramatically in size and scope as a result
of both increased supply and stronger demand. On the supply side, banks interests in minimizing
reserves held against loans and in the active management of their loan portfolios has induced them to
structure a greater volume of loans in ways that make them easier to trade. On the demand side,
investors have recognized loans as a distinct asset class with their own set of return correlations and
their own place within many asset diversification schemes.
Moodys believes that access to analysis and information relating to the credit aspects of bank loans
is critical to the continued development of the secondary market for loans. As bank loans have historically been private lending instruments, detailed public information on the past performance of loans
under various business conditions remains obscured. Consequently, many investors are at an informational disadvantage relative to banks and have come to rely on a simple rule of thumb for measuring
the credit risks associated with loans not rated by Moodys: A loans credit rating is one notch higher
than the firms senior unsecured bond rating. While this rule of thumb may not be grossly mistaken,
for any particular loan it can be. Moodys began rating bank loans publicly according to the traditional Aaa to C rating scale in 1995 in order to help investors overcome this difficulty. As of October 24,
1996, we had assigned 663 ratings to over $150 billion of bank debt.
Moodys focuses a great deal of attention on an analysis of the collateral and seniority of a bank loan
in assigning a credit rating. While an obligor is roughly just as likely to commit a payment default on its
bank loans as on its long-term bonds, the post-default recovery that bank loan investors realize can be
very different from that realized by bondholders. In cases in which Moodys believes that the seniority
and security of a loan will be sufficient to generate a significantly greater recovery than would be experienced by bondholders, the reduced credit risk of the bank loan is reflected in a higher rating.
Chart 1 shows the differences between Moodys bank loan and senior unsecured debt ratings for
the 171 firms for which both ratings were available as of August 15, 1996. In about half of the cases,
the bank facility rating is one or more notches higher than the senior unsecured bond rating of the
firm, reflecting Moodys opinion of the superior value of the loans collateral position. However, in
many cases the bank loan carries the same Moodys rating as the firms senior unsecured debt. In such
situations, it is Moodys opinion that any post-default advantage that the loan may enjoy would be of
so little economic consequence that the loan would be essentially indistinguishable from the senior
unsecured debt in terms of credit risk. The dispersion among the rating differences between zero and
four reflects differences in Moodys opinions of the incremental protections the various bank loans
provide against credit loss.

Chart 1: Senior Unsecured Ratings vs. Bank Loan Ratings of the Same Company
(as of August 15, 1996, 171 bank loans = 100%)
100
90

49%

80
70
60

32%

50
40
16%

30
20
10

2%

1%

0
0

Differences Between Loan and Bond Ratings (notches)

Moodys Special Report

This report estimates average and median recovery rates for defaulted bank loans using two distinct
methodologies and datasets and examines the similarities and differences between the recoveries on
public debt and on private debt in default. The first section of the report analyzes secondary market
defaulted loan pricing to derive estimates of defaulted loan recovery rates. Because this sample necessarily consists of traded loans, the results of this analysis are more representative of the recovery characteristics of syndicated loans. The second section examines actual post-default payments made on a
sample of defaulted bank loans from Loan Pricing Corporations Loan Loss Database to generate
recovery rate estimates. The chosen sample consists of loans largely made to small and mid-size commercial and industrial firms and the results of this section are more representative of the recovery characteristics of non-traded loans. The third section compares and contrasts Moodys results from the
first two sections with those of Edwards and Asarnows 1995 research on the topic.1

Recovery Rate Estimates Based on Secondary Market Prices

Just as in any market for performing assets, the prices of distressed and defaulted bank loans reflect
the expectations of market participants as to the present value of the securitys future cash flows,
adjusted for its risk characteristics. Many investors feel that because of poor liquidity, the use of pricing grids, and regulation-induced pricing distortions, secondary market pricing does not accurately
reflect the intrinsic value of bank loans in default. Despite these market characteristics, in this section
we consider defaulted loan prices as proxies for loan recovery rates. While the trading price is only an
estimate of recovery and can be expected to be lower than any actual payout by the amount of the
markets risk premium, it has the distinct advantage of being the definitive measure of recovery for
those investors who liquidate positions after default.
Methodology and Data
Credit agreements tend to be more easily and privately renegotiated. Credit agreement amendments that
would classify as defaults in Moodys corporate bond default research, such as maturity extensions and
reduced interest rates, are difficult to discover and document for unrated facilities. Consequently, reliable
information on the incidence of a wide range of bank loan defaults is not available. For this part of our
research, we have chosen a practical definition of default: a default on publicly held bonds as recorded in
Moodys corporate bond default database or a filing for bankruptcy protection.2 While this definition
fails to include many credit agreement modifications, it has the advantages of being more easily understandable, verifiable and comparable to our public bond default research.
Actual transaction prices are not available for most of the defaulted bank loans in our database.
Instead, we rely on bid prices as reflected on loan dealers quote sheets. Additionally, defaulted bank
loan price quotes tend to remain nearly constant for extended periods of time reflecting the pace of the
defaults resolution and the corresponding information flow, the relatively high transaction costs associated with a loan sale and the markets illiquidity. In order to mitigate the effects of using bid quotes
on illiquid securities, we collected these prices from several different sources.3 We take the average of
these dealers bid quotes as our price for the defaulted loan.
The date of the price is also important. We choose to let some time pass after the default date so
that investors have time to assess accurately, and therefore to price, the severity of the default. On the
other hand, as time passes, the time value of money generates an upward bias in the price as an indicator of the loans eventual recovery. As a compromise, for each dealer and loan we take one price
which is the earliest price recorded between two and eight weeks after the default.
Our sample consists of 58 U.S. borrowers with one loan per borrower.4 The earliest default for
which we have full information occurred in September of 1989 (the mortgage financier Lomas
Financial Corporation) while the most recent occurred in July of 1996 (the construction company
Morrison Knudsen Corporation). In terms of industry affiliations, the majority (52) of the defaulting
borrowers were industrial firms, with the balance spread amongst the financial, transportation and
utilities industries. Within the group of defaulting industrial firms, a noteworthy 17 borrowers were
retailers.
Edwards, David and Asarnow, Elliot, Measuring Loss on Defaulted Bank Loans A Twenty Four Year Study, Journal of Commercial
Lending, Vol. 77, No. 7, March 1995, p.11-23.
Moodys defines default as any missed or delayed disbursement of interest and/or principal. We include as defaults distressed exchanges where (i)
the issuer offered bondholders a new security or package of securities that amount to a diminished financial obligation (such as preferred or common stock, or debt with a lower coupon or par amount) and (ii) the exchange had the apparent purpose of helping the borrower avoid default.
3
Among the contributors were BDS Securities, Goldman Sachs, Merrill Lynch, Lehman Brothers and Citibank.
4
Including prices for different loans involved in the same default would have the effect of biasing our results towards the characteristics of borrowers with multiple loans outstanding and artificially lowering our estimates of the standard deviation of the defaulted loan prices.
1

Moodys Special Report

Defaulted Loan Price Distribution

The average of these 58 defaulted bank loan prices comes to $71 per $100 par amount of the defaulted loan. However, the average price is only one measure of what could be considered the most typical
value that would be realized for a defaulted bank loan. As Chart 2 portrays, the distribution of prices
is skewed towards the high end of the price scale. This indicates that while the average of these prices
is $71, more than half are greater than that value. The median price, $77, is a more robust measure of
the center of an asymmetric distribution such as Chart 2 portrays. It indicates that a loan drawn at
random from this sample would yield a price of at least $77 half of the time.

Chart 2: Defaulted Bank Loan Price Distribution

Frequency (58 = 100%)

14
12
10
8
6
4
2

$91-$100

$81-$90

$71-$80

$61-$70

$51-$60

$41-$50

$31-$40

$21-$30

$11-$20

$0-$10

Price Per $100 Par Amount

Another characteristic of the price distribution above is its tremendous dispersion. The lowest price
observed is $15 while the greatest is $98. One of the most common measures of dispersion is the standard deviation, $21. A 95% confidence interval for the mean defaulted loan price based on this standard deviation extends from $29 dollars on the low side to $112 on the high side.5 The size of the confidence interval conveys that the likelihood that an individual loan will be worth much more or much
less than the average is high.
While useful to some extent, confidence intervals derived from averages and standard deviations
can generate misperceptions when applied to small samples from skewed distributions. The large
upper bound, $112, seems unlikely ever to be reached as bank loans are overwhelmingly floating rate
instruments resetting interest rates each quarter making prices over $100 rare. An alternative approach
is to consider a band into which 95% of the observations actually fell. Of the 58 prices, 56 (97%) fell
between the second lowest price, $15, and the second highest price, $97. Using this distribution-free
methodology one can see that prices greater than $97 are rare.
Comparison with Price-Based Defaulted Bond Recovery Estimates
Moodys also maintains a database of prices for defaulted bonds from which we derive estimates of
bond recoveries. The defaulted bond prices are highly variable (as are default resolutions), and so we
typically calculate averages over as long a time horizon as possible in order to increase the sample size
and our confidence in the results. However, in this case, we intend to compare our average defaulted
bond prices with those for defaulted bank loans. Because the defaulted loan price sample includes only
U.S. loans that have defaulted since 1991, we limit our sample of defaulted bond prices to include only
bonds issued by U.S. companies that defaulted during these years. The cover chart presents the averages of these prices segmented by seniority of claim.
The average defaulted senior secured bank loan price is, $71, $14 higher than the average price for
senior secured bonds that defaulted in the same time period, $57. Defaulting senior unsecured bonds
The 95% confidence interval is calculated as the average 1.96 * standard deviation. Conceptually, one can be 95% confident that the true mean
of defaulted loan prices lies between the upper and lower confidence limits.

Moodys Special Report

were worth, on average, even less, $46 and defaulted subordinated bonds were worth only $34 on
average. The average price for defaulted bank loans' is higher than for senior unsecured bonds or subordinated bonds because of the loans security and, in the case of subordinated bonds, seniority. It is
noteworthy that the average price of defaulted bank loans is much higher than that of senior secured
bonds of similar seniority. A statistical test reveals that this difference is significant at the 3% level of
confidence. Hence, the data indicate that, on average, bank loan recoveries tend to be both economically and statistically significantly higher than senior secured bond recoveries.
While the cover chart provides a useful summary of the relative effects of seniority and security on
recovery, a more precise calculation is the average difference in prices for various defaulted debt
instruments of the same obligor. This controls for the particular circumstances of each default and
focuses more narrowly on the role seniority plays in determining recovery. There are twelve instances
for which we have post-default pricing for both the bank debt and senior unsecured bond debt of the
same defaulter. The average difference between the loan and bond prices comes to $13. There are
another 47 instances for which we have post-default pricing for both the bank debt and subordinated
bond debt of the same defaulter. The average difference between the senior secured bank loan and
subordinated bond prices in these cases comes to $37. This indicates that by controlling for the
specifics of each default, we can see that senior bank loans may be expected to be worth 37% of par
more than subordinated bonds upon default.

Recovery Rate Estimates Based on Post-Default Payment Data

Another approach to determining the value of defaulted bank loans is to consider the actual payments
that defaulted loan holders ultimately received in the defaults resolution. Conceptually, the proportion of the par amount of the defaulting loan represented by the discounted value of the post-default
payments is the recovery earned on the defaulted loan. This methodology requires detailed information on the timing, size of defaults and on the payments loan holders actually received, as well as
determination of the discount rate to apply to post-default payments.
Data and Methodology
Loan Pricing Corporation (LPC) maintains a proprietary database designed to track and measure the
credit performance of bank loans. The database details loan and borrower characteristics and payment
histories including payments made on loans subsequent to default. With the stated goal of contributing
to the further development of the secondary market for loans, LPC agreed to allow Moodys to
explore certain aspects of this database in order to ascertain reliable recovery rate estimates.6 The data
Moodys analyzed derived from the U.S. loan portfolios of 24 U.S. and non-U.S. banks and contains
payment information on a sample of defaulted bank loans that comprises over 70% of the par amount
of these banks distressed loan portfolios.
We limited our sample to include only senior secured loans that the contributing banks had placed
on Non-Accrual and that had completed their workouts. This limited the number of post-default
payment streams considered to those of 229 bank loans. Not surprisingly the sample is predominantly
comprised of small to mid-sized borrowers since they form the bulk of borrowings in the overall commercial loan market. The borrowers represent a geographically and industrially diversified cross-section of United States borrowers. All but one had defaulted since 1990.
In examining this data, Moodys considered as defaulted any loan that a contributing bank placed
on non-accrual. For each defaulted loan, Moodys examined three payment streams in estimating ultimate recovery: interest payments, principal payments, and post-default draw-downs on the loan. The
recovery rate estimate for a particular loan is then the sum of present values of each of these cash
flows as of the time of default divided by the outstanding par amount of the loan at the time of
default. Symbolically,

Recovery =

PV (I) + PV(P) + PV(D)

Par Amount as of Default Date

Loan Pricing Corporation did not provide Moodys with access to any information that would allow Moodys to identify lender or borrower
names. LPC also requires Moodys to adhere to their strict security and confidentiality policies.

Moodys Special Report

where I, P, and D represent the post-default streams of interest payments, principal payments, and
draw-downs, respectively, and PV(.) represents the present value of the stream of payments. The present value calculation for the stream of interest payments, for example, takes the form,

where n is the number of interest payments made between the date of default and the date of resolution, r is the discount rate, and Tk is the number of days from the date of default that the borrower
made the payment. The factor of 4 in the denominator reflects the fact that this loan was to pay
interest on a quarterly basis.
The estimated recovery rate calculation above relies also upon the discount rate. While the contractual lending rate is the most appropriate discount rate to apply for the sake of calculating present values, it is unavailable for these loans. To proceed we rely on estimates of what the contractual lending
rates for these loans were. Our estimates are based on LPCs market-based model of the spread over
LIBOR paid by borrowers. The model incorporates a wealth of borrower- and loan-specific information into a flexible estimation structure to explain approximately 70% of the variation in loan spreads
over LIBOR. Moodys believes this model presents reasonably accurate estimates of the loans contractual lending rates.

Recovery Rate Estimate Distribution

The recovery estimates for senior secured bank loans derived from the post-default payment data in
Loan Pricing Corporations database generate a distribution that is significantly more skewed towards
the high end of the recovery scale than the distribution of defaulted bond prices. The average recovery
estimate is 79%; however, the median value is much higher at 92%.

Chart 3: Recovery Rate Estimates for Senior Secured Bank Loans

(Drawn from LPCs Loan Loss Database)

Number of Loans

120

90

60

30

0
0%-20%

20%-40%

40%-60%

60%-80%

80%-100%

>100%

Recovery Rate

As noted previously, the average and the median, while estimates of the center of the distribution of
recoveries, are not particularly indicative of what any particular realization from this distribution
would be. The concentration at the high end of the recovery scale generates both a high average recovery and a high standard deviation of recoveries, 29%. A distribution-free 95% confidence interval
extends from 6% on the low side to 107% on the high side.

Moodys Special Report

Additional results derived from the LPC database confirm intuitive notions about the value of a
loans collateral in bankruptcy. For example, there is a clear increase in recovery rates for loans
secured with current assets over loans collateralized by property plant and equipment. An additional,
but perhaps unintuitive result is the virtual invariance of average recovery rates with the asset size of
the borrower. The average given above, 79%, is based on all 229 defaulted loans. However, limiting
the sample to borrowers with total assets exceeding $25 million lowers the average recovery estimate
to 78% (see Table 1). Considering just borrowers with total assets exceeding $100 million generates
an average recovery rate of 77%. However, given the distributions high standard deviation, the difference between the 79% average recovery rate for all firms and the 77% average recovery rate for firms
with total assets exceeding $100 million can not be considered meaningful in a statistical sense.

Table 1 - Recovery Rate Estimates by Size of Borrower

Total Assets
Greater than:

Number
of Loans

Average
Recovery
Estimate

Median
Recovery
Estimate

$0
$25 million
$50 million
$100 million

229
39
30
25

79%
78%
77%
77%

92%
86%
84%
85%

Standard
Deviation
29%
27%
28%
30%

Comparisons with the Edwards and Asarnow Loan Loss Study

Edwards and Asarnow (E&A) conducted one of the most complete publicly available studies on the
value of bank loans in default. The study examines a set of 831 commercial and industrial loans that
Citibank had classified as either Doubtful or had placed on non-accrual over the period extending
from 1970 to 1993. Citibank has detailed accounting and payment information on each of these loans
and on the expenses incurred in recovery. These values are discounted to the date of default using
Citibanks yearly average interest rate on domestic commercial and industrial loans. Their present
value, as a percentage of the par amount of the loan at the time of default, represents E&As estimate
of the recovery earned on each loan. The average of these recoveries is $65. The distribution of E&As
recoveries is also skewed towards the high end, with a median of $79.
The variation of E&As recovery estimates from the average value is also significant. While the
authors did not calculate the standard deviation of their recoveries directly from the data, a reasonable
approximation may be derived from the published results. Moodys estimates the standard deviation
to be $32. Similarly, one can infer that about 95% of the recoveries fall between 1% and 99%. The
high variability of the data implies that while the mean recovery is $65 per $100 par amount, that
value is not very informative about the most likely experience of any one loan that defaults.
Secondary Market Price-Based Recovery Rate Estimates versus E&As Results
Given the significant differences between the methodologies employed, it is surprising that their averages, $71 and $65, and medians, $76 and $77, for Moodys and E&A respectively, are so similar.
However, the distributions of the two sets of recovery estimates are dissimilar. The E&A recovery distribution is bi-modal, with significantly more recoveries at the high end of the scale and a larger incidence of near-zero recoveries. This generates the higher estimated standard deviation for their sample
vis--vis the sample of secondary market prices $32 versus $21.
At least part of the reason for the higher dispersion of recovery rates exhibited by the E&A data is
that the data set includes loans positioned at different levels within firms capital structures and of differing security. The greater frequency of loans with near-zero recoveries in the E&A data could be
attributable to the inclusion of subordinated and unsecured loans in the sample.
Also, the greater frequency of loans experiencing nearly complete recoveries in the E&A recovery
data could be due to differences between the definitions of default employed. E&As definition of
default any loan considered doubtful or placed on non-accrual, may include a greater proportion of
loans that experienced mild defaults relative to those experienced by the loans for which we have secondary market pricing. The sample of loans considered doubtful may include a considerable number
8

Moodys Special Report

of loans that never actually missed a payment or altered, in any economically important way, the
terms of the credit agreement. On the other hand, those loans for which we have secondary market
pricing are considered as defaulted only if they have filed for Chapter 11 protection or if they have
defaulted on long-term public bonds according to Moodys definition of default in that sector. In these
cases, the default would typically involve a realized economic loss to investors. It seems likely that private workouts involving only a small group of private lenders entail defaults of lesser average severity
than public defaults involving a large and fractious group of creditors.
Recovery Rate Estimates Based on LPCs Database versus E&As Results
One major difference between these two studies lies again in the default definitions used. In contrast to
E&As methodology, the estimates derived from LPCs database do not include, as defaults, loans considered doubtful. Because doubtful loans have not necessarily committed actual payment defaults,
they may be expected to be less severe and therefore to enjoy greater recoveries on average.
Additionally, our selection criteria assured that the recoveries we estimated from the LPC database are
based only upon the experience of senior secured loans while E&As data may include senior secured,
subordinated and unsecured loans. This is consistent with the fact that the distribution of our recovery
estimates derived from the LPC data does not include the large mass of near-zero recovery loans that
E&As dataset does, but does share the large mass of nearly zero losses. Finally, the E&A data is conditioned upon Citibanks lending culture and covers a 23-year period. LPCs data is drawn from a large
and more diverse sample of lending institutions and is drawn primarily from the post-1990 period.

Conclusion
Moodys believes that access to analysis and information relating to the credit aspects of bank loans is
critical to the continued growth of the secondary market for such loans. This study begins to address
loan investors needs for more complete understanding of the recovery characteristics of defaulted
bank loans.
Moodys results of the first two sections were derived from two distinct datasets; one consisting of
secondary market prices and the other based on post-default loan cash flows. The empirical recovery
rate data generated by the secondary market prices are more representative of the recovery characteristics of syndicated loans. These data generate an average recovery rate estimate of 71% and a median
estimate of 77%. The recovery rate estimates generated by the sample taken from Loan Pricing
Corporations post-default payment data, the majority of which are non-syndicated loans, generate an
average recovery rate estimate of 79% and a median estimate of 92%. Differences between these estimates and those of the previous section are attributable to differing methodologies and samples.
The recovery rate for defaulted bank loans can be expected to be higher than for defaulted unsecured long-term public debt issues. After controlling for the individual circumstances of default, the
seniority and security of bank loans generates an average recovery rate that is 13% of par greater than
the recovery rate for senior unsecured bonds and 37% of par greater than the recovery rate for subordinated bonds. The higher recovery rates of loans relative to senior unsecured and subordinated debt
obligations are on average captured in higher credit ratings.
Both datasets reveal a high degree of dispersion in the recovery rate estimates. Such dispersion is
attributable in great part to variations in the quality of the security behind these loans. It demonstrates
the inappropriateness of simple rules for analyzing the credit risks associated with any particular loan
(e.g., that a loans credit risk is accurately reflected by the Moodys credit rating that is one notch
higher than the rating on the firms senior unsecured debt). Because of the significant variation in the
quality of the security behind loans, Moodys concentrates a great deal of attention to the security
behind a bank credit facility in assigning a credit rating.

Moodys Special Report

20641

Special Report

Defaulted Bank Loan Recoveries