Special Comment
December 2003
Contact
Phone
New York
Praveen Varma
Richard Cantor
David Hamilton
1.212.553.1653
Recovery Rates on Defaulted Corporate Bonds and
Preferred Stocks, 1982–2003
This Special Comment presents an extensive set of recovery rate statistics on defaulted bonds and preferred stocks,
incorporating the vast amount of recovery data that has become available as a result of the large number of corporate
defaults that occurred between 2000 and 2003.
Recovery rate statistics are presented in three ways – issuer-weighted means, issuer-weighted medians, and valueweighted means – each of which may be relevant to certain investors.
We report relative recovery and loss rates on bonds of different levels of security and priority of claim based on
portfolio aggregates and contrast the results with those obtained from samples limited to issuers that have multiple
classes of debt outstanding at the time of default. In addition, we report recovery rates by type of default, by year of
default, by industry, by country of domicile, and by historical rating.
Figure 1 illustrates one of the report’s findings: the higher the rating prior to default, the higher the average
recovery rate in default. By implication, Moody’s credit ratings provide greater differentiation with respect to credit
risk than revealed just through differences in default rates by rating category because differences in loss rates (default
rates times loss severity rates) by rating category are even larger.
Figure 1
Holding Period senior Unsecured Issuer-Weighted Mean Recovery Rates
Baa
50%
Ba
B
Caa-C
45%
Recovery Rates
40%
35%
30%
25%
20%
15%
10%
5%
0%
1
3
5
Holding Period in Years
*The holding period recovery rate is the average recovery rate on bonds with a particular rating that subsequently
defaulted at some time over a specified time-period.
Table of Contents
Page
Introduction ....................................................................................................................................................... 3
Sample and Methodology .................................................................................................................................. 3
Summary Results .............................................................................................................................................. 4
Issuer-Level Relative Recovery Rates................................................................................................................. 4
Recovery Rates Based on Initial Default Events .................................................................................................. 5
Trends in Recovery Rates over Time .................................................................................................................. 6
Correlation Between Recovery Rates and Default Rates ..................................................................................... 7
Recovery Rates and Issuers’ Rating History........................................................................................................ 8
Recovery Rates by Industry.............................................................................................................................. 10
Recovery Rates By Domicile............................................................................................................................. 11
Appendix ......................................................................................................................................................... 13
2
Moody’s Special Comment
Introduction
One by-product of the extraordinary number of corporate defaults over the last few years has been a wealth of new
data on recovery rates on defaulted bonds and preferred stocks, which allows for the expanded recovery research presented in this Special Comment.1 The approach taken here is to be fairly expansive in the presentation of data but to be
concise in the discussion of the results. Results lacking general interest, but of interest to particular investors are presented in the Appendix.
The organization of paper is as follows. After a brief discussion of our dataset and methodology, we examine how
recovery rates have varied by seniority and security, time period, prior rating history, industry and geography.
Sample and Methodology
SAMPLE DESCRIPTION AND TIME PERIOD
The sample is drawn from defaults that occurred between 1982 and 2003. Over this time period, Moody’s recorded
roughly 2,800 default events. After eliminating companies that defaulted on instruments other than bonds and/or preferred stock, roughly 2,100 defaulting issuers were left in the sample. The sample also excluded certain defaulted
instruments, such as preferred stock issued by trust companies or bonds that were initially issued at deep discount to
par. The final sample consisted of 1,237 issuers that have price data on roughly 2,500 defaulted bonds and 79 issuers
that defaulted on 111 defaulted preferred stocks.
DEFINITION OF DEFAULT
For a security to be included in our sample of defaulted bonds and preferred stocks the issuer of the security must have
been categorized as being in “default” as defined by Moody’s.
Moody’s default definition includes three types of events:
1. A missed or delayed disbursement of interest and/or principal, including delayed payments made within a
grace period;
2. Filing for bankruptcy, administration, legal receivership, or other legal blocks (perhaps by regulators) to the
timely payment of interest and/or principal; or
3. Consummation of a distressed exchange where: (i) the issuer offers 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, lower seniority, or longer maturity); or (ii) the exchange had the
apparent purpose of helping the borrower avoid default.
The definition of a distressed exchange is intended to capture events that change the relationship between the
bondholder and bond issuer from the relationship that was originally contracted and that subjects the bondholder to
an economic loss. Technical defaults (covenant violations, etc.) are not included in Moody’s definition of default.
DEFINITION OF THE RECOVERY RATE
The recovery rates reported in this Special Comment are based on the 30-day post-default bid prices, except in the case
of distressed exchanges in which we use the trading prices of exchanged instruments two weeks prior to the exchange.
This allows us to capture the recovery price prior to exchange, but after the announcement of the offer.2
ESTIMATION OF AVERAGE RECOVERY RATES
There are various ways to calculate recovery rates. We present three different estimates of recovery rates throughout:
1. Value-Weighted Mean Recovery Rates: Value-weighted recovery rates represent the average of recovery
rates on all defaulted issuers, weighted by the face value of those issues. These estimates are useful for
predicting recovery rates on a market portfolio.
2. Issuer-Weighted Mean Recovery Rates: Issuer-weighted recovery rates are established by estimating
recovery rates for each issuer and then aggregating them across all issuers. They are useful for predicting
the recovery rates for portfolios that are highly diversified.
1.
2.
See Moody’s Special Comment, “Debt Recoveries for Corporate Bankruptcies, June 1999. For an analysis of earlier recovery data, see also, “Everything You Ever
Wanted to Know About Recoveries on Defaulted Bonds”, Altman, E. and V. Kishore (1996) Financial Analysts Journal 57-64.
After the exchange process is complete, the defaulted instrument may stop trading or may trade very thinly.
Moody’s Special Comment
3
Issuer-Weighted Median Recovery Rates: Median recovery rates are estimated as median of issuer-weighted
recovery rates and are used for predicting the most likely recovery rates on an arbitrarily selected issuer.
Each of these measures of the recovery rate can in turn be calculated in at least three different ways, depending
upon the choice of data aggregation across time. The most common way to aggregate is simply to pool data across
time and take a simple average. This is the approach taken in this Special Comment. Alternatively, one can calculate
average recovery rates for each year and take either a simple mean across years or find the median value across years.
These latter measures tend to produce higher average recovery rate estimates because they place less weight on years
with a high volume of defaulted instruments, which (as we show later on) are most often years with low average recovery rates.
3.
Summary Results
In Figure 2, we present all measures of average recovery rates — issuer-weighted and value-weighted — for bonds with
levels of security and priority and for preferred stock.
Figure 2
Defaulted Bond Recovery Rates by Security and Priority of Claim: 1982-2003
Issuer Weighted
Defaulted Instrument
Value Weighted
Mean
All Bonds
- Equipment Trust
- Sr. Secured
- Sr. Unsecured
- Sr. Sub.
- Sub.
- Jr. Sub.
Preferred Stock
Preferred Stock and Bonds
33.8
61.0
50.3
32.9
29.0
27.1
22.9
6.5
31.8
Mean
35.4
62.1
51.6
36.1
32.5
31.1
24.5
15.3
33.9
Median
Standard
Deviation
30.9
71.0
50.0
31.0
27.0
29.0
16.9
9.3
28
23.9
25.7
26.9
25.0
23.9
21.4
19.7
17.7
25.6
Number of
Observations
1237
11
145
484
372
362
20
79
1316
Default
Amounts in
Billion USD
403.0
2.6
34.1
258.6
65.8
39.3
2.6
11.0
414.0
Figure 2 shows that recovery rates, on average, are monotonically related to priority in capital structure. Equipment trust bonds have the highest recovery followed by senior secured bonds, senior unsecured and so on.
Another interesting observation that emerges from Figure 2 is that value-weighted recovery rates are always less
than equally-weighted recovery rates. This suggests that issuers that defaulted on larger amounts of debt had lower
recovery rates on average.
Issuer-Level Relative Recovery Rates
Figure 2 implies differential recovery rates for different security classes when averages are measured across the market
portfolio of defaulted securities. However, the differences observed in the market portfolio may be very different from
the relative recovery rates (or relative severity rates) on different components of the capital structure for individual
issuers.
Calculation of relative recovery rates based on the market portfolio is appropriate for an investor that is trying to
decide which asset class to invest in — whether to focus on secured bonds, senior unsecured bonds, or subordinated
bonds, etc. However, if an investor is trying to select among different instruments issued by a single issuer, she will need
to know the average relative recovery rates across the security classes for issuers that have all these security classes outstanding. The two approaches can yield very different results. Moody’s practices with respect to notching for priority of
claim and security, for example, are based in large part upon the latter method of calculating relative recovery rates.3
Relative severity can be defined as the loss experienced by creditors of a seniority class with respect to creditors
with different seniority class for the same issuer. Estimation of relative severity requires a matched sample to eliminate
the impact of factors like region, industry, and time period. This requirement substantially reduces the number of
observations available for estimation purposes.
3.
4
See “Summary Guidance for Notching Secured Bonds, Subordinated Bonds, and Preferred Stocks of Corporate Issuers,” Moody’s Special Comment, September
2001.
Moody’s Special Comment
Figure 3 presents median recovery and loss severity rates for secured and subordinated bonds relative to senior
unsecured bond recovery and loss rates.4
Figure 3
Issuer-Level Relative Recovery and Loss Severity Rates by Seniority
140%
124%
Recovery Rate Relative to Senior Unsecured
Loss Rate Relative to Senior Unsecured
125%
120%
100%
81%
80%
51%
60%
40%
20%
0%
Secured Bonds
Subordinated Bonds
Results presented in Figure 3 show that the median recovery for senior secured bonds is 124% of median recovery
for senior unsecured bonds. Conversely, median loss severity for senior secured bonds is 81% of the loss severity of
senior unsecured bonds.5 Similarly, median recovery for subordinated bonds is only 51% of senior unsecured bonds
while median loss severity is 125% of median loss for senior unsecured bonds.
Recovery Rates Based on Initial Default Events
Moody’s definition of default includes several different types of credit events, but the central feature of the definition is
that bondholders suffer loss because of the default. Some of them, like Chapter 11 filings, prepackaged Chapter 11 filings, and Chapter 7 filings are clear-cut public events in the sense that the company filing for bankruptcy frequently
issues a press release. However, Moody’s default definition includes many other events such as missed interest payments or distressed exchanges.
Issuers with different initial default events have experienced different average recovery rates as displayed in Figure 4. 6
Figure 4
Bond Recovery Rates Based on Initial Default Events
Issuer Weighted
Default Event
Missed principal payment
Grace period default
Distressed Exchange
Regulatory Seizures*
Missed interest payment
Bankruptcy
- Prepackaged Chapter 11
- Chapter 11
- Chapter 7
All Events
Value Weighted
Mean
61.4
54.8
43.2
38.3
26.8
31.7
39.6
30.7
7.1
32.6
Mean
62.5
51.8
52.7
38.9
32.4
32.3
38.6
31.1
17.4
35.4
Median
69.0
50.0
47.0
32.0
28.1
25.0
36.3
25.0
17.0
31.0
Standard
Deviation
25.0
23.4
23.5
24.0
21.3
25.5
27.7
24.3
13.0
23.9
Number of
Observations
Default
Amounts in
BillionUSD
30
41
106
49
652
347
22
334
3
1237
12.5
7.5
77.5
22.3
178.2
105.1
5.7
98.5
0.9
403.0
*Regulatory seizures include placements under receivership, administration, and conservatorship that prevented issuers from making debt service
payments and other government imposed payment moratoria.
4.
5.
6.
All classes of subordinated debt are combined for this analysis.
For example, if the average relative loss on senior secured obligations is 50% then the loss suffered by the holders of senior unsecured bonds is, on average,
50%*124%=62%. Note that there is not a “one-to-one” relationship between median relative recovery rates and median relative loss rates in a traditional sense that
Relative Recovery = 1- Relative Severity.
These initial default events may or may not lead to eventual filing for bankruptcy court protection.
Moody’s Special Comment
5
The majority of default events are missed interest payments (over 50%) followed by Chapter 11 filings (25%) and
distressed exchanges (9%). Grace period defaults, which are 3.4% of all defaults, are missed interest and/or principal
payments that were cured within the grace period.
When estimated on an issuer basis, the highest mean recovery rates are roughly 62% observed in the case of
“missed principal payment.” While readers may find it curious that this type of default would have recovery rates
almost twice as high as missed interest payments, a majority of the “missed principal payments” default events were
observed in the utility industry which, as we show later, had the highest recovery rate. Moreover, the very fact that a
company was able to postpone default or bankruptcy right up to the very day its principal payments became due may
indicate that its ultimate recovery prospects were quite strong.
Grace period defaults and distressed exchanges have the second highest recovery rates at about 50%. High recovery rates for grace period defaults may be unsurprising because these defaults are often temporary though often followed later on by further new payment defaults or even bankruptcies. The mean recovery rate on distressed exchanges
at roughly 50% is substantially higher than the 32% mean recovery rate on all bankruptcies. Higher recovery rates on
distressed exchanges may help explain why bondholders are willing to accept exchanges rather than risk even lower
recoveries in bankruptcy. Though the sample is limited to three issuers, recovery rates on Chapter 7 liquidation have,
unsurprisingly, been very low.
Trends in Recovery Rates over Time
Recovery rates exhibit significant variation over time and exhibit pro-cyclicality, as shown in Figure 5. In times of economic boom, recovery rates are higher. They tend to be lower in times of declining economic activity.
Figure 5
Bond Recovery Rates by Year
Issuer Weighted
Default Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
Mean
Median
Value Weighted
Mean
38.6
56.5
47.9
37.0
49.9
51.4
37.5
37.9
27.2
34.3
44.3
52.7
48.5
49.1
54.0
54.7
38.0
35.1
25.6
31.0
31.8
40.2
42.0
39.4
Mean
36.3
44.4
47.2
45.8
48.2
48.7
39.3
32.7
27.1
36.9
46.8
42.9
44.1
45.0
47.1
52.5
40.4
36.9
26.3
25.2
31.9
39.5
40.2
41.6
Median
38.8
45.3
43.8
43.3
45.4
47.0
35.6
30.2
24.0
35.5
46.0
38.0
39.9
44.5
43.1
50.8
36.0
32.0
20.0
18.4
26.0
33.5
37.1
38.4
Standard
Deviation
14.4
14.0
19.2
18.5
21.5
25.6
18.1
22.6
18.3
24.2
23.4
24.1
20.7
23.2
24.9
30.8
27.0
24.4
21.3
21.7
23.4
25.4
n.a.
n.a.
Number of
Observations
22
17
16
19
36
31
29
59
88
82
36
23
20
34
19
32
64
106
120
179
151
54
n.a.
n.a.
Default
Amounts in
Billion USD
0.8
1.1
0.7
1.7
3.5
9.1
4.2
10.0
16.7
14.7
5.3
2.5
2.3
6.5
2.6
4.8
10.8
26.3
27.7
116.6
118.2
16.8
n.a.
n.a.
Default Rate
1.04%
0.96%
0.92%
1.00%
1.89%
1.50%
1.36%
2.34%
3.58%
3.23%
1.31%
0.94%
0.56%
1.03%
0.51%
0.65%
1.23%
2.16%
2.36%
3.80%
2.97%
1.73%
n.a.
n.a.
From a high of 52.5% in 1997 to a low of 25.2% in 2001, the issuer-weighted mean recovery rates exhibit significant variation. Interestingly, these recovery rates also exhibit a mean reversion process where the long-term mean is
roughly 40%. The three measures of annual average recovery rates are plotted in Figure 6. While there is significant
variation over time, as can be observed from the figure, all three measures of average recovery rates tend to move
closely together over time.
6
Moody’s Special Comment
Figure 6
Annual Recovery Rates: Value-Weighted Mean, Issuer-Weighted Mean and Median
60%
Recovery Rate
50%
40%
30%
20%
Value-Weighted Mean
Issuer-Weighted Mean
Issuer-Weighted Median
10%
0%
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
Year
Correlation Between Recovery Rates and Default Rates
Recovery rates and default rates are negatively related: higher default rates are associated with lower recovery rates and
lower default rates are correlated with higher recovery rates.7 This relationship is depicted in Figure 7.
Figure 7
The Correlation Between Default Rates and Issuer-Weighted Mean Recovery
Recovery Rate= 50.284 - 6.302*Default Rate
R2 = 0.5956
60%
1997
50%
1984
Recovery Rate
1996
1998
1982
30%
1986
1992
1994
40%
1987
1999
1991
1988
1989
2002
2000
1990
20%
2001
10%
0%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
Default Rate
There are at least two — possibly related — explanations for this observed correlation even though they may be
impossible to distinguish empirically. One, higher default rates increase the supply of defaulted debt and thereby
decrease its price. Two, when the market price for distressed debt is low, distressed companies find it more difficult to
obtain financing and default more frequently as a result.
7.
This relationship has been documented for many years in Moody’s annual corporate bond and default studies, and it has recently received academic attention due to
its portfolio risk correlation implications. See, for example, Altman, E., B. Brady, A. Resti and A. Sironi, “The Link Between Default And Recovery Rates: Implications
For Credit Risk Models And Procyclicality, “ forthcoming in the Journal of Business and Hu, Y. and W. Perraudin, “The Dependence of Recovery Rates and Defaults,”
Working Paper, Birkbeck College, February 2002.
Moody’s Special Comment
7
Recovery Rates and Issuers’ Rating History
While it is expected that ratings differentiate relative default risk, it is less clear whether they differentiate relative loss
rates in the event of default. Moody’s ratings are intended to rank securities with respect to relative expected loss rates,
not simply to relative expected default rates. When an issuer has more than one class of debt in the market, Moody’s
obviously rates more secure and higher priority claims higher than less secure and lower priority claims. However,
within the same security class it is significantly more difficult to predict which issuers in the event of default are likely
to have higher or lower recovery rates.
In this section, we investigate to the extent that ratings not only differentiate relative default risk, but also distinguish relative recovery rates on defaulted senior unsecured debt issues.8
(In the Appendix, we provide number of observations for each cell in Figure 8(a).)
One way to analyze the relationship between the recovery rates and issuers’ rating is to simply estimate recoveries
based on issuers’ ratings at some time prior to default. Figure 8 (a) shows recovery rates based on issuer ratings at various points in time prior to defaulting. Evidently, bonds of investment-grade issuers at any point in time up to five years
prior to default have higher recovery rates than bonds of the issuers rated speculative grade.
Figure 8(a)
Realized Senior Unsecured Issuer-Weighted Mean Recovery Rates
One Year Before
Default
Prior Rating
Aaa
Aa
A
Baa
Ba
B
Caa-Ca
Investment Grade
Speculative Grade
All Issuers
Two Years Before
Default
n.a.
95.4
51.3
43.3
37.7
35.9
28.4
46.2
34.6
35.7
n.a.
62.1
47.1
42.3
40.3
35.0
22.4
44.5
33.9
35.4
Three Years
Before Default
Four Years
Before Default
n.a.
30.8
42.0
43.7
39.7
36.2
17.9
42.7
34.5
36.3
97.0
44.4
44.5
40.8
41.9
35.2
27.6
43.0
36.8
38.6
Five Years Before
Default
74.1
41.1
45.7
38.1
42.6
37.8
12.3
41.3
39.2
39.9
However, these recovery rates are point-in-time recovery rates. For example, the table simply shows that the average recovery rate for senior unsecured bonds of an issuer rated single ‘A’ exactly three years prior to default is 42%.
Realistically, a bond can default any time during its holding period. Investors purchasing a single ‘A’ rated bond
and expecting to hold it for five years would like to know at the time of purchase the likelihood of it defaulting in any
of those five years and the likely recovery rate in each of those years. The five-year holding period loss rate depends on
the both the time profile of default rates and recovery rates.
As such, we need to know both the time profile of recovery rates and default rates to calculate holding period
recovery rates.9
8.
9.
Additional data are presented in the Appendix.
Holding period recovery Rate (HPR) for any given rating category and any horizon T can be estimated as following:
∑
=
T
HPRT
t =1
DRt * RRt
CDRT
Where,
DRt is the default rate for this rating category in year t, where t=1, 2 through T, and,
DRt = CDRT - CDRT-1 for T > 1
DRt = CDRT for T=1
RRt is the recovery rate for defaulting issuers holding this particular rating in t years prior to default
CDRT is the cumulative default rate for issuers beginning with this rating at the start of holding period T
And consequently, holding period loss (HPL) can be calculated as follows:
HPL = 1 − HPR
T
T
Where,
HPLT is holding period loss rate for time period T
HPRT is the holding period recovery rate estimated above.
8
Moody’s Special Comment
Figure 8 (b) presents cumulative default rates by rating category over different horizons.
Figure 8(b)
Weighted Average Cumulative Default Rates
Prior Rating
Year 1
Aaa
Aa
A
Baa
Ba
B
Caa-C
Investment Grade
Speculative Grade
All Issuers
Year 2
0.0%
0.0%
0.0%
0.3%
1.4%
6.5%
23.9%
0.1%
5.5%
1.9%
Year 3
0.0%
0.0%
0.1%
0.7%
4.1%
15.1%
41.7%
0.2%
11.6%
3.9%
0.0%
0.0%
0.3%
1.2%
7.4%
23.5%
58.6%
0.5%
17.6%
5.7%
Year 4
Year 5
0.1%
0.1%
0.4%
2.0%
11.0%
31.6%
73.9%
0.8%
23.3%
7.4%
0.2%
0.2%
0.6%
2.6%
14.5%
39.5%
86.3%
1.0%
28.6%
8.8%
Holding period recovery rates presented in Figure 8 (c) are obtained by calculating the default rate contributions
of every year to any given holding period’s cumulative default rate. These contributions provide the weights needed to
calculate the weighted average holding period recovery rates presented in 8(c) that are extrapolated from 8(a) and 8(b).
Figure 8(c)
Holding Period Mean Recovery Rates by Rating Category
Prior Rating
Aaa
Aa
A
Baa
Ba
B
Caa-C
Investment Grade
Speculative Grade
All Issuers
One Year Before
Default
n.a.
95.4
51.3
43.3
37.7
35.9
28.4
46.2
34.6
35.7
Two Years Before
Default
n.a.
78.7
48.3
42.7
39.4
35.4
25.8
45.2
34.2
35.5
Three Years
Before Default
n.a.
30.8
44.4
43.1
39.5
35.7
23.5
44.0
34.3
35.8
Four Years
Before Default
97.0
40.5
44.5
42.3
40.3
35.6
24.4
43.6
34.9
36.4
Five Years
Before Default
81.3
40.8
44.8
41.2
40.9
36.0
22.6
43.0
35.7
37.0
A comparison of Figure 8 (a) and Figure 8 (c) shows certain differences between realized and expected recovery
rates. For example, while the average realized recovery rate for bonds of issuer rated single ‘A’ that defaulted in three
years is 42%, the holding period recovery rate was 44.4%. In some instances, the difference is larger. For example, the
five-year recovery rate for issuers rated Caa-C is 12.3% compared to the expected holding period recovery rate of
22.6% for the same period.
Figure 8 illustrates that:
1. Rating prior to default is directly correlated with expected recovery rates. The higher the rating preceding
default, the higher the recovery rate.
2. Over a five-year horizon, defaulted issues rated investment-grade at any time prior to default have a better
rate of recovery regardless of the rating at default. This is reflected in both point-in-time recovery rates as
in Figure 8(a) and in holding-period recovery rates as in Figure 8(c).
3. Most holding period recovery rates are roughly the same regardless of the length of the holding period for
a given rating category. For example, the expected holding period recovery rates for issuers rated Baa range
from 43.3% for one year to 41.2% for five years.
Figure 1 presents this same information graphically.
Moody’s Special Comment
9
Recovery Rates by Industry
Recovery rates also vary from industry to industry. Defaults in a particular industry are often correlated due to sectorwide economic shocks such as sensitivity to energy prices or the introduction of new technologies. Cyclically correlated defaults may increase the supply of distressed debt over some time period, thus — everything else being equal —
reducing recovery rates. Increasing sector-wide default rates may also reduce the value of assets in that industry, further aggravating the problem. Recovery rates may be impacted by quality and type of assets, life of assets and their ability to continue generating revenues.
Figure 9 shows the historical variation of recovery rates by industry. Defaulted bonds in certain industries such as
public utilities have recovered more, and defaulted bonds in industries like telecom have recovered far less than the
average recovery rates overall. The range of issuer-weighted recovery rates is wide: from roughly 23% in telecom
industry to over 50% in the public utility industry.
Figure 9 reveals the results of a simple statistical significance test that examine whether an industry’s issuerweighted mean recovery rate differs significantly from the broader average for the full sample.
Figure 9
Average Recovery Rates by Industry
Industry
Value Weighted Mean
Utility-Gas
Oil and Oil Services
Hospitality
Utility-Electric
Miscellaneous
Transport-Ocean
Media, Broadcasting and Cable
Transport-Surface
Finance and Banking
Industrial
Retail
Transport - Air
Automotive
Healthcare
Consumer Goods
Construction
Technology
Real Estate
Steel
Telecom
All Industries
56.4
53.4
45.7
47.3
43.3
42.3
43.5
35.1
51.6
35.1
34.8
33.3
27.4
28.2
26.7
29.5
25.5
24.3
26.0
17.1
33.8
Issuer Weighted Mean
51.5*
44.5*
42.5*
41.4*
39.5*
38.8*
38.2
36.6
36.3
35.4
34.4
34.3
33.4
32.7
32.5*
31.9*
29.5*
28.8*
27.4*
23.2*
35.4
* Statistically different from the All Industries issuer-weighted mean recovery rate at 95%
confidence level.
Among the 20 industries listed in the exhibit, six had statistically higher recovery rates than the overall mean for
the whole sample as indicated by an asterisk. Six had lower recovery rates while for seven industries the difference is
not significant.10 Clearly, bond of defaulting issuers in some industries like utilities (industries with hard assets that
have longevity and revenue producing capacity) have higher recovery rates compared to other industries like telecom
and steel, two industries at the opposite end of the spectrum but facing very similar challenges: overcapacity and
declining prices.
10. Note that this test is biased towards not finding significant differences.
10
Moody’s Special Comment
Recovery Rates By Domicile
In Figure 10, we present recovery rates based on domicile of the defaulting issuer. The determination of domicile is
not made simply on the basis of the country of incorporation. Moody’s uses a set of criteria to determine the effective
— as opposed to the legal — domicile of rated corporate bond issuers. In addition to legal domicile, we also consider
the jurisdiction under which a firm might file for bankruptcy and the location of the majority of its revenues and assets.
The increasing use of domiciles of convenience such as Bermuda, and the growth of transnational corporate families
makes the determination of the ‘true’ country of domicile both complex and vital.11
Figure 10
Recovery Rates by Issuers’ Domiciles
Value
Weighted
Domicile
United States
Canada
United Kingdom
Argentina
Other Europe
Indonesia
Mexico
Other Asia
Australia
Other Latin America
Netherlands
Hong Kong
Brazil
Korea
Sovereign
All Countries
Issuer Weighted
Mean
Mean
Median
35.4
27.0
23.9
33.5
23.7
23.5
47.6
26.3
37.6
48.3
21.7
32.9
23.0
81.6
31.2
33.8
35.4
34.5
28.9
30.2
30.2
25.3
50.7
34.8
43.0
52.9
20.7
31.2
29.4
93.1
34.4
35.4
31.6
30.0
21.0
25.0
26.0
24.0
45.0
33.5
29.0
40.0
22.2
32.0
29.5
102.5
39.8
29.8
Standard
Deviation
23.8
24.4
27.5
18.4
25.8
9.8
20.2
17.7
26.1
27.0
12.8
11.3
8.7
17.2
24.3
23.8
Number of
Observations
1055
41
29
25
21
11
11
7
7
6
6
5
4
3
6
1237
On the surface, readers may find it curious that recovery rates are significantly different for defaulted bonds in different countries. However, regional factors can have a large impact on recovery rates. For example, a liquid market for
distressed debt as found in the United States can result in higher average recovery rates. Additionally, the presence of
different bankruptcy regimes can impact recovery rates. For example, many countries’ bankruptcy codes do not permit
distressed companies to restructure their debts within bankruptcy and avoid liquidation. As a result, though defaults in
those countries may be less frequent, recoveries in the event of default may be lower.
The amount of bank debt typically found on a companies’ balance sheets, which varies from country to country,
can also affect average recovery rates since bank debt is often in a better position than bond debt in bankruptcy.
Alone or together, these factors affect recovery rates in a very meaningful fashion. For example, in Europe, issuerweighted recovery rates have been significantly lower than those in the United States, perhaps because of differences
in bankruptcy regimes and more limited market in Europe for distressed debt.
Based on a small sample, defaulted sovereign bonds issued by countries have historically recovered roughly 34%
of the face value, which is in line with corporate recovery rates.12
The Appendix presents issue-weighted (as opposed to issuer-weighted and value-weighted) recovery rates based on
year, initial default event, industry, issue ratings- initial as well as immediately prior to default, and domicile for investors interested in such statistics. However, issuer- and value-weighted recovery statistics presented in the main text are
probably more meaningful for most investors and probably more indicative of expected future recovery rates.
11. For example, this study treats two Bermuda-based issuers – Viatel and RSL Communications – as European defaults because the majority of revenues and assets
of these firms are located in Europe. Similarly, Asia Paper and Pulp is considered an Asian default though it is domiciled in The Netherlands.
12. For more information see “Sovereign Bond Defaults, Rating Transition, And Recoveries (1985-2002),” Moody’s Special Comment, February 2003.
Moody’s Special Comment
11
Related Research
Special Comments
Default & Recovery Rates of Canadian Corporate Bond Issuers, 1989-2002, February 2003 (#79420)
Default and Recovery Rates of European Corporate Bond Issuers, 1985-2002, May 2003 (#78087)
Sovereign Bond Rating Defaults, Transitions, And Recoveries (1985-2002), February 2003 (#77350)
Debt Recoveries For Corporate Bankruptcies, June 1999 (#46119)
12
Moody’s Special Comment
Appendix
Figure 11
Issue-Weighted Annual Recovery Rates Recovery Rates by Security and Priority of Claim
Senior Secured
Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
Mean
72.5
40.2
n.a.
83.6
56.4
71.3
66.7
46.5
41.7
47.3
62.4
n.a.
67.1
67.5
63.6
89.0
54.0
45.7
41.7
59.4
53.3
61.7
Number of
Observations
1
1
n.a.
1
10
2
14
4
14
12
12
n.a.
3
9
4
4
15
17
18
42
28
9
Senior Unsecured
Mean
35.2
58.6
49.4
54.5
53.8
50.7
40.1
42.4
34.7
50.0
52.2
49.7
53.7
52.9
63.6
60.5
42.6
42.1
25.5
38.2
36.4
43.0
Number of
Observations
14
15
2
3
11
13
20
15
38
56
15
8
8
13
4
18
43
69
75
229
283
64
Senior Subordinate
Mean
48.1
43.5
67.9
30.4
48.2
44.4
33.3
34.2
27.2
45.3
45.5
51.6
33.7
36.8
45.0
44.5
48.4
29.4
20.2
20.4
24.9
37.9
Number of
Observations
2
1
1
11
8
8
9
27
53
44
22
8
5
17
7
8
13
41
55
68
28
19
Subordinated
Mean
31.7
41.6
42.4
40.3
37.2
48.6
34.5
28.1
20.6
24.6
31.1
45.4
33.7
39.4
26.2
43.7
13.5
38.7
31.4
16.4
30.4
11.8
Junior Subordinated
Number of
Observations
20
13
17
24
43
17
24
49
53
62
23
17
8
7
5
8
14
14
13
23
13
8
Mean
n.a.
n.a.
n.a.
48.5
n.a.
n.a.
36.5
16.6
10.7
7.8
n.a.
n.a.
n.a.
n.a.
n.a.
30.6
62.0
n.a.
15.5
47.0
n.a.
n.a.
Number of
Observations
n.a
n.a
n.a
1
n.a
n.a
1
3
5
3
n.a
n.a
n.a
n.a
n.a
3
2
n.a
2
1
n.a
n.a
Moody’s Special Comment
13
14
Figure 12
Issue-Weighted Recovery Rates for Various Seniorities Based on Initial Default Event
Moody’s Special Comment
Senior Secured
Default Event
Mean
Bankruptcy
- Chapter 7
- Chapter 11
- Prepackaged Chapter 11
- Others
Distressed Exchange
Grace Period Default
Missed Interest Payment
Missed Principal Payment
Regulatory Seizures*
All Events
Senior Unsecured
Number of
Observations
56.2
n.a.
58.0
41.3
19.0
74.1
70.3
48.4
83.6
40.3
49.4
Number of
Observations
Mean
86
n.a.
78
7
1
11
11
89
13
7
303
Senior Subordinate
35.7
16.3
35.6
35.6
73.5
55.6
49.4
32.1
63.8
59.7
43.8
Mean
328
4
312
13
3
81
14
418
56
87
1316
Number of
Observations
24.0
0.0
22.1
39.2
26.7
53.3
50.9
30.2
53.4
50.5
46.4
110
0
93
11
6
42
15
275
10
2
564
Subordinated
Mean
27.1
7.8
26.4
45.9
37.2
41.1
43.8
29.5
41.4
15.2
31.8
Number of
Observations
Junior Subordinated
Mean
121
5
108
5
8
55
11
247
13
6
579
39.4
n.a.
38.4
41.5
n.a.
17.5
n.a.
18.0
n.a.
n.a.
27.6
Number of
Observations
6
4
2
n.a.
1
14
n.a.
n.a.
27
*Regulatory seizures include placements under receivership, administration, and conservatorship that prevented issuers from making debt service payments that would prevent issuers from making timely
debt service payments and other government imposed moratoria.
Figure 13
Issue-Weighted Recovery Rates Based on Initial Issue Rating
Issue-Weighted
Initial Rating
Aaa
Aa
A
Baa
Ba
B
Caa
Ca
C
Mean
Median
50.1
54.4
50.9
45.7
41.1
34.4
24.8
21.9
17.5
50.1
52.0
51.3
40.3
33.5
29.0
18.0
22.8
17.5
Minimum
47.2
5.8
1.5
1.3
1.0
0.5
0.1
0.5
17.5
Maximum
53.0
96.8
103.5
96.9
103.9
122.6
98.0
48.0
17.5
Standard Deviation
4.2
23.7
27.0
27.1
26.3
24.3
22.0
17.7
n.a.
Number of Observations
2
52
134
297
250
1106
111
8
1
Figure 14
Issue-Weighted Recovery Statistics for Various Seniorities Based on Initial Issue Ratings
Senior Secured
Rating
Aaa
Aa
A
Baa
Ba
B
Caa
Ca
C
Mean
Number of
observations
50.1
63.6
67.5
61.2
52.2
54.5
42.7
n.a.
n.a.
2
17
23
23
23
91
17
n.a.
n.a.
Senior Unsecured
Mean
n.a.
40.2
52.7
46.0
41.1
33.7
18.2
19.4
17.5
Number of
observations
Senior Subordinate
Mean
n.a.
14
78
218
122
358
49
5
1
n.a.
n.a.
n.a.
2
37
31
32
8
n.a.
Number of
observations
n.a.
n.a.
n.a.
2
14
359
27
1
n.a.
Subordinated
Junior Subordinated
Number of
observations
Mean
n.a.
n.a.
39.0
31.3
31.5
32.9
13.3
27.1
n.a.
n.a.
n.a.
10
12
66
246
15
1
n.a.
Mean
n.a.
n.a.
n.a.
n.a.
22.4
15.6
24.4
43.5
n.a.
Figure 15
Recovery Rates Based on Issue Rating Immediately Prior to Default
Issue-Weighted
Prior to Def
Rating
Mean
Median
Min
Max
Aaa
Aa
A
Baa
Ba
B
Caa
Ca
C
n.a
n.a
46.8
101.5
69.2
48.3
39.7
40.4
25.0
n.a
n.a
46.8
101.5
76.6
47
36
20
10.0
n.a
n.a
41.9
99.1
6.8
1.0
0.5
0.5
0.1
n.a
n.a
51.8
103.9
101.8
103.0
123.0
107.1
122.6
Standard
Deviation
n.a
n.a
7.0
3.4
22.4
24.8
25.5
25.2
20.23
Number of
Observations
n.a
n.a
2
2
55
296
1020
441
97
Number of
observations
n.a.
n.a.
n.a.
n.a.
2
12
3
1
n.a.
Moody’s Special Comment
15
16
Figure 16
Issue-Weighted Recovery Rates by Security and Priority of Claim Based on Rating Immediately Prior to Default
Moody’s Special Comment
Senior Secured
Rating
Aaa
Aa
A
Baa
Ba
B
Caa
Ca
C
Mean
Senior Unsecured
Number of
observations
n.a.
n.a.
n.a.
n.a.
67.9
64.3
56.1
41
22
Mean
n.a.
n.a.
n.a.
n.a.
2
53
106
27
n.a.
Number of
observations
n.a.
n.a.
46.8
103.9
69.4
43.5
43.8
23.1
17.9
Senior Subordinate
Mean
n.a.
n.a.
2
1
47
108
418
227
31
Number of
observations
n.a.
n.a.
n.a.
n.a.
89
42
34
26
15
Subordinated
Mean
n.a.
n.a.
n.a.
n.a.
1
47
203
104
43
Junior Subordinated
Number of
observations
n.a.
n.a.
n.a.
n.a.
64.0
40.7
32.1
23.8
14.9
Mean
n.a.
n.a.
n.a.
n.a.
5
62
189
68
18
Number of
observations
n.a.
n.a.
n.a.
n.a.
n.a.
28.5
20.0
13.0
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
12
7
8
n.a.
Figure 17
Issue-Weighted Recovery Rates by Security and Priority of Claim Based on Industry
Senior Secured
Industry
Automotive
Construction
Consumer Goods
Finance and Banking
Healthcare
Hospitality
Industrial
Media, Broadcasting & Cable
Miscellaneous
Oil and Oil Services
Real Estate
Retail
Steel
Technology
Telecom
Transport - Air
Transport-Ocean
Transport-Surface
Utility-Electric
Utility-Gas
Mean
81.3
56.0
50.0
46.4
n.a
58.4
48.9
48.1
52.7
55.6
n.a
56.7
32.6
56.1
58.0
60.2
47.8
62.2
70.3
64.4
Number of
Observations
1
2
3
24
n.a
26
35
3
4
13
n.a
12
18
5
3
7
6
3
50
5
Senior Unsecured
Mean
35.4
42.5
28.9
55.0
42.1
47.8
41.4
49.2
49.8
47.5
34.3
49.0
24.4
30.3
20.0
21.6
34.8
41.4
31.4
76.5
Number of
Observations
22
34
32
177
11
16
100
53
21
28
19
87
24
19
186
43
10
33
52
27
Senior Subordinate
Mean
22.3
16.5
32.1
36.5
26.7
41.3
37.5
30.3
33.9
41.7
22.4
33.1
29.0
33.9
26.4
25.9
57.5
62.2
70.2
40.4
Number of
Observations
19
13
27
29
32
30
95
29
20
24
7
42
4
24
16
9
1
16
8
5
Subordinated
Mean
37.1
27.3
27.5
25.3
35.8
25.7
41.9
29.1
36.9
35.3
20.6
23.0
37.0
26.1
29.3
32.8
41.8
34.4
33.2
44.0
Number of
Observations
5
19
17
55
34
24
53
21
21
46
14
43
15
46
16
24
2
7
8
1
Junior Subordinated
Mean
n.a
7.0
12.1
24.0
3.6
31.1
43.5
50.0
n.a
n.a
21.8
29.4
n.a
n.a
n.a
11.3
n.a
48.5
n.a
n.a
Number of
Observations
n.a
2
3
1
1
2
1
1
n.a
n.a
3
5
n.a
n.a
n.a
1
n.a
1
n.a
n.a
Figure 18
Issue-Weighted Recovery Rates by Security and Priority of Claim Based on Domicile
Senior Secured
Domicile
Argentina
Australia
Brazil
Canada
Hong Kong
Indonesia
Korea
Mexico
Netherlands
Other Asia
Other Europe
Other Latin America
United Kingdom
United States
Mean
n.a.
23.6
n.a.
55.4
50.0
28.0
73.2
72.8
55.0
45.0
62.7
40.0
38.6
57.3
Number of
Observations
n.a.
2
n.a.
10
1
9
2
3
1
1
4
1
5
181
Senior Unsecured
Mean
Senior Subordinate
Number of
Observations
31.8
50.0
25.8
30.6
40.5
19.4
103.5
43.4
20.7
43.2
19.4
40.0
24.5
42.3
Mean
38
1
6
59
6
7
2
14
11
6
21
6
38
782
Number of
Observations
30.0
23.6
n.a.
55.4
50.0
28.0
73.2
72.8
55.0
45.0
62.7
40.0
38.6
57.3
Subordinated
Number of
Observations
Mean
1
2
n.a.
10
1
9
2
3
1
1
4
1
5
181
n.a.
38.2
n.a.
16.8
n.a.
n.a.
102.5
n.a.
13.0
n.a.
n.a.
n.a.
20.1
31.0
n.a.
3
n.a.
10
n.a.
n.a.
1
n.a.
1
n.a.
n.a.
n.a.
2
455
Figure 19
Number of Observations Used in Estimation of Holding Period Recovery Rates
Prior Rating
Moody’s Special Comment
Aaa
Aa
A
Baa
Ba
B
Caa-C
Investment Grade
Speculative Grade
All Issuers
One Year
Before
Default
N.A.
1
7
30
41
245
73
38
359
397
Two Years
Before
Default
N.A.
2
18
36
62
218
56
56
336
392
Three Years
Before
Default
N.A.
2
25
42
58
154
35
69
247
316
Four Years
Before
Default
1
7
18
42
48
103
9
69
160
229
Five Years
Before
Default
1
6
17
34
45
68
2
58
115
173
Junior Subordinated
Mean
n.a.
n.a.
n.a.
n.a.
24.0
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
24.1
Number of
Observations
n.a.
n.a.
n.a.
n.a.
1
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
20
17
PAGE INTENTIONALLY LEFT BLANK
PAGE INTENTIONALLY LEFT BLANK
To order reprints of this report (100 copies minimum), please call 1.212.553.1658.
Report Number: 80272
Author
Production Specialist
Praveen Varma
Ida Chan
© Copyright 2004, Moody’s Investors Service, Inc. and/or its licensors including Moody’s Assurance Company, Inc. (together, “MOODY’S”). 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 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 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 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. 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 $1,500 to $1,800,000.
20
Moody’s Special Comment