Default, Transition, and Recovery: Recovery Study (U.S.): Piecing Together The Performance Of Defaulted Instruments After The Recent Credit Cycle

1. As more companies emerge from bankruptcy after defaulting during 2008 and 2009, we have put together a clearer picture of the defaults and recoveries from the recent credit downturn. We believe that ultimate recovery rates have rebounded after hitting a trough during the fourth quarter of 2008 and the first quarter of 2009. However, we still expect that recovery rates will continue to vary widely, as they have historically, and we don't expect that they will, on average, return to the highs experienced during the heyday of 2006-2007 anytime soon.

2. From 2008-2010, recovery rates averaged 49.5% across all financial instruments--slightly lower than the average for 1987-2007. We note that ultimate recovery rates (the values of securities when they emerge from default) vary widely around the average and are affected by instrument-, company-, and macroeconomic-level factors. Instrument-level factors include the type of instrument, its seniority within the capital structure, and its collateral type. Company-level factors include the overall capital structure, the debt cushion, and the equity levels. Macroeconomic-level factors include whether the economy is in a boom or a recession, the market liquidity available to fund companies as they emerge from bankruptcy, and the overall default rate. Within the most recent credit cycle, another variable--distressed exchanges--added a new dimension to our recovery data set because a significant number of companies were able to renegotiate their credit terms through distressed exchanges instead of through more traditional bankruptcies. From 2008-2010, distressed exchanges on bonds had an average recovery rate of 42%, compared with 30% for post-bankruptcy exit values.

An Overview Of Our Study's Results

3. We compare the variance in recovery rates for defaulted instruments by first looking at the position of the instrument within the capital structure. From 1987 to 2011 (our sample period), default instrument recoveries improved as the instrument's seniority increases within the capital structure. The revolving credit facilities, which are typically the most senior instruments, had the highest recovery amount, with 88%, on average, of the defaulted principal amount. The average recoveries decline as the level of seniority decreases in the capital structure to term loans, senior bonds, and nonsenior subordinated notes, until it reaches 28% for "all other subordinated bonds" (see chart 1). There is a wide range of outcomes within these average recoveries by instrument type, which we measure by the coefficient of variation (the standard deviation scaled by the mean). Debt cushion, which is the amount of debt below a defaulted instrument, is another variable that affects defaulted instruments' recoveries. The presence of and the quality of the collateral backing a defaulted instrument is also a key variable for analyzing the variance in recovery rates.

Chart 1

4. Recovery rates also vary in response to macroeconomic factors such as business cycle and default rate. We have found that during economic downturns, such as the 1990-1991, 2000-2001, and 2008-2009 downturns, recovery rates were generally lower than the long-run averages. During periods of strong liquidity, such as in 2003-2007, recovery rates were generally above the average. For example, recovery rates hit their peak in 2007 at the height of the credit boom, and the trough occurred during the recession from fourth-quarter 2008 through first-quarter 2009. These peaks and troughs followed the ebb and flow of liquidity in the market quite closely. We also noted that exit valuations tend to vary significantly in any given year, especially during times of economic stress.

5. Over the past few years, there has been a proliferation of distressed exchange offers. Exchange offers, which were mostly made on unsecured bonds, accounted for almost one-third of defaults in the U.S. from 2008 to 2009. Since our study last year, we've added many instruments that have emerged from bankruptcy since 2008 and have drawn some initial conclusions on the recoveries of instruments following both bankruptcies and distressed exchanges during this stress period. We made three key observations on distressed exchanges offers. First, exchanges offers are higher, on average, than bankruptcy exit values for unsecured bonds. This is no surprise since investors would refuse the exchange offers if the expected outcome in bankruptcy of their total exposure to the company were higher than the offer. Second, although distressed exchanges have higher recovery values than bankruptcies, their actual outcomes vary substantially. Third, exchange offers also exhibited significant cyclicality: They generally hit a trough at the height of a recession or financial crisis before rebounding. Note that these conclusions only apply to the performance of unsecured bonds in bankruptcy and in distressed exchanges. Although we have a few examples of secured loans that went through distressed exchanges, the numbers from 2008 and 2009 are too small in relation to the bankruptcy emergences to draw specific conclusions.

6. These key points have emerged from our analysis:

• For all issuers that emerged from default during 2008-2010, the average discounted recovery rate across all instruments was 49.5%, less than the long-term (1987-2007) average of 51.1%. Loan and revolving credit facility recovery values averaged 71.9% on a discounted basis, which was 2% lower than the average for 1987-2007. By contrast, unsecured and subordinated bond recovery values averaged 37%, which was greater than the long-term average of 35%. This resulted from the preponderance of distressed exchange offers in 2008 and 2009 within this asset class. Distressed exchanges are normally associated with higher recovery rates, on average, than emergences from bankruptcy.
• Distressed exchanges dominated the sample of bond emergences recorded from 2008 to 2009, based on the particularly high ratio of exchanges to bankruptcies during those years. In our sample for 2008, there are nearly five distressed exchanges for every bankruptcy. Distressed exchanges accounted for approximately 60% of all bond emergences in 2008 and 2009--but this number dropped to zero in 2010. Historically, within Standard & Poor's LossStats® database, distressed exchanges have comprised 10% of emergences since 1987. From 2008 to 2010, the average distressed exchange value for bonds was 42%, compared with 30% for post-bankruptcy valuations.
• The long-term data continue to underscore the importance of rank in the capital structure as a key determinant of recovery. The more senior the instrument, the higher the recovery rate. From 1987 to 2011, on a discounted basis, lenders achieved recovery rates of 78% on revolving credit lines, 69 % on term loans, 56% on senior secured bonds, 44% on senior unsecured bonds, and 26% on subordinated bonds. For all instruments in our sample, the long-term discounted recovery rate was 51.0%.
• On a dollar-weighted basis, the average recoveries for loans and facilities declined for the period between 2008 and 2010 to 60%, well below the discounted mean of 72% from that period and somewhat below the long-term historical dollar-weighted average of 65%. Much of this discrepancy resulted from one issuer, Lyondell Chemical Co., which declared bankruptcy in 2009. If we were to exclude Lyondell's loans from this sample, then the dollar-weighted recovery rate from 2008-2010 would rise to 70%.
• We believe that it is important to be cautious about forming any analytical conclusions based on averages across all instruments because recoveries vary significantly, and their distribution is not normal. Values tend to gather at the tails of the distribution, with a significant number of instruments having little or no recovery value and another cluster at close to full recovery. Moreover, recovery rates vary even more in years of economic stress. Rank in the capital structure, collateral type, company-level considerations such as leverage and tangibility ratios, and macroeconomic conditions are largely responsible for the high degree of variation.

7. Standard & Poor's Ratings Services provides a comprehensive, fundamental approach to estimating post-default recovery prospects on instruments through its recovery ratings service. We based all of the recovery analysis in this study on Standard & Poor's LossStats database, which is available in Standard & Poor's CreditPro® database and contains ultimate recovery values for more than 3,800 defaulted instruments in the U.S. between 1987 and December 2011. We've included a "Definitions" section at the end of this study, which also includes details about our calculation method.

Recovery Rates Vary Widely Around The 51% Average

8. Recovery rates, which we define as the value of securities at the point of emergence from default, have averaged 51% on a discounted basis and 60% on a nominal basis, based on our sample from 1987 to 2011 (see table 1). If measured on a dollar-weighted basis, which is the sum of all defaulted debt in the sample divided by the sum of the dollar amount of debt recovered, the averages are slightly lower: 48% on a discounted basis and 55% on a nominal basis.

9. Loans and revolving credit facilities, that have seniority in the capital structure and are often secured, have, on average, recovered 74% on a discounted basis and 84% on a nominal basis. When measured on a dollar-weighted basis, the average recovery for loans and facilities is noticeably lower at 65% on a discounted basis and 74% on a nominal basis. These dollar-weighted loan recoveries are 9%-10% lower than the simple averages. Dollar-weighted recoveries can easily be skewed by a small number of large deals. For example, Lyondell, a subsidiary of LyondellBasell Industries AF S.C.A. had a large impact on the sample following its bankruptcy in 2009. Upon its emergence from bankruptcy, Lyondell's $20 billion in bank debt and loans recovered 44% overall on a dollar-weighted basis, and specifically, LyondellBasell Finance Co.'s $8 billion second-lien bridge loan recovered less than 5%. If we exclude Lyondell from our sample, the dollar-weighted average recovery for loans or facilities from 1987-2011 would increase to 68% on a discounted basis.

10. Since bonds are subordinate to loans within the capital structure, it is not surprising that they generally have lower average recoveries. The long-run discounted average recovery for bonds is 38% (46% on a nominal basis), and the average recoveries are approximately the same on a dollar-weighted basis: 39% on a discounted basis or 46% on a nominal basis. As we move down the capital structure, from loans to secured bonds to unsecured and subordinated bonds, the mean recovery decreases while the variance around that mean increases. We measure this variance by determining the coefficient of variation, which is the standard deviation scaled by the mean. Since subordinated creditors receive payment on their claims only after all secured and unsecured senior creditors have been paid, subordinated issues generally have a lower recovery average and a wider variety of recovery outcomes than the more senior issues.

Table 1

The Distressed Exchange Aftermath Following The Credit Boom

11. During 2003-2007, recovery rates increased rapidly as the buoyant credit market provided ample liquidity. This allowed companies to receive attractive valuations upon emergence from bankruptcy and pushed recovery rates above their historical average of 51%. However, recovery rates declined substantially in 2008 with the onset of the recession. From 2008 to 2011, 111 issuers emerged from default with a total of 441 instruments, accounting for a principal amount of approximately $173 billion. Sixty-five of the issuers were bankruptcy cases and 46 were distressed exchange offers. Although the 111 issuers are just a sample of the number of issuers that defaulted during these years, we expect that the sample is representative of entity recoveries. In the U.S., 76 rated entities defaulted in 2008, 168 in 2009, and 58 in 2010. These defaults resulted from factors including bankruptcies and distressed exchanges, and the entities included financial and nonfinancial companies.

12. The large number of distressed exchanges in 2008-2009 is rather unprecedented. Distressed exchanges represented approximately 66% of the instruments in our sample in 2008 and 35% in 2009, compared with approximately 20% from 2000 to 2003--the last cycle of rising defaults. With liquidity generally unavailable for high-risk companies in 2008 and 2009 and debtor-in-possession facilities unattainable for many bankrupt companies, bondholders were more amenable to negotiate terms with distressed companies out of the fear that a bankruptcy would result in liquidation or a near-zero recovery on the defaulted instrument. This year's report has a more balanced mix of bankruptcies and distressed exchanges than last year's report, since we've added more companies to our sample (see chart 2).

Chart 2

13. Overall, the average discounted and nominal recovery rates across all instruments from 2008 to 2010 were 49.5% and 51.9%, respectively (see table 2). On a discounted basis, unsecured and subordinated bonds recovered 37% on average, while loans and revolving credit facilities recovered 72%.

Table 2

14. The 37% average recovery rate for unsecured and subordinated bonds seems surprising for the period, since this is the only instrument group that experienced discounted recovery above its long term average. However, when we exclude distressed exchanges and look only at bankruptcies, the average discounted recovery rate for this group falls to 28%, which is below the long-run average recovery for unsecured bonds. To put the current credit cycle into perspective, we compared the average post-bankruptcy ultimate recoveries across instruments (chart 3). The results show that recoveries are lower across all instruments except for subordinated bonds during the recent cycle, but this above-trend recovery could be a result the relatively small sample size of 20 instruments.

Chart 3

Table 3

15. In our analysis, we use the discounted recovery rate to more accurately compare bankruptcies and distressed exchanges. We also note the issuer-level recovery rate for both bankruptcies and distressed exchanges. In the case of distressed exchanges, the value reflects only the issues that were exchanged.

16. Although the variance remained wide, recovery rates generally declined from the beginning of 2008 through first-quarter 2009 (see chart 4). Recoveries reached their cyclical low during fourth-quarter 2008 and first-quarter 2009. The timing of this low seems reasonable because it aligns with the trough in liquidity, which subsequently rose rapidly. As the U.S. government's Troubled Asset Relief Program and the Federal Reserve's large scale asset purchases brought additional liquidity into the marketplace, this drought subsided. This shift caused exchange values to increase by the beginning of second-quarter 2009, but the rise tapered off by the second-half of 2009 as bankruptcies started to increase (see chart 5; note the distressed exchange cluster in the fourth-quarter 2008 through third-quarter 2009). Ninety-two percent (43) of the distressed exchanges in our sample from 2008 to 2010 occurred in the 11 months ending in September 2009. These 43 distressed exchanges had an average discounted recovery of 43%, compared with an average recovery of 54% for the 34 issuers that filed bankruptcy during the same period

17. We found the recoveries for unsecured bonds in distressed exchanges during 2008 and 2009 were quite different than the recoveries for unsecured bonds in bankruptcy cases (see table 4). The distressed exchanges resulted in a wide range of outcomes for unsecured bonds, which largely experienced discounted recoveries between 10% and 90%. Meanwhile bonds experienced near-zero recoveries following a bankruptcy during fourth-quarter 2008 through first-quarter 2009, as shown in chart 4. Through 2008, unsecured bonds generally recovered more through a distressed exchange than from emergence from bankruptcy. However, this trend reversed in first-quarter 2009, when bankruptcy emergence began to offer higher recovery values than distressed exchanges. Nonetheless, recoveries from both types of default began to increase through second- and third-quarter 2009.

Chart 4

Chart 5

Table 4

Recovery Rates, Credit Cycles, And Default Rates

18. Recovery rates tend to decline in and around years of economic stress (see chart 6), and they exhibit a negative correlation with default rates (see charts 6, 7, and 8).

19. In years when the speculative-grade default rate was higher than 7% (1990, 1991, 2000, 2001, 2002, and 2009), the recovery rate was lower than the long-term average of 50% (see chart 8). One reason for this is that an increase in the supply of defaulted debt during periods of high default rates tends to depress valuations of post-default securities. Furthermore, default rates usually peak during or immediately following contractions in economic growth. In such periods, asset values are generally depressed, leaving the collateral values at lower levels than during times of economic expansion.

Chart 6

Chart 7

Chart 8

Chart 9

20. While the average recovery levels show some correlation to default rates, the variation in recoveries also shows a connection to default rates and the business cycle. To highlight the prevalence of subpar recoveries by year, we also determined the percent of instruments that have recovery rates below the long-term (1987-2010) mean, which we calculated for each basic instrument class in our data set (term loans, revolving credit facilities, senior secured bonds, senior unsecured bonds, senior subordinated bonds, subordinated bonds, and junior subordinated bonds) (see chart 10). The result show that more than half of the instruments recovered less than their long-term average in 11 of the 23 years, reaching peaks during or following recessions (in 1990, 2003, and 2009). Alternatively, in 2007, 85% of all emerging instruments experienced recoveries at least equal to their long-term average by instrument class. We measured this variance by calculating the coefficient of variation of the discounted recovery rate for all instruments by year (associated with time of emergence). We note that variation in recoveries tends to be lower when economic growth is stable and default rates are declining, which the coefficient of variation demonstrates (see chart 11). We also note that changes in the speculative-grade default rate tend to be positively correlated to the coefficient of variation. During periods when the speculative-grade default rate is increasing, the variation among recoveries increases as well.

Chart 10

Chart 11

21. We investigated the extent of cyclical variation by looking at recovery rates by default vintage instead of year of emergence from bankruptcy. This helped us gain insight into the ultimate recovery rates of a pool of issuers that defaulted during similar economic scenarios. We facilitate this comparison by arranging the recovery data into five groups: three major bear market periods during which credit quality declined (1990-1991, 2000-2001, 2008-2009), a bull market from 2003 to 2007, and all remaining years. From this data distribution, we note the following key points:

• During the most recent recession (2008-2009), the average instrument recovery rate of 48.6% across all instruments was higher than during the other recent economic downturns. In comparison, recovery rates averaged 45.3% in 2000-2001 and 45.1% in 1990-1991 (see chart 12).
• Unsecured bonds defaulting experienced a higher rate of recovery of 37.6% in 2008-2009 than the long-term average of 36.5% during nonstressed periods. When taking only unsecured bonds from bankruptcy cases into account, the average recovery rate from 2008-2009 falls to 28% (see chart 13).
• We note the dichotomy in recovery rates on unsecured bonds between bankruptcies and distressed exchanges in 2008-2009 (see chart 14). Distressed exchanges experienced higher average recoveries than instruments emerging from bankruptcy. In addition, investors participating in a distressed exchange could be relatively certain that they would recover at least a portion of their principal investment, while this was less certain following a bankruptcy. Forty-five percent of bankruptcy cases for unsecured bonds resulted in recoveries of 0%-10%, compared with 2% from distressed exchanges.
• Secured term loan recovery rates were lower during 2008-2009 than in 2003-2007, but were higher than in 2000-2001 (see chart 15). In 2008-2009, 50% of secured loans recovered 90% or greater. However, in 2001-2001, more than 30% of secured loans recovered par or greater, compared with less than 2% in 2008-2009. The sample size for 1990-1991 is limited and not comparable.
• The increase in secured term loan debt as a percentage of the total capital structure is part of the reason for the depressed loan recoveries. Indeed, the debt cushion, which measures the percentage of debt below an issue, decreased to 32% in 2008-2010 from the 43% in 2000-2001 (see chart 16).
• Based on our historical data sample, recovery rates were exceptionally high during 2003-2007. The recovery performance during this time is higher relative to other bull-market periods, attesting to the significant impact that the surge in liquidity had on defaulted security valuations. On average, loans recovered 81% on a discounted basis and bonds recovered 57%. On a dollar-weighted basis, loans recovered at a comparable level, while bonds recovered at an even higher 64%. We do not expect that recovery rates will return to these levels in the near future. We believe that these rates were elevated because of the looser credit protections and the easy liquidity that were available during the credit boom in 2003-2007.
• A spotlight on unsecured bonds highlights their anomalous performance in 2003-2007 (see chart 17). Our data show that 40% of securities that defaulted in 2003-2007 had ultimate recovery rates of more than 70%. This was an uncharacteristically high level of recoveries for unsecured bonds, as only 17% of defaulted unsecured bonds recovered 70% or more in all other years. At the opposite end of the spectrum, 29% of unsecured bonds recovered less than 40% in 2003-2007--a much smaller share than the 66% typical during all other years.
• By emergence year, secured instruments' recovery values increased to 75% in 2010, just above their long-term discounted average of 73%. Meanwhile, unsecured instruments remained below their long term average at 30% in 2010 (see chart 18).

Chart 12

Chart 13

Chart 14

Chart 15

Chart 16

Chart 17

Chart 18

Debt Cushion, Collateral Types, And Other Factors Affect Variation In Recovery

22. Recovery rates among instruments vary because of several factors. These include instrument-specific factors such as seniority in the capital structure and the value of the collateral backing an instrument. For example, low-seniority issues that are subordinate to 80% or more of the total debt in the capital structure (that is, debt cushion of 20% or less) vary greatly and recover slightly less than 40% on average, while issues at the top of the capital structure tend to recover much more and see less variation (see chart 19). Even controlling for these factors, we observe a variation in the recovery rate dispersion in different time periods. It is key to evaluate each deal's debt cushion in the context of the quality and the extent to which the company's assets have been pledged to its creditors. A large debt cushion may not provide much of a buffer to creditors if the company's assets have not been valued correctly. Business cycles, which include factors such as the state of liquidity in the market, the supply of distressed debt, and investors' expectations for the economy, are all factors which affect recovery rates.

23. The type of collateral used to secure a bond or loan can affect the recovery rate. Instruments secured by second-lien collateral, in particular, have shown average recoveries lower than other secured collateral types. In 2003-2011, instruments backed by second liens recovered 54.5% on average, compared with 77.2% for all other secured instruments (see table 5). These second-liens had lower debt cushions on average than other types of secured loans. However, even in years when the average second-lien debt cushion approached that of other collateral types, the second lien instrument's average recovery was lower than that of other collateral types (see chart 20).

24. We believe that using the arithmetic mean and standard deviation to describe recovery values can be misleading because recovery rates are not normally distributed. A simple histogram of recovery values for all bonds shows that the distribution is not normal: a large proportion of cases fall into the lowest bucket (0%-10%), while far fewer cases fall into the various ranges above the mean (see chart 21). Recovery experience for loans and revolvers is quite different (see chart 19). Notably, on a discounted basis, 28% of bonds in our sample recovered less than 10%, with only 4.6% of bonds recovering par or better and 7% recovering 0%.

25. The distribution of recoveries for loans and revolving credit facilities is nearly the mirror image to that of bonds. Only 4.3% of loans and revolving credit facilities recovered less than 10%, while 28.6% had at least a par recovery. Loans or revolvers had a recovery of zero in less than 1% of cases,. The difference between bond and loan or revolver recoveries results not only from position in the capital structure, but also from security. Ninety-three percent of loans and revolving facilities were secured compared to approximately 14% of bonds. The standard deviation of recovery by instrument type also shows a wide range by year (see chart 22).

Chart 19

Table 5

Chart 20

Chart 21

Chart 22

Chart 23

26. For more details on recovery rates and potential modeling methods, see "Ultimate Recoveries," by Craig Friedman and Sven Sandow, published in Risk magazine on Aug. 10, 2003, and "Estimating Conditional Probability Distributions Of Recovery Rates: A Utility-Based Approach" by Craig Friedman and Sven Sandow, published in "Recovery Risk: The Next Challenge In Credit Risk Management," Risk Books, 2005, edited by Edward Altman, Andrea Resti and Andrea Sironi.

Definitions

27. We define recovery as the ultimate recovery rates following emergence from three types of default: bankruptcy filings, distressed exchanges, and nonbankruptcy restructurings. Unless specified otherwise, we base recoveries at the instrument level, and we discount them by using each instrument's effective interest rate.

28. We define the effective interest rate as the prepetition rate at the time the last coupon was paid. For fixed-coupon instruments, this is the fixed rate; and for floating-rate instruments, it is the floating rate used at the time of default. Nominal recovery rates are also reported, which are the nondiscounted values received at settlement.

29. We prefer discounted rates in this study because they allow us to better compare bankruptcies of different lengths. For example, the nominal rate on a distressed exchange could be the same as that on a bankruptcy case that takes two years to resolve. However, investors in the bankruptcy case are significantly worse off because they could lose significant time value while waiting for the final settlement. On the other hand, a distressed exchange could take only a day. In a historical study, discounted recovery rates offer the major benefit of making different time periods more comparable by preventing any major bias that could occur if time between default and emergence differs greatly. Note that Standard & Poor's provides recovery ratings that map to nominal values. This is appropriate because it lets investors choose their own discount rates when making decisions based on their nominal estimates.

30. Recovery is the value creditors receive on defaulted debt. Companies that have defaulted and moved into bankruptcy will usually either emerge from the bankruptcy or will be liquidated. On emergence from bankruptcy, creditors often receive a cash settlement, new instruments (possibly debt or equity), assets or proceeds from sale of assets, or some combination.

Ultimate recovery

31. Ultimate recovery is the value of the settlement a lender receives by holding an instrument through its emergence from default. The recovery is based on the amount received in the settlement divided by the principal default amount. Within Standard & Poor's LossStats database, three recovery valuation methods are used to calculate ultimate recovery:

• Trading price at emergence. We can determine the recovery value of an instrument by using the trading price or market value of the prepetition debt instruments upon emergence from bankruptcy. Of the three methodologies, this one is the most readily available because most debt instruments continue to trade during bankruptcy proceedings.
• Settlement pricing. The settlement pricing includes the earliest public market values of the new instruments that a debtholder receives in exchange for the prepetition instruments. It is similar to the trading price method, except that it is applied to the new (settlement) instrument instead of the old (prepetition) instrument.
• Liquidity-event pricing. The liquidity event price is the final cash value of the new instruments or cash from the sale of assets that the lender acquires in exchange for the prepetition instrument.

Discounted recovery

32. We calculated the discounted recovery values by discounting instruments or cash received in the final settlement on the valuation date back to the last date that a cash payment was made on the prepetition instrument. The last-cash-pay date represents the true starting point for the interest accrual, which is why this date is used as the starting point for the discounting rather than the default date of the instrument or the bankruptcy date of the company. In Standard & Poor's LossStats database, the coupon rate for the instrument is used as the discount factor.

Thirty-day trading price recovery

33. Market participants have historically used a 30-day after-default recovery method, though trading prices 30 days after default may not accurately reflect the financial strength of the debtor upon emergence. Actual trading prices on the 30th day (calendar days) after default are not always available, and when they are, they may be affected by market volatility on that date. To correct this and to ensure consistency within Standard & Poor's LossStats database, we used the average trading price of all available prices between 15 days and 45 days after default.

Related Research

• U.S. Recovery Study: Capital Structure, Cyclicality, and Distressed Exchanges Are Key to Recovery Performance, Dec. 7, 2010
• U.S. Recovery Study: Increased Defaults And Low Liquidity Depressed Recovery Rates In Early 2009 (Premium), Feb. 24, 2010
• Distressed Exchanges Are Increasing And Affecting Recovery (Premium), Jan. 28, 2009
• Liquidation Increasingly Likely In Some Post-Default Recoveries, Jan. 7, 2009
• As Defaults Increase, The Debt Market Spotlight Shifts To Recovery, June 20, 2008
• U.S. Recovery Study: Swollen Recovery Rates Conceal Underlying Trauma, Nov. 21, 2008