NEW YORK (Standard & Poor's) Sept. 26, 2011—The proliferation of distressed exchanges in recent years has had a major impact on default rates but only a minor effect on ratings performance. As Standard & Poor's Ratings Services' default studies have shown, default rates can fluctuate wildly over time. This is particularly apparent in the buildup and fall from peaks in the default cycle that typically occur during and immediately after recessions. We've examined the impact of selective defaults on the overall default rate, as well as how they affect ratings performance as measured by the Gini ratio, which is a measure of the rank-ordering power of the rating system over a given time horizon. Distressed exchanges are one example of a selective default. We've seen that while selective defaults have accounted for a substantial proportion of the overall speculative-grade default rate at times, this trend has been mostly limited to the past 10 years because selective defaults were relatively uncommon prior to that time. Also, distressed exchanges have had a more muted impact on ratings performance than on the overall default rate.

Distressed Exchanges Account For Most Selective Defaults In Recent Years

For the most part, distressed exchanges have been geographically centered in the U.S., which experienced a surge in these transactions during 2009 (see table 1). All other countries had at most a handful of companies within their borders that used this mechanism for payment alleviation in the past three and a half years.

Distressed exchanges also appear to be correlated to both economic pressures and higher borrowing costs. The incidence of distressed exchanges hit a peak in 2009-–the same year that general defaults peaked. However, as a proportion of total defaults, distressed exchanges constitute nearly a third of the defaults seen in the year to date (through May 31). During 2009--the peak of the recent default cycle--distressed exchanges peaked in popularity, constituting just under 39% of all defaults recorded in that year. That was more than four times the share seen in 2008. The stigma of default is a strong motivator for many companies to do all that they can to avoid it, at least its general definition. From an investor's perspective a selective default is also preferable to a general default because distressed exchanges tend to have higher recovery rates among bonds. For more information on corporate recovery rates, see "U.S. Recovery Study: Capital Structure, Cyclicality, And Distressed Exchanges Are Key To Recovery Performance," published on Dec. 7, 2010.

During the most recent U.S. recession--the worst since the Great Depression--credit markets froze to a halt, and high-yield bond spreads in the U.S. averaged roughly 1,029 basis points (bps) during 2009 (see table 1). Despite these difficult conditions, more than 100 companies managed to avoid total default on all of their obligations that year and were able to exercise distressed exchanges, thus reducing their bondholders' potential losses. While they are still selective defaults, distressed exchanges have shown to be far less onerous to bondholders than if a company defaults on all of its outstanding obligations.

Table 1

When broken out by sector, it is clear that the majority of instances of distressed exchanges since 2008 occurred in the leisure time/media sector (34). Although this sector may have seen the highest number of distressed exchanges, the proportion of total defaults in the leisure time/media sector that were attributable to distressed exchanges over this time period is roughly in line with the total, coming in at 29.8%. By this measure, the leading sector is utilities, where all of the defaults in the industry were distressed exchanges. Obviously, this is an outlying sector with regard to all measures of default, with only two defaulters in the past three and a half years. Still, other sectors, such as the forest and building products/homebuilders and aerospace, automotive, capital goods, and metal sectors, experienced both a considerable number of distressed exchanges and a higher-than-average proportion of these defaults (see table 2).

Table 2

How Selective Defaults Affect The Speculative-Grade Default Rate

We consider distressed exchanges to be a form of selective default. While these events generally occur far less frequently then general defaults, they have increased in recent years and have particularly affected the speculative-grade default rate. In our default studies, we consider instances of selective default to be equivalent to a general default for our issuer-based default rates and transition matrices. For more information on ratings criteria related to distressed exchanges, see "Rating Implications Of Exchange Offers And Similar Restructurings, Update," published on May 12, 2009.

When separating defaults by these delineations, or types, we observe the considerable effect that selective defaults had on pushing the speculative-grade default rate to its recent peak. Chart 1 compares speculative-grade and adjusted speculative-grade default rates in the U.S. going back to 1981. While the first selective default recorded in the database occurred in May 1991, these defaults did not noticeably affect the overall rate until roughly 1999. Beginning in the recession of 2001, they became more numerous, producing, on average, a 1.12% difference over the period of October 1999 through March 2004. During the relatively calm years of 2004 through 2007, the number of selective defaults declined again, bringing the adjusted default rate back in line with the overall rate.

However, as the most recent recession became deeper and more prolonged, selective defaults, particularly distressed exchanges, surged, accounting for nearly half of the defaulters during the period of exponential increase in the default rate. At its peak in November 2009, the speculative-grade default rate in the U.S. climbed to 11.5%, a higher point than during the prior recession. However, contributing to this increase was a sizable influx of selective defaults, which when taken out of the total number of defaulters, reduces the default rate to a peak of 6.9% during the same month-–a difference of 4.6%. The same holds true globally, with the difference in default rates during November 2009 at a comparable 4.3%.

Chart 1

How Selective Defaults Affect Ratings Performance

Clearly selective defaults have had a substantial effect on the overall speculative-grade default rate, particularly toward the peak in the most recent cycle. But while the presence of selective defaults may increase the default rate by nearly five percentage points, their effects on ratings performance is more muted. Charts 2 and 3, as well as table 3 present Gini ratio statistics and time series globally.

Over the one-year horizon, the largest differences between the original Gini series (which appear in our default studies) and the modified series (which exclude selective defaults), occur in the years 2004 and 2007. At the end of 2004, the difference in the two Gini ratios was 3.62, and the difference at the end of 2007 was 3.24. This implies that more defaulters in the overall universe that were ultimately selective defaults began these years with lower ratings relative to their cohorts who were classified as general defaulters. While the largest differences in the global Gini ratios occurred in 2004 and 2007, overall the years with the largest differences are clustered in the period from 2002 to 2010 (see chart 2). This is largely due to the increase in distressed exchanges over the past few years, and we think that this relationship will continue if distressed exchanges remain a popular option among issuers in financial trouble.

Chart 2

The adjusted three-year Gini ratio also displays the same relationship relative to the standard figure. Chart 3 shows that the largest difference between the standard Gini ratio and the adjusted measure also occurs with the 2004 cohort, producing a difference of 3.62 in the three years ending 2006. Similarly, the largest differences between the two series occur across the cohorts beginning 2002 and after. This is consistent with expectations as the incidence of selective default has increased steadily since this period.

Chart 3

Despite there being some differences in the Gini ratios between rating populations inclusive and exclusive of these events, the overall magnitude of these differences are muted. Table 3 presents summary statistics for multiple time horizon Gini ratios. Whether measuring the weighted long-term averages, or the standard averages, the differences between populations inclusive and exclusive of selective defaults is no greater than 1.45%, seen in the weighted averages for the one-year Gini ratios. In fact, these divergences collapse over longer time frames.

Since selective defaults have not been present throughout the entire time frame beginning Jan. 1, 1981, we think the only fair comparison would be to compare the performance of the time frames in which they existed relative to all default events. The bottom half of table 3 shows the same summary statistics, but covering only the years ending 1998-2010. For example, in the three-year time horizon, this would begin with the 1996 cohort, with the three years ending Dec. 31, 1998. Although 1998 was not the first year in which a selective default occurred, it is the first year these defaults became more than an isolated anomaly. The bottom portion of the table shows that on average, differences between Gini ratios still exist, but they are relatively minor. The Gini ratios for the rated population exclusive of selective defaults are still lower than those for the entire rated population, but the largest gap amounts to only 2.15% in the one-year weighted average. Variability between the two Gini measures remains roughly equivalent, either when looking at the entire time frame, or the abbreviated 1998-2010 horizon.

Table 3

Table 4

Related Criteria And Research

• 2010 Annual Global Corporate Default Study And Rating Transitions, March 30, 2011
• U.S. Recovery Study: Capital Structure, Cyclicality, And Distressed Exchanges Are Key To Recovery Performance, Dec. 7, 2010
• Till Debt Do Us Part: A Study Of Serial Defaulters, Nov. 10, 2010
• Rating Implications Of Exchange Offers And Similar Restructuring, Update, May 12, 2009