Quarterly Global Corporate Default Update And Rating Transitions

Globally, nine companies (eight public and one confidentially rated) defaulted in the third quarter of 2011. The volume of rated debt affected by defaulters in the third quarter was $6.3 billion, down from 15 defaults in the second quarter with $39.5 billion in debt.

These salient points emerged from the third-quarter 2011 default update and rating transitions report:

• Of the nine defaulters in the third quarter, seven are domiciled in the U.S. region, one is from Israel, and one is based in New Zealand. The seven U.S. defaulters accounted for 92.7% of the total $6.3 billion in debt affected.
• The trailing-12-month global speculative-grade default rate as of September 2011 was 1.65%, down slightly from 2% as of June and 3.4% as of the same period in 2010. The default rate is now at its lowest point since June 2008--before the global financial crisis and the ensuing recession in the U.S.
• Overall credit quality has, in our view, remained stable over the past 21 months as the number of upgrades outpaced downgrades globally. The downgrade-to-upgrade ratio rose to 0.98% in the third quarter from 0.64% in the previous quarter, but it's still below 1%. The increase in this ratio is partially attributable to the many downgrades of 'AAA' rated U.S. home loan banks. Standard & Poor's Ratings Services downgraded 22 U.S. home loan banks to 'AA+' following the U.S. sovereign downgrade on Aug. 5.
• By rating category, default activity over the past four quarters relative to long-term averages indicates a prolonged period of favorable lending conditions globally. Even the 'CCC' to 'C' rating category had a default rate over the past four quarters that was nearly 10 percentage points lower than its long-term average.
• An analysis of the transition rates during the four quarters ended Sept. 30 suggests that ratings behavior remains consistent with long-term trends, showing a clear negative correspondence between credit quality and default probability, in our opinion. The notable exception to this is the 'AAA' rating category, which saw a proportionately large number of entities in the U.S. downgraded to 'AA+' following the U.S. sovereign downgrade.

Table 1

Quarterly Default And Rating Trends

A quantitative measure of ratings' rank-ordering power over a given time horizon--the Gini ratio--shows a Gini coefficient of 82% at one year, 75% at three years, 72% at five years, and 70% at seven years. If corporate ratings' rank orders only randomly approximated default risk, the Gini coefficient would be zero. Alternatively, if corporate ratings were perfectly rank ordered so that all defaults occurred only among the lowest-rated entities, the curve would capture all of the area on the graph above the diagonal, and its Gini coefficient would be 100% (for more details, see "Appendix: Gini Methodology" and chart 12). For the four quarters ended Sept. 30, the one-year global Gini coefficient was 91.37%.

Table 2

In the third quarter, nine companies defaulted, affecting debt worth $6.3 billion. This is a modest amount historically--both by the number of defaulters and the amount of affected debt--even though it's up from five defaulting entities with $3.6 billion in debt in the first quarter (see charts 1 and 2). This also compares with 19 defaults, affecting rated debt worth $57.3 billion, in third-quarter 2010. If the pace of defaults set so far in 2011 continues throughout the rest of the year, defaults will total 39 for full-year 2011--a marked decline from 81 in 2010 and 265 in 2009 (see table 1). However, fears of contagion are beginning to build as the European debt crisis continues. Greece's default is becoming more likely, and other peripheral countries, as well as Italy, are facing higher borrowing costs. Although the count of defaults is still relatively low, the financial landscape portends for an eventful remainder of 2011.

Chart 1

Lending conditions remained positive in the third quarter, as indicated by the modest number of defaults relative to the elevated levels during the same period in 2009. The default count fell to nine in the third quarter from 15 in the second quarter. Similarly, the amount of debt affected by defaulting entities fell to $6.3 billion from $39.5 billion. However, we do note that one issuer--U.S.-based NewPage Corp.--accounted for $3.2 billion of the third-quarter total. By both default count and affected dollar amounts, the majority of the activity was in the U.S. (see chart 2).

Chart 2

Defaults in the trailing four quarters were concentrated in the 'CCC' to 'C' rating category (see table 3). The 'B' and 'BB' rating categories also saw some defaults in the last four quarters, but at levels far below their long-term averages. Even the expectedly weak 'CCC' to 'C' category displayed a default rate over the last four quarters that was nearly 10 percentage points below its average. The 'B' category is also typically a weak performer, marked by elevated default rates. However, its default rate over the past four quarters is more than one standard deviation lower than the long-term average--a telling sign of the positive credit market conditions for corporate borrowers as of late.

Table 3

Data that we collected from the pool of global defaulters indicate that defaults among speculative-grade entities tend to be clustered in the third year after the initial rating, particularly in the 'BB' and 'B' rating categories (see chart 3). For example, among defaulters that were rated 'B' at origination, the default rate climbs to a high of 19.2% in the first three years and falls thereafter. Defaulted issuers rated 'BB' at origination show a similar pattern, but they peak a little later--in the fourth year. Conversely, for defaulting entities initially rated 'CCC', we observed the highest default rate in the first year, which is not surprising given their weaker credit profiles.

Chart 3

The average time to default (from original rating) for the nine defaulting issuers in the third quarter was 5.52 years, roughly equivalent to the 5.36 years on average for the defaulters in the second quarter and slightly faster than the 5.76-year average for the entire database, which covers 30.5 years. However, it should be noted that among the nine defaulters, PMI Mortgage Insurance Co. took 26.5 years to reach default, pushing up the average time to default during the third quarter to 5.52 years. If PMI was not included, the average time to default would only be 2.9 years. PMI was also the only defaulted entity in the third quarter that was originally rated investment-grade ('BBB-' or higher). The rating path among defaulters in the trailing 12 quarters broadly reflects the long-term ratings trend, which shows that both the average rating and the median rating on all defaulting entities were in the speculative-grade category in the five years preceding default (see chart 4).

Chart 4

Generally, there is a negative correlation between the initial rating on an issuer and its time to default, if that were to occur. For example, for the entire pool of defaulters (1981 through September 2011), the average times to default for issuers that were originally rated in the 'A' and 'B' categories were 12.5 years and 4.7 years, respectively, from the initial rating date (or from Dec. 31, 1980, the start date of the study), whereas issuers rated in the 'CCC' category or lower had an average time to default of only 2.6 years. In cases when an entity emerges from a previous default (including distressed exchanges), we considered it a separate entity, with the original rating being the first rating after the default event. In our analysis, we consider the median, average, and standard deviations for the time to default based on the original rating (see table 4; the "range" column shows the differences between each rating category's minimum and maximum times to default). We also note the average and median times to default from each rating category and both the rating originations and the transitions for each category (see table 5). In both tables, the standard deviation of the times to default shrinks as the rating gets lower.

Table 4

Table 5

Very often, defaulting entities fall victim to what can be considered a rapid decline in credit quality before default, marked by a series of downgrades. These downgrades often occur faster leading up to the default event. We present this downgrade momentum in table 6, which displays time to default statistics similar to those in tables 4 and 5, but are calculated from the last day that an entity possesses a particular rating (before an upgrade or downgrade), rather than from the first day it was assigned a rating or from the first day it receives an upgrade or downgrade.

Not surprisingly, the average and median times to default are shorter when measured from the last day the defaulter possessed a particular rating rather than based on either an initial rating or from the first day of receiving a rating following a default event. In particular, this perspective shows the extent to which a few of the nine 'AAA' defaulters have been outliers because their median time to default is much shorter than the average, as well as the median times to default from tables 4 and 5.

Table 6

During the third quarter, ratings stability increased relative to previous two quarters as the number defaults, downgrades, upgrades, and ratings withdrawals fell. In the three months ended Sept. 30, the global default rate fell to 0.15% from an already relative low of 0.24% in the second quarter. Meanwhile, the proportion of upgrades and downgrades reached virtual parity in the third quarter, coming in at 2.47% and 2.42%, respectively--in contrast to the increased proportion of upgrades relative to downgrades in the second quarter. As a result, the downgrade-to-upgrade ratio increased to 0.98% for the third quarter. (A ratio of 1% indicates that the number of downgrades is equal to the number of upgrades.) The quarterly rate of downgrades has been below the 4% long-term average since fourth-quarter 2009, reflecting a prolonged period of favorable conditions for corporate borrowers globally.

Table 7

Quarterly Defaults By Sector And Region

Of the eight publicly rated defaults reported in the third quarter, seven were domiciled in the U.S. and one in New Zealand (see table 8). Three were from the transportation sector, two from forest products and homebuilders sector, and one each was from the insurance, consumer and service, and financial institutions sectors. (For a more detailed description, see "Third-Quarter 2011 Default Synopses," published Nov. 9, 2011, on RatingsDirect on the Global Credit Portal.)

Historically, defaults in the financial sector precede those in the nonfinancial sectors by about three months (see chart 5). However, during the most recent downturn, the two sectors began to move in unison, though financial defaults started to pick up in third-quarter 2010. The nine defaults in third-quarter 2011 led to a default rate of 0.18% for nonfinancial entities and 0.10% for financial entities. Of the nine defaulters, two were classified as financial issuers--belonging to either the insurance or financial institutions sectors. The default rate for the financial sector clearly exhibits more volatility than that of the nonfinancial sector, but the number of financial issuer defaults is smaller relative to nonfinancials.

Table 8

Chart 5

Further reflecting the favorable global credit market conditions over the past 12 months, quarterly speculative-grade default rates have continued to fall in all regions from a year ago (see chart 6). The global speculative-grade default rate finished the third quarter at 0.34%, down from 0.68% as of the same period a year ago. The U.S. default rate also decreased to 0.46% from 0.77% in third-quarter 2010, and the European default rate was 0% for both third-quarter 2011 and third-quarter 2010. However, the quarterly default rates in all regions are slightly higher relative to the first quarter, though lower than those of the second quarter. Quarterly default rates tend to provide a more current picture of the current default trends because we calculate them independently from default activity from 12 months ago, which is reflected in the headline numbers (the trailing-12-month figures). Considering this, the most notable finding of our study is that when measured either on a trailing 12-month or quarterly basis, default rates are at or near their lowest points in the past few years.

Chart 6

In the past few months, both the global and regional default rates have started to settle at the low levels seen before the recent recession. Through September, the trailing-12-month global speculative-grade default rate was at 1.65%, compared with 3.4% in third-quarter 2010 (see table 9 and chart 7). In the U.S., the speculative-grade default rate declined to 1.9% in September 2011 from 4% in September 2010. The most recent readings for the default rates are now the lowest seen in more than three years. In November 2009, the trailing 12-month default rates were 10% globally and 11.5% in the U.S. Although the U.S. economy has not resumed a healthy pace of growth as it did after the previous recessions, credit conditions for corporations have become far more sanguine with the help of the Federal Reserve's quantitative easing policies and maintenance of low interest rates. Corporate borrowing costs increased around the time of the U.S. sovereign downgrade but have been falling sharply since early October. Despite this, new issuance within the U.S. has been minimal through this period, particularly in the speculative-grade domain, owing to increased investor skepticism towards the U.S. economic recovery and a fear of Europe's debt troubles having contagion effects across the Atlantic.

Chart 7

Table 9

Typically, a growing share of speculative-grade-rated issuers foreshadows an increase in the speculative-grade default rate. An increase in the proportion of lower-rated issuers during the five years starting in the spring of 2002 provided the basis for the recent increase in the speculative-grade default rate, which peaked in November 2009 (see charts 8, 9, and 10). It's also interesting to note that globally, as well as within the U.S., the long-term trend of the proportion of issuers with speculative-grade ratings is increasing, while the speculative-grade default rates have been reaching less severe peaks with each subsequent recession. Also, the proportion of issuers with speculative-grade ratings has been steadily increasing in the U.S., as well as globally. In fact, within the U.S., 52% of active corporate ratings are speculative grade.

Chart 8

Chart 9

Chart 10

With the downgrade of the U.S. to 'AA+' from 'AAA' on Aug. 5, nearly two dozen regional home loan banks and insurance companies whose ratings were tied to that of the U.S. were subsequently downgraded to 'AA+' from 'AAA' shortly thereafter. This is reflected in the corporate transition matrices for both global and U.S. regions (see table 10). Because the 'AAA' rated universe of corporate borrowers is relatively small, the impact of the downgrades has been substantial, with the 'AAA' stability rates finishing the last four quarters at 46.7% and 17.7% for global and U.S. issuers, respectively.

Table 10

Despite this shock to the highest rated entities, an analysis of transition rates during the four quarters ended Sept. 30, suggests that ratings behavior remains consistent over the long term, showing a clear negative correspondence between credit quality and default probability. Globally, investment-grade issuers tend to exhibit less ratings volatility than their speculative-grade counterparts (see table 10). For instance, the probability that any issuer rated 'A' at the beginning of the period (for example, Oct. 1, 2010) will still be rated 'A' at the end of the period (Sept. 30, 2011) is 91.8%, whereas the probability that a 'B' rated issuer will still be rated 'B' at the end of the four quarters is 76.3%. The same relationship holds true when we analyze transition rates separately for the U.S. and Europe. This pattern generally coincides with the long-term (1981-2010) ratings behavior among all global rated issuers (see table 11).

Table 11

U.S. Speculative-Grade Default Forecast

Our baseline forecast (with a 60% probability) for the U.S. corporate speculative-grade default rate for the 12 months ending in September 2012 is 3.1%. This prediction is lower than the long-term average of 4.59% and the 2.25% recorded as September 2011. To realize the baseline projection, a total of 48 speculative-grade-rated issuers would need to default in the next 12 months. This implies an average of four defaults per month, more than the average of about 2.3 per month in the last 12 months. (For more details, see "U.S. Corporate Default Rate Expected To Rise To 3.1% By September 2012," published Nov 4, 2011.)

We draw the forecast results from a variety of factors, including Standard & Poor's proprietary default model for the U.S. speculative-grade bond market. The main components of the model include economic variables such as the unemployment rate; financial variables such as corporate profits, the Federal Reserve's Senior Loan Officer Opinion Survey on Bank Lending Practices, interest burden, and the slope of the yield curve; and credit-related variables such as negative bias. The interaction between the dependent variable--the U.S. speculative-grade default rate--and the input variables is in line with expectations. For instance, increases in the unemployment rate and negative bias historically have been positively correlated with the speculative-grade default rate. This means that as unemployment rates increase or as the proportion of entities with negative outlooks or ratings on CreditWatch negative rises, default rates generally accelerate.

The default projections are consistent with our in-house economic projections, which represent our economists' best estimate of where the U.S. economy could be heading. However, many factors could contribute to substantial variability. In addition to the baseline, we test for two alternative scenarios, one pessimistic and the other optimistic. The pessimistic scenario yields a mean 12-month projection of 5.1%; a total of 79 issuers would need to default to realize the pessimistic scenario. Conversely, the optimistic scenario yields a mean 12-month projection of 1.6%, which would indicate 25 issuers defaulting in the next 12 months.

Chart 11

Appendix: Gini Methodology

To measure ratings performance or ratings accuracy, the cumulative share of issuers by rating is plotted against the cumulative share of defaulters in a Lorenz curve to show the accuracy of its rank ordering. Max O. Lorenz developed the Lorenz curve as a graphical representation of the proportionality of a distribution. To build the Lorenz curve, the observations are ordered from the low end of the ratings scale ('CC') to the high end ('AAA'). If Standard & Poor's corporate ratings' rank orders only randomly approximated default risk, the curves would fall along the diagonal. Thus, the Gini coefficient--which is a summary statistic of the Lorenz curve--would be zero. If corporate ratings were perfectly rank ordered so that all defaults occurred only among the lowest-rated entities, the curve would capture all of the area above the diagonal on the graph, and the Gini coefficient would be one (see chart 12). To calculate the Gini coefficients, as shown in chart 12, we would divide area B by the total area A plus B. In other words, the Gini coefficient captures the extent to which actual ratings accuracy diverges from the random scenario.

Chart 12

Related Criteria And Research

• Third-Quarter 2011 Default Synopses, Nov. 9, 2011
• U.S. Corporate Default Rate Expected To Rise To 3.1% By September 2012, Nov 4, 2011
• 2010 Annual Global Corporate Default Study And Rating Transitions, March 30, 2011