Credit quality for global structured finance securities deteriorated for the fifth consecutive year in 2011. The downgrade and upgrade rates in 2011 were significantly higher than long-term historical averages. In 2012, we expect overall downgrade and default rates to drop on improving underlying collateral performance for key sectors. However, because structured finance collateral performance usually lags that of the overall economy, we still expect to see moderate rating volatility and defaults overall. Potential effects of criteria changes could also contribute to rating volatility this year.

U.S. collateralized loan obligation (CLO) credit performance was strong, continuing a trend that started in late 2010 and through 2011, resulting in a number of upgrades and positive CreditWatch placements. U.S. consumer asset-backed securities (ABS) generally performed well. However, we lowered a large portion of 'AAA' rated ABS FFELP student loans to 'AA+' due to the U.S. sovereign downgrade. Last year, upgrades rose and downgrades dropped among European structured finance. Asia-Pacific structured finance securities also performed relatively well in 2011, and Australian RMBS has continued to have solid performance and strong market fundamentals. In recent years, we've seen criteria changes prompt a significant number of rating actions. Because transitions can reflect both criteria changes and a transaction's actual credit performance, we believe rating transition rates may not be as indicative of credit trends as they once were.

In 2011, we lowered our ratings an average of 1.2 notches on global structured finance securities (a one-notch downgrade for example, would be from 'AA' to 'AA-'), significantly less than the 3.5 notches in 2009 and less than the 1.4 notches in 2010. We downgraded approximately $2.5 trillion of the roughly $9.6 trillion original issuance amount of outstanding global structured securities at the beginning of 2011. These downgrades reflect a combination of criteria updates and our assessment of the deteriorating credit quality of the underlying mortgage and other collateral in some countries. Notably, adverse performance in U.S. collateral also caused large numbers of defaults among investment-grade ('BBB-' or higher) and speculative-grade ('BB+' or lower) global structured securities in 2011.

Nevertheless, bonds with higher ratings continue to experience lower average default rates, and lower-rated bonds continue to default more frequently. We empirically tested this behavior and the overall ranking power of ratings through rank-ordering statistics, such as the Gini coefficient. At the same time, the annual 'AAA' absolute default rate in 2011 slightly declined to 0.19% from 0.28% in 2010, but declined more significantly from 0.54% in 2008. Approximately $450 billion in securities defaulted last year, which represents less than 5% of the total outstanding original issuance amount of global structured securities.

Looking forward, we believe improving collateral performance—particularly in consumer ABS and CLOs—is likely to aid global structured securities performance in 2012 and beyond. However, elevated unemployment and the slow recovery in U.S. housing may continue to weigh on collateral performance, especially among residential mortgage-related collateral. We expect the large "shadow" housing inventory--or the volume of distressed U.S. residential mortgages--to remain a concern for the U.S. market. In most European countries, Asia, Japan, and Australia, delinquency rates have either stabilized or started to decline in recent years.

We observed the following key global credit trends during 2011:

• Overall, global structured ratings declined by roughly 1.2 notches, when taking all rating changes and even stable ones into account.
• We downgraded approximately $2.5 trillion (26%) of the roughly $9.6 trillion original issuance amount of outstanding global structured securities in 2011, and the amount of defaulted securities was $450 billion (5%) during the year.
• The one-year Gini coefficient for global structured finance ratings, which provides a measure of the ratings' ability to rank-order default risk, was 78.8% in 2011, up from 71.8% in 2010. However, the latest five-year Gini coefficient statistic was 45.6% at the end of 2011, showing the relative weakness in the rank-ordering power for global structured finance ratings over that period.
• The percent of classes we upgraded rose to 5.4% in 2011 from 2.6% in 2010.
• We lowered 36% of all global structured finance ratings in 2011, up from 33.5% during 2010.
• Rating volatility increased in 2011; roughly, 41.5% of global structured securities experienced rating transitions, up from 36% in 2010.

We also noted the following trends at the sector and subsector levels during 2011:

• We lowered U.S. commercial mortgage-backed securities (CMBS) ratings primarily due to adverse collateral performance, and the sector experienced a downgrade rate of 40.8% (based on the number of outstanding ratings).
• The credit quality of U.S. ABS declined as well, but this sector still performed relatively better than other U.S. collateral. The upgrade and downgrade rates were 5% and 9.9% of the number of ratings outstanding at the beginning of 2011, respectively.
• Declines in credit quality for U.S. FFELP student loan ABS primarily came on the heels of the U.S. sovereign downgrade.
• Globally, collateralized debt obligations (CDOs) backed by mortgage securities experienced significant downgrades due to the continued negative rating actions on the underlying RMBS.
• Corporate CDO transactions, including cash flow CLO transactions and synthetic corporate investment-grade CDOs, saw significant downgrades in late 2009 and early 2010 due to changes in the criteria we use to rate these transactions combined with a stressed economic environment. Late in 2010 and 2011, however, upgrades have reigned in the U.S. CLO sector as collateral performance rebounded.
• The ratings on a significant number of U.K. residential mortgage-backed securities (RMBS) and U.S. student loan ABS are currently on CreditWatch with negative implications.
• In Europe, significant declines in commercial real estate prices have led to refinancing risk for commercial mortgage loans, thus contributing to a downgrade rate of 39.2% in European CMBS.
• European RMBS and ABS saw downgrade rates of about 28.1% and 21.6%, respectively, and collateral performance varied significantly between subsectors.

Elevated Default And Transition Rates In 2011

U.S. housing market-related sectors continued to show credit deterioration and rating volatility in 2011, despite signs of the sector's recovery, including recent lower delinquency and foreclosure rates and stabilizing home sales. Performance among certain key sectors during 2011 was noticeable. In particular, FFELP student loan ABS downgrades were mainly related to the U.S. sovereign downgrades.

The downgrades we took on CMBS primarily reflected actual and expected credit deteriorations. On the other hand, U.S. CLOs backed by leveraged loans saw an unprecedented number of upgrades resulting from solid performance among speculative-grade corporate loan issuers, very low default rates on leveraged loans during the year, and pay downs to the senior tranches of earlier vintage CLO transactions.

Consumer loan delinquencies and losses stabilized in 2011, reaching record low levels for credit cards. In 2011, ratings were relatively more stable among consumer ABS—particularly among transactions backed by auto loans and credit cards. However, a large number of U.S. private-label student loan ABS ratings are currently on CreditWatch negative, reflecting the elevated level of the U.S. unemployment rate, among other factors. Many of the ABS downgrades in recent years have reflected rating actions on related monoline insurers and counterparty financial institutions.

Rising Upgrades And Falling Downgrades For Euro And APAC Structured Finance

The number of downgrades among European structured finance transactions continued to decrease in 2011 amid the region's slow recovery over much of the year. The downgrade rate fell slightly for the second consecutive year to 22.3% from 25.4%. However, the number of downgrades was still over 2.5 times (x) the number of upgrades. We raised the ratings on 8.5% of tranches outstanding at the beginning of 2011, up from 5.6% in 2010. Excluding 'AAA' ratings, the upgrade rate was 10.7%. The average magnitude of downgrades fell to 2.9 notches in 2011 from 4.0 in the previous year. The average magnitude of upgrades has begun to edge up over the past two years, and reached 2.2 notches in 2011. The average change in credit quality rose to -0.5 notches from -0.9 notches in 2010.

We downgraded a third of tranches rated 'AAA' at the beginning of 2011 during the year. We lowered the ratings on some 25% of RMBS and CMBS tranches after our revised criteria for assessing counterparty risk took effect in January 2011. (See "Counterparty And Supporting Obligations Methodology And Assumptions," published Dec. 6, 2010, and "2010 Counterparty Criteria—What Happened Next For Global Structured Finance Ratings??" published July 22, 2011). Almost 70% of downgrades relating to the application of our revised counterparty criteria affected tranches that we previously rated 'AAA'. Successive downgrades of the Irish, Portuguese, and Greek sovereigns had a domino effect on the ratings of structured finance securities from these countries.

By asset class, European CMBS exhibited the highest downgrade of 39.2%, followed by 28.1% for European RMBS in 2011 due to depressed real estate markets and tight refinancing conditions. The annual default rate in European structured finance has increased every year since 2005, reaching 2.3% in 2011. Overall in Europe, the performance of most consumer-related collateral remained relatively robust, and credit deterioration was most evident among Spanish RMBS and ABS, as well as in the nonconforming niche of U.K. RMBS. European transactions backed by corporate risk saw greater deterioration, and European CMBS suffered from unprecedented declines in commercial real estate values, leading to a significant rise in upcoming refinancing risk.

Asia-Pacific structured finance securities performed relatively well in 2011. RMBS continues to dominate the Australian market with good credit performance and strong market fundamentals. Australian ABS has also experienced positive rating changes two years in a row. The strong RMBS performance also prevails in Asia-Pacific, with Japan experiencing upgrades and other countries seeing rating stability. The credit quality deterioration has slowed for Japanese CMBS, compared with 2010. Japanese CDOs reversed course with more upgrades in 2011, while Japanese ABS experienced moderate downgrades.

Overall, the rating transition statistics for 2011 represent the combination of criteria updates and adverse collateral performance. As a result, the historical rating transition trend may be less informative of credit trends in recent years than in previous periods, when credit performance primarily drove the rating changes.

Higher Gini Coefficients Suggest Improving Performance

Higher ratings are associated with lower default rates, and lower ratings typically see higher defaults overall. Charts 1 and 2 and table 1 document that annual defaults for both investment- and speculative-grade securities were much higher from 2008-2011 than in any previous years in the history of the structured market (except for the spike in 1988 due to a low sample size). The default rates for 'AAA' securities were near 0.3% in 2009 and 2010, but dropped to 0.19% in 2011. The historical weighted average one-year 'AAA' default rate is 12 basis points (bps, with a standard deviation of 12 bps), compared with a default rate of approximately 54 bps in 2008. Longer-horizon default rates (e.g., three-, five-, and 10-years) are all provided in Appendices 2 through 4. For example, the three-year 'AAA' default rate was 3.8% at the end of 2011, down from 6.04% in the previous year. Five-year 'AAA' default rates also increased in the periods ending in 2011, but the 10-year period does not yet fully capture the adverse behavior of 2005-2007 vintage mortgage-related securities.

Chart 1

Chart 2

Table 1

We assess how well our ratings have performed as indicators of relative default risk over time. For this reason, we are calculating historical Gini coefficients, or measures of statistical dispersion, to analyze the relative rank-ordering performance of our ratings in various categories (see chart 3).

To calculate the Gini coefficients, we plotted the cumulative percent of structured finance ratings (sorted from low to high ratings) against the cumulative proportion of defaults in a Lorenz curve (a graphical representation of the proportionality of a distribution) to help demonstrate how our ratings rank-order default risk. In the chart, the "random" line—for which the cumulative proportions of ratings and defaults are the same, indicating no ability to rank risk—provides a benchmark for measuring the ranking test of our ratings. The "ideal" curve represents the best possible results at rank-ordering default risk.

In general, the higher the Gini coefficient--which represents the area between the random line and the Lorenz curve--the greater the correlation between our ratings and the securities' default behavior. We calculate the Gini coefficient using a scale from 0% to 100%. In 2011, the one-year Gini coefficient for defaults of rated global structured securities was 78.8%, which indicates an acceptable level of correlation, in our opinion. This value is up from 71.8% in 2010, but down significantly from about 93.4% in 2006. In 2012, we expect the one-year Gini coefficient to continue improving over last year.

We also calculate sector-specific historical one-year Gini coefficients, as well as coefficients over longer periods of time, such as three and five years. As shown in chart 4, Gini coefficients were in the 90% area for each sector before 2006, whereas performance varied significantly among sectors in 2011. For example, the one-year Gini coefficient for CDOs improved to 91% in 2011 from 67% in 2010. The primary cause of the overall decline in the Gini coefficient for global structured securities in 2009 and 2010 was the drop in this coefficient for CDOs. However, the overall one-year Gini coefficients have also increased for the ABS sector. CMBS and RMBS one-year Gini coefficients are relatively low, improving only slightly.

Gini statistics are generally useful, but they have shortcomings. For example, speculative-grade securities are likely to default within a one-year time frame, leading to relatively higher but less useful Gini statistics. On the other hand, we believe the five-year Gini coefficient has its own shortcomings, in that it may not fully reflect the poor performance of 2005-2007 vintage mortgage-related structured securities. Given these limitations, we believe the three-year Gini coefficient is the most meaningful in this study because it mainly reflects the poor performance of 2005-2007 vintages.

Overall, the three-year Gini coefficients are showing some signs of improvement, as they reached 72.8% in the latest three-year period from 2009 to the end of 2011, up from 45.7% in 2010 and up from 43.6% in 2009 (see chart 5). Excluding the latest few years, the Gini coefficients generally indicate that structured finance ratings remained highly correlated with default rates over three-year windows.

Finally, the five-year Gini coefficients have declined further to 41.6% in 2011, down from 56.9% in 2010. While the historical variation in Gini coefficients highlights the cyclical behavior of defaults over time, the recent decline indicates the relative weakness in the rank-ordering power of structured finance ratings.

Chart 4

Chart 5

Sector's High Rating Volatility Reaches Fifth-Year Mark

Despite some signs of global economic recovery, Euro sovereign debt crises remain a key concern for the global economic recovery. At the same time, the slow recovery continues to hurt the performance of structured finance securities for the fifth consecutive year on a lagged basis, meaning that we may continue to see deterioration in credit quality even if economic recovery picks up. But the pace of deterioration slowed in 2011.

Mortgage-related securities continued to drive downgrades and defaults in 2011, following similar trends a year earlier. Residential mortgage collateral is generally performing better in Europe, Asia, Japan, and Australia than in the U.S.

U.S. RMBS still experienced a high number of downgrades and defaults in the past few years. In addition, CMBS experienced downgrades in early 2010 and 2011. These rating actions generally followed Standard & Poor's revised criteria assumptions to account for the slow global economic recovery and its outlook for U.S. unemployment levels. In the meantime, CLOs have seen steady upgrades since late 2010, which has continued into 2011 because of the improving credit performance of underlying collateral.

Both the overall downgrade and upgrade rate rose slightly in 2011. The one-year downgrade rate for global structured securities slightly increased to 36% from 33.5% in 2010, while the upgrade rate increased to 5.4% from 2.6% in 2010. Additionally, in terms of severity of rating changes, the average downgrade was about 1.2 notches when we consider the magnitude and frequency of rating transitions, as well as the stability of the ratings.

The rating transitions for 2011 revealed areas of stability, variations in default rates, and correlation with the initial rating levels: That is, securities with higher ratings tended to be more stable than their lower-rated counterparts, or experience upgrades (see chart 6 and table 2).

Chart 6

Table 2

Chart 7 shows the upgrade and downgrade rates for each rating category in each sector during 2011. The results show higher downgrade rates for RMBS, CMBS, and CDOs, and lower downgrade rates for ABS across the rating categories.

Chart 7

Chart 8 compares rating stability with default and near-default rates for the various rating categories during 2011 and illustrates that less-stable rating categories also tend to experience higher default rates. In other words, clear correlations exist between the rating level, the stability of the rating, and the frequency of defaults and near-defaults.

Chart 8

Table 3 shows the number of ratings outstanding at the beginning of 2011 for each sector and region and provides rating transition rates for the year.

Table 3

Chart 9 shows the frequency of upgrades and downgrades for our global structured finance ratings during 2011 and illustrates aggregate credit performance for the combined regions and sectors during several previous years. The 5.4% upgrade rate in 2011 was up from 2.6% in 2010, while the downgrade rate dropped to 33.5% from 36% at the end of 2010.

Chart 9

Tables 4 and 5 show the historical performance of each sector and region. In the U.S., the number of ABS upgrades during 2011 was higher than in the previous year: we raised 4.7% of our outstanding U.S. ABS ratings last year, up from 4.3% in 2010. At the same time, U.S. ABS downgrades dropped to 9.9% in 2011 from 13.2% in 2010. A few ABS subsectors—including student loan ABS, small-business loans, and franchise loans—accounted for most of the downgrades. Student loan ABS downgrades were mainly related to the U.S. sovereign downgrade. In Europe, however, the level of ABS downgrades increased to 21.5% in 2011 from 12% in 2010.

Ratings volatility has also risen in recent years in markets outside the U.S. and Europe. Although, in some instances the high upgrade or downgrade rates must be placed in the context of a relatively small data set: a case in point is the high downgrade rate in Asia, excluding Japan. We also provide standard deviations for annual upgrade and downgrade rates; high variability over time indicates cyclical components of rating transitions.

Table 4

Table 5

The magnitude of rating changes hit record highs in 2008 and 2009. Approximately, 41% of the downgrades in 2008 and 25% of downgrades in 2009 involved 10 notches or more, the highest annual percentages we've ever observed. While the rates for the past two years are high, the percent did improve to 9.5% in 2010 and 6.6% in 2011. At the same time, roughly 56.5% of upgraded securities and 55.5% of the downgraded securities experienced no more than a two-notch rating change in 2011. Additionally, about 43.5% of upgraded securities and 44.5% of downgraded securities had transitions of three notches or more in 2011 (see chart 10).

Chart 10

The year of issuance or vintage continues to explain the variation in performance. The fundamental credit characteristics and economic environment of a transaction's vintage year—for example, the recession in the U.S. and globally between 2000 and mid-2003, which brought changes in underwriting standards—may influence its current credit behavior. Table 6 illustrates the impact of the issuance (vintage) year on credit performance of global structured securities in 2011.

Transactions issued in 2005 through 2007 experienced significantly higher downgrade and default rates in 2011 than other vintages, owing to credit deterioration among mortgage securities and CDOs backed by these mortgage securities. Default rates were also uncharacteristically high for these more recent vintages, particularly securities originated in 2007, which posted a 19.3% default rate and a 5.7% near-default rate.

Table 6

Auto And Corporate-Related Subsectors Dominate 2011 Upgrades

Transactions collateralized by auto and corporate debt constituted a significant percentage of the positive rating actions overall during 2011, especially CLO transactions. Downgrades also affected high percentages of the ratings on securities from many subsectors in 2011. U.S. cash flow CDOs backed by mezzanine mortgage securities led downgrades among key categories, at 74.5%, and U.S. Alt-A RMBS came in second at 65.5% (see table 7).

Table 7 also shows sector-specific default rates. For example, Alt-A RMBS saw 30.2% default rate in 2011 and hybrid high-grade CDOs of ABS default rate is (76.5%).

Table 7

Default Rates Trend Higher

Cumulative and vintage-based default rates, as well as the time to default are alternative ways to approach the default performance. Higher-rated securities usually take longer to default, while lower-rated securities default sooner. However, this expectation may not always play out if the sample size is small, especially among securities that were highly rated before defaulting.

Table 8 provides the distribution of the times to default—which represent the number of years between the origination date and the default date for each defaulted security—according to the original rating categories.

Table 8

At the end of 2011, the cumulative original issuance amount of defaulted global securities had reached approximately $1.3 trillion since the inception of the structured market. To put this number in perspective, the total outstanding balance of global structured securities in 2011 was about $8 trillion and total ever rated was $19 trillion in terms of original issuance amount.

Table 9 further breaks down the cumulative defaults by issuance year for investment-and speculative-grade ratings. Overall lifetime default rates for investment- and speculative-grade securities were 21.9% and 47.7% through the end of 2011, respectively. As shown in chart 11, speculative-grade default rates were higher for issuance years 1994-2001 and 2004-2008.

Table 9

Chart 11

Cumulative default rates (CDRs), which are summed across many years, provide another way to summarize the behavior of static pools. Table 10 shows the default rates for global structured finance using this methodology for up to 10 years, updated through 2011. We calculated the CDRs by dividing the number of ratings in a static pool that defaulted at the end of a specific time horizon by the number of ratings that survived or did not default up to that point (the statistical term is "conditional on survival"). We assigned all ratings in the study to one or more static pools (i.e., the constituents in each pool didn't change) according to their age. We used the number of ratings to weight each static pool. After creating the pools, we averaged and weighted the first-year default rates for all the pools, and then the second-year default rates, and so forth. The CDRs equal one minus the product of the proportion of survivors.

The results of our CDR analysis again suggest that lower ratings are generally associated with higher default rates, and vice versa. The table also provides standard deviations in default rates; high standard deviations indicate that the default rates are spread out over a large range of values. Speculative-grade ratings, in particular, show high standard deviations over longer periods.

Table 10

Improving Credit Quality Trends

Positive performance in the auto and corporate-related sectors helped the drop in the global structured finance's credit quality to slow in 2011. The credit quality of global structured securities significantly increased between 2004 and 2006, later reversed starting in 2007, and lasted to date (see charts 12 and 13).

Average credit quality has declined in the RMBS, CDO, and CMBS sectors between July 2007 and December 2011, and the declines were more significant in 2008 and 2009. The criteria changes caused a significant number of downgrades in 2009 and early 2010, especially among CMBS and CLOs. The deterioration was not as significant for ABS.

The average credit quality of structured securities fell more significantly in 2008 and 2009 than in previous periods. When we took both the magnitude and frequency of rating changes into account, the average credit rating for all structured finance securities dropped 3.5 notches during 2009 and 1.4 notches in 2010. The decline was only 1.2 notches in 2011.

There has been divergence in performance of different asset classes within structured finance, with global RMBS and CMBS turning in net downgrades of 1.68 and 1.44 notches, respectively. Comparatively, the average downgrade in the ABS sector was about 0.21 notches and CDOs were 0.11 notches for the 12 months ended Dec. 31, 2011. Chart 12 shows the average number of notches changed for each sector in each 12-month period. This calculation considers our entire portfolio of structured ratings (both public and confidential) and gives equal weight to each rating outstanding at the beginning of each 12-month period.

Chart 12

Chart 13 shows the average changes in credit quality for structured finance securities in each rating grade. Investment-grade securities, which had been more stable during the past few years, saw greater declines in credit quality during the latest 12-month period, dropping an average of 1.15 notches, compared with a 1.19-notch decrease for speculative-grade securities.

Chart 13

Table 11 provides descriptive statistics for the changes in credit quality shown in charts 12 and 13.

Table 11

High Correlation Between CreditWatch Placements and Future Rating Changes

A number of U.S. cash flow CLO tranches (392) on CreditWatch positive as of March 12, 2012, are likely to see upgrades in the upcoming months. A moderate number of securities--including U.K. RMBS and U.S. student loan ABS--are currently on CreditWatch with negative implications, suggesting that these securities could experience downgrades over the next few months following our review of their performance. Table 12 summarizes the number of ratings on CreditWatch negative, positive and developing among the global structured finance subsectors.

Table 12

The CreditWatch data provide insight into the potential near-term behavior of those securities over the next few months. However, the credit performance of global structured securities for 2012 will ultimately depend primarily on the health of global economy—in particular, the performance of global corporate credits, the unemployment picture, and interest rates, as well as any changes in housing market fundamentals.

Appendix I: Methodology And Terminology

This section provides a detailed discussion of the rating transition and default methodology. It also explains the study's terminology, including definitions of transition windows, rating modifiers versus full rating categories, near-defaults, treatment of rating withdrawals and identical ratings, and region and sector combinations.

Default rates are the key focus in our study and we continue to emphasize the volatility of historical default rates. In particular, we provide standard deviations of default and transition rates for each rating category, and we highlight their historical volatility and provide their long-term averages.

We also analyze rating volatility (as measured by the standard deviation of transition and default rates) because we believe it provides additional insight into the range of historical default and transition history. The economic environment, particularly in this credit cycle, is a significant cause of the variations, but other factors—such as the availability of data, the length of the performance period, and the small sample sizes in certain years—may explain additional variability in historical default and transition rates.

We also focus on the rating transition and default trends for long-term issue ratings of global structured securities in this study. This includes ABS, CDOs, CMBS, RMBS, and single-name synthetics. We do not discuss short-term ratings, such as those on asset-backed commercial paper conduits or issuer credit ratings in this report.

In this study, "default rate" refers to the percentage of ratings we lower to 'D' over a certain period. "Near-default rate" refers to the percentage of securities we downgraded to the 'CC' or 'C' rating categories. "Transition" refers to how much a Standard & Poor's Ratings Services rating has changed, either up or down, over a certain period. To measure the relative performance of our ratings over time, we report both default and transition rates, and provided rank-ordering, or Gini, statistics.

In addition, we quantify ratings movements using transition rates, which measure the proportion of ratings that have changed over a given time period. Readers might find transition rates useful as benchmarks of how stable or volatile our ratings have been historically, and compare with the experience of securities with similar characteristics in their own portfolios across sectors.

Rating transition

The rating transition of a single static pool (cohort approach) reflects the rating of each security at the beginning and end of the observed transition window. For instance, we calculated the transition rates by determining the ratings of each security at the beginning and end of the period. We then tabulated these ratings in a two-dimensional table to calculate the percentage of ratings that stayed the same and the percentage of ratings that changed. During this process, we count every security only once, even if it had more than one rating change during the period. In other words, we use a security's rating on the first day of the period and the last day of the period to calculate the transition rates, while disregarding the interim rating changes.

Weighted average transition

For weighted average transition rates, we calculate the individual transition rates of static pools as described earlier. We then create a single averaged matrix weighted by the number of ratings in each static pool.

Transition window

A transition window refers to a defined period in which we observe a security's rating transition. For example, the 2010 performance transition window starts at the beginning of 2010 and ends at the end of 2010. For historical data, transition windows follow annual windows, running from January to December of the same year.

Rating modifiers

We use rating modifiers ("+" and "-") to calculate the upgrade and downgrade percentages, as well as the magnitude of rating changes throughout this study. However, some transition tables may use full rating categories for practical reasons. In other words, the use of a full rating category suggests that transitions from, for example, 'AA' to 'AA-' or from 'BBB+' to 'BBB-' are not considered to be rating transitions because the rating remained within the full rating category.

Rating withdrawal

We do include a security with a withdrawn rating at the beginning of a transition window in the transition and default rate calculations for that period.

Identical 'AAA' ratings

In assessing rating transitions, this study collapses the rating history of identical 'AAA' ratings in a given transaction. Where we treat original 'AAA' ratings from the same transaction as a single rating, we include only the rating with the longest expected maturity in the calculation of rating transitions.

Near-defaults

In this study, near-defaults include securities downgraded to 'CC' or 'C' during the period or in any other transition window. If a security's rating is already 'CC' or 'C' at the beginning of a transition window, we don't include it in the near-default category because that event did not take place during the transition window analyzed. The study's tables show the default and near-default rates. We consider securities rated 'CC' or 'C' to be highly vulnerable to nonpayment risk.

Structured finance region

We do not include Canadian transactions in the U.S. region. Europe includes transactions mainly in the countries of the European Union. Asia (non-Japan) includes all transactions in Hong Kong and Taiwan, as well as in other countries in the region. Japan includes transactions only in Japan. The Latin America and emerging markets segment includes emerging market countries, such as Brazil, Mexico, Turkey, and South Africa. We use a transaction's issuer country as the primary variable to determine its region. However, when the issuer was established in a tax-haven territory, such as the Cayman Islands, then we use its domicile of assets as the secondary variable for defining the region. In some cases, the domicile of assets may fail to provide additional insight into the region, especially when a transaction's domicile of assets includes more than one country. In those cases, we use our surveillance office and/or the transaction's primary currency to complete the region classification for all structured finance securities we rate.

Structured finance sector

We further segment the U.S. and European regions according to the structured finance sectors. We don't use the sector to segment the data in other, smaller regions because the sample size would typically be too small to calculate meaningful transition rates. ABS includes typical collateral types, such as credit cards, student loans, and auto loans/leases. ABS also includes manufactured housing, franchise loans, 12b-1 transactions, and corporate securitizations. We include home equity loan transactions in the RMBS sector. We include single-name synthetics, also known as "repacks" in Europe, in the U.S. and European single-name synthetic categories. However, the CDO sector includes all synthetic CDO ratings, in addition to cash and market value CDOs and leveraged funds.

Appendix II: Default And Transition Tables

Appendix tables 1-18 provide various default and transition rates for global structured securities.

Appendix table 1

Appendix table 2

Appendix table 3

Appendix table 4

Appendix table 5

Appendix table 6

Appendix table 7