U.S. Defaults More Than Doubled In 2023 As Credit Pressures Mounted

The number of corporate defaults more than doubled in 2023 to 82, from 32 in the previous year, as inflation and higher interest rates squeezed some issuers' cash flows. The consumer services sector led defaults, and consumer-reliant sectors such as consumer services and leisure time/media contributed the most to the rise in defaults, with 50% of the total.

Downgrades increased in 2023 and outpaced upgrades--a reversal in the trend from the previous year. Overall credit quality deteriorated across all speculative-grade rating categories. The U.S. speculative-grade default rate rose to 4.47% from 1.66% in 2022. This is the highest speculative-grade default rate since 2020 (see chart 1 and table 1).

Table 1

For the purposes of this study, we consider issuers that reemerge from default--including after a distressed exchange, according to our criteria--to be new entities and consider the ratings on these entities to be initial ratings. We make our best effort to capture these defaults in the database, and we include them in the annual default rate calculations if the entity was rated by Jan. 1 of the year it defaulted.

However, if S&P Global Ratings withdrew the rating before Jan. 1 of the year the issuer defaulted, we do not include the issuer in the default rate calculation for that year.

Of the 96 total U.S. corporate defaulters in 2023, 82 had active ratings at the start of the year and 14 were not rated when the year began. Eight of these 14 issuers followed prior defaults earlier in 2023, after these issuers received new initial ratings following an earlier default, and one was newly rated and defaulted in 2023 (see table 2).

Our study of corporate defaults in the U.S. region (which includes the tax havens Bermuda and the Cayman Islands) identified a clear correlation between low ratings and the probability of default. The one-year Gini ratio increased to 89.4% in 2023 from 84.6% in 2022. It remained higher than the 80.5% long-term weighted average one-year Gini ratio for U.S. corporate ratings going back to 1981.

The Gini ratio is a measure of the rank-ordering power of ratings over a given time horizon. High Gini scores reflect a high proportion of the lowest-rated issuers among defaulters in a given period.

This study includes industrials, utilities, financial institutions (including banks, brokerages, asset managers, and other financial entities), and insurance companies from the U.S. region.

We calculated default and transition rates based on the number of issuers in the sample period. The weighted average default rates in this study use the number of issuers at the beginning of each year as the basis for each year's weight (see Appendix 1 for terms and definitions).

The data we present in this study, unless labeled otherwise, refers to only public and confidential issuer credit ratings on nonfinancial and financial issuers and excludes credit estimates.

Table 2

Consumer-Reliant Sectors Propelled U.S. Defaults In 2023

With sectors exposed to consumer spending challenged by shifting demand, the consumer/service sector had the most defaults in the U.S. with 29, following by the leisure time/media sector with 20. Together, they made up 51% of the total U.S. default tally. This marks the seventh consecutive year that the consumer/services sector led the U.S. default tally, although leisure time/media had the highest default rate at 7.34%.

Six nonfinancial industries had a higher default rate in 2023 than in 2022, and all of them had a default rate above their long-term averages (see chart 2).

The U.S. health care default rate had the third-highest percentage point increase over its long-term average (4.48% compared with 1.71%). Overall, ratings in this sector have seen above-average deterioration, which we expect to persist into at least the first half of 2024.

The deterioration is largely contained in the very low-end of the speculative-grade ratings portfolio, which is populated mainly by sponsor-owned health care services companies. These highly leveraged companies are struggling to generate adequate free cash flows given high inflation, particularly in terms of labor, as well as the high interest rates (see "Industry Credit Outlook 2024: Health Care," Jan. 9, 2024).

Most of the time, nonfinancial issuers have higher cumulative default rates than financial issuers, which reflects the higher proportion of speculative-grade issuers within U.S. nonfinancial sectors (table 3).

Table 3

The energy and natural resources industry was the only corporate sector that had more upgrades than downgrades or defaults in 2023. In 2023, the sector remained resilient because producers and refiners have used two years of strong prices and margins to lower their debt levels (see "Industry Credit Outlook 2024: Oil and Gas," Jan. 9, 2024).

Most sectors experienced net negative rating actions in 2023 (more issuer downgrades or defaults than upgrades, see table 4). The consumer/service sector had the most net negative rating actions, with 53, as input cost prices remained high, consumers shifted toward value and cheaper private-label alternatives, and refinancing conditions for highly leveraged, lower-rated companies remained tough given tight capital markets (see "Industry Credit Outlook 2024: Consumer Products," Jan. 9, 2024).

Table 4

All But Two Defaults In 2023 Were Speculative Grade

We rated two defaulters--U.S.-based issuer Silicon Valley Bank and its nonoperating holding company, SVB Financial Group--in the 'BBB' category at the beginning of 2023. These were the first investment-grade (rated 'BBB-' or above) defaults since 2019.

We lowered our ratings on Silicon Valley Bank and SVB Financial Group in March 2023, after the Federal Deposit Insurance Corp. was appointed as receiver. This occurred following a rapid acceleration of customer deposit outflows, made worse by the bank's concentration in corporate deposits from investor-funded technology companies and its unrealized loss position in its held-to-maturity securities portfolio.

Despite these two defaults, most defaults came from issuers rated in the lowest rating categories, supporting the view that ratings are effective indicators of relative credit risk. Defaults in the 'CCC'/'C' speculative-grade rating category in 2023 were above the long-term averages (see table 5).

Given the default of Silicon Valley Bank and SVB Financial Group in 2023, the 'BBB+' and 'BBB' investment-grade rating categories were above the long-term averages.

Table 5

Distressed Exchanges Led Defaults

Distressed exchanges (which are typically selective defaults), were the most common reason for default in 2023, with 48, and accounted for the highest share of affected debt, at 55%. By sector, leisure time/media led distressed exchanges with 12, followed by consumer services with 11 and aerospace/automotive/capital goods/metal with 10.

These distressed exchanges typically involve entities in distress that restructure their obligations in a way that offers less than the original promise. A distressed exchange is an alternative to a conventional default, in which the investor or counterparty stands to fare even worse, and this motivates (at least partially) their acceptance of such an offer.

S&P Global Ratings treats such offers and buybacks analytically as de facto restructuring--and, accordingly, as equivalent to a default on the part of the issuer. Given this analytical treatment, we include distressed exchanges in default rates and other relevant statistics.

Missed principal and interest payments were the second most common cause of default in 2023 (with 24 defaults). Eight issuers each from the consumer services and leisure time/media sectors defaulted for this reason.

The largest default in 2023 came from Community Health Systems Inc., which accounted for $13.9 billion in debt outstanding in March 2023. We lowered the issuer credit rating to 'SD' (selective default) from 'B-' after the company completed open-market and privately negotiated repurchases at below par of a number of notes in March 2023. We consider this tantamount to a default under our criteria given that lenders received less value than they were initially promised under the original securities.

After being re-rated soon after the initial distressed exchange, the issuer was again downgraded to 'SD' from 'CCC+' because it completed open-market and privately negotiated below-par repurchases of both its first- and second-lien notes, which we viewed as tantamount to a default under our criteria. In December 2023, this issuer defaulted for a third time.

The U.S. Accounts For Most Of The World's Speculative-Grade Issuers

Speculative-grade issuers have grown as a share of rated U.S. corporate issuers in recent years as investors have shown willingness to accept higher credit risk for additional yield. The share of U.S. corporate ratings that are speculative grade ended 2023 at 55.1%, below the all-time high (59% in September 2021) and up nearly 10 percentage points since 2004.

Newly assigned ratings have contributed to most of this growth. Since 1980, 67% of new ratings annually have been speculative grade on average, and this proportion has been rising. Since 2009, speculative-grade ratings have averaged about 86% of new ratings annually.

In 2023, we assigned 163 corporate issuers new ratings in the U.S., and 84% of these were speculative grade.

The U.S. accounts for 45% of all global corporate issuer ratings. The high proportion of speculative-grade ratings in the U.S. also accounts for most global speculative-grade corporate issuers (see chart 3).

Globally, there were 3,289 speculative-grade issuers at the end of 2023, with 51% of these based in the U.S. Furthermore, the U.S. accounts for 63% of issuers rated 'B- 'or lower (the lowest-rated speculative-grade issuers) globally.

With more than half of the world's speculative-grade issuers and nearly two-thirds of the weakest-rated issuers, the U.S. typically experiences a higher number of rated corporate defaults annually than other regions. In 2023, 63% of global defaults were from issuers based in the U.S, up from 44% in 2022.

Since 1992, the global speculative-grade default rate has been greater than the U.S. speculative-grade default rate in only seven calendar years (see chart 4).

Corporate Ratings Follow A Gradual Path To Default

As an issuer's credit quality weakens and it moves toward default, the ratings should reflect that.

When we look at the path to default for U.S. corporate issuers since 1981, we see that the median rating for an issuer five years prior to default was 'B+' (for all defaulters since 1981). The median U.S. corporate rating falls to 'B' 30 months prior to default, to 'B-' eight months prior to default, and to 'CCC+' three months prior to default (see chart 5).

U.S. corporate issuers that have defaulted within the past 12 quarters have shown lower median ratings along the path to default. These issuers had a median rating of 'B' five years prior to default. This median rating drops to 'B-' 36 months prior to default, to 'CCC+' 13 months prior to default, and to 'CCC' three months prior to default. The number of issuer defaults over the trailing 12 quarters is considerably smaller (173) than the full pool from 1981 to 2023 (2,447).

The median path to default for nonfinancial issuers is similar to the total sample, given most defaults in this study are from nonfinancial issuers (see chart 6). Nonfinancial issuers have historically had a smoother, more gradual path to default than financial issuers, albeit with a much larger sample.

There were 2,270 nonfinancial issuer defaults from 1981 to 2023, while the total from the most recent 12-quarter period is 167.

By comparison, financial services defaults are less frequent, yet some of these issuers have historically experienced relatively swift transitions to default. Financial services companies are typically more confidence-sensitive than nonfinancial companies, and the loss of confidence from stakeholders (such as counterparties or funding sources) can contribute to a rapid decline in liquidity and credit quality. This was evident during the global financial crisis, when many highly rated financial services issuers defaulted within a short time.

Since 1981, the median rating for financial services issuers has been 'BBB-' five years prior to default, notably higher than the 'B+' median rating for a nonfinancial corporate issuer five years prior to default.

Financial entities that defaulted over the past 12 quarters (from 2021 to 2023) experienced a much more gradual path to default from a lower rating level. These issuers displayed a median rating of 'BB-' five years prior to default, which is two notches higher than the median rating ('B') for a nonfinancial corporate issuer that defaulted over the same period. Notably, for financial services entities, these median ratings are based on a much smaller sample of just 177 issuer defaults (for the period from 1981 to 2023).

To measure upgrades and downgrades in this study, we compared the rating on an issuer as of Jan. 1 with that as of Dec. 31 of the same year. Using this approach, issuers downgraded (or upgraded) multiple times during the year are counted as one downgrade (or upgrade).

The U.S. default rate rose and credit quality weakened, with more downgrades than upgrades in 2023. The upgrade rate fell to 6.95% while the downgrade rate rose to 9.38% year over year (see table 6). The downgrade-to-upgrade ratio increased in 2023 to 1.35x.

Table 6

Transition Tables And Cumulative Default Rates Demonstrate Rating Performance

One-year rating transitions in the U.S. were generally consistent with rating transitions globally in 2023. U.S. investment-grade ratings tend to have more stability (as measured by transition rates) than speculative-grade ratings (see table 7).

For example, 93.84% of U.S. issuers rated 'BBB' on Jan. 1 were still rated 'BBB' on Dec. 31, whereas 85.83% of 'BB' rated issuers maintained a 'BB' rating over the same period. As you move diagonally from 'BBB' across the table, the transition rate should generally fall because the likelihood that ratings move down or up increases further down the ratings scale.

Table 7

In any given year, this relationship may not hold true, but it is clear when multiple samples are averaged together (see table 8). Of the U.S. issuers rated 'AAA', 87.54% retained the rating after one year, whereas only 75.78% of issuers rated 'B' maintained the rating after one year, on average. The stability of higher-rated issuers in the U.S. is largely consistent with global corporate ratings performance (see tables 8-10).

The lowest transition rate (one minus the diagonal rate) is observed not in the 'AAA' rating category but the 'A' rating category, and this becomes more pronounced as the time horizon lengthens. The smaller sample sizes of issuers in the 'AAA' and 'AA' categories likely contribute to this.

Table 8

Table 9

Table 10

When we look at rating transitions in more detail by including the rating modifier (the plus [+] or minus [-] following the rating), the same relationship generally holds true.

However, differences in stability rates frequently exist among rating modifiers within the same rating category. For example, within the 'AA' category, the 'AA+' rating has a stability rate of 82.2%, which is slightly lower than the stability rate of 82.3% for debt rated 'AA' (see table 11). Differences in the sample size of issuers for each rating contribute to this.

Table 11

In addition to greater ratings stability, higher-rated issuers also have lower default rates over time. In the U.S., defaults are less frequent among higher-rated entities than they are for lower-rated entities. This relationship remains true over time, as the cumulative average default rates illustrate (see tables 12 and 13 and chart 8, which illustrates the data in the top half of table 12).

Table 12

Table 13

The average number of years between the initial rating and default for the 96 U.S. corporate defaulters in 2023 was 4.9 years, less than the average 7.9 years in 2022.

The defaulter with the longest time to default in 2023 was Audacy Inc. at 24.0 years. We initially rated the company 'BB-' in 1999. Audacy Inc. has not made the interest payments on its senior secured first-lien revolving credit facility and term loan both due 2024, senior secured second-lien notes due 2027, or senior secured second-lien notes due 2029.

Bed Bath & Beyond Inc. had the shortest time to default in 2023, at 18 days. On Jan. 27, 2023, the U.S.-based specialty retailer defaulted for a first time in 2023 after failing to repay the outstanding loans and additional obligations. The issuer then defaulted for a second time in April 2023 after filing for Chapter 11.

Historically, issuers rated in higher rating categories exhibit longer times to default on average than issuers rated in lower rating categories (see chart 9 and tables 14-15). For example, from 1981 to 2023, the average time to default from the initial rating for U.S. issuers rated 'B' averaged 5.5 years, while issuers rated 'BB' had an average time to default of 7.9 years.

This relationship is true for every sequential rating category from 'AAA' to 'CCC'/'C'. The relationship also holds when the time to default is measured using post-original issuer credit ratings.

In the 43 years covered in this study, from 1981 to 2023, seven U.S. issuers initially rated 'AAA' defaulted: Macy's Inc., Ally Financial Inc. (formerly known as GMAC Financial--a subsidiary of General Motors Corp.), Ambac Assurance Corp., Mutual Benefit Life Insurance Co., Executive Life Insurance Co. CA, Motors Liquidation Co. (formerly known as General Motors Corp.), and Eastman Kodak Co. The average time between initial rating and default for these issuers was 19.4 years.

Table 14

Table 15

Gini Ratios And Lorenz Curves

A quantitative analysis of the performance of S&P Global Ratings' corporate issuer credit ratings shows they continue to correlate with the level of default risk across several time horizons. To measure ratings performance, the cumulative share of defaulters is plotted against the cumulative share of issuers by rating in a Lorenz curve to visually render the accuracy of the rank ordering (see charts 10-12 and Appendix 3).

Over the long term, the U.S. weighted average one-year transition to default has a one-year Gini coefficient of 80.51%, a three-year Gini coefficient of 72.39%, a five-year Gini coefficient of 68.85%, and a seven-year Gini coefficient of 66.34% (see table 16). These weighted average Gini ratios are weighted by yearly issuer counts since 1981 (see Appendix 2).

Table 16

As expected, Gini coefficients decline as the time horizon lengthens because longer time horizons increase the likelihood of credit degradation among higher-rated entities.

In the one-year U.S. Lorenz curve, for example, speculative-grade issuers accounted for 96.7% of total corporate defaults from 1981 to 2023 but only 44.5% of total issuers during that time (see chart 10). The five-year Lorenz curve shows that over a longer time horizon, speculative-grade issuers accounted for 91.9% of defaulters but just 42.8% of total issuers (see chart 12).

If the rank ordering of ratings had little predictive value, the cumulative share of defaulting corporate entities and the cumulative share of all entities at each rating level would be nearly the same, producing a Gini ratio of zero.

Trends in the one-year Gini ratio emerge during periods of extreme highs and lows in the default cycle. In periods with a high number of defaults, there tends to be greater variation in the distribution of defaults across the ratings spectrum, which reduces the Gini ratio--that is, when default pressure is high, economic conditions are such that issuers across the rating spectrum are more likely to suffer a rapid deterioration of credit quality.

The one-year Gini ratio for U.S. corporate ratings increased to 89.4% in 2023 from 84.6% in 2022 (the lowest recorded was 57.6% in 2008).

Appendix 1: Methodology And Definitions

This long-term corporate default and rating transition study uses the CreditPro database of long-term local currency issuer credit ratings. The analysis excludes public information (pi) ratings and ratings based on the guarantee of another company or government entity. We also do not include short-term issuer credit ratings.

S&P Global Ratings does not require all issuers with rated debt to have an issuer credit rating. Therefore, if an issuer has rated debt but not an issuer credit rating, we assign a proxy rating so the CreditPro corporate dataset accurately represents the complete universe of ratings. The local currency senior unsecured rating is the preferred debt rating for the proxy because it is usually consistent with the issuer credit rating. In a small number of cases, we use the subordinated debt rating or the senior secured rating as the proxy.

An S&P Global Ratings issuer credit rating is a forward-looking opinion about an obligor's overall creditworthiness. This opinion focuses on the obligor's capacity and willingness to meet its financial commitments as they come due. It does not apply to any specific financial obligation because it does not take into account the nature and provisions of any single obligation, its standing in bankruptcy or liquidation, statutory preferences, or the legality and enforceability of the obligation.

Counterparty credit ratings, corporate credit ratings, and sovereign credit ratings are all forms of issuer credit ratings. Issuer credit ratings can be either long-term or short-term.

Our ongoing enhancement of the database used to generate this study could lead to outcomes that differ to some degree from those reported in previous studies. However, this poses no continuity problem because each study reports statistics back to Dec. 31, 1980. Therefore, each annual default study is self-contained and effectively supersedes all previous versions.

Issuers included in this study

For the purposes of this study, the U.S. region includes the tax havens Bermuda and the Cayman Islands.

This study analyzes the rating histories of 11,756 corporate issuers in the U.S. that S&P Global Ratings rated as of Dec. 31, 1980, or that were first rated between that date and Dec. 31, 2023. These include industrials, utilities, financial institutions, and insurance companies with long-term local currency ratings. Structured finance vehicles, public-sector issuers, and sovereign issuers are the subjects of separate default and transition studies, and we exclude them from this study.

In this study, the insurance industry includes life insurance, health insurance, property/casualty insurance, reinsurance, bond insurance, mortgage insurance, and title insurance. In addition to these subsectors, this study also group insurance service providers (such as insurance brokers and third-party administrators that are rated according to corporate criteria) with the insurance industry.

To avoid overcounting, we exclude subsidiaries with debt that is fully guaranteed by a parent or with default risk that is considered identical to that of a parent. The latter are issuers with obligations that are not legally guaranteed by a parent but that have operating or financing activities that are so inextricably entwined with those of the parent that it would be impossible to imagine the default of one and not the other. At times, however, some of these subsidiaries might not yet have been covered by a parent's guarantee, or the relationship that combines the default risk of parent and subsidiary might have come to an end or might not have begun. We included such subsidiaries for the period during which they had distinct and separate risk of default.

Issuers with withdrawn ratings

S&P Global Ratings withdraws ratings when an entity's entire debt is paid off or when the program or programs rated are terminated and the relevant debt extinguished. For the purposes of this study, a corporate rating may also be withdrawn as a result of mergers and acquisitions. Others are withdrawn because of a lack of cooperation, particularly when an issuer is experiencing financial difficulties and refuses to provide all the information needed to continue surveillance on the ratings, or at the entity's request.

Definition of default

An obligor rated 'SD' (selective default) or 'D' (default) is in default on one or more of its financial obligations, including rated and unrated financial obligations but excluding hybrid instruments classified as regulatory capital or in nonpayment according to terms. An obligor is considered in default unless S&P Global Ratings believes that such payments will be made within five business days of the due date in the absence of a stated grace period, or within the earlier of the stated grace period or 30 calendar days.

A 'D' rating is assigned when S&P Global Ratings believes that the default will be a general default and that the obligor will fail to pay all or substantially all of its obligations as they come due. An 'SD' rating is assigned when S&P Global Ratings believes the obligor has selectively defaulted on a specific issue or class of obligations but will continue to meet its payment obligations on other issues or classes of obligations in a timely manner. S&P Global Ratings lowers its rating on an obligor to 'D' or 'SD' if the obligor is conducting a distressed exchange offer.

'R' (regulatory intervention) indicates that an obligor is under regulatory supervision owing to its financial condition. This does not necessarily indicate a default event, but during the period of regulatory supervision, the regulators may have the power to favor one class of obligations over others or pay some obligations and not others. Preferred stock is not considered a financial obligation; thus, a missed preferred stock dividend is not normally equated with default. On July 5, 2019, we removed 'R' from all rating scales.

We deem 'D', 'SD', and 'R' issuer credit ratings to be defaults for the purposes of this study. A default is assumed to take place on the earliest of:

• The date S&P Global Ratings revised the rating(s) to 'D', 'SD', or 'R';
• The date a debt payment was missed;
• The date a distressed exchange offer was announced; or
• The date the debtor filed for, or was forced into, bankruptcy.

When an issuer defaults, it is not uncommon for S&P Global Ratings to subsequently withdraw the 'D' rating. For the purposes of this study, if an issuer defaults, we end its rating history at 'D'. If any defaulting entity reemerges from bankruptcy--or otherwise restructures its defaulted debt instruments, thereby reestablishing regular, timely payment of all its debts--we reenter this issuer into the database as a new entity. Its rating history after the default event is included in all calculations as entirely separate from its experience leading up to its earlier default.

Many practitioners use statistics from this default study to estimate the "probability of default" and "probability of rating transition." It is important to note that S&P Global Ratings' credit ratings do not imply a specific probability of default.

Calculations

Static pool methodology.  We conduct our default studies on the basis of groupings called static pools. For the purposes of this study, we form static pools by grouping issuers (for example, by rating category) at the beginning of each year, quarter, or month that the database covers. Each static pool is followed from that point forward. All issuers included in the study are assigned to one or more static pools. When an issuer defaults, we assign that default to all of the static pools to which the issuer belonged.

S&P Global Ratings uses the static pool methodology to avoid certain pitfalls in estimating default rates. For example, this methodology ensures that default rates account for rating migration and allows them to be calculated across multiperiod time horizons.

Some methods for calculating default and rating transition rates might charge defaults against only the initial rating on the issuer, ignoring more recent rating changes that supply more current information. Other methods may calculate default rates using only the most recent year's default and rating data, which may yield comparatively low default rates during periods of high rating activity because they ignore prior years' default activity.

The pools are static in the sense that their membership remains constant over time. Each static pool can be interpreted as a buy-and-hold portfolio. Because errors, if any, are corrected by every new update and because the criteria for inclusion or exclusion of issuers in the default study are subject to minor revisions as time goes by, it is not possible to compare static pools across different studies. Therefore, every update revises results to the same starting date of Dec. 31, 1980, to avoid continuity problems.

Entities that have had ratings withdrawn--that is, revised to not rated (NR)--are surveilled with the aim of capturing a potential default. Because static pools only include entities with active ratings as of the beginning date of a given pool, we exclude issuers with withdrawn ratings, as well as those that have defaulted, from subsequent static pools. If the rating on an entity is withdrawn after the start date of a particular static pool and the issuer subsequently defaults, we will include it in that static pool as a default and categorize it in the rating category of which it was a member at that time.

For instance, the 1981 static pool consists of all issuers rated as of 12:00 a.m. on Jan. 1, 1981. Adding those issuers first rated in 1981 to the surviving members of the 1981 static pool forms the 1982 static pool. All rating changes that took place are reflected in the newly formed 1982 static pool through the ratings on these entities as of 12:00 a.m. on Jan. 1, 1982.

We used the same method to form static pools for 1983 through 2023. From Jan. 1, 1981, to Dec. 31, 2023, a total of 12,619 first-time-rated organizations were added to form new static pools, while we excluded 2,447 defaulting issuers and 7,112 issuers with last ratings of NR.

Consider the following example: An issuer is originally rated 'BB' in mid-1986 and is downgraded to 'B' in 1988. This is followed by a rating withdrawal in 1990 and a default in 1993. We would include this hypothetical issuer in the 1987 and 1988 pools with the 'BB' rating, which was the rating at the beginning of those years. Likewise, it would be included in the 1989 and 1990 pools with the 'B' rating.

It would not be part of the 1986 pool because it was not rated as of the first day of that year, and it would not be included in any pool after the last day of 1990 because the rating had been withdrawn by then. Yet each of the four pools in which this issuer was included (1987-1990) would record its 1993 default at the appropriate time horizon.

Default rate calculation.  We calculated annual default rates for each static pool, first in units and later as percentages with respect to the number of issuers in each rating category. We combined these percentages to obtain cumulative default rates for the 43 years the study covers.

Issuer-weighted default rates.  All default rates that appear in this study are based on the number of issuers rather than the dollar amounts affected by defaults or rating changes. Although dollar amounts provide information about the portion of the market that is affected by defaults or rating changes, issuer-weighted averages are more useful measures of the performance of ratings.

Average cumulative default rate calculation.  The cumulative default rates in this study average the experience of all static pools by first calculating marginal default rates for each possible time horizon and for each static pool, weight-averaging the marginal default rates conditional on survival (survivors being nondefaulters) and accumulating the average conditional marginal default rates (see tables 5, 13-14, and 20-22). We calculated conditional default rates by dividing the number of issuers in a static pool that default at a specific time horizon by the number of issuers that survived (did not default) to that point in time. Weights are based on the number of issuers in each static pool. Cumulative default rates are one minus the product of the proportion of survivors (nondefaulters).

For instance, in table 22, the weighted average first-year default rate for all speculative-grade-rated issuers in the U.S. for all 43 pools was 3.95%, meaning an average of 96.05% survived one year. Similarly, the second- and third-year conditional marginal averages--shown in the summary statistics at the bottom of the table--were 3.91% for the first 42 pools (96.09% of issuers that did not default in the first year survived the second year) and 3.52% for the first 41 pools (96.48% of issuers that did not default by the second year survived the third year), respectively.

Multiplying 96.05% by 96.09% results in a 92.30% survival rate to the end of the second year, which is a two-year average cumulative default rate of 7.7%. Multiplying 92.30% by 96.48% results in an 89.05% survival rate to the end of the third year, which is a three-year average cumulative default rate of 10.95%.

Transition analysis

Transition rates compare issuer credit ratings at the beginning of a period with the ratings at the end of the period. To compute one-year rating transition rates by rating category, we compared the rating on each entity at the end of a particular year with the rating at the beginning of the same year. An issuer that remained rated for more than one year was counted as many times as the number of years it was rated.

For instance, an issuer continually rated from the middle of 1984 to the middle of 1991 would appear in the seven consecutive one-year transition matrices from 1985 to 1991. If the rating on the issuer was withdrawn in the middle of 1991, it would be included in the column representing transitions to NR in the 1991 transition matrix. Similarly, if it defaulted in the middle of 1991, it would be included in the column representing transitions to 'D' in the 1991 one-year transition matrix.

All 1981 static pool members still rated on Jan. 1, 2023, had 43 one-year transitions, while issuers first rated on Jan. 1, 2023, had only one. Each one-year transition matrix displays all rating movements between letter categories from the beginning of the year through year-end. For each rating listed in the matrix's leftmost column, there are nine ratios listed in the rows, corresponding to the ratings from 'AAA' to 'D', plus an entry for NR (see table 25).

The only ratings considered in these calculations are those on entities at the beginning of each static pool and those at the end. All rating changes that occur in between are ignored. For example, if an entity was rated 'A' on Jan. 1, 2023, and downgraded to 'BBB' in the middle of the year and then upgraded to 'A' later in the year (with no other subsequent rating changes), this entity would be included only in the percentage of issuers that began the year as 'A' and ended the year as 'A'. This also applies to transition matrices that span longer time horizons. If an issuer defaults or the rating is withdrawn in the middle of the year, then we would consider it to be rated 'D' or not rated as of Dec. 31 of that particular year.

Multiyear transitions

To calculate multiyear transition matrices, we compared the ratings at the beginning of the multiyear period with the ratings at the end. For example, three-year transition matrices were the result of comparing ratings at the beginning of the years 1981-2021 with the ratings at the end of the years 1983-2023. Otherwise, the methodology was identical to that used for single-year transitions.

We calculated average transition matrices on the basis of the multiyear matrices just described. These average matrices are a true summary, the ratios of which represent the historical incidence of the ratings listed in the first column changing to the ones listed in the top row over the course of the multiyear period (see tables 25-28). Transition matrices that present averages over multiple time horizons are also calculated as issuer-weighted averages.

Rating modifiers

We use rating modifiers (plus and minus signs) 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 rating category suggests that transitions to, for example, 'AA' from 'AA-' or to 'BBB+' from 'BBB-' are not considered to be rating changes because the rating remained within the rating category.

Comparing transition rates with default rates

Rating transition rates may be compared with the marginal and cumulative default rates described in the previous sections. For example, the one-year default rate column of table 13 is equivalent to column 'D' of the average one-year transition matrix in table 9 and the cumulative average in the summary statistics of the year one column in table 20.

However, the three-year default rate column in table 13 is not the same as column 'D' of the average three-year transition matrix in table 10. This difference results from the different methods of calculating default rates.

The default rates in table 10 are calculated as not conditional on survival, while those in table 13 are average default rates, conditional on survival. The three-year default rates in table 13 are calculated in the same way as those in the cumulative average section for the year three column in table 20, while those in the 'D' column of table 10 are equivalent to adding up all the defaults behind the year three column's annual default rates in table 20, divided by the sum of all the issuers in table 20 for the years 1981-2021.

The links between transition matrices and average cumulative default rates are best illustrated through tables 20-22. The default rates in the columns of these tables, associated with each static pool year, are calculated in the same way as they would be for individual years' one-year transition matrices. Tables 20-22 are broken out by the broadest rating classifications (all rated, investment-grade, and speculative-grade, respectively). These tables can also be constructed for each rating category.

As an example, the year two column in table 20 shows the two-year default rates (conditional on survival) for each static pool. These are calculated in the same way as the default column in table 7, though table 7 shows the one-year default rates exclusively.

In the summary section at the bottom of tables 20-22, the first row shows the issuer-weighted averages of the marginal default rates. These marginal averages are then used to calculate the cumulative average default rates in the row directly beneath them, as explained in the average cumulative default rate section above. These default rates are the same that appear in table 13 and are average cumulative default rates conditional on survival.

Standard deviations

Many of the tables and charts in this study display averages of default rates, transition rates, and Gini ratios. Often these are issuer-weighted averages. Prior studies have shown that fluctuations in default rates and transitions can vary greatly depending on many circumstances specific to particular time frames, industries, and geographic regions. As a supplement to many of the averages and time series presented in this study, standard deviations are also shown to provide a gauge of the dispersion of data behind these averages.

For the transition matrices in tables 9-12 and 25-28, the standard deviation for each cell in a given matrix is a weighted standard deviation, calculated using the data from each of the underlying cohort years that contribute to the averages, weighted by that cohort year's issuer base for each rating level.

For example, in the average one-year global transition matrix in table 9, each cell's weighted standard deviation is calculated from the series of that particular cell in each of the 43 cohorts beginning with the 1981 cohort and ending with the 2023 cohort. The squared difference between each cohort's transition rate and the weighted average--which is the data point in each cell--is multiplied by each cohort's weight. These weights are based on each cohort's rating level's contribution to the 43-year total issuer base for each rating level. We then divide this by the ratio of the total number of nonzero weights minus one and the total number of nonzero weights.

For the Gini ratios in table 17, the standard deviations are derived from the time series of Gini ratios for all of their constituent annual cohorts. As an example, the standard deviation applied to the seven-year weighted average U.S. Gini ratio in table 17 (6.03) was calculated from the time series of all available seven-year Gini ratios by cohort. In this case, these are the seven-year Gini ratios from the 1981 cohort through the 2017 seven-year cohort.

We calculated standard deviations for Gini ratios in this study as the standard deviations of a sample, and not those of a population.

Time sample

This study limits the reporting of default rates to the 15-year time horizon. However, the data was gathered for 43 years, and all calculations are based on the rating experience of that period. In addition, average default statistics become less reliable at longer time horizons as the sample size becomes smaller and the cyclical nature of default rates has a bigger effect on averages.

Appendix 2: Additional Tables

Table 17

Table 18

Table 19

Table 20

Table 21

Table 22

Table 23

Table 24

Table 25

Table 26

Table 27

Appendix 3: Gini Methodology

To measure relative ratings performance, we utilize the Lorenz curve as a graphical representation of the proportionality of a distribution, and we summarize this via the Gini coefficient. For this study, the Lorenz curve is plotted with the x-axis showing the cumulative share of issuers, arranged by rating, while the y-axis represents the cumulative share of defaulters, also arranged by rating. On both axes of the Lorenz curve, the observations are ordered from the low end of the ratings scale ('CCC'/'C') to the high end ('AAA').

As an example, if 'CCC'/'C' rated entities made up 10% of the total population of issuers at the start of the time frame examined (x-axis) and 50% of the defaulters (y-axis), then the coordinate (10, 50) would be the first point on the curve. If S&P Global Ratings' corporate ratings only randomly approximated default risk, the Lorenz curve would fall along the diagonal. Its Gini coefficient--which is a summary statistic of the Lorenz curve--would thus 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 (the ideal curve), and its Gini coefficient would be 1.

The procedure for calculating the Gini coefficients is illustrated in chart 14: Area B is bounded by the random curve and the Lorenz curve, while area A is bounded by the Lorenz curve and the ideal curve. The Gini coefficient is defined as area B divided by the total of area A plus area B. In other words, the Gini coefficient captures the extent to which actual ratings accuracy diverges from the random scenario and aspires to the ideal scenario.

Appendix 4: Defaults In Profile

In 2023, there were 96 issuer defaults (including six confidential defaults) on $146.5 billion of debt. Seven issuers defaulted twice: Moran Foods LLC (SAL Acquisition Corp.), Bed Bath & Beyond Inc., Rising Tide Holdings Inc., Community Health Systems Inc., LendingTree Inc., WeWork Cos. LLC, and Range Parent Inc.

Related Research

2023 Annual Global Corporate Default And Rating Transition Study, March 28, 2024