Standard & Poor's Ratings Services has updated its data on the performance and default rates of sovereign ratings through year-end 2008. (1) We believe the data continue to support our view that:
The sovereign rating default experience is in line with reference default rates proposed under the Basel II guidelines, even though default rates for individual years vary widely.
Sovereign ratings are more stable at higher rating levels.
Upgrades have outnumbered downgrades of both foreign- and local-currency sovereign credit ratings over the last decade, though downgrades far outnumber upgrades in 2008 and will likely continue to do so in 2009.
The relative rank ordering of sovereign ratings is consistent.
Sovereign ratings are no more volatile than other credit ratings; large rating movements in either direction are the exception and not the rule, even over several years.
Local- and foreign-currency default rates of sovereigns with speculative-grade ratings will likely rise from current levels now that spreads on external financing rates are widening and the world enters a synchronized recession.
Background And Methodology
Standard & Poor's published its first sovereign rating transition study for foreign-currency ratings in 1999. (2) We have enhanced, expanded, and updated the study annually, following the same methodology used for our corporate and structured finance default studies. (3) We believe that examining default rates is valuable because all of Standard & Poor's default studies have found a clear correlation between credit quality and default remoteness: As a general matter, the higher the rating, the lower the probability of default, and vice versa. Many default studies, including this one, also look at transition rates, which gauge the degree to which ratings change—either up or down—over a particular time. Transition studies have repeatedly confirmed that higher ratings tend to be more stable and that speculative-grade debt generally experiences more rating volatility.
This study is based on long-term sovereign credit ratings. The methodology tracks rating changes over time, including changes to 'D' for default or, more commonly for sovereign issuers, to 'SD' for selective default, as sovereigns often continue to service some of their debt to commercial and official creditors. The study is based on issuer (in this case, central government) ratings, not issue ratings. Issuers are not adjusted for size (i.e., they are not dollar-weighted), so for the purposes of this study, a default by Argentina counts the same as a default by the Seychelles. Our study tracks defaults on a sovereign's commercial debt, including both bonds and bank loans (see Appendix 1). When tallying defaults for the study, we use both public and confidential ratings.
Standard & Poor's only rates sovereign governments upon request for the initial rating. Some governments elect not to publish an initial rating. After a government consents to publish the initial rating, Standard & Poor's changes ratings publicly, up or down, as credit events warrant. Standard & Poor's has withdrawn a public sovereign rating once, that on the Republic of Mali in July 2008 at the government's request. Today, all 124 sovereigns that Standard & Poor's rates have public local- and foreign-currency ratings.
This study excludes ratings when the government did not accept both the foreign- and the local-currency unpublished ratings Standard & Poor's initially assigned because Standard & Poor's did not surveil the ratings after the initial analysis. This has occurred three times. Two of these governments have since published their ratings, and their public ratings are included here. One of the three governments defaulted after its initial approach to Standard & Poor's and rejection of the rating. We do not include this default in the data. However, its inclusion would not, in our view, materially alter these rating transitions or default statistics.
We compile the data using static pools. A static pool is a portfolio of issuers defined on Jan. 1 of each year. The study tracks rating migrations and defaults for each static pool going forward from that point to the end of a specified time horizon, and the results for each static pool are averaged to calculate the average rating migrations for each rating category over specific time horizons. In other words, transitions measure rating changes at one period end versus a subsequent period end. In addition, the grouping of obligors in a specific static pool remains constant, and the pools are aggregated and weighted by the number of issuers by rating category. Thus, in our analysis, static pools from earlier years, when there were fewer sovereign ratings, count less in the study than do pools from more recent years, when there are many more sovereign ratings.
Universe dynamics
Chart 1 shows the growth in the number of sovereigns rated by Standard & Poor's.
Chart 1
As described in Appendix 2, Standard & Poor's predecessor institutions started rating sovereign governments in the 1920s. However, after the Great Depression, World War II, and the imposition of the Interest Equalization Tax (IET) in the U.S. in 1963, ratings lapsed on all sovereigns but the U.S. and Canada, both of which had 'AAA' ratings. The Yankee bond market again became attractive for foreign governments with the repeal of IET in 1974, and demand for sovereign ratings resumed from American investors (Standard & Poor's original base of demand). The number of rated sovereigns rose to 12 in 1980 from two at the time of IET's 1974 repeal. The figure almost tripled by 1990, to 31, with all but one sovereign rated investment grade.
From that point on, there was a marked increase in the number of ratings and an expansion into lower rating categories. The increase derived from the successful restructuring of Mexico's defaulted bank debt into Brady bonds, thus opening up speculative-grade sovereigns as an attractive asset class. Thus, in our view, the growth in speculative-grade issuers is not a function of downgrades but rather a result of less-creditworthy issuers seeking ratings to enhance their access to private capital (see Chart 2). Today, Standard & Poor's rates governments of virtually every major economy in the world.
Chart 2
Standard & Poor's had assigned only a few local-currency sovereign debt ratings before 1992; we aligned most of these with the foreign-currency debt rating on the sovereign. Today, all rated sovereigns have both foreign- and local-currency credit ratings. In 1992, Standard & Poor's revised its view of a sovereign's greater inherent flexibility in servicing its local-currency debt based on the unique powers a sovereign enjoys within its own borders—particularly in its control of domestic financial and monetary systems. (4) As a result, many sovereigns now have local-currency credit ratings that are higher than their foreign-currency credit ratings (see Chart 3).
Chart 3
Sovereign Foreign-Currency Rating Transition Rates
Turning to the data for rating changes, Table 1 shows the weighted average one-year transition matrix for foreign-currency ratings by rating category during the 34-year period beginning in 1975. The vertical axis shows the rating at the beginning of the year, the horizontal axis the rating at the end of the year. If ratings never changed, 100% would appear along the diagonal. In Table 1, 'AAA' ratings are the most stable; on average, 98.06% of these 'AAA' ratings began and ended the year at that level. Credit quality became slightly less stable at lower rating levels, with 94.61% of 'AA' ratings beginning and ending the year at that level, 93.22% of 'A', 88.06% of 'BBB', 87.04% of 'BB', 84.54% of 'B', and 38.89% of 'CCC'/'CC'.
Table 1
Sovereign Foreign-Currency Average One-Year Transition Rates (1975-2008)
Upgrades exceed downgrades in most rating categories. For example, in Table 1, one can see that on average, 7.96% of 'BBB' rated sovereigns were raised to 'A' and 3.99% were lowered to speculative-grade in calendar-year 2008. The exceptions were the 'AAA' category (Standard & Poor's highest level) and the 'CCC'/'CC' level. At 'CCC'/'CC', the lowest rating categories for solvent entities, we raised 22.22% of sovereign ratings on average to 'B', and 38.99% defaulted. The improvement of emerging market sovereign fundamentals, which persisted through 2007, partly explains the modest bias toward upgrades. (5)
Tables 2, 3, 4, and 5 record the weighted average rating transitions for longer periods for foreign-currency sovereign credit ratings by rating category since 1975. For inclusion in the matrix, the issuer must be in at least one static pool; in other words, we must have rated it by the start of the last period. In the case of the three-year matrix, we must have rated the sovereign by Jan. 1, 2006; in the case of the 10-year matrix, by Jan. 1, 1999. Thus, there are progressively fewer counts (or observations) the longer the transition period, given the growth of sovereign ratings and, of course, given that there will be one fewer static pool for each added year in the matrix.
Table 2
Sovereign Foreign-Currency Average Three-Year Transition Rates (1975-2008)
Tables 2, 3, 4, and 5 illustrate that many of the observations drawn from examining the one-year rating transition period also hold true for the longer periods. For example, higher ratings are usually more stable than lower ones. In addition, upgrades outnumber downgrades at the 'AA', 'BBB', and 'B' levels and are broadly balanced at the 'A' and 'BB' levels. These tables also demonstrate that large movements of sovereign ratings are uncommon, even over long periods. For example, in the 10-year rating transition in Table 5, only 1.88% of ratings in the 'AAA' category fell by more than one category over the three decades of the survey, 0.71% at the 'AA' category, and 1.35% at the 'A' category. The ratios of two or more category movements remain modest over the 10-year transition periods, even at the 'BBB' category (14.67% changed, 2.67% up and 12.00% down), the 'BB' category (32.94%, 12.94% up and 20.00% down), and the 'B' category (48.15%, 11.11% up and 37.04% down to default).
Standard & Poor's has also updated its weighted-average one-year and three-year rating matrices for foreign-currency sovereign credit ratings with rating modifiers (i.e., pluses and minuses). These data, contained in Tables 6 and 7, confirm an upward bias on rating actions.
Table 6
Sovereign Foreign-Currency Average One-Year Transition Rates With Rating Modifier (1975-2008)
Chart 4 shows the balance of foreign-currency rating actions since 1995. The chart shows rating actions (both across and within rating categories) between Jan. 1 and Dec. 31 for each year covered. We count multiple rating actions during a calendar year as a single action. In the last 14 years, upgrades have outnumbered downgrades 12 times. The two exceptions were 1999 and 2008. We expect downgrades to outnumber upgrades again in 2009: At year-end 2008, we had assigned negative outlooks to 26 foreign-currency sovereign ratings and positive outlooks to only two. (6)
Chart 4
Chart 5 presents the severity of rating transitions on a one-year horizon during 1975-2008. Although there are more upgrade transitions than downgrade transitions, downgrade transitions are more pronounced. We believe this is because sovereign creditworthiness usually improves gradually with time but that policy mistakes made in a difficult external environment can lead to a sharp deterioration in a sovereign's capacity to pay its debt.
Chart 5
Correlation Of Successive Rating Actions
Table 8 shows the correlation between two subsequent identical rating actions over a two-year horizon. Between 1975 and 2008, there were 168 foreign-currency sovereign credit rating upgrades. Of this set, 39% were raised again over the course of the next two years, 53% remained unchanged, and 7% were downgraded. Similarly, during these 34 years, we lowered foreign-currency sovereign credit ratings 141 times. Of this set, we lowered 57% again within two years, 36% remained the same, and we raised 7%. The correlation between two subsequent identical rating actions is a number between 1, which means a perfect positive correlation (every rating action is followed by the same rating action), and negative 1, which means a perfect negative correlation (every rating action is followed by the opposite rating action). A correlation of zero would mean perfectly random rating behavior. There is a small correlation within two years between an upgrade and a subsequent upgrade of 0.34, and there is a moderate correlation between a downgrade and a subsequent downgrade of 0.46.
We believe that the nature of credit explains the moderate correlation for downgrades: In our view, it results from the use of discrete measures in the form of credit ratings to model a continuous real-world variable of long-term credit quality. When a sovereign's creditworthiness deteriorates, Standard & Poor's could lower a rating more than once over two years to keep up with the relatively sharp movement in credit quality as rating committee analysts attempt to distinguish the long-term trend, which Standard & Poor's ratings seek to address, from the cycle, which ratings do not address. (7) Rating committees analyze data that national governments collect with a lag and frequently revise. In addition, policy execution often trails policy intent. Surprises are usually on the downside. On the other hand, when a sovereign's creditworthiness is increasing, we have observed that improvements are usually incremental and tended to be more easily anticipated. If one lengthened the rating horizon beyond two years, our data show that the correlation between sovereign ratings from one rating action to the next would rise, so that the correlation of successive upgrades would exceed that of successive downgrades.
Table 8
Sovereign Foreign-Currency Rating Action Correlation During A Two-Year Horizon (1975-2008)
(% unless otherwise indicated)
Total (number)
Followed by rating action in same direction
Stayed the same
Followed by a rating action in opposite direction
Upgrades*
168
39¶
53
7
Downgrades§
141
57**
36
7
*Excludes upgrades from SD. ¶Correlation between an upgrade and a subsequent upgrade: 0.34. §Includes downgrades to SD. **Correlation between a downgrade and a subsequent downgrade: 0.46.
Sovereign Local-Currency Rating Transition Rates
Tables 9, 10, 11, 12, and 13 contain weighted-average transition matrices for local-currency sovereign credit ratings over various time periods. There are fewer data points for local-currency sovereign ratings than for foreign-currency sovereign credit ratings, so the results are less robust in our view, particularly for the transition matrices for longer periods. Like foreign-currency sovereign credit ratings, local-currency ones are generally more stable at higher rating levels than at lower levels. However, the absolute level of stability for local-currency sovereign credit ratings is usually slightly less at most levels than for foreign-currency sovereign credit ratings.
Rating actions that crossed rating categories also appear more skewed toward downgrades in the 'A' through 'AAA' categories. We believe that this bias toward downgrades resulted, in part, from the convergence of foreign- and local-currency ratings for sovereigns adopting (or about to adopt) the euro. When ratings converged at the higher level, it resulted in an upgrade of the foreign-currency sovereign credit rating; when ratings more often converged at the lower level, it resulted in a downgrade of the local-currency sovereign credit rating. We've also observed that the bias toward downgrades also stems from changing criteria following two disruptions of local-currency public finance: Mexico's near default in 1995 on its tesobonos and Russia's 1998 default on its GKOs.
Table 9
Sovereign Local-Currency Average One-Year Transition Rates (1993-2008)
Tables 14 and Table 15 contain weighted-average one-year and three-year, respectively, transition matrices for local-currency sovereign credit ratings with rating modifiers. Our views reached regarding local-currency rating movements across rating categories hold for these tables, which show single-notch rating actions from one calendar year to the next year and to three years later, although upgrades and downgrades appear more balanced. A time series (see Chart 6) shows that we have raised more local-currency ratings than we lowered in seven of the past 14 years for the one-year rating horizon.
Table 14
Sovereign Local-Currency Average One-Year Transition Rates With Rating Modifier (1993-2008)
Table 16 displays the cumulative foreign-currency default rates for 1975-2008 by rating category, for both public and confidential sovereign ratings. Cumulative default rates average the experience of all static pools in a select period of time by calculating marginal weighted-average default rates conditional on survival (survivors being nondefaulters) for each possible time horizon and accumulating marginal default rates. We calculate conditional-on-survival default rates by dividing the number of issuers in a static pool that default within a specific time horizon by the number of issuers that had survived (had not defaulted) up to that point in time. In the context of sovereign ratings, we treat governments that selectively default as a complete default. When the sovereign emerges from selective default, we then treat it as a new entity altogether. We based weights on the number of issuers in each static pool.
The cumulative default rate is 1 minus the proportion of survivors (nondefaulters). At some point, the cumulative default rate will not increase as the horizon is expanded because the lengthened horizon captures no additional defaults. The cumulative default rate (conditional on survival) differs from the transition-to-default rate (seen in the right-hand column of the transition matrices) in that the former counts defaults only once, while transition rates count defaults once per observation horizon and averages them. In addition, cumulative default rates capture defaults that have occurred before a cohort period ends. (8)
Table 16
Sovereign Foreign-Currency Cumulative Average Default Rate Without Rating Modifiers (1975-2008)
The foreign-currency sovereign defaults included in the data occurred in Argentina, Belize, the Dominican Republic, Ecuador, Grenada, Indonesia, Pakistan, Paraguay, Russia, Seychelles, Uruguay, and Venezuela (see Table 17).
Table 17
Sovereign Foreign-Currency Selective Defaults
Country
Selective default date
Emergence date
Time in selective default
Rating one year before selective default
Rating at emergence from selective default
Russia
Jan. 27, 1999
Dec. 8, 2000
22 months
BB-
B-
Pakistan
Jan. 29, 1999
Dec. 21, 1999
11 months
B+
B-
Indonesia
March 30, 1999
March 31, 1999
One day
B-
CCC+
Indonesia
April 17, 2000
Oct. 2, 2000
Six months
CCC+
B-
Argentina
Nov. 6, 2001
June 1, 2005
54 months
BB
B-
Indonesia
April 23, 2002
Sept. 5, 2002
Four months
B-
CCC+
Paraguay
Feb. 13, 2003
July 26, 2004
18 months
B
B-
Uruguay
May 16, 2003
June 2, 2003
One month
BB-
B-
Grenada
Dec. 30, 2004
Nov. 18, 2005
11 months
BB-
B-
Venezuela
Jan. 18, 2005
March 3, 2005
One month
B-
B
Dominican Republic
Feb. 1, 2005
June 29, 2005
Five months
CCC
B
Belize
Dec. 7, 2006
Feb. 20, 2007
Three months
CCC-
B
Seychelles
Aug. 7, 2008
-
B
-
Ecuador
Dec. 15, 2008
-
B-
-
Sources: Standard & Poor's Sovereign Ratings.
As a result, given the small universe, conclusions must only be tentative for both sets of data. For foreign-currency sovereign defaults, no sovereign rated in the 'A' category or higher on Jan. 1 has defaulted within a decade (see Table 16). Cumulative default rates increase progressively—although not linearly—as ratings decline. Thus, for 1975-2008, in the three-year cumulative default column, 1.68% of sovereigns rated in the 'BBB' category on Jan. 1, on average annually, defaulted within three years after Jan. 1. (The 1.68% pertains to Indonesia and Uruguay.) The figure rises to 4.09% for the 'BB' category, 7.68% for the 'B' category, and 61.11% for the 'CCC'/'CC' categories. Ten-year cumulative default rates for sovereigns averaged 1.17% for investment-grade sovereigns and 23.53% for speculative-grade ones.
Tables 18 through 21 show the components of the foreign-currency default rates broken down by rating category and by year. As there have been no defaults within 10 years in the 'AAA', 'AA', and 'A' categories, the tables only show the lower categories. The data fields contain the transition-to-default data, which are also weight-averaged in the 'SD' column of Tables 1-5. The cumulative default rate (conditioned on survival) is presented at the bottom. We observe that these component data highlight the small size of the universe: The numerator in each yearly calculation comprises only one or two defaults. We also note that the default rate in a single year can vary a great deal. Thus, in Table 18, the five-year cumulative default rate for 'BBB' is 14.29% in 1995, 25.00% in 1996, 11.11% in 1997, 0.00% in 1998, and 6.25% in 1999. The first three calculations represent the default of Indonesia, and the last is Uruguay's default.
Table 18
Foreign-Currency 'BBB' Static Pool Default Rates By Rating (1989 - 2008)
Transition-to-default rates (%)*
—Time horizon (years)—
Date¶
Number of issuers
1
2
3
4
5
6
7
8
9
10
1989
2
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1990
2
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1991
3
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1992
2
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1993
6
0.00
0.00
0.00
0.00
0.00
0.00
16.67
16.67
16.67
16.67
1994
8
0.00
0.00
0.00
0.00
0.00
12.50
12.50
12.50
12.50
12.50
1995
7
0.00
0.00
0.00
0.00
14.29
14.29
14.29
14.29
14.29
14.29
1996
4
0.00
0.00
0.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
1997
9
0.00
0.00
11.11
11.11
11.11
11.11
11.11
11.11
11.11
11.11
1998
16
0.00
0.00
0.00
0.00
0.00
6.25
6.25
6.25
6.25
6.25
1999
16
0.00
0.00
0.00
0.00
6.25
6.25
6.25
6.25
6.25
6.25
2000
16
0.00
0.00
0.00
6.25
6.25
6.25
6.25
6.25
6.25
2001
16
0.00
0.00
6.25
6.25
6.25
6.25
6.25
6.25
2002
16
0.00
6.25
6.25
6.25
6.25
6.25
6.25
2003
13
0.00
0.00
0.00
0.00
0.00
0.00
2004
12
0.00
0.00
0.00
0.00
0.00
2005
13
0.00
0.00
0.00
0.00
2006
13
0.00
0.00
0.00
2007
13
0.00
0.00
2008
14
0.00
Marginal default rate (%)§
0.00
0.53
1.16
1.27
1.40
1.55
0.88
0.00
0.00
0.00
Cumulative default rate (%)**
0.00
0.53
1.68
2.93
4.29
5.77
6.60
6.60
6.60
6.60
*The weighted average of these transition-to-default rates appears in the SD column of the transition matrices. ¶First year in the cohort beginning Jan 1. No observations between 1975 and 1988. §Using conditional-on-survival methodology, where defaults of the previous horizon are subtracted out of the numerator and denominator of the default rate for subsequent horizons. **Calculated by multiplying nondefault marginal rates, which are then subtracted from 1 to get the cumulative default rate.
Table 19
Foreign-Currency 'BB' Static Pool Default Rates By Rating (1984 - 2008)
Transition-to-default rates (%)*
—Time horizon (years)—
Date¶
Number of issuers
1
2
3
4
5
6
7
8
9
10
1984
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1985
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1986
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1987
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1988
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1989
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1990
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1991
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1992
2
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1993
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1994
6
0.00
0.00
0.00
0.00
0.00
0.00
0.00
16.67
16.67
16.67
1995
8
0.00
0.00
0.00
0.00
0.00
0.00
12.50
12.50
25.00
25.00
1996
10
0.00
0.00
0.00
0.00
0.00
10.00
10.00
30.00
30.00
30.00
1997
13
0.00
0.00
7.69
7.69
15.38
15.38
30.77
30.77
30.77
30.77
1998
18
0.00
11.11
11.11
16.67
16.67
22.22
22.22
22.22
22.22
22.22
1999
18
0.00
0.00
5.56
5.56
11.11
11.11
11.11
11.11
11.11
11.11
2000
15
0.00
6.67
6.67
6.67
6.67
6.67
6.67
6.67
6.67
2001
14
7.14
7.14
7.14
7.14
7.14
14.29
14.29
14.29
2002
16
0.00
0.00
0.00
6.25
12.50
12.50
12.50
2003
17
0.00
5.88
11.76
11.76
11.76
11.76
2004
17
5.88
5.88
5.88
5.88
5.88
2005
19
0.00
0.00
0.00
0.00
2006
19
0.00
0.00
0.00
2007
22
0.00
0.00
2008
23
0.00
Marginal default rate (%)§
0.81
1.80
1.53
1.15
1.96
2.24
2.59
3.03
1.19
0.00
Cumulative default rate (%)**
0.81
2.60
4.09
5.19
7.05
9.13
11.48
14.16
15.18
15.18
N/A—Not applicable; no observations. *The weighted average of these transition-to-default rates appears in the SD column of the transition matrices. ¶First year in the cohort beginning Jan 1. No observations between 1975 and 1983. §Using conditional-on-survival methodology, where defaults of the previous horizon are subtracted out of the numerator and denominator of the default rate for subsequent horizons. **Calculated by multiplying nondefault marginal rates, which are then subtracted from 1 to get the cumulative default rate.
Table 20
Foreign-Currency 'B' Static Pool Default Rates By Rating (1990 - 2008)
Transition-to-default rates (%)*
—Time horizon (years)—
Date¶
Number of issuers
1
2
3
4
5
6
7
8
9
10
1990
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1991
1
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1992
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1993
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1994
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1995
4
0.00
0.00
0.00
0.00
25.00
25.00
25.00
25.00
25.00
25.00
1996
5
0.00
0.00
0.00
20.00
20.00
20.00
20.00
20.00
20.00
40.00
1997
4
0.00
0.00
25.00
25.00
25.00
25.00
25.00
25.00
50.00
50.00
1998
5
0.00
20.00
20.00
20.00
20.00
20.00
20.00
60.00
60.00
60.00
1999
7
0.00
0.00
0.00
0.00
0.00
0.00
28.57
28.57
28.57
28.57
2000
14
0.00
0.00
0.00
7.14
7.14
21.43
21.43
21.43
21.43
2001
19
0.00
5.26
10.53
10.53
21.05
21.05
21.05
26.32
2002
15
0.00
6.67
6.67
13.33
13.33
13.33
13.33
2003
16
12.50
12.50
12.50
18.75
18.75
18.75
2004
19
0.00
5.26
10.53
10.53
10.53
2005
21
4.76
9.52
9.52
9.52
2006
23
0.00
0.00
0.00
2007
25
0.00
4.00
2008
28
7.14
Marginal default rate (%)§
2.42
3.41
2.05
3.33
3.09
2.60
3.23
6.38
3.33
5.56
Cumulative default rate (%)**
2.42
5.74
7.68
10.76
13.52
15.76
18.48
23.68
26.23
30.33
N/A—Not applicable; no observations. *The weighted average of these transition to default rates appears in the SD column of the transition matrices.¶First year in the cohort beginning Jan 1. No observations between 1975 and 1989. §Using conditional-on-survival methodology, where defaults of the previous horizon are subtracted out of the numerator and denominator of the default rate for subsequent horizons. **Calculated by multiplying nondefault marginal rates, which are then subtracted from 1 to get the cumulative default rate.
Table 21
Foreign-Currency 'CCC'/'CC' Static Pool Default Rates By Rating (1999 - 2008)
Transition-to-default rates (%)*
—Time horizon (years)—
Date¶
Number of issuers
1
2
3
4
5
6
7
8
9
10
1999
3
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
2000
1
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
2001
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
2002
2
50.00
50.00
50.00
50.00
50.00
50.00
100.00
2003
3
0.00
0.00
33.33
33.33
33.33
66.67
2004
2
0.00
50.00
50.00
50.00
100.00
2005
3
33.33
33.33
33.33
66.67
2006
3
33.33
33.33
66.67
2007
1
0.00
100.00
2008
0
N/A
Marginal default rate (%)§
38.89
18.18
22.22
16.67
25.00
33.33
100.00
N/A
N/A
N/A
Cumulative default rate (%)**
38.89
50.00
61.11
67.59
75.69
83.80
100.00
N/A
N/A
N/A
N/A—Not applicable; no observations. *The weighted average of these transition to default rates appears in the SD column of the transition matrices.¹First year in the cohort beginning Jan 1. No observations between 1975 and 1998. §Using conditional-on-survival methodology, where defaults of the previous horizon are subtracted out of the numerator and denominator of the default rate for subsequent horizons. **Calculated by multiplying nondefault marginal rates, which are then subtracted from 1 to get the cumulative default rate.
Looking at the 34 years of data disaggregated by rating modifier (see Table 22), we believe that most of the same observations made regarding defaults by rating category are valid. In the column for one-year cumulative default rates, no sovereign rated 'BB' or higher on Jan. 1 defaulted on its foreign-currency debt within the calendar year. Moving to the next column, no sovereign rated 'BBB' or higher defaulted within two calendar years, and no sovereign rated 'BBB+' or higher defaulted within a decade. By and large, we observe that cumulative default rates for ratings with modifiers have tended to increase progressively as ratings decline (with exceptions at the 'BBB-' and 'BB-' levels and a few other minor exceptions). There is a significant increase in default rates between 'BB' and lower ratings.
Table 22
Sovereign Foreign-Currency Cumulative Average Default Rate With Rating Modifiers (1975-2008)
