Taiwan Ratings' Corporate Default & Rating Transition Study 2004

2005/06/16

Susan Chu, Serene Hsieh

Taiwan Ratings Corp. is pleased to release its inaugural default and rating transition study, which examines its ratings track record since it first assigned credit ratings in 1998. The study shows that the performance of its ratings has generally been in line with global norms, as evidenced by the observations of Standard & Poor's Ratings Services. The study primarily measures rating movements over time and provides a quantitative measure of ratings performance, implying Taiwan Ratings' historical ability to predict the likelihood of default.

The study covers solicited issuer ratings, totally 183 entities, rated by Taiwan Ratings between Jan. 1, 1998 and Dec. 31, 2004. The study also covers confidential ratings and withdrawn ratings.

Taiwan Ratings' pool of 183 credit ratings is statistically small and covers a short period, implying that the data may not be well seasoned. The pool is small compared with Standard & Poor's Asia ex-Japan study of 439 issuers ("Asia ex-Japan Default and Rating Transition Study" published by Standard & Poor's on Apr. 4, 2005)) and Standard & Poor's global study of more than 10,000 issuers ("Annual Global Corporate Default Study" published by Standard & Poor's on Jan. 26, 2005). Furthermore, Taiwan Ratings' pool only covers ratings history data for seven years from year-end 1997 to year-end 2004 (see the methodology section for details of the issuers included in this study), compared with 13 years (from year-end 1991 to year-end 2004) for Standard & Poor's Asia ex-Japan study and  24 years (from year-end 1980 to year-end 2004) for Standard & Poor's global study. While the time to default averages about five years by Standard & Poor's global norms, about half of Taiwan Ratings' issuer ratings were not assigned until 2001.

Key Findings

  •                One default was recorded in 2001.
  •                Improving economic and market factors have helped strengthen rated issuers' credit profiles since 2002.
  •                Taiwan Ratings' study generally mirrors the observations of Standard & Poor's global transition studies. Higher rated issuers tend to exhibit less rating volatility than their lower rated counterparts.
  •                The volatility of Taiwan Ratings' ratings is, however, higher than Standard & Poor's global observations, especially in the lower rating categories. This can be largely attributed to Taiwan Ratings' small sample and short track record. Nevertheless, Taiwan Ratings' average downgrade-to-upgrade ratio appears comparable with Standard & Poor's observations.

Only one default

There is just one default record in Taiwan Rating's rated pool (see table 1), but this would not count after adjusting for issuers not rated [N.R.] after rating withdrawn. This issuer was rated by Taiwan Ratings in 1999, it withdrew its ratings in 2000, and it defaulted in 2001. The result partly reflects the global norm by Standard & Poor's that lower ratings tend to have a higher likelihood of default. However, as this is the only default in Taiwan Ratings' pool, the default ratios by Taiwan Ratings could further develop.

Table 1 Taiwan Cumulative Average Default Rates, 1999-2004 (%)

ˇXTime HorizonˇX

Rating category

Year 1

Year 2

Year 3

Year 4

Year 5

Year 6

twAAA

0.00

0.00

0.00

0.00

0.00

0.00

twAA

0.00

0.00

0.00

0.00

0.00

0.00

twA

0.00

0.00

0.00

0.00

0.00

0.00

twBBB

0.00

0.00

0.00

0.00

0.00

0.00

twBB

0.00

0.00

0.00

0.00

0.00

0.00

twB

0.00

5.26

5.26

5.26

5.26

N/A

twCCC/twCC

0.00

0.00

0.00

0.00

N/A

N/A

All rated

0.00

0.24

0.24

0.24

0.24

0.24

N/A--not applicable (no issuer ratings in that category at that time).Source: Standard & Poor's CreditPro ® 6.6. and Taiwan Ratings' database


One possible reason for Taiwan Ratings' limited default record is that about 70% of the pool consists of financial institutions (see table 2), which have received direct and indirect support from the Taiwan government. Also, the first group of manufacturing companies that were assigned ratings by Taiwan Ratings generally had strong financial profiles. Taiwan Ratings only expanded most of its ratings coverage of manufacturers since 2001 and so the pool has not had time to season. The proportion of higher rated companies has been on the rise since 2002 in line with improving economic and market factors (see table 3).

Table 2 Taiwan Rating Classification of New Issuers

-- Sector --

Year

Financial Institutions

Industrials & Utilities

Total

1998

31

2

33

1999

22

6

28

2000

24

5

29

2001

11

15

26

2002

27

9

36

2003

8

9

17

2004

6

8

14

Total

129

54

183

% of total

70.5

29.5

100.0

Source: Standard & Poor's CreditPro® 6.6. and Taiwan Ratings' database


Table 3 Taiwan Issuer Credit Ratings at Year End, 1998-2004 (%)

Rating category

1998

1999

2000

2001

2002

2003

2004

Weighted average

twAAA

6.1

5.0

4.7

4.7

4.5

4.4

5.9

4.9

twAA

6.1

11.7

10.6

11.2

17.3

22.8

27.4

17.6

twA

21.2

18.3

17.6

19.6

24.1

20.6

39.3

24.2

twBBB

42.4

35.0

28.2

29.0

25.6

25.0

15.6

26.0

twBB

24.2

23.3

31.8

30.8

21.1

21.3

8.9

21.9

twB

0.0

6.7

5.9

3.7

4.5

2.9

3.0

3.9

twCCC/twCC

0.0

0.0

1.2

0.9

3.0

2.9

0.0

1.5

Total

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

Source: Standard & Poor's CreditPro® 6.6.


Rating appropriateness is best determined after a few business cycles and on a sizeable pool of issuers. In light of Taiwan's increasing deregulation and market liberalization, Taiwan Ratings expects to see the development of more mature financial and capital markets. The transition to greater market discipline may possibly result in defaults by fundamentally weaker companies. Defaults are part of the process of market efficiency as demonstrated in the developed economies of North America and Europe.

Taiwan Ratings and Standard & Poor's performance correlation

Taiwan Ratings' ratings system is primarily based on the methodology and criteria used by Standard & Poor's, with minor adjustments made to better reflect domestic conditions. Taiwan Ratings uses a different rating scale from Standard & Poor's.

Taiwan Ratings' rating transition study (see table 4) supports Standard & Poor's global observations (see table 5 and table 6) that higher credit ratings tend to be more stable than lower ratings. For example, there is a 95% (after adjusting for N.R.) possibility for ratings in the category of 'twAA' to remain unchanged over a one year period (see the methodology section for the definition of transition analysis).

Taiwan Ratings' higher rated categories appear extremely stable, but Taiwan Ratings' scale registered faster rating movements in the lower rating categories compared with Standard & Poor's scale. This is largely a result of the smaller number of issuers and shorter study period in Taiwan Ratings' study compared with Standard & Poor's Asia ex Japan and global transition studies. Statistically, a smaller number of observations and a shorter time period may potentially contribute to wider deviations of estimation from the findings of the larger number of observations. The higher volatility in Taiwan Ratings' lower rating categories also reflects the characteristics of a domestic rating, which have finer gradations and thus make Taiwan Ratings' ratings more sensitive to changes in the credit profiles of rated issuers.

Table 4 Taiwan Average One-Year N.R.-Adjusted Transition Rates, 1999-2004 (%)

--- Rating category at year-end (%) --

Rating category at start of year

twAAA

twAA

twA

twBBB

twBB

twB

twCCC/twCC

twAAA

96.2

3.9

0.0

0.0

0.0

0.0

0.0

twAA

2.4

95.1

2.4

0.0

0.0

0.0

0.0

twA

0.0

13.8

83.5

2.8

0.0

0.0

0.0

twBBB

0.0

0.7

19.2

77.4

2.7

0.0

0.0

twBB

0.0

0.0

1.7

21.4

75.2

0.9

0.9

twB

0.0

0.0

0.0

0.0

17.7

64.7

17.7

twCCC/twCC

0.0

0.0

0.0

0.0

40.0

0.0

60.0

N.R.-issuer not rated after rating withdrawn.
Source: Standard & Poor's CreditPro® 6.6. and Taiwan Ratings' database


Table 5 Asia ex-Japan Average One-Year N.R.-Adjusted Transition Rates, 1992-2004 (%)

--- Rating category at year-end (%) --

Rating category at start of year

AAA

AA

A

BBB

BB

B

CCC/CC

D

AAA

87.5

12.5

0.0

0.0

0.0

0.0

0.0

0.0

AA

2.4

80.5

14.6

0.0

0.0

2.4

0.0

0.0

A

0.0

2.2

88.0

7.7

0.9

1.3

0.0

0.0

BBB

0.0

0.0

3.3

83.9

9.9

2.4

0.3

0.3

BB

0.0

0.0

0.0

5.1

89.3

2.8

2.1

0.7

B

0.0

0.0

0.4

0.0

10.7

81.9

4.1

2.9

CCC/CC

0.0

0.0

0.0

0.0

0.0

10.3

66.7

23.1

N.R.-issuer not rated after rating withdrawn.
Source: Standard & Poor's CreditPro® 6.6.


Table 6 Global Average One-Year N.R.-Adjusted Transition Rates, 1981-2004 (%)

--- Rating category at year-end (%) --

Rating category at start of year

AAA

AA

A

BBB

BB

B

CCC/CC

D

AAA

91.7

7.7

0.5

0.1

0.1

0.0

0.0

0.0

AA

0.6

90.5

8.1

0.60

0.1

0.1

0.0

0.0

A

0.1

2.2

91.3

5.8

0.4

0.2

0.0

0.0

BBB

0.0

0.2

4.1

89.7

4.7

0.8

0.2

0.3

BB

0.0

0.1

0.4

5.8

83.4

8.1

1.0

1.3

B

0.0

0.1

0.2

0.3

5.8

82.5

4.8

6.2

CCC/CC

0.1

0.0

0.4

0.5

1.5

11.2

54.8

32.4

N.R.-Entity not rated after rating withdrawn.
Source: Standard & Poor's CreditPro® 6.6.


It should be noted that Taiwan Ratings conducted large-scale rating adjustments after the completion of a major portfolio review in December 2004. In particular, significant rating adjustments were made to financial institutions, as their risk profiles had improved beyond Taiwan Ratings' original expectations, mainly as a result of accelerating integration within financial holding company groups, improvements in risk management, and continuing indirect system support. As a result, unlike global observations where transition rates generally move toward lower categories at lower ratings, the overall transition rates for lower ratings (or even most categories) in Taiwan Ratings' pool tended to move up rather than down.

Higher rating volatility in Taiwan Ratings' pool is reiterated in the frequency of annual rating changes, which ranged between 10% and 70% in 1999-2004 and averaged 40% in the period (see table 7). This contrast with the annual rating changes in Standard & Poor's global pool, which ranged between 20% and 30% in 1981-2004 and averaged 28% in the period (see table 9). Standard & Poor's Asia ex Japan study also highlighted higher rating volatility, which was offset to some extent by its larger pool and longer length of observation (see table 8). The high ratio of annual rating changes in Taiwan Ratings' pool also reflects the large-scale rating adjustments conducted in December 2004. Nevertheless, the downgrade-to-upgrade ratio for Taiwan Ratings' pool averaged 1.4x for 1999-2004, which is comparable to Standard & Poor's regional and global observations.

Table 7 Taiwan Summary of Annual Rating Changes*, 1999-2004 (%)

Year

Issuers as at Jan. 1

Upgrades (%)

Downgrades (%)

Defaults (%)

Withdrawn ratings (%)

Changed ratings (%)

Unchanged ratings (%)

Downgrade/ upgrade ratio

1999

33

6.1

0.0

0.0

3.0

9.1

90.9

0.0

2000

60

3.3

11.6

0.0

6.7

21.7

78.3

3.5

2001

85

2.4

11.8

0.0

4.7

18.8

81.2

5.0

2002

107

18.7

17.8

0.0

9.4

45.8

54.2

1.0

2003

133

16.5

3.8

0.0

10.5

30.8

69.2

0.2

2004

136

58.8

2.9

0.0

10.3

72.1

27.9

0.1

Weighted average (1999-2003)

11.5

9.8

0.0

7.9

29.2

70.8

1.8

Weighted average (1999-2004)

23.1

8.1

0.0

8.5

39.7

60.3

1.4

* This table compares the net change in ratings from the first to the last day of each year. All intermediate ratings are disregarded.
Source: Standard & Poor's CreditPro® 6.6. and Taiwan Ratings' database


Table 8 Asia ex Japan Summary of Annual Rating Changes*, 1992-2004 (%)

Year

Issuers as of Jan. 1

Upgrades (%)

Downgrades (%)**

Defaults (%)

Withdrawn ratings (%)

Changed ratings (%)

Unchanged ratings (%)

Downgrade/ upgrade ratio

1992

1

0.0

0.0

0.0

0.0

0.0

100.0

0.0

1993

6

0.0

0.0

0.0

0.0

0.0

100.0

0.0

1994

14

7.1

0.0

0.0

0.0

7.1

92.9

0.0

1995

33

12.1

3.0

0.0

3.0

18.2

81.8

0.3

1996

42

9.5

2.4

0.0

7.1

19.1

81.0

0.3

1997

71

4.2

31.0

1.4

4.2

40.9

59.2

7.3

1998

125

4.0

45.6

7.2

11.2

68.0

32.0

11.4

1999

126

3.2

12.7

4.8

15.1

35.7

64.3

4.0

2000

156

8.3

3.9

1.3

5.1

18.6

81.4

0.5

2001

165

8.5

4.9

2.4

6.1

21.8

78.2

0.6

2002

175

18.3

13.7

0.6

6.9

39.4

60.6

0.8

2003

246

9.8

5.3

0.0

7.7

22.8

77.2

0.5

2004

275

18.2

2.9

0.0

4.7

25.8

74.2

0.2

Weighted average

9.0

12.8

2.0

7.7

31.4

68.6

2.5

*This table compares the net change in ratings from the first to the last day of each year. All intermediate ratings are disregarded.
** excludes downgrades to 'D', which are shown separately in the default column.

Source: Standard & Poor's Asia ex-Japan Default and Rating Transition Study 2004


Table 9 Global Summary of Annual Rating Changes*, 1981-2004 (%)

Year

Issuers as at Jan. 1

Upgrades (%)

Downgrades (%)**

Defaults (%)

Withdrawn ratings (%)

Changed ratings (%)

Unchanged ratings (%)

Downgrade/ upgrade ratio

1981

1,394

9.97

12.98

0.14

2.01

25.11

74.89

1.30

1982

1,442

5.96

12.76

1.18

5.20

25.10

74.90

2.14

1983

1,475

7.12

11.66

0.81

5.69

25.29

74.71

1.64

1984

1,578

10.96

10.08

0.89

4.31

26.24

73.76

0.92

1985

1,650

7.82

13.64

1.09

4.30

26.85

73.15

1.74

1986

1,880

7.23

15.90

1.70

6.86

31.70

68.30

2.20

1987

2,037

7.07

11.78

0.93

9.43

29.21

70.79

1.67

1988

2,125

9.04

11.86

1.41

7.91

30.21

69.79

1.31

1989

2,170

9.54

10.97

1.52

7.88

29.91

70.09

1.15

1990

2,161

6.25

15.32

2.59

6.39

30.54

69.46

2.45

1991

2,109

6.07

14.46

3.13

3.46

27.12

72.88

2.38

1992

2,210

9.50

11.36

1.36

4.03

26.24

73.76

1.20

1993

2,418

8.40

9.10

0.54

8.23

26.26

73.74

1.08

1994

2,683

6.93

9.21

0.60

4.73

21.47

78.53

1.33

1995

3,070

9.06

9.25

0.94

4.14

23.39

76.61

1.02

1996

3,347

9.32

7.11

0.48

7.35

24.26

75.74

0.76

1997

3,672

8.69

7.60

0.60

8.80

25.68

74.32

0.87

1998

4,178

7.18

11.37

1.27

7.97

27.79

72.21

1.58

1999

4,599

5.63

10.72

2.13

9.02

27.51

72.49

1.90

2000

4,754

6.63

11.65

2.27

7.05

27.60

72.40

1.76

2001

4,851

5.65

15.07

3.69

7.19

31.60

68.40

2.67

2002

4,888

5.18

18.25

3.52

6.59

33.53

66.47

3.53

2003

4,951

6.48

13.84

1.84

7.37

29.53

70.47

2.13

2004

5,189

8.58

7.03

0.69

6.98

23.28

76.72

0.82

Weighted average

7.41

11.72

1.64

6.76

27.53

72.47

1.71

*This table compares the net change in ratings from the first to the last day of each year. All intermediate ratings are disregarded. **;Excludes downgrades to 'D', shown separately in the default column.

Source: Standard & Poor's Asia ex-Japan Default and Rating Transition Study 2004

Rating transition implications for defaults

As Taiwan Ratings continues to develop its rating performance evaluation, including rated pool size and rating history, the default and rating transition of Taiwan Ratings' scale is likely to move more towards the Standard & Poor's norm. However, more economic or business cycle tests may occur before rating performance of Taiwan Ratings' scale move towards the international experience.

Based on Standard & Poor's historic observations, cumulative default rates may be projected into the future, based on the assumption that the rating transition rates are stable. Based on Standard & Poor's observations, the one-year default rate column of 'cumulative average default rates table (see table 11)' is equivalent to the D (default) column of 'one-year transition rate' (see table 10), while the three-year default rate column of 'cumulated average default rates table' is about the level of the D (default) column of 'three-year transition rate'. The slight difference in results between the two tables mainly stems from the different static pools used to calculate transition to default and cumulative average default rates. Cumulative average default rates are the summary of all static pools, while the number of pools used in the average transition rate is limited by the transition's time horizon.

Table 10 Global Average One-Year & Three-Year Transition Matrices, 1981-2004 (%)

One-Year Transition Rates

From/To

AAA

AA

A

BBB

BB

B

CCC/CC

D

N.R.

AAA

87.44

7.37

0.46

0.09

0.06

0.00

0.00

0.00

4.59

AA

0.60

86.65

7.78

0.58

0.06

0.11

0.02

0.01

4.21

A

0.05

2.05

86.96

5.50

0.43

0.16

0.03

0.04

4.79

BBB

0.02

0.21

3.85

84.13

4.39

0.77

0.19

0.29

6.14

BB

0.04

0.08

0.33

5.27

75.73

7.36

0.94

1.20

9.06

B

0.00

0.07

0.20

0.28

5.21

72.95

4.23

5.71

11.36

CCC/CC

0.08

0.00

0.31

0.39

1.31

9.74

46.83

28.83

12.52

Three-Year Transition Rates

From/To

AAA

AA

A

BBB

BB

B

CCC/CC

D

N.R.

AAA

68.00

17.31

2.43

0.40

0.09

0.03

0.00

0.03

11.70

AA

1.43

64.99

18.69

2.30

0.41

0.28

0.02

0.09

11.79

A

0.12

4.91

66.08

12.15

1.64

0.68

0.11

0.25

14.06

BBB

0.08

0.60

9.22

59.62

8.48

2.52

0.57

1.49

17.42

BB

0.03

0.22

1.05

11.81

42.89

11.72

1.72

6.73

23.84

B

0.01

0.09

0.51

1.23

10.14

37.18

4.34

18.96

27.55

CCC/CC

0.10

0.00

0.50

1.21

2.52

12.98

13.58

44.77

24.35

Sources: Standard & Poor's Global Fixed Income Research; Annual Global Corporate Default Study: Corporate Defaults Poised to Rise in 2005; Standard & Poor's CreditPro® 7.0.


Table 11 Global Cumulative Average Default Rates, 1981-2004 (%)

ˇXTime HorizonˇX

Rating category

Year 1

Year 2

Year 3

Year 4

Year 5

Year 6

Year 7

AAA

0.00

0.00

0.03

0.06

0.10

0.17

0.24

AA

0.01

0.04

0.09

0.19

0.30

0.41

0.54

A

0.04

0.13

0.24

0.40

0.61

0.84

1.11

BBB

0.29

0.81

1.40

2.19

2.99

3.73

4.34

BB

1.20

3.58

6.39

8.97

11.25

13.47

15.25

B

5.71

12.49

18.09

22.37

25.40

27.77

29.76

CCC/CC

28.83

37.97

43.52

47.44

50.85

52.13

53.39

Investment grade

0.11

0.31

0.55

0.86

1.20

1.53

1.84

Noninvestment grade

4.91

9.76

14.05

17.52

20.22

22.46

24.32

All rated

1.64

3.29

4.78

6.04

7.08

7.97

8.71

Source: Standard & Poor's Global Fixed Income Research; "Annual Global Corporate Default Study: Corporate Defaults Poised to Rise in 2005"; Standard & Poor's CreditPro ® 7.0.


There is a major difference in the implicit default risk between Standard & Poor's global scale and Taiwan Ratings' scale, which is primarily positioned as the national scale and does not reflect a degree of Taiwan's sovereign risk (rated AA-/Negative/A-1+ by Standard & Poor's Rating Services).

Ratings performance survey methodology

This study is based on Taiwan Ratings' database of long-term issuer credit ratings. Taiwan Ratings' credit rating is a current opinion of an obligor's creditworthiness (that is, issuer, corporate, or counterparty credit rating), or the overall capacity to meet specific financial obligations (that is, issue credit ratings), relative to other obligors active in Taiwan's markets. Taiwan Ratings' credit rating symbols are accompanied by a 'tw' prefix to denote Taiwan and the rating scale focuses on Taiwan's financial markets. Taiwan Ratings' rating scale does not address sovereign risk and is not directly comparable to Standard & Poor's global scale.

Issuers included in this study

The study analyzes the rating histories of 183 long-term credit ratings from Dec. 31, 1997 to Dec. 31, 2004. These obligors include Taiwan industrials, utilities, insurance companies, financial holding companies, banks, securities firms, and other financial institutions. The global data presented in this report refer the Standard & Poor's ratings histories of all 11,150 long-term rated issuers rated by Standard & Poor's from Dec. 31, 1980 to Dec. 31, 2004. And the Asia ex-Japan data refers to Standard & Poor's ratings histories of 439 long-term rated issuers rated by Standard & Poor's from Dec. 31, 1991 to Dec. 31, 2004.

Static pool methodology

Taiwan Ratings conducted its default and transition study on the basis of groupings called static pools. Static pools are formed by grouping issuers by rating category at the beginning of each year covered by the study. Each static pool is followed from that point forward. All companies included in the study are assigned to one or more static pools.

Cumulative average default rate calculation

A default is recorded upon the first occurrence of a payment default on any financial obligation, rated or unrated. Annual default rates were calculated for each static pool. Cumulative default rates that average the experience of all static pools were estimated. Weights are based on the number of issuers in each static pool.

Transition analysis

Transition rates compare issuer ratings at the beginning of a time period with ratings at the end of the period. To compute one-year rating transition rates by rating category, the rating on each entity at the end of a particular year is compared with the rating at the beginning of the same year. An issuer that is rated for more than one year is counted as many times as the number of years it is rated. For instance, an issuer continually rated from mid-1998 to mid- 2003 would appear in the four consecutive one-year transition matrices from 1999 to 2002. All 1999 static pool members still rated on Dec. 31, 2004, had six one-year transitions, while companies first rated between Jan. 1, 2004, and Dec. 31, 2004 had only one.

Each one-year transition matrix displays all the rating movements between categories from the beginning of the year through to the year end. For each rating listed in the matrix's left-most column, there are seven ratios listed in the rows, corresponding to the ratings from 'twAAA' to 'twCCC/twCC'. For instance, table 4 shows that of all 'twA' rated companies over the 1999-2004 time period, 83.5% were rated the same at year-end, while 13.8% had been upgraded to 'twAA', and 2.8% had been downgraded to 'twBBB', and so on.

Practical application of transition rates

Rating transition rates are useful to investors and credit professionals for whom rating stability is important. For instance, investors restricted by regulations, or inclined to hold top-grade bonds, may want to assess the likelihood that Taiwan Ratings analysts continue to assign top ratings to their investments. Conversely, investors buying high-yield bonds in the hope of profiting from a rating upgrade would be able to gauge that expectation realistically.

The credit community might also use rating transition information to help determine maturity exposure limits or to measure credit risk in the context of value-at-risk models. Cumulative default rates can be projected for any number of years into the future, based on the assumption that the rating transition rates are stable and follow a first-order Markov process. Rating transition matrices can also be constructed to produce stressed default rates. Such matrices are often used in the area of credit risk measurement.

N.R.-adjusted transition rates

The difference between an N.R-adjusted table and one that is not is that the former is based on pools whose denominators have been gradually pared down by dropping those obligors whose ratings have been withdrawn (set to N.R.). The number of withdrawn ratings grows particularly large in the case of lower ratings categories after just a few years. Little is known about obligors after their ratings are set to N.R.