‚ 1 The New Capital Adequacy Framework for Credit Risk Possible Impact on the Austrian Banking Sector and Banking Supervision Franz Partsch Credit Division.

‚ 1

The New Capital Adequacy Framework for Credit Risk

Possible Impact on the Austrian Banking Sector and Banking Supervision

Franz PartschCredit Division

Oesterreichische NationalbankVienna, 1 February, 2001

‚ 2

OverviewEmpirical analysis: sample 37 larger Austrian banks

Magnitude and variability of credit risk• Data sources: annual bank supervision audit report, monthly

statistical returns Portfolio structure

• Data sources: central credit register, rating data, sector default data

Conclusions"Road map" for the implementation of the new

Accord based on empirical evidence and judgement

Disclaimer: available data do not come from credit risk management

sources and can only serve as more or less suitable

proxies for credit risk

‚ 3

Watch loans

3737373737N =

Watch 12/99Watch 12/98Watch 12/97Watch 12/96Watch 12/95

40

30

20

10

0

-10

31

1232

28

3132

28

12

13

428

12

254

28

4

28

Extreme values Outliers MedianBox Interquartile Range Highest Non-Outlier Lowest Non-Outlier

‚ 4

Doubtful loans

3737373737N =

Doubtful 12/99

Doubtful 12/98

Doubtful 12/97

Doubtful 12/96

Doubtful 12/95

20

10

0

-10

12

28148

13

15

148

13

84

13

13

413

‚ 5

Loss loans

3737373737N =

Loss 12/99Loss 12/98Loss 12/97Loss 12/96Loss 12/95

8

6

4

2

0

-2

28

22

22

28

22

226

28

6

28

‚ 6

Value adjustments to operating result

3434333334N =

12/9912/9812/9712/9612/95

Va

lue

ad

just

me

nts

to

op

era

ting

re

sult

120

100

80

60

40

20

0

-20

67

Value adjustments of claims and allocations to provisions for contingent claims and for credit risks

‚ 7

Value adjustments to loans

3434333334N =

12/9912/9812/9712/9612/95

Va

lue

ad

just

me

nts

to

loa

ns

2,5

2,0

1,5

1,0

,5

0,0

-,5

9

107

‚ 8

Value adjustments to assets

3434333334N =

12/9912/9812/9712/9612/95

Va

lue

ad

just

me

nts

to

ass

ets

1,4

1,2

1,0

,8

,6

,4

,2

0,0

-,2

65

107

7262

‚ 9

Conclusions (I)

Credit risk is by no means immaterial for the average large

Austrian bank has been, on average, fairly stable over the last

years shows significant and increasing differences

between banks

‚ 10

Austrian Central Credit RegisterDescription

Register of all borrowers from financial institutions (banks, leasing companies, insurance companies) with more than ATS 5m in total loans outstanding or credit lines

Purpose service for reporting institutions source of information for supervisory authorities

Content structural data on borrowers (name , address, legal form

etc.) monthly reporting by types of loans quality check and aggregation regular and ad-hoc information on total indebtedness of

borrowers

‚ 11

Loans by borrower type (I)

57,11% 57,27%

44,16% 41,34%37,89%

7,51% 6,89%

5,40%5,32%

5,10%

4,74% 4,39%

4,19%3,97%

3,62%

10,48% 10,23%

11,72% 16,04%15,89%

12,55% 13,98%

16,80% 14,38%18,05%

7,61% 7,24%

17,74% 18,94% 19,46%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1995 1996 1997 1998 1999

Domestic corporate Domestic private Domestic public

Domestic financial Foreign others Foreign financial

‚ 12

Loans by borrower type (II)

343434343434N =

Foreign f inancial

Foreign others

Domestic f inancial

Domestic public

Domestic private

Domestic corporate

80

60

40

20

0

-20

2302524

23

30

13

‚ 13

Loans by country type

3333N =

Non-OECDOECD

120

100

80

60

40

20

0

-20

1729

1729

‚ 14

Country Risk Weights (Rating agency)

22222222N =

100%50%20%0%

120

100

80

60

40

20

0

-20

1

29

29

17

‚ 15

Country Risk Weights (Export Credit Agency)

44444N =

150%100%50%20%0%

100

80

60

40

20

0

-20

‚ 16

Hypothetical default rates (corporate sector risk)

4242N =

BankruptcyPayment incidence

8

6

4

2

0

-2

27

7

27

Payment incidence: reported non-payment of commercial or financial debt

‚ 17

Corporate sector risk distributions

373737N =

High riskMedium riskLow risk

100

80

60

40

20

0

-20

15

26

1337

15

13

37

36

Payment incidence

‚ 18

Corporate sector risk distributions

363636N =

High riskMedium riskLow risk

100

80

60

40

20

0

-20

15

37

15

3724

Bankruptcy

‚ 19

Borrower number by borrower type

373737373737N =

Foreign f inancial

Foreign others

Domestic f inancial

Domestic public

Domestic private

Domestic corporate

5000

4000

3000

2000

1000

0

-1000

31361910

21210

12021029

362

1102

4

10

2

1

‚ 20

Effective number of loans

37N =

700

600

500

400

300

200

100

0

-100

4

37

10

‚ 21

Hypothetical Granularity Scaling Factor

37N =

1,0

,8

,6

,4

,2

0,0

-,2

28

363012

‚ 22

Conclusions (II)Credit risk

is by no means immaterial for the average Austrian large bank

has been, on average, fairly stable over the last years shows significant and increasing differences between banks

Portfolio structure large banks have significant domestic and foreign lending in

all exposure classes (corporates, public, financial) country risk is concentrated in highly rated areas, but lending

to countries with low ratings is material for some banks corporate exposures are concentrated in medium risk sectors,

but lending to corporates in high risk sectors is material for some banks

the number of borrowers in some exposure classes (public, financial) is fairly small and (lack of) granularity will be an issue for some banks

‚ 23

"Road map" for Basel II

Data own time series on ratings, defaults, losses data pooling mapping to external data check against other data sources

Estimation of risk parameters robust methods using relatively few data transparency for tests by risk managers,

supervisors and market participants

First: sound rating system and risk management frameworkThen: