Refine
Year of publication
- 2001 (5) (remove)
Document Type
- Working Paper (3)
- Conference Proceeding (1)
- Report (1)
Language
- English (5)
Has Fulltext
- yes (5)
Is part of the Bibliography
- no (5) (remove)
Keywords
- Kreditrisiko (5) (remove)
Institute
Evaluating the quality of credit portfolio risk models is an important question for both banks and regulators. Lopez and Saidenberg (2000) suggest cross-sectional resampling techniques in order to make efficient use of available data and to produce measures of forecast accuracy. We first show that their proposal disregards crosssectional dependence in simulated subportfolios, which renders standard statistical inference invalid. We proceed by suggesting another evaluation methodology which draws on the concept of likelihood ratio tests. Specifically, we compare the predictive quality of alternative models by comparing the probabilities that observed data have been generated by these models. The distribution of the test statistic can be derived through Monte Carlo simulation. To exploit differences in cross-sectional predictions of alternative models, the test can be based on a linear combination of subportfolio statistics. In the construction of the test, the weight of a subportfolio depends on the difference in the loss distributions which alternative models predict for this particular portfolio. This makes efficient use of the data, and reduces computational burden. Monte Carlo simulations suggest that the power of the tests is satisfactory.
JEL classification: G2; G28; C52
Evaluating the quality of credit portfolio risk models is an important issue for both banks and regulators. Lopez and Saidenberg (2000) suggest cross-sectional resampling techniques in order to make efficient use of available data. We show that their proposal disregards cross-sectional dependence in resampled portfolios, which renders standard statistical inference invalid. We proceed by suggesting the Berkowitz (1999) procedure, which relies on standard likelihood ratio tests performed on transformed default data. We simulate the power of this approach in various settings including one in which the test is extended to incorporate cross-sectional information. To compare the predictive ability of alternative models, we propose to use either Bonferroni bounds or the likelihood-ratio of the two models. Monte Carlo simulations show that a default history of ten years can be sufficient to resolve uncertainties currently present in credit risk modeling.
This paper uses a unique data set from credit files of six leading German banks to provide some empirical insights into their rating systems used to classify corporate borrowers. On the basis of the New Basle Capital Accord, which allows banks to use their internal rating systems to compute their minimum capital requirements, the relations between potential risk factors, rating decisions and the default probabilities are analysed to answer the question whether German banks are ready for the internal ratings-based approach. The results suggests that the answer is not affirmative at this stage. We find internal rating systems not comparable over banks and furthermore we reveal differences between credit rating determining and default probability determining factors respectively. Klassifikation: G21, G33, G38
In the recent theoretical literature on lending risk, the coordination problem in multi-creditor relationships have been analyzed extensively. We address this topic empirically, relying on a unique panel data set that includes detailed credit-file information on distressed lending relationships in Germany. In particular, it includes information on creditor pools, a legal institution aiming at coordinating lender interests in borrower distress. We report three major findings. First, the existence of creditor pools increases the probability of workout success. Second, the results are consistent with coordination costs being positively related to pool size. Third, major determinants of pool formation are found to be the number of banks, the distribution of lending shares, and the severity of the distress shock.
In recent years new methods and models have been developed to quantify credit risk on a portfolio basis. CreditMetrics (tm), CreditRisk+, CreditPortfolio (tm) are among the best known and many others are similar to them. At first glance they are quite different in their approaches and methodologies. A comparison of these models especially with regard to their applicability on typical middle market loan portfolios is in the focus of this study. The analysis shows that differences in the results of an application of the models on a certain loan portfolio is mainly due to different approaches in approximating default correlations. That is especially true for typically non-rated medium-sized counterparties. On the other hand distributional assumptions or different solution techniques in the models are more or less compatible.