Humboldt-Universität zu Berlin - Statistik

Statistical Aspects of Credit Rating (VL)

M. Müller


Credit rating is the assessment of the quality of a credit with respect to its risk of default. Typical terms which are estimated in this context are the probability of default (PD) and the recovery rate (RR) or the loss given default (LGD). Obviously, numerous factors can influence the default risk of a credit and thus its rating. Data sources are historical data (e.g. balance sheet data for corporates) or market data (bond spreads or credit default swaps).


  1. Introduction components of a rating system, Basel 2
  2. Statistical Classification discriminant analysis (LDA, QDA), logistic regression, ordered regression
  3. Evaluation of Rating Systems discriminatory power, goodness of fit, hypothesis tests for estimated PDs, sample selection
  4. Estimaton from Market Data distance to default, transition matrices, hazard rates, credit risk models