Abstract
This paper evaluates the capacity of five structural credit risk models to forecast default rates. In contrast to previous studies with similar objectives, the paper employs firm-level data and finds that model-based forecasts of default rates tend to be unbiased and to deliver point-in-time errors that are small in both statistical and economic terms. In addition, in- and out-of-sample regression analysis reveals that the models account for a significant portion of the variability of credit risk over time but fail to fully reflect its dependence on macroeconomic cycles.
Authors
- Nikola A. Tarashev
JEL codes
- G33
- E44
- G28
- C13