Abstract
The significance of credit risk models has increased with the introduction of the New Basel Accord, known as Basel II. The aim of this study is to examine default rate modeling.
This paper follows two possible approaches to macro credit risk modeling, empirical models and a latent factor model based on Merton. We employ data over the time period from 1988 to 2003 for the Finnish economy, including time series of bankruptcy, numbers of firms and industry-specific data.
Linear vector autoregressive models are used in the case of a dynamic empirical model. We examine how significant macroeconomic indicators determine the default rate in the whole economy and in industry-specific sectors.
Since these models cannot provide microeconomic foundations, we employ a model with one latent factor, although multi-factor models are also considered. This estimation helps us to understand the relationships between credit risk and macroeconomic indicators.
Both models can be used for default rate prediction or stress testing by central authorities.