Modelling Scenario Analysis and Macro Stress-testing

Macro stress-testing has become an important tool to assess financial stability. This paper describes a tool kit for scenario analysis and macro stress-testing. It is based on a model which maps multivariate scenarios to banks' credit and interest rate risks by deterministic and stochastic simulations. Our approach is an extension of existing macro stress-testing models as it distinguishes between probability of default on the one hand and loss given default on the other and allows for separate models for domestic and foreign portfolios. Another contribution of the paper is that the stochastic simulations generate loss distributions which provide insight in the extreme losses and allow for changing correlations between risk factors in stress situations. The methodology is applied to the Dutch banking sector.