COVID-19, Credit Risk and Macro Fundamentals

We investigate the relationship between macro fundamentals and credit risk, rating migrations and defaults during the start of the COVID-19 pandemic. We find that credit risk models that use macro fundamentals as covariates overestimate credit risk incidence due to the unprecedented drops in economic activity in the first lockdowns. We argue that this break in the macro-credit linkage is less affected if we take an unobserved components modeling framework, both at shorter and longer credit risk horizons. An additional advantage of these models is that they automatically provide an integrated forecasting approach for both the credit and macro variables in the model. An effort to repair the macro-credit link via the addition of government subsidy expenses, though better in-sample, provides a worse fit to credits if implemented pre-covid.