Evaluation and Comparison of Market and Rating Based Country Default Risk Assessment

We compare two different approaches to assess country default risk by evaluating their forecast accuracy. In particular we analyze whether market based or rating based risk assessment is superior. To evaluate the forecast accuracy we analyze the differences between several default risk measures and realizations of defaults/non defaults in the forecast period. The considered risk measures are, on the one hand, default probabilities and binary crisis forecasts (default/non default), on the other. Within a sample of 19 emerging market countries in 1998 to 2007 we find that risk measures derived by reduced form credit risk models from market data outperform ratings of S&P