Does risk aversion drive financial crises? Testing the predictive power of empirical indicators

There are several types of risk aversion indicators used by financial institutions. These indicators, which are estimated in diverse ways, often show differing developments, although it is not possible to directly assess which is the most appropriate. Here, we consider the most well-known of these indicators and construct others with standard methods. As financial crises generally coincide with periods in which risk aversion increases, we try to check if these indicators rise just before the crises and also if they are able to forecast crises. We estimate logit and multilogit models of financial crises - exchange rate and stock market crises - using control variables and each of the risk aversion indicators. In-sample simulations allow us to assess their respective predictive powers. Risk aversion indicators are found to be good leading indicators of stock market crises, but less so for currency crises.