Can uncertainty predict stock markets? A cross country analysis.

Abstract

Economic intuition suggests that uncertainty could predict stock markets. We consider two uncertainty measures: implied volatility and economic policy uncertainty (EPU). It is wellknown implied volatility is negatively correlated with the stock market returns, but its ability to predict returns is limited. Much less is known about the relationship between the EPU indices proposed by Baker et al (2012) and respective stock market returns. We therefore study the impact of implied volatility and EPU on stock markets, utilizing a dataset consisting of 12 countries. We study this relationship for each country separately, and also together utilizing panel regressions with standard errors adjusted for heteroskedasticity, autocorrelation and cross-sectional dependency using Driscoll and Kraay (1998) standard errors for linear panel models. We find evidence of a negative concurrent relationship between implied volatility and stock market returns that is common across countries and holds during various economic states. In addition, evidence of EPU having predictive capabilities of stock-market returns is present across countries and hold during ordinary times of the economy. Economic significance of our results is illustrated by a very profitable trading strategy, delivering over 15% annualized abnormal return.