Measuring Time-Varying Market and Currency Risks with Stochastic Dominance : Evidence from Country Level Stock Returns

In this paper, time-varying market and currency risks among a selected set of developed and emerging economies are compared in terms of stochastic dominance. For this purpose, time-varying exchange rate exposure and market betas are obtained through a multivariate model that explicitly allows for time-varying second moments. Two betas are not assumed to be orthogonal and we explicitly allow for non-orthogonality. The cumulative distribution functions of time-varying betas in the sample indicate that stock returns in emerging economies are more exposed to currency risk, though their exposure to market risk is moderate. On the contrary, the stock returns in developed economies are more exposed to market risk while their exposure to currency risk is remarkably low. There is also evidence to establish the notion that, during the post-currency crisis period, currency risk in Korea is fading out over time