Estimating Time-Varying Currency Betas : New Evidence from Nine Developed and Emerging Markets

This paper examines the conditional time-varying currency betas from five developed markets and four emerging markets. We employ BEKK multivariate GARCH models of Engle and Kroner (1995) to estimate the time-varying conditional variance and covariance of returns of stock index, the world market portfolio and changes in bilateral exchange rate between the US dollar and the local currency. It is found that currency betas are more volatile than those of the world market betas. Currency betas in emerging markets are more volatile than those in the developed markets. Moreover, we find evidence of long-memory in currency betas. The usefulness of time-varying currency betas are illustrated by two applications