Understanding how equity returns change over time is of great use in understanding macroeconomy as well as the dynamics in the global equity markets. Time-varying volatility modeled through the process of conditional variance proves to be an effective measure of the changes in financial returns. In particular, the ARCH (Autoregressive Conditional Heteroskasticity) class volatility models are shown to adequately capture the persistence in volatility. Among the various extensions of the ARCH/GARCH family, EGARCHJt (Exponential Generalized ARCH with jumps and heavy-tails) is a model that incorporates heavy-tailness and asymmetric impact from returns in the estimation process.
In this paper, we explore the volatility behavior in Emerging Markets (EM) and how it is different from that in the U.S., as EM gain increasing attention from global investors. We apply the EGARCHJt model to a broad set of data that comprises both U.S. and EM equities and compare volatility behaviors across countries in these markets. The results suggest that EM tends to have lower level of persistence, less degree of leverage effects, more jumps and heavier tails than the U.S. market.
Keywords: GARCH; Bayesian Analysis; Heavy-tails; Jumps; Emerging Markets; Equities; Markov Chain Monte Carlo