Heavy-Tailed Distributions, GARCH Model and the Stock Market Returns in South Korea

Abstract

As other developed economies over the world, the stock market plays a crucial role in facilitating the economic growth. In this paper, we compare two different types of heavy-tailed distribution, the Student’s t distribution and the normal reciprocal inverse Gaussian distribution, within the generalized autoregressive conditional heteroskedasticity (GARCH) framework for the daily stock market returns of South Korea (KOSPI). Our results show two important findings: i) the daily KOSPI returns exhibit conditional heavy tails even after volatility clustering effect has been accounted for; and ii) the NRIG distribution has a better in-sample performance than the Student’s t distribution.