DETERMINATION OF VOLATILITY AND MEAN RETURNS: AN EVIDENCE FROM AN EMERGING STOCK MARKET

In the present research we work with excess returns for an emerging stock market i.e. Jamaican Stock Price Index for the determination of volatility persistence and persistence in the mean returns series. We model excess returns in this stock market using state space or unobserved component models, which is a signal extraction approach. Our model encompass stable distributions to account for fat tails and GARCH-like effects to account for time varying volatility that may be present in the series. The study results that are obtained using the most general as well as the restricted versions of the state space models reveal statistically significant evidence of volatility persistence in the excess returns series. Further, there exist persistent predictable signals in returns series at 5 percent level of significance, and the value of an efficiently estimated excess returns series is percent per month (percent per annum). Further, the series encompass a stable characteristic exponent of showing a non-normal behavior in this market.