A Simple Power-Law Tail Estimation of Financial Stock Return

Abstract

This study proposes a simple methodology to estimate the power-law tail index of the Malaysian stock exchange by using the maximum likelihood Hill’s estimator. Recursive procedures base on empirical distribution tests are use to determine the threshold number of observations in the tail estimation. The threshold extreme values can be selected bases on the desired level of p-value in the goodness-of-fit tests. Finally, these procedures are apply to three indices in the Malaysian stock exchange.