Empirical study on efficiency of Thailand stock market based on confidence interval of hurst index using DFA

Abstract

We are interested in investigating the efficiency of Thailand stock markets (SET and MAI markets) by using the Hurst index based on the Detrended Fluctuation Analysis (DFA). This study is an extension of the work of Sukpitak and Hengpunya (2016) by using Monte Carlo simulation and statistical analysis to construct a confidence interval of Hurst index for the efficient market based on the assumption that the sample of Brownian motion represents assets prices in efficient markets, then applying the interval as an indicator for the efficiency of Thailand stock markets. We found that there is a consistency in the behavior of Hurst index among different time windows sizes based on a constructed confidence interval. Based on this study, the result shows that both SET and MAI markets have some inefficient periods and we fail to reject that the markets are efficient in the recent years. To get more accurate conclusion on the efficiency of Thailand markets, other tools or techniques are required.