Fractal properties of stock markets: the experience of Southeast Asian countries

Abstract

The existing base of empirical research in the field of fractal analysis of time series proves that many financial markets do not obey the efficient market hypothesis. In this paper attempts to find inefficient markets in Southeast Asia by estimating the Hurst index for stock indices. The purpose of the article is to reveal the dynamics of the Hurst exponent based on data on the returns of well-known stock indices on the stock markets of Southeast Asia. Two hypotheses are tested: stock markets in Malaysia, Indonesia, Thailand, and Singapore are examples of inefficient markets; the Hearst exponent on Asian stock index returns can predict full-scale financial collapses. The hypothesis testing methodology is based on the methods of fractal analysis of time series. Using the sliding window method the dynamics of the Hurst exponent is revealed which provides insight into market persistence, degree of predictability and memory in the stock markets of Malaysia, Indonesia, Thailand and Singapore. The used method of rescaled range (R/S analysis) made it possible to distinguish a stochastic process from a non-stochastic one and reveal signs of self-similarity in the time series. Thus, this study aims to identify fractal properties in the time series of the FTSE Bursa Malaysia KLCI (Malaysia), IDX Com­posite (Indonesia), SET Index (Thailand), STI Index (Singapore). The obtained results can be useful when choosing a market for invest­ments, taking into account its fractal struc­ture, especially during the onset of full-scale financial collapses.