Information Efficiency of Stock Markets

Abstract

The efficient market hypothesis is one of the possible approaches when explaining the movements of asset prices in financial markets. In this thesis, the author deals with the hypothesis of whether the main indexes of selected Asian stock markets (mainland China, Hong Kong, Japan) follow the random walk hypothesis or not. The basic testing period was defined from 2010 to 2020 using the data of daily frequency. The predictability of returns in the form of random walk models is investigated with the help of selected statistical tests. The application framework is based on a combination of mathematical models of efficient markets with linear and nonlinear testing procedures. The random walk hypothesis is rejected most often in the case of the SSE50 index using the whole data sample. In addition, the dynamics of the results of the BDS independence test and variance ratio test in time are evaluated as well. To evaluate the significance of applied tests, the methods of bootstrapping are utilized. Based on dynamic results of statistical tests, the investment strategy proposal was discussed as well.