TESTING THE RANDOM WALK HYPOTHESIS FOR EMERGING MARKETS: EVIDENCE FROM LINEAR AND NON-LINEAR UNIT ROOT TESTS

In this study, the random walk hypothesis for emerging markets has been tested. First of all,Harvey et. al. (2008) linearity test was made in this study where different time intervals were handled. ADF (1979) unit root test was made to the linear series in order to test the efficiency of the market based on the results of the linearity test and in stock exchanges in India and Russia where Brazil and China stock markets are not efficient, it was concluded that the efficient market hypothesis is valid. In addition, these results were supported by Kapetanios (2005) unit root test that allowed more than two breaks and structural breaks were endogenously. For non-linear series,Kapetanios, Shin and Snell (2003) (KSS) unit root tests were made and the efficient market hypothesis in Argentina, Indonesia, Mexico and Turkey is valid. In addition, it was found with the help of a time varying KSS unit root test which also takes the seasonal changes into account that activity at different periods for all markets was concerned but this efficiency was not valid at all periods for any market. This result supports the idea which was about the findings on the validity of the random walk hypothesis for many developing countries in previous studies that were not clear.