Testing Random Walk Hypothesis and Volatilities of SRI and Traditional Indices

This study deals with the various asymmetric effects and market efficiency of the traditional and sustainable indices based on daily returns during the period from December 1998 to March 2015. Here, Dow Jones Sustainability Index US [socially responsible investment (SRI) index] and Dow Jones World US (conventional index) indices are considered. The well-known volatility forecasting measures like autoregressive conditional heteroskedasticity (ARCH), Generalised Autoregressive Conditional Heteroskedasticity, Exponential Generalised Autoregressive Conditional Heteroskedasticity and threshold ARCH (TARCH) are applied to attain the objectives. ARCH measure confirms about volatility clustering in the indices. Significant asymmetric shocks and persistence of conditional volatilities are present in the daily returns of the indices. According to the EGARCH measure, the return of the DJSI US is free from the leverage effect, whereas in TARCH measure, leverage effect exists in both the indices. Here, TARCH measure is the best forecasting measure for the DJSI US index because it provides lowest root mean squared error, mean absolute error and mean absolute percent error. But in case of DJ World, EGARCH measure is superior. Finally, the returns of the SRI and conventional indices follow the characteristics of random walk hypothesis that means the indices are informationally efficient at their weak forms, and no one can predict the stock price movements and cannot earn abnormal profits by applying technical analysis.