Testing Weak Form of Efficient Market Hypothesis: Evidence from Bombay Stock Exchange (BSE) & National Stock Exchange (NSE)

Abstract

The study examines the validity of Weak form of Efficient Market Hypothesis in Indian stock market with special reference to Bombay Stock Exchange (BSE) and National Stock Exchange (NSE). For testing the presence of random walk, log returns of daily, weekly and monthly closing prices from July 1997 to Dec 2013 are employed. The market efficiency is tested by using both parametric and non-parametric test including run test, Ljung-Box Q-statistic test, unit root test, variance ratio test and GARCH (1, 1). In order to verify the normality of return series, Jarque Bera test was used and descriptive statistics were examined. The results of the same revealed non-normality of return series and further exhibits negatively skewed and leptokurtic returns. The run test results reveal that successive market returns are not independent for daily and weekly returns but presence of random walk is found in monthly returns for both BSE &NSE. Daily returns are serially correlated but for weekly and monthly returns insignificant serial correlation is found. The unit root test is used for testing the stationarity of the return series and the results show that daily, weekly and monthly returns are non-stationary when tested at level but become stationary at first difference. The variance ratio test reveals traces of weak form of market inefficiency under assumption of homoscedasticity and heteroscedasticity. The GARCH (1, 1) test reveals high persistence and slow reversion to the mean. The implications of rejection of random walk hypothesis supports the predictability of future stock prices by studying the pattern of historical prices and providing an opportunity to investors to earn supernormal profits by holding a well-diversified portfolio.