Testing the Martingale Hypothesis in the Indian stock market: evidence from multiple variance ratio tests

Abstract

This paper tests the martingale hypothesis and the weak form efficiency in the stock prices of the Indian market. We have used the multiple variance ratio test and its improved modifications which include the power-transformed test, joint rank tests and the Wald-type test. These tests are finite sample tests and do not have the shortcomings of small sample size. This paper also applies the wild bootstrap procedure on all the tests used because it shows desirable small sample properties under conditional heteroskedasticity. We have used weekly data covering pre- and post-subprime crisis. We have also used moving subsample windows to examine the dynamic behavior of stock prices. This can help to study the sensitivity of the results to a particular sample period and to obtain inferential findings robust to possible structural changes and presence of influential outliers. Our results provide evidence in support of the weak-form efficiency of S&P CNX Nifty, CNX 100 and S&P CNX 500 indices. But CNX Nifty Junior, Nifty Midcap 50 and CNX Smallcap show some deviation from the martingale hypothesis.