A search for long-range dependence and chaotic structure in Indian stock market

This study tests for the presence of nonlinear dependence and deterministic chaos in the rate of returns series for six Indian stock market indices. The overall result of our analysis suggests that the returns series do not follow a random walk process. Rather it appears that the daily increments in stock returns are serially correlated and the estimated Hurst exponents are indicative of marginal persistence in equity returns. Result from the test of independence on filtered residuals suggests that the existence of nonlinear dependence, at least to some extent, can be attributed to the presence of conditional heteroskedasticity. It appears, therefore, that low order GARCH-type models can adequately explain some, but not all, of the observed nonlinear dependence in the data. Further, we find very little evidence to support the proposition that returns are generated by a chaotic system. Only in two out of six cases the results are supportive of sensitive dependence on initial condition, which indicates chaos. Presence of chaos in market indices implies that profitable nonlinearity based trading rules may exist at least in the short-run. Finally, fairly contrary to the findings of previous studies, rejection of random walk hypothesis offers some possibility of returns predictability.