Non-Linear Dependence of Indian Shariah Market

Abstract

The work focuses on linearity or otherwise of Indian Shariah market i.e. CNX NIFTY, CNX500 and S&P BSE TASIS 50 during the period spanning from 01/January/2008 to 31/June/2013. In order to detect the presence, BDS test of Brock et al (1996) was employed and null hypothesis was strongly rejected the existence of Independent Identical Distribution (IID). On further investigation on presence of IID with the help of GARCH (1, 1) and enquiry on whether the non-linear dependence was caused by predictable conditional volatility, it was found that the non-linearity was caused not by serial dependence or linear dependence rather it was caused by volatility clustering and GARCH effect in the return series. This leads to the inference that the market lacks efficiency and rejects the Random Walk Hypothesis. Hence, alerts the policy makers on the predictability of the market even for short horizon which in principle should not exist, as Shariah market is expected not to give opportunity for abnormal return.