Non-Linear Dependence in Stock Returns: Evidences from India
Journal of Quantitative Economics, Vol. 8, No. 1, January 2010
17 Pages Posted: 7 Mar 2011 Last revised: 5 Aug 2011
Date Written: January 1, 2010
This paper examines non-linear dependence in Indian stock returns using a set of non-linearity tests. The daily data between 1997 and 2009 for eight indices from National Stock Exchange (NSE) and six indices from Bombay Stock Exchange (BSE) are used. The results suggest strong evidence of non-linear structure in stock returns. The non-linear dependence, however, is not consistent throughout the sample period as indicated by windowed Hinich test [1996, Journal of Non-parametric Statistics, 6, 205-221] suggesting episodic non-linear dependence in Indian stock returns. The existence of episodic non-linear dependency is associated with events such as uncertainties in international oil prices, sub-prime crisis followed by global economic meltdown, and political uncertainties among others.
Keywords: Nonlinearity, preditiability, market efficiency, random walk, episodic dependence, Indian stock market, NSE, BSE Nifty, Sensex
JEL Classification: G15, C49
Suggested Citation: Suggested Citation