Intertemporal Capital Asset Pricing Model (ICAPM) of Merton (1973) postulates a positive relationship between time-varying conditional risk and conditional return on securities. However, empirical evidence is inconclusive on this issue. On the other hand, there is strong evidence that volatility increases disproportionately with negative shocks in stock returns. This paper examines the relationship between time-varying return and volatility in two NSE-based stock indices-S&P CNX Nifty and CNX Nifty Junior, using four alternative model specifications i.e., GARCH-M, EGARCH, TGARCH and the Hamilton's Two-Regime Markov-Switching Model. Although this study finds strong evidence of asymmetric volatility adjustments, there is no significant relationship between conditional volatility and expected returns. Unconditional volatility and returns in two switching regimes are found negatively associated. In contrary to what is suggested by volatility-feedback hypothesis, the negative association between volatility and return is persistent rather than transitory and reversible. 

The notion that risk and returns are positively associated is so fundamental that Ghysels et al. (2005) call it ‘the first fundamental law of finance’. The Capital Asset Pricing Model (CAPM) (Sharpe, 1964; Lintner, 1965; and Mossin, 1966) formally presents this relationship in a static equilibrium context. According to this model, the expected return from an investment is directly proportional to its systematic risk. Merton’s (1973, 1980) Intertemporal Capital Assets Pricing Model (ICAPM) extends this notion in a dynamic framework. According to ICAPM, the time-varying expected excess return on an asset is a positive linear function of its time-varying (conditional) variance. However, subsequent studies draw conflicting conclusions regarding the sign of the relationship between conditional volatility and return. In general, they find a weak or negative causality of conditional volatility on conditional return.

On the other hand, as French et al. (1987) emphasize, there is much stronger evidence that positive innovations to volatility are correlated with negative innovations to return. This phenomenon is often explained in the light of ‘leverage effect’ and the ‘volatility feedback hypothesis’.Unfortunately, a sound theoretical framework is not available to define the time-varying expected returns and volatility. Different studies have used different specifications based on their empirical suitability. As Harvey (2001) points out, the empirical findings on risk-return relationship are highly sensitive to model specification of return dynamics. Most of the support for a zero or positive relation has come from the studies that have used the standard GARCH-M model of conditional volatility.

Applied Finance Journal, Intertemporal Capital Asset Pricing Model, ICAPM, National Stock Exchange, NSE, Time-Series, Statistical Techniques, Stock Market Returns, GARCH Model, American Statistical Association, International Stock Market.