Idiosyncratic Risk Matters: Evidence from India

Mohinder Parkash and Rajeev Singhal

For a sample of US firms, Goyal and Santa-Clara (2003) find that the average stock variance is positively associated with higher returns in the subsequent month. The authors interpret the evidence as supporting the hypothesis that idiosyncratic risk matters. Several subsequent papers employing different methodologies present evidence conflicting with the findings of Goyal and Santa-Clara. We adopt the approach of Goyal and Santa-Clara in the Indian context. The results we obtain when we measure volatility on a monthly basis do not offer support for evidence documented in Goyal and Santa-Clara. However, when we compute volatility over weekly intervals, we find results consistent with those presented by Goyal and Santa- Clara. Our results in the Indian context are in line with the generally supportive evidence obtained using international data about the positive relationship between conditional volatilities and expected returns with an important difference. Supporting the arguments of Andersen, Bollerslev, Diebold, and Labys (2003), our results suggest that higher frequency data is more powerful in predicting future volatilities.

DOI : https://doi.org/10.21102/graf.2017.03.81.09