Reduced Form Estimation Of The Noise Trading Model

Abstract

In this paper, we develop and implement the reduced form version of the Noise Trading Model in the Indian stock market, which we had proposed in an earlier paper. We show how to estimate the model without making any assumptions regarding the cross-sectional dependence that may exist among the individual stocks. This allows us to come up with precise estimates of the share of information versus noise in the opening stock price. To be specific, information accounts for 52% of the variance of the opening stock price and noise contributes to the rest. When we split the overall sample into weekend versus weekday, we find that the information share of the opening price is significantly higher after a weekend at 72% compared to 46% during the rest of the week.