Do industries predict the stock market due to slow diffusion of information?

The slow diffusion of information hypothesis has emerged as a more convincing explanation for lead-lag patterns in assets returns compared to traditional explanations such as non-synchronous or thin trading, liquidity factor, or size factor, etc. We provide further support to slow diffusion of information hypothesis from an emerging market. We use a rich data set of weekly returns of 34 industries listed at the Karachi Stock Exchange over the period 1998 to 2011. In a separate regression for each industry, we regress KSE-100 Index current returns on the lagged industry returns and a set of control variables. Our results indicate that a large number of industries predict the market returns up to 3 weeks. The predictive power of industries decreases as we increase the prediction horizon. These findings are robust even after we control for known predictors of market return such as size of an industry, trading volume of an industry, and the lagged trading volume of the market. Our results support the slow diffusion of information hypothesis.