Information in Noise : Strategic Trading under Autocorrelated Uninformed Orders

This paper develops a strategic trading model in which uninformed orders are allowed to exhibit a general correlation structure that generates autocorrelation in the order flow. Since the order flow is predictable, informed traders and the market maker not only need to infer information about the asset value, but also forecast future order flows. The correlation structure of uninformed orders has significant effects on trading strategies, market liquidity, and the informativeness of prices. Since the empirical autocorrelation in order flows is likely to come from uninformed traders, strategic trading models should not assume them to be simply noise