I examine the informational content of the limit order book and the degree to which algorithmic traders strategically use limit and market orders for informed trading. To this end, I study intraday return predictability from market and limit orders for all NYSE stocks over 2002-2010. I distinguish between two sources of return predictability: inventory management and information. My first finding is that informed limit orders are the dominant source of intraday return predictability. I exploit a quasi-natural experiment (the NYSE Hybrid introduction) to identify the effect of algorithmic trading on limit order book informativeness. My evidence indicates that algorithmic trading leads to an increase in overall return predictability. However, the increase in predictive power is smaller for limit order book variables than for market order imbalance. Overall, my results suggest that although algorithmic trading leads to increase in adverse selection by liquidity demanders, this negative effect is partially offset by the increase in informed liquidity provision.

Discussant: Wenqian Huang (VU University Amsterdam)