Inattention in Crowded Markets

Abstract:

We present a novel model of crowded markets and rational inattention. Our main application is a financial market where early entrants can buy low (far from fundamentals) and sell high (close to fundamentals). Early entrants bid up the price for late entrants, while late entrants increase the exit price for early entrants. The order when an investor learns the existence of this opportunity is her type. Before entry, the investor can also learn on about how many are already entered which determines the attractiveness of the opportunity. However, such learning is costly. Our analysis emphasize two sources of inefficiency: the inefficiency of the entry decision (which can lead to crowding), and the inefficiency of the learning decision. Our model predicts under-entry in transparent markets, over-entry in opaque markets and over-learning in all types of markets. We show that sometimes more opaque markets provide higher welfare. We also show that as the proportion of sophisticated traders is increasing in global markets, crowding might not increase. Also, welfare might be higher in a market with very few sophisticated investors, even if they do not eliminate arbitrage opportunities. Our results also provide a rational for the increasing popularity of exclusive, and opaque trading venues as black pools.