Learning in Crowded Markets

Adam Zawadowski Peter Kondor - Mar 12, 2018

Working Paper No.  00031-00

We study a capital reallocation problem in which  investors can enter into a new market where they compete with each other in identifying the best deals. While ex ante investors are uncertain about their relative advantage in identifying the best deals, they can devote costly resources to learning about their relative advantage in a fully flexible way. We find that investors might allocate too much or too little capital to the new market.  Increasing competition between investors induces them to learn more,  shifting the distribution of entrants towards those with a relative advantage. However, competition does not change overall entry, so it does not affect the efficiency of capital allocation.  Thus, learning induced by competition turns out to be wasteful and welfare decreasing.  Allowing investors to learn in a fully flexible way -- as opposed to requiring Gaussian signals -- is what makes our argument transparent. We study several extension. 


Download Paper

big data Market Closures Feedback effect