Learning in crowded markets

We present a novel entry-game with endogenous information acquisition to study the welfare effects of opacity and competition. Potential entrants to an opaque market are uncertain about their competitive advantage relative to other investors, i.e. their type. They construct optimal costly signals to learn about their types, where the marginal cost of learning captures the opacity of the market. In general, the individually optimal entry and learning decisions are socially suboptimal. Players over-invest in learning and more opaque markets are associated with more crowding. Nevertheless, more opaque markets might still lead to higher welfare by implying a better trade-off between the degree of crowding and the total cost of learning. Similarly, decreasing the share of smart investors in the market might also improve welfare. However, fierce competition is always detrimental to welfare as it leads to more wasteful learning without changing the level of crowding.