Learning in Crowded Markets

We develop a model of capital reallocation to analyze whether the presence of more arbitrageurs improves capital allocation and welfare. While trades can become crowded due to imperfect information and externalities, arbitrageurs can devote resources to flexibly learn about the number of earlier entrants. Above a threshold, increasing the number of arbitrageurs does not affect capital allocation: whether there is eventually too little or too much capital allocated to the trade is solely determined by the parameters of the market. The flexibility in the learning technology is key to this insight. However, the presence of more arbitrageurs decreases welfare, as they use more aggregate resources to learn about each others’ position. When both sophisticated and unsophisticated arbitrageurs are present, increasing the share of sophisticated arbitrageurs might be welfare reducing. ∗For comments and suggestions, we are grateful to Patrick Bolton, Thomas Chemmanur, Johannes Hörner, Victoria Vanasco and seminar participants at Boston University, Central European University, ESSET 2014 (Gerzensee), Brigham Young University, GLMM 2015 (Boston), Paul Woolley Conference 2015 (LSE). We thank Pellumb Reshidi for his excellent research assistance. The support of the European Research Council Starting Grant #336585 is acknowledged. †Central European University, Department of Economics, Nádor u. 9., Budapest, H-1051, Hungary, e-mail: [email protected], http://www.personal.ceu.hu/staff/Peter_Kondor/index.php ‡Central European University, Department of Economics, Nádor u. 9., Budapest, H-1051, Hungary, e-mail: [email protected], http://www.people.bu.edu/zawa