Advantageous Selection in Fintech Loans

Using data from the largest peer-to-peer (P2P) lender in the United States, I document advantageous selection in loan amount. By exploiting a natural experiment within the platform, I show that borrowers who select larger loans are less likely to default. This selection is driven by households who live in states with bankruptcy-friendly laws, where borrowers' default costs are lower. Standard models where borrowers maximize their utility cannot rationalize my results and make the opposite prediction. In a simple model of household borrowing, I show that my results can be explained by the fact that borrowers facing higher loan prices search more intensively for cheaper loans. This effect is stronger for the safest borrowers, as they enjoy the greatest benefits from the switching.