Is the Crowd Wise Enough to Recognize Creditworthy Borrowers?

We study whether the crowd can apply its collective wisdom in a context of high information asymmetry ? a crowdlending platform that matches lenders from the global north with borrowers from developing countries, and observe whether the lenders screen the borrowers correctly. We exploit interventions by administrators of a platform who covered missing funding for campaigns to observe the counterfactual ? repayment history of loans that did not raise the required sums from the crowd of investors. Our findings are threefold. First, we find that the borrowers which the crowd chose not to fund were less likely to repay by 11 percentage points - but only when the borrowers voluntarily shared financial information of their business with potential lenders. Second, this ability to predict borrower repayment is lost once the loan offers a non-pecuniary, social return in addition to the monetary one. Lastly, financial experts among the crowd did not prove to have significantly superior predictive ability over amateur members. Since crowdlending platforms rely on the screening competency of its supply side members to know a creditworthy borrower from one who is likely to default, these results are important to reveal the conditions under which the crowd is able to prove its predictive power.