Do Social Networks Solve Information Problems for Peer-to-PeerLending?Evidence from Prosper.com

This paper studies peer-to-peer (p2p) lending on the Internet.Prosper.com, the first p2p lending website in the US, matches individuallenders and borrowers for unsecured consumer loans. Using transactiondata from June 1, 2006 to July 31, 2008, we examine what informationproblems exist on Prosper and whether social networks help alleviate theinformation problems. As we expect, data identifies three informationproblems on Prosper.com. First, Prosper lenders face extra adverseselection because they observe categories of credit grades rather thanthe actual credit scores. This selection is partially offset whenProsper posts more detailed credit information on the website. Second,many Prosper lenders have made mistakes in loan selection but they learnvigorously over time. Third, as Stiglitz and Weiss (1981) predict, ahigher interest rate can imply lower rate of return because higherinterest attracts lower quality borrowers. Micro-finance theories arguethat social networks may identify good risks either because friends andcolleagues observe the intrinsic type of borrowers ex ante or becausethe monitoring within social networks provides a stronger incentive topay off loans ex post. We find evidence both for and against thisargument. For example, loans with friend endorsements and friend bidshave fewer missed payments and yield significantly higher rates ofreturn than other loans. On the other hand, the estimated returns ofgroup loans are significantly lower than those of non-group loans. Thatbeing said, the return gap between group and non-group loans is closingover time. This convergence is partially due to lender learning andpartially due to Prosper eliminating group leader rewards whichmotivated leaders to fund lower quality loans in order to earn the rewards.