P2P CREDIT MARKETS · Insights from Person-to-Person Credit Markets

Insights from Person-to-Person Credit Markets

Traditional models of credit markets have been challenged in recent years by the proliferation of new business models for credit allocation. One example is the person-to-person credit market model which allows individual borrowers and lenders to interact directly, thus, eliminating the traditional bank middlemen. The research project provided new insights into the economics of alternative credit markets by addressing the following open questions: 1. What are the effects of interest rate restrictions? 2. To what extent do these markets exhibit local-bias? 3. Is non-verified information valuable? 4. What are the effects of transaction costs and social distance in such a market? The research project addressed these questions by utilizing new rich data from two big person-to-person credit markets – Prosper.com and Kiva.org. Essentially each of these markets is a platform for borrowers and lenders to interact and originate loans. In recent years person-to-person credit markets have gained increasing popularity. In such markets, all market makers employ a similar business model encompassing an online platform that is devised to allow borrowers and lenders to interact and to originate loans. The market maker is responsible for transferring funds between market participants and to initiate collection efforts in case of missing repayments. In addition, some market makers match between lenders and borrowers based on the lenders’ preferences. In return, borrowers and lenders pay fees to the market maker. The research that came out of the first project was recently published in the Review of Economics and Statistics (95(4), 1238-1248, October 2013). The main findings featured in this article are that higher interest rate caps increase the probability that a loan will be funded, especially if the borrower was previously just “outside the money.” I do not find, however, changes in loan amounts and default probability. The interest rate paid rises slightly, probably because online lending is substantially, yet imperfectly, integrated with the general credit market. The research that came out of the second and the third projects have been summarized in the Master Thesis of two of my research students. The main findings from the local-bias project reveals that individuals are likely to lend more to individuals reside in their state, or in a neighboring state, rather than to individuals from states that are further away. Moreover, the bias is inversely proportional to the distance between borrower and lender states. In addition, the bias is slightly greater among lenders who choose medium-risk borrowers. The data do not support that conjecture that the bias is driven by asymmetric information. Apparently, the reason for the bias is likely not a rational one, and probably caused by preferring the near and familiar. The main findings in the verifiable-vs-non-verifiable project reveal that non-verifiable information affects lenders' lending decisions. Moreover, the effect of this information is found to depend on the sentiment expressed by the information. For example, optimistic sentiment is found to have a greater effect than pessimistic sentiment. The research that came out of the fourth project was recently published in the B.E. Journal of Economic Analysis & Policy (Contributions) (14(1), 271-96, July 2013). In this project we use a field experiment at Kiva, the online microfinance platform, to examine the role of transactions costs and social distance in decision-making. Requests for loans are either written in English or another language, and our treatment consists of posting requests in the latter category with or without translation. The data provide evidence that relatively small transactions costs have a large effect on the share of funding coming from speakers of languages other than that in which the request was written. Social distance plays a smaller role in funding decisions.

Data: CORDIS, © European Union