The Micromarket for Prosper Loans

Objectives:

1.      To briefly explain the mechanics of the Prosper Peer-to-Peer loan marketplace from the point of view of the borrower (demander of credit) and lender (supplier of credit).

2.      To describe the rich database of Prosper loans which has been created over the past 5 years.

3.      To  specify a simultaneous equations model of this marketplace which is exactly identified.

4.      To estimate the reduced-form parameters of this model using OLS regression techniques.

5.      To offer some tentative observations regarding the demand and supply of such loans for the 195,000 Prosper members.

Data/Methods

      The data for this study is available in the public domain at www.prosper.com.  For this study, the level of analysis is taken to be the individual lender’s portfolio of loans and the individual successful borrower’s loan specifics.  Each subject lender will be a successful bidder on one subject borrower’s loan.  The Prosper marketplace changed radically during an 8-month period prior to August 1, 2009.   Prosper originated no loans because they were in a“quiet perod” as they pursued  SEC permission to create a third-party platform to trade existing notes.

      After the quiet period ended, Prosper introduced a new Rating System which classifies borrowers’ probability of default in 2% increments ranging from  AA (safest, 0%-2%) to HR (riskiest, 14%-16%).  Thirty funded Prosper loans in each of the seven ratings will be chosen for analysis, with 15 loans originated before the quiet period began and 15 after.  A lender on this loan with the same average Rating as the borrower will be selected (ex.:  a “B” borrower will be matched to a “B” lender on her loan).

      The tentative model is specified as follows.  The quantity demanded (average bid size) is a function of the interest rate and the surplus loan supply (ratio of accepted /total bids).   The surplus loan supply is intended to reflect the relative attractiveness of the loan listing from the point of view of potential lenders.  The quantity supplied (average loan size) is a function of the interest rate and the Herfindahl index.  The Herfindahl index captures the relative degree to which a given lender’s funds are concentrated in just a few Prosper Ratings.

Results/Expected Results

      In the demand equation, both the interest rate and the surplus loan supply are expected to have negative coefficients.  In the supply equation, both the interest rate and the Herfindahl index are expected to have positive coefficients.

Discussion

      It is hoped that this initial foray into the Prosper Marketplace microdynamics will yield some interesting conclusions.  The “quiet period” has made it possible to observe what effect, if any, Prosper’s ability to provide a secondary market   will have.  While we anticipate that the increased liquidity should increase loan supply, other economy-wide changes have occurred during this period which may confound results.