Reputation in lending marketplaces

For marketplaces to be successful, generating trust between the participants is key. We know this intuitively, for any transaction to occur there has to be a modicum of trust! An interesting way to build trust is via signaling “reputation” which is doubly important in lending marketplaces as there is a required level of anonymity between the participants.  As a product manager, always think like the user :). What are the key concerns of the users (Borrowers and Investors) in a lending marketplace?

For the investor:

  • Should I trust the borrower to whom I’m lending to? How legitimate is this borrower, is this an outright scam?
  • How should I price default risk? As an investor, I understand that defaults will happen. I can compensate for that by demanding a higher interest rate, so that on an average I will still make the return that I need after defaults.

For the borrower:

  • How can I generate trust in an open marketplace system?
  • How do I minimize my interest rate?
  • How do I optimize for time to close? Speed is key for borrowers, how do I signal the interest rate that I’m willing to pay as well combined with my reputation to get funded quickly?

Are these even the right things that users care about, is there any data that indicates that reputation really matters? Luckily there has been some research done in this topic. I would recommend reading this paper ” The Invisible Value of Information Systems Reputation Building in an Online P2P Lending System” . The researchers got a rather larger dataset from a p2p lending site (probably is my hunch, they do not disclose the company in the paper) and built some models to test out certain hypotheses regarding reputation and interest rate sensitivity, reputation and loan default rates and reputation and loan close rates.

A few key conclusions, aka the TLDR version.

Careful inspection of the coefficients for each reputation state reveals some interesting patterns. The intercept for low-reputation borrowers is negative and significant, while the intercept for the high-reputation borrowers is positive and significant. It means that a low  reputation borrower’s listing has a lower chance of getting funding, ceteris paribus. Also the coefficient for amount for low-reputation borrowers is negative and significant meaning that the listings that ask for a higher amount are less likely to be funded. However, interestingly, the opposite is true for the medium and high reputation borrowers.

Reputation does matter and intact better reputation leads to larger loan sizes being funded.

The intercepts in the interest rate model, form instance, are all significant, such that low-reputation borrowers have the least reduction in the final interest rate and the high-reputation borrowers have the highest. Also interestingly, credit grade and the number of delinquencies of the borrowers are only significant in the interest rate model for the low-reputation borrowers.

Reputation matters. as a borrower, if you want a low interest rate.

The paper also provides some good data on specific features that provide reputation signaling.

Another surprising finding is the significant and negative sign of the coefficients for the number of questions and answers for low and medium reputation borrowers. This could be because listings that receive more questions are listings that provide less information, so fewer lenders are willing to invest in them, resulting in higher interest rates

Having a mechanism for borrowers and investors to communicate with each other and displaying the results for the entire population of users has a signaling effect.

Whilst insignificant for medium and high reputation borrowers, the coefficient for inclusion of images in listings is positive and significant for low-reputation borrowers and improves their chances of getting a loan.

Images in listings also convey information that enables borrowers to enhance their reputation with investors.

While building marketplace, product managers need a strategy for reputation management. In lending marketplaces the implementation is a bit different than traditional marketplaces as there is a certain required level of anonymity between the two parties (borrower and investor). This adds a layer of complexity as the end goal is to convey signaling information without compromising anonymity. I’ll explore certain reputation management strategies and tactics in a future post. What do my fellow PM’s recommend?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s