Learn More *Top*


Credit Variables Explained: Public Records

By Angela Ceresnie
August 12, 2013

Public Records are provided by the credit reporting agencies that lenders use to understand borrowers’ credit history (Lending Club currently uses Transunion, and Prosper uses Experian).  They include financial related judgments, foreclosures, suits, wage attachments, bankruptcies, state/federal tax liens, and past-due child support.  They receive this information from various county, state and federal courts.  It does not include criminal activity; that information is not on a consumer credit report.  These records are generally on a credit bureau for 7 years, although it can vary sometimes (bankruptcies last for 10 years).

Currently, Lending Club does not use Public Records specifically in its underwriting of prospective borrowers, but it does look at FICO which is heavily influenced by the presence of a public record (especially if it is recent.)


Is the Public Records variable predictive?

To answer this, we first need to look at the volume, since we know that Lending Club does aggressive underwriting work up front.  From the below graph, we see that less than 12% of the in-funding loans even have a Public Record, so it’s not a big group to review.  Additionally, 96% of the population that has been funded has 0 public records, so historical data on this population is limited.




Here is the performance of the population with Public Records.  As we know from the above graph, the volume is small and certainly negligible for those with 2+ Public Records, but keeping them separate illustrates an important point.  Obviously, based on the below numbers, investing in a loan with any number of Public Records is likely going to result in a low return.  However, if volumes were ignored, one might believe they should invest in those with 3 or 4 Public Records, because the return on these jumps up to 15%+.  This is where intuition is important when developing a credit strategy.  Why would it make sense that more public records is bad, and what is the likelihood that this is a true reflection of what individuals with 3+ Public Records behave like, and not an anomaly due to low sample size.  We’d venture to say, stick with your gut on this one and don’t invest in these notes.


In Conclusion

Lending Club has some variables in their extract that allow you to parse out Public Records by the number of months/years since they were first reported and their size (e.g. only those greater than $100.)  But, based on the volume and performance we showed above, further dissecting this variable is not likely to produce much improvement to your strategy.  Additionally, those variables are not available on “browse notes”, so implementing a strategy segmenting the Public Records would be a challenge.

LendingClub’s “browse notes” interface allows you to filter based on the existence of Public Records (yes or no).   Depending on your strategy, it probably makes sense to take a look at this variable given how easy it is to use.


Interested in how you can invest in P2P just like a credit risk expert? The Orchard team is building something amazing to supercharge your Peer to Peer Lending results. Join the Orchard waiting list, and we’ll save you an invite!