Credit Variables Explained: Employment Tenure

Borrower employment status is a big topic in the P2P investor community.  The reason: we’re all trying to understand the cash flows that will eventually pay back our investments and interest.  As in any investment strategy, it’s important to understand the size of the cash flows (income amount) and the stability.  Another related factor is the amount of debt that the borrower is already servicing from this income.  In this post we’ll discuss employment length, which can be considered a proxy for stability.

Is Employment Tenure Predictive?

First off, Lending Club does not explicitly verify employment tenure, so this variable is self-reported.  As discussed in earlier blog posts, self-reported data is not necessarily bad.  The investor simply needs to understand that it is self-reported and therefore may not predict performance in the same way as verified data.

Despite being unverified, many investment strategies being advertised today by P2P investors include employment length as a component.  Many look at 4+ years, others 5+ years.

Currently, LendingClub filters only enable loans to be filtered up to 5 years of employment, so our analysis lumps all loans with 5+ years of employment together (which includes 52.4% of the current in-funding population.) We did take a look, and the performance is stable across the greater than 5 year population.


Below is the performance based on employment tenure.  Based on these numbers, the stability of income is not as strong a predictor of performance as other variables we've reviewed, but there is some weakness in the lower end.  Specifically, we do not like the “n/a” category.  It isn’t entirely clear what this means, but from the performance below, it isn't good and should likely be omitted from any filtering strategy.  Any further exclusion decisions would have to depend on the investor’s risk/return preferences and expectations.


In Conclusion

LendingClub allows a user to filter out loans with employment tenure < 5 years (by year) on their current platform.  Based on our analysis, we always remove the “n/a” loans given that (1) they have sub-par performance and (2) we aren't entirely clear what this means, and we only invest in loans where the data is clear and understandable (even if self-reported).