Andrey Kamenov, Ph.D. Probability and Statistics
One particularly interesting field in the peer-to-peer lending data provided by Prosper (https://www.prosper.com/tools/DataExport.aspx) is borrowers’ occupations. About a quarter of all loan records (25.4 percent, to be precise) list the borrower’s occupation as “Other” — we will ignore such loans for now.
On the following chart, you will see the distribution of the total number of loans between major occupations (for which there are at least a thousand records).
It is clear that the most popular occupation for Prosper borrowers is “Professional,” due in part to its generality.
What should be interesting for any prospective investor, however, is the safety of loans to borrowers of each occupation. Nationwide, the top three occupations are engineers, analysts and programmers, while sales and clerical workers are at the bottom of the list. The difference is impressive: while 36.5 percent of all loans to people employed in sales were eventually charged off or defaulted, the number drops to 15.7 percent for programmers.
Of course, it would be interesting to see if there is any difference between states. To do this, we group occupations by industry, since it appears that the nationwide numbers are usually quite similar within a group.
To compare the resulting six groups, we calculate their Wilson score: the upper estimate of the actual proportion of defaulted or charged-off loans.
As you may have expected, more than half of the states (which comprise 80 percent of the nation’s total population) show the same result as seen nationwide. Among the states that did not follow the nationwide trend, Pennsylvania is the most populous. The safest borrowers here are those with administrative or executive jobs.
About Andrey Kamenov
Andrey Kamenov, Ph.D. Probability and Statistics
Andrey Kamenov is a data scientist working for Advameg Inc. His background includes teaching statistics, stochastic processes and financial mathematics in Moscow State University and working for a hedge fund. His academic interests range from statistical data analysis to optimal stopping theory. Andrey also enjoys his hobbies of photography, reading and powerlifting.
Other posts by Andrey Kamenov:
This is the holy grail of p2p lending – determining which loans will end up in default. I was unable to get anywhere with your link; perhaps you need to be a Prosper member, and I’ve been using Lending Club for several years. I’d be interested in seeing how do the 36 month borrowers differ from the 60 month borrowers, how much weight should be placed on earlier defaults (maybe once they’ve defaulted they try very hard not to default again?) and also if owners of businesses make for better or worse borrowers.
Please ignore, I didn’t scroll down far enough to see that I could be notified of follow up comments or new posts, so I’m sending this one with those fields checked.