Andrey Kamenov, Ph.D. Probability and Statistics
For a regular investor, the single most important field in the Prosper dataset is the Prosper borrower rating. This is an estimate of a borrower’s creditworthiness based on various other metrics, including employment length, number of open accounts, prior delinquencies and many others. Based on the aggregate score, the borrowers are then split into seven groups: AA, A, B and so on. Alternatively, an investor may use numerical ratings ranging from 1 to 7, with 7 corresponding to the safest, most trustworthy borrowers.
How good is this rating? You can think of it as an estimate of the default probability, so it would be natural to assume that out of any two borrowers, the one with the higher rating would be less likely to default. Indeed, take a look at the following chart:
As you can see, the larger the difference between two borrowers’ ratings, the more likely it is that the person with a lower rating will default first.
Another advantage of this visualization is that it allows us to see how important the differences between ratings are. The borrowers rated “AA” (or 7) are actually in a class of their own, with significantly lower default rates even compared to the next grade, “A.”
On the other hand, as you can see, the difference between grades 1 and 2, as well as 3 and 4, is barely visible (although still correctly estimated by the rating).
Where do the high-graded borrowers live? The highest percentage of all borrowers from a single state is registered in Alaska (26 percent), followed by Delaware (25 percent) and Illinois (24 percent). Several other states, including California and New York, show numbers around 22 percent.
Only 13 percent of all borrowers receive one of the two highest ratings in Alabama and Mississippi, and the lowest percentage is seen in New Mexico (11 percent).
Source: Prosper downloadable data
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:
I usually find your graphics compelling, but really didn’t understand this graphic. I understand your text fine, just don’t understand first lender vs second lender.