Monday, February 2, 2009

Prosper.com - 02/01/09 late loan stats update

Here's the February 1, 2009 update to to my Prosper.com late loan statistics charts.

These charts show statistics for the performance of all prosper.com loans. Each curve represents the set of loans that were created in one calendar month. The vertical axis is the fraction of those loans that have "gone bad", in other words are 1 month late or worse (up to and including default). The horizontal axis is the observation date. All data comes from Prosper.com's performance web page.

Click on the chart to see a larger clearer version.


Here's a chart of the same data in which each curve has been slid to the left to a common origin. The horizontal axis is now days since loan origination month.



Explanation of methodology can be found in my prior postings in this blog, and in forum discussions on the old prosper forum, now archived at www.prosperreport.com

I attempted to update my lendingclub late loan charts this month, but the stats page on the lendingclub web site is displaying bad data. They promise to fix it soon.

The best discussion among prosper.com lenders takes place on prospers.org.

2 comments:

  1. There is a sharp down-curve in one of the lines at the extreme right, near the top. Any idea what that indicats?

    ReplyDelete
  2. The dip on the right edge of the July'06 curve is a glitch in prosper's data. This happens from time to time. Some loans disappear for a day or so and then reappear. Even tho I keep the "observation date" setting on their web site constant, the numbers change slightly from day to day. In the case of July '06, the numbers have already moved back up. On Feb 2, when I read the data, they showed 3 loans in the "1-3 month late" category as of 2/1/09. Today, Feb 3, they show 15 loans in the "1-3 mont late" category as of 2/1/09. This would make that curve consistent with all the others. This correction will show up in next month's graphs. I had an exchange with Prosper about a similar looking bug a few months ago, which they corrected. I believe they have some database cacheing logic that isn't perfect. When I see numbers that look like glitches, I check back later to see if they've changed.

    ReplyDelete