New Analysis Site Prosper Stats Launches for Investors

This is the second in my weekly series of new p2p lending analysis sites and today we turn our attention to Prosper. Since Lendstats went down in July Prosper investors have been left in the dark with nowhere to go for data analysis. Until now.

Prosper Stats launched last month and after playing around with it for the last couple of weeks I think it is even more useful than the old Lendstats site. Rocco Galgano is the creator of the site and he decided to build it after he went looking for a Lendstats alternative for Prosper investors and could not find anything. He initially did it just for his own use but soon realized he had created something others would value.

Methodology for Calculating ROI

When you go to the main page at Prosper stats you see some search options and a count of all the loans along with many fields including the all important estimated return. The overall estimated return (for all Prosper 2.0 loans) is seems to be slightly higher than Lendstats ROI numbers so I asked Rocco to explicitly detail how he calculates this number. Here is his response:

What I did was calculate the average percentage of loan value that is lost during a default (this was about 85%). Then I calculated how often a loan goes to charge off after going late (calculated this separately for each late value). I multiplied the two together to come up with a loss estimate. Then I lowered it a little to compensate for charge offs that eventually get paid, then finally I went ahead and rounded stuff to even numbers to make it easier to work with (this is of course just an estimate).

The numbers I am currently using in my calculation for chances of defaulting based on late status are:

Late: 60%
1 Month late: 85%
2 Month late: 90%
3+ Month late: 95%

I multiply that by 85% for the average amount of principal lost then I multiple that by the principal remaining to get my dollar amount loss estimate.

So if you had a loan that had $10.00 in principal and it was Late, I would estimate a loss of $5.10 ($10.00 * .6 * .85). The same loan at 3 Months late would be an estimated loss of $8.08 ($10.00 * .95 *.85). Anything that has defaulted gets completely written off.

Maybe some of the mathematicians here can comment on these calculations but the end results seems to be within a percentage point of where Lendstats was. Regardless, it does provide a way for investors to compare filtering strategies – something we have been unable to do for several months now.

A Comprehensive Set of Tools for Investors

There is a lot to like about Prosper Stats. Many of the favorite parts of Lendstats have been included plus several unique and useful new features. Rather than list them all out here I decided to do a quick video so people can get a more hands-on idea about how this site work. There are three main components of Prosper Stats (Loan Analysis, Previously Late Loans and Top Lenders) and each section is covered in the video below. If you can’t see the video below here is the URL on Youtube.

[leadplayer_vid id=”510966BE32991″]

So what you do you think Prosper investors? Please let me know in the comments below.

Notify of
Newest Most Voted
Inline Feedbacks
View all comments
Oct. 13, 2012 7:21 pm

This is terrific. Thanks, Rocco for sharing this. And kudos to Peter for the video review. Now I don’t have to pick the loans off my memory of Lendstats’s Prosper data prior to its shutdown.

Oct. 14, 2012 1:59 pm

Do you use some type of “scoring system” to decide how much to allocate to each note?
Besides the known parameters available through this new tool, do you have any other criteria, such as “intuition” that you factor into the equation? In Prosper, there are fields that a borrower can fill out <> and <>. Do you assign a higher “score” to those that fill this out and it makes sense vs. those who leave it blank? Thanks, Mark.

Oct. 22, 2012 7:56 am
Reply to  Peter Renton

Peter, for some reason the web site cut out my text. The text in quotes in my post was Purpose of Loan and My Financial Situation. Some people fill it out, some people don’t. What I was trying to ask is do you weigh people that fill it out a little higher than people who don’t?

Oct. 14, 2012 2:30 pm

Interesting observation:
A lot of so-called “experts” suggest excluding the State of California. According to the new tool, this is incorrect. The tool is calculating 13.36% (all my criteria + CA) and 12.97% if excluding California.

B. Mason
Oct. 14, 2012 8:39 pm
Reply to  Mark

States that were adversely affected by the housing bubble (FL/CA) have been declining in their negative effect on loan defaults over the last year.

Oct. 15, 2012 1:53 am
Reply to  B. Mason

Yeah I know, but that’s what my question was about… looks like statistics don’t reflect that. The new tool mentioned in this post calculates 13.36% (all my criteria + CA) and 12.97% if excluding California USING DATA FROM THE PAST YEAR.

Oct. 15, 2012 12:24 pm
Reply to  Peter Renton


Any word from Rocco that he would include loan purposes and borrowers’ field of employment in the future? I have found these to be important criteria for loan selection.

Oct. 16, 2012 2:34 am


I added 4 new search criteria today that Propser had but I didn’t. They are Loan category, Min expected return %, Credit score change, and Occupation. I used the same names that Prosper uses on their site for consistency, but Occupation and Loan category should be the two you are looking for.

Oct. 17, 2012 7:57 am
Reply to  Rocco

Thanks again, Rocco. You are a lifesaver!

Oct. 17, 2012 7:54 am

Thanks, Rocco – great tool. Thanks for the video, Peter.

When using the “previously late” filter to find and then sell notes, what is the fastest way to put the note up for sale on folio fn after you find it using your “previously late” filter? It seems like folio fn doesn’t let us search for a specific listing number and I have been scrolling through folio fn to find the loan. I’m not sure if that is the fastest way and thought you might have found a better solution.

How would you quickly/efficiently go about selling one or many notes that you find on your “previously late” filter?

Oct. 17, 2012 8:58 am

I spent a bit more time using your previously late tool and it’s very helpful. When selling previously late notes, searching by date within works well and I noticed 2 things:

1.) makes you search for notes by “purchase date” – slightly different than what prosper-stats shows you, which is the “origination date”

2.) shows you the loan# and the note#, which seem to be different than the “listing number” on prosper-stats, so I have been focusing on the loan size to identify the loan

Again, thank you Rocco for this great tool. Any advice/feedback on the best way to use it is greatly appreciated!

Oct. 17, 2012 11:51 pm
Reply to  Thomas

Sorry, no real useful advice/feedback. As Peter mentiond Prosper does not provide the note id so I can’t link directly to the notes like I can for listings. Just like you I go to my note list on Prosper and manually look through matching things up. I do keep track of the notes in a spreadsheet so that I am not looking up the same notes over and over though.