Prosper Loan Types vs. Returns

Thought I'd look at ROI by loan type on Prosper, as well. Once again, these numbers are from Lendstats.com.

And, for comparison, the Lending Club data I posted a few days ago.

Story Telling

A very interesting article about the effects of story telling on loan rates during the Prosper 1.0 time period out of the University of Delaware and Rice University.

They analyzed each description in terms of the identities it portrayed, including being trustworthy, successful, hardworking, moral, religious or having economic hardship.

They found that the more identities portrayed, the less likely a loan was to default:

The more identities the borrowers constructed, the more likely lenders were to fund the loan and reduce the interest rate but the less likely the borrowers were to repay the loan – 29 percent of borrowers with four identities defaulted, where 24 percent with two identities and 12 percent with no identities defaulted.

This seems in line with the data I've seen in my Needs Series. It really does seem like we might be able to pull out useable information for determining future defaults from a Loan's Title and Description.

Thanks to lovinglifestyle on the lendstats forum for publishing the link.

So many new things...

Feels like Christmas morning at Prosper... I see:

• Automated Quick Invest
• A view of their annualized return calculation
• Pretty, personalized charts
• Lender promotion percentages

Now we spend the rest of the day playing. :)

More Loan Types vs. Returns

Expanding on my charts from yesterday:

This is all ROIs for Lending Club, split by loan type, for loans starting between 2008 and 2010. It looks to me like Credit Card and Debt Consolidation types were the only types consistently above average. Everything else looks to have bounced around:

Loan Type vs. Returns

One of the things that struck me about Ken's blog post that I referenced yesterday when I looked at Inquiries vs. Returns was that some loan types performed so much more poorly than others. I know when I first read it I thought it seemed unlikely. Indeed, even now my investing criteria don't filter by any particular loan type.

In his post, he saw that credit card loans had the highest ROI, just over 5%. Business loans had a low ROI just under -3% and education loans had the worst return on investment at under -4%.

I looked at Lending Club loans issued from 2008 through today to see if this trend continued or was just random noise and found these results:

It's worth noting that there are no new Educational Loans for 2011. But, indeed, year over year we see educational (blue) and business (red) loans performing more poorly than the average loan (grey) and Credit Card loans performing better than average.

At the moment it looks like the ROI is converging, but this could be due to the recency of many 2011 loans. It will be interesting to see a similar graph a year from now and compare.

Inquiries vs. Returns

Just over a year ago Ken from lendstats.com made a graph showing Returns vs. Credit grade (among other interesting graphs.) I found this fascinating and since I read it I have started investing in listings with only small numbers of inquiries.

I wanted to see if this trend is continuing so, using lendstats' complete performance breakdown I put together the following graphs for both Prosper and Lending Club:

To me it looks like, in general, the trend is holding true. Loans with more inquiries appear to have a lower ROI than loans with fewer inquiries. There may be some bobbling at 5+ inquiries but there seems to be much more variation. Indeed, looking at these four graphs the deviation for the 0-inquiry loans is always the smallest and, as we move to more inquiries, the deviation seems to get larger and larger.

It's also nice to see, in Lending Club loans, how the 0-inquiry loans appear to gain a 1%-2% ROI each year that passes for the past 4 years. Looking at the spread between all loans and D's and below, it looks like much of that gain has come from the lower-rated loans.

AI Series: The Wrong Way?

Up until now I've been imagining a computer learning algorithm to determine whether or not I should invest in any given loan. I'm starting to wonder if that's the best question I can be asking.

I'm a small lender, without a lot of money invested into either Prosper or Lending Club. As my loans are paid off, I reinvest that money in more loans. I wonder if a better question for an AI for a lender like me would be:

Given that I have $200 to invest and these are the open loan requests, which requests will likely yield the highest returns on my investment.

Of course, if I look at the problem this way, there still needs to be a way for the program to determine that my money is better left uninvested to wait for another set of listings.


Incidentally, for those interested in Machine Lending, Smart Peer Lending is doing a series on their own efforts to make a Recommendation Engine. Very interesting stuff.