Tepom.com

Personal finance advice for the average American.

Tuesday, September 16, 2008

Prosper.com: Convincing My Wife, Part 1

My friend and frequent tepom.com commentator Steve Butcher has an active account with Prosper.com and is doing quite well with it. His current rate of return is about 16.9%. I'm interested in joining him and trying my hand at peer-to-peer lending, however, my wife takes a very conservative approach to lending and will definitely need some "talking-into" if I ever want to open an account.

I will write a small series of posts that will attempt to justify peer-to-peer investing and convince my wife at the end that it's a good idea. The end goal of my analysis will be to prove my hypothesis that Prosper.com provides an excellent lending opportunity by using their empirical loan data. If my hypothesis is proven, I will convince my wife to start investing and share with you all my plan of attack.

Day one analysis: The correlation between debt-to-income ratio and loan delinquency
Prosper.com evaluates borrowers' credit reports and assigns them an alphabetical risk rating, from AA (the best) to HR (high risk -- the worst), with A, B, C, D, and E in the middle. In today's analysis, I decided to evaluate the correlation between a borrower's debt-to-income ratio (DTI) and the delinquency rate for each credit classification. My analysis assumes that any amount that is not current (>15 days late) is considered delinquent. It should be noted that these numbers do not suggest a loan default, just the probability of delinquency. Here I evaluate both the number of default dollars and the number of default loans. In my opinion, the number of delinquent dollars are more important than the number of delinquent loans because we intend to diversify our money as much as possible. The more you diversify your money, the more important the number of deliquent dollars become. The less you diversify, the more important the number of delinquent loans become.

Though my analysis evaluates delinquencies as a whole, it should be noted that those with poorer credit that go into delinquency are much more likely to have their loans written off than those with better credit. Essentially, those that have better credit are better at getting themselves out of trouble.

Click a chart to enlarge:
Looking at all loans regardless of the borrower's DTI (the blue line), it comes as no surprise that as the borrower's credit rating declines, there is a higher probability of the loan becoming delinquent.

However, when three bands of DTI are considered, the equation changes. With a DTI of anywhere from zero to 50%, the curve doesn't shift or change significantly for any classification of borrower. However, with a DTI above 50%, the curve is flipped nearly upside down. This suggests that a dollar loaned to a person with great credit but a high DTI ratio is more likely to be late than a dollar loaned to a person with poor credit and a similarly high DTI.

Similarly, when looking at the number of delinquent loans, the curves are closely aligned for borrowers with DTI less than 50%. But once the DTI rises above 50%, the difference in likelihood of the loan's delinquency rises substantially, especially for borrowers with great credit.

To summarize, the number of loans that will become delinquent do not vary significantly as the borrower's DTI changes except when the borrower has great credit. A borrower with great credit and a high DTI is much more likely to become delinquent than a borrower with great credit and low DTI. A borrower with poor credit and a high DTI is just as likely as a borrower with poor credit and a low DTI to become delinquent on his or her loan. When you consider a dollar loaned, a dollar loaned to a person with great credit and high DTI is much, much more likely to be late as opposed to a dollar loaned to a person with great credit and a low DTI. The reverse is true for poor credit borrowers. A dollar loaned to a person with poor credit and a low DTI is more likely to be late than a dollar loaned to a poor credit borrower with a high DTI.

So what do we learn from this? When lending to people with great credit, be wary of those with a high DTI requesting a lot of money. Don't lend to them unless you can obtain a premium rate. When considering lending to a borrower with poor credit, be sure to diversify your money across different loans. If you're torn between two poor credit borrowers, if the only difference between their profiles is their DTI, consider lending your money to the one with the higher DTI, especially if their rate is higher!

More to come on other Prosper.com lending strategies.

Labels: , , , , , , , , , ,

3 Comments:

  • At September 16, 2008 10:47 AM , Blogger Steve said...

    As Scott said, I do currently use Prosper.com as a low level investment. I haven't put much into it yet, just enough to try it out. Compared to my stocks, 401(k), and high interest savings, Prosper has been by far the best investment - currently the only one other than savings with a positive return due to the market downturns. However, it is not a liquid asset like savings or almost liquid like stocks - it has to be money I'm not going to need to access in a bind. Judging by my 401(k) performance, I am considering shifting some of my regular retirement investment to Prosper. I'd love it if Scott did an analysis on liquidity of assets to help me figure out how to best use Prosper (hint, hint).

    Your quick default analysis is pretty accurate in my experience, and agrees with Prosper's analyses. Out of 11 loans, I have 1 that is about a month late - and they have a high credit rating and high DTI. The big red flag on that load was that on the listing, they didn't list a stated income (which is verified by Prosper), but did list a high income in their loan description. That was shady, but since I am just trying Prosper out, I invested in a full range of loans, 1 or 2 each of every credit rating AA to HR.

    One interesting thing to note is that when you bid, Prosper gives you an estimate on your return based on typical performance of similar loans, essentially building Scott-like analysis into the bidding process. This is done by subtracting from the rate you are bidding on. For someone with good credit, DTI, etc, they might only subtract 1-2% from your bid rate to estimate your return. For a less ideal borrower, 10-20% or more might be subtracted, sometimes enough to make your expected return negative - a good indicator that the rate you are bidding is not worth the risk associated with the borrower.

    I'll post more about my Prosper experience as Scott continues his analysis, and will try to keep it mostly on topic with his posts.

    Steve
    stevescookingjournal.blogspot.com
    www.iHateWheat.com

     
  • At September 16, 2008 11:45 AM , Blogger Scott Bliss said...

    Thanks for the comment, Steve. I didn't realize that they'll give you an assessment of your risk at the time you bid!

    It's interesting too how you can be distracted by the borrower's face and story!

    More analysis tomorrow...

     
  • At September 19, 2008 7:49 PM , Anonymous OnTheFence said...

    Drop the saucer eyes from Steve's estimated ROI. His average portfolio age is only 173 days. A more realistic expectation of what his ROI will really be like won't show up until his portfolio matures to an average of 365 days.

    And based on his current lending strategy, it's not going to be so rosey.

     

Post a Comment

Links to this post:

Create a Link

<< Home