Prosper Revealed

I made my first prosper loan in May 2006 and have funded 40 loans. For about 1.5 years I funded one or two loans a month at $50 per loan. I was watching and learning. My method was to look for loans through solid groups with interest rates of 15% or better and read the stories and ask questions. I spent 1-2 hours acquiring one loan. Prosper lending looked good, and early on, borrowers were fairly consistent at repaying loans.

As I watched people repay loans, I started to realize some things:

1. The group leader and dynamics had little influence on repayment
2. I wouldn’t know if a borrower would be good at repaying for at least 6 months
3. Borrowers who promised they had “learned their lesson” still were late in repayment and non-payment
4. Some borrowers tried to keep up their payments but soon gave up
5. Borrower stories could be partially or completely fabricated and false
6. People’s good intentions and promises can mean nothing and you have no way to know who will be reliable solely from their story, budget calculations, and answers to questions.
7. Credit scores of A and B went delinquent and defaulted about as much as C and D. Credit scores of E and HR and NC (no credit history) were worse.

One day I watched a video clip from Prosper Days. I believe it was their CFO. He made an important statement. He said that the best predictor of someone’s future behavior is his or her past behavior. That had a significant impact on me and I started to change my methods. I became more concerned with reliability of repayment than rate or stories or lender questions and borrower answers.

I began using standing orders which are now called portfolio plans. I experimented with standing orders for months. I would adjust the criteria always looking for a set of criteria that would produce low delinquency rates. I finally arrived at a set of criteria that I felt quite confident would produce good results. I expected to have very low delinquency rates. Within one month of my new prosper lending program, I bid $50 on a loan where the borrower had a B credit rating with a loan amount of $8,500. The loan funded and the borrower has not made one payment. It rattled my confidence in my loan criteria, but I also wondered if the borrower was fraudulent and stole someone’s identity. Despite that bad loan, I still feel fairly confident that my criteria will produce good financial results.

Then Prosper made one of their unannounced and apparently non-field tested changes. They eliminated standing orders and wrapped them into portfolio plans. Essentially, I can accomplish the same thing, but it is much more difficult and cumbersome. I wrote several emails to Prosper politely proposing further refinements and to simplify the portfolio plans. Their responses were formal, vague, uninformative, and placating. That’s when I decided to leave Prosper.

As of 4/21/08, I have loaned over $1800 to 40 loans, realized a net income of over $217, and have the following loan status:

3 – 4+ month late
2 – 3 months late
1 – 2 months late
3 – 1 month late
1 – late
2 – late (<15d)
4 – paid
24 – current

No loans have defaulted. I would like to know why Prosper does not sell the 4+ month late loans or determine they are defaulted. Regardless, my lates account for 12 loans for about $600. If they don’t recover, I may still come out of this thing with a zero loss. But I’m not holding my breath. I expect a loss.

I’m sure other Prosper lenders have better results, but if you watch over time, the lates and defaults continue to climb. See my review of prosper results for results based on average lending behavior. For now, I’ve stopped prospering at Prosper.com.