Monday, October 30, 2006
Prosper.com Borrower Policy Trumps My New Years Resolution?
Recently, I have been making initial bids on these loans 3-5 days prior to their closing. You ask, why 5 days or less? Fundamentally, I believe in just in-time bidding/funding. This strategy best helps me minimize the amount of unobligated funds in Prosper's 0% interest bearing accounts.
Recently, i've been noticing that some non-auto bid loans have been funding and closing in well less than 10 days. Here's an example:
Loan 53970.
This loan closed within about 2 days and 15hrs of opening (the typical period is 10 days). Frustrated w/ the lost bidding opportunity, I called Prosper.com. The Prosper rep notified me that borrowers have an account feature that allows them to accept a loan interest rate at any time once the loan reaches 100% funding.
Those lenders who piled in early on this loan, got a loan w/ an ROI of 14.6% (a respectable rate). Now, I'm in a quandary as to whether or not embark on my above New Years resolution. The one silver lining to the New Years resolution is that I'd avoid some borrowers who are in such a hurry for funds that they may have underlying emergent issues (hidden risks) not clearly communicated in their loan requests. Of course, this is an inherent risk that may still exist in auto-fund requests that I continue to fund.
Hmm, what will I do? I'll likely stick w/ my New Years Resolution and leave it to my readers on how they time their bids.
Prosper Lending: A Big Surge of Borrowers About to Hit the Market
It then dawned on me that the Christmas/New Year season will be a good period for lenders. Here's why:
(1) Who ever came up w/ "It's better to give than receive?" LOL, perhaps a retailer, i'm not sure. However, many people are going to spend more money than they have on hand. This will create more shortfalls when emergencies hit.
(2) Some of the other lenders you're competing against may get a little bit shorter on funds if they're not giving frugally, but making a splash w/ the purchase of big Christmas presents for others (perhaps a big flat screen for themselves).
(3) Other people may be making New Years resolutions that cost money.
(4) Others may be spending too much money on travel to see family this season.
(5) Others may be spending too much money on Christmas/New Years parties.
You should expect the beginning of a surge of Christmas related listings starting about 10 days prior to Thanksgiving. The one above may be one of the first ones.
I haven't yet decided if i'm bidding on the listing above. I've emailed her a couple of questions and am waiting on a response. The one challenging thing for her is that English must be her second language after I suspect Mung/Lao or Thai. I hope she doesn't think it's smart to "SAVE" at 25.48% interest. It's only saving if she would have otherwise spent the money on a 32% Bank of America or Chase credit card. If that's the case, on the one hand I applaud her for saving about 6.5% but on the other hand, she needs to work on money management.
Saturday, October 28, 2006
My Prosper Loan Portfolio Performance
Bottom line: Prosper over states my loan portfolio performance by about 3%. Instead of earning a 20.28% return, I've earned an annualized yield 17.1%.
My Prosper.com loan portfolio (Portfolio display provided by Eric's Credit Community).
a) Number of loans: 85
b) Total Loaned: $4,320.23*
c) Number delinquent: 2
- Loan 314XX: 2 months late
- Loan 311XX: 7 days late
*Notes:
1) Prosper states my effective interest rate is 20.28%
2) Prosper doesn't factor in the impact of group rewards and other fees. Instead, THE SITE MISREPRESENTS effective yields. YOU HAVE TO CALCULATE THIS ON YOUR OWN.
3) Meanwhile, if the delinquent accounts in paragraph "c" actually default, my effective interest rate would be reduced to 16.7% and 16.3% respectively (approximately 0.4% reduction for each default).
4) In calculating my effective interest rate, I assumed all interest income would have been reinvested at an effective 17.1% interest rate. I also ignored the 0% periods where interest income sits w/o being in an active loan.
5) When using "Eric's Credit Community," one will almost always get slight irregularities in the reporting of "Eric's Credit Community" and Prosper.com.
Can bid style (competitive bid vs. autobid) on Prosper.com help determine likelihood of late payment?
Perhaps, people with higher likelihoods of making late payments or defaulting on prosper loans are more desperate and in greater need for fast loan funding. As a result, borrowers may use the autobid feature to help ensure their loan request(s) get funded.
Like my last analysis of pictures in delinquent loans, I went back to the prosper borrower stats at "Eric's Credit Community." I looked at all of the group sponsored delinquent loans and found that there were 161 delinquent group loans with valid hyperlinks.
Of these, I found that there were only 36 competitively bid delinquent loans and a whopping 125 delinquent autobid loans. ALL OTHER THINGS EQUAL, it appears that the bidding style has some merit in determining likelihood of late payment.
Based on the above numbers, late payments occurred with 347% greater frequency when the borrower used the autobid feature.
Disclaimer: This analysis is based on ALL OTHER THINGS EQUAL (ceterus paribus for statisticians). In all likelihood, there are a number of other extenuating circumstances that also contribute to greater likelihoods of late payment.
Recommendation: Consider using this as a supplement to prosper.com credit report analysis.
Thursday, October 26, 2006
Can Photos on Prosper Loan Requests Help Determine Default Risk?
I did some back of the envelope math and found some possible evidence that pictures matter on Prosper loan requests.
I reviewed the prosper borrower stats at "Eric's Credit Community." Eric shows that there are 210 prosper loans that are late in payment. The accuracy of this number is limited in part due to the fact that his number of late non-group affiliated loans is based only on what is reported to him by prosper lenders.
Anyways, I looked at pictures posted (or not posted) on each one of the late loans and found that loan requests without pictures of the borrower occurred 31% more frequently than requests w/ one or more pictures of the borrower.
Pictures are only a piece of the puzzle available to lenders. I recommend continued use of credit report analysis and email dialogue w/ borrowers to determine favorable risk-reward lending cases.