Diggin’ For Gold Around The Portfolios

12/7/07 posted by Mike from Prosperous Land

When Prosper created their portfolio plans, they added a tool to help new lenders enter the marketplace. It assembles a portfolio of loans whose only objective is to track the average loan return in the specific market slices, much like index funds have done with the stock market. Want to play it conservative, you buy a S&P 500 Index. Feeling aggressive? Try the EAEF Foreign Stock Index. If a lender is interested in tracking the loan market, new lenders can establish portfolios without the learning curve of manual bidding.

Those of us who take a managed fund approach to our lending by using manual bids will have to be careful because the portfolio plans will change the bidding environment. If the portfolios are widely adopted, we should expect loans within the portfolio’s umbrella to be in higher demand. Higher demand for certain loans will result in lower interest rates for those loans, lowering the possible return on investment. A similar effect has been found when the S&P 500 adds a new stock to the portfolio.

At the same time, it creates opportunities to find nuggets that are outside of Prosper’s four portfolio plans. This is risky territory (otherwise it would be inside the portfolios), but risk is a problem for lenders when they’re not paid appropriately to take on the risk. Careful efforts to mitigate risk can drive down the net defaults and appropriate bidding can compensate for the remaining risk. For example, non-autofunded C grade loans with 0 delinquencies and 2 to 3 inquiries are more dangerous than those with the balanced portfolio preferred 0 or 1 inquiries (-9.05% vs. -5.02% net defaults respectively). If you take those 2 to 3 inquiry loans and further constrain bidding based on other risk factors like DTI and loan amount ( DTI less than 20%, amount less than $10k), it’s possible to find listings that have better net defaults (-3.97%) than those found in the portfolio criteria.

This was just one of many opportunities. Prosper doesn’t add these more finely tailored criteria to the portfolios because there aren’t enough listings to support the volume of loans portfolios demand. For individuals willing to dig deep around the periphery however, it’s a veritable gold mine.

—

Mike writes for Prosperous Land.

Posted in Lend To Others, Prosper Lenders

10 Responses

PersonalLoanPortfolio | December 7th, 2007 at 9:48 am

Mike, that is a great summary of that functionality.

Prosper, I like the look and feel of the new forum. Very clean and crisp!

LoanChimp | December 7th, 2007 at 10:26 am

Interesting post, but I’m still grasping for the effectiveness of your strategy.

The Performance Page is a very usefull tool, but also very dangerous to use when comparing different strategies. If you’re going to make a comparison to a portfolio plan, you should use the same date ranges and observation date.

What I’m getting at is this: When I run the performance page with the link you provide (DTI less than 20%, amount less than $10k), and keep the same date ranges and observation date, then *change* the criteria to mimic that of the “Balanced Portfolio” C slice (with 0 DQ and 0-1 inq), the default rate is indeed slightly higher for the portfolio plan C slice, but because the average interest rate is also higher, the net ROI figure is actually *better* for the portfolio plan C slice compared to your suggested criteria. (9.86% for the PP vs. your 9.45% 2-3 inq search)

Moral of the story? Do your own research and don’t believe everything you read on the internet…

RateLadder | December 7th, 2007 at 10:38 am

I completely agree… I wrote a post with a similar take a couple of weeks ago…

When digging in the periphery… high rates make up for a lot of sins…

http://www.rateladder.com/2007/11/18/portfolio-plans-the-good-the-bad-and-the-ugly/

Mike | December 7th, 2007 at 3:15 pm

@LC: You got me there. I would argue that this is the historical rate before the lender bidding guidance and portfolios entered the scenes. The portfolios don’t buy not-autofunded C loans with 2+ inquiries (reducing demand), and the bidding guidance gives a 13.58% net default rate. I would expect the average loan interest rate to rise for newly funded loans in the 2-3 inquiry group going forward.

LoanChimp | December 10th, 2007 at 2:03 pm

@Mike, I agree.

I would also think that for every bucket there will be certain loan types that outperform the net default rate provided by the bidding guidance, even those included in the Portfolio Plans.