Re: Miserly Bastard about Prosper Lending
Miserly Bastard over at Yet Another Blog about Money had a post a while back that was critical about prosper. He raises some good issues, but I don't agree with all of them.
He raises six negatives about lending on prosper (bold are his original points):
Lack of real diversification
Although Prosper gives a nod to diversification, the truth is that a portfolio of Prosper.com loans is very poorly diversified...When credit markets talk about diversification of default risk, they are talking about diversification across thousands of borrowersI agree with him on this one. People need to understand that to achieve any measure of diversification they need to be constantly reinvesting the proceeds from the prosper payments they receive monthly.
Non-transparent risk of default
Prosper puts a lot of faith in the FICO score of the applicantThe reason why the entire loan industry puts faith in FICO scores (or some variant) is because they work. Research has shown that while people's FICO scores vacillate over time, higher FICO scores are more stable. The caveat, of course, is that once again the sample size needs to be large enough. For all this stuff to wash out, you need a large pool of borrowers.
He also makes the point that a good FICO score is easier (that you would think) to achieve. Age is taken into account into the FICO scores. Plus, defaults are mitigated by the collection agencies that are employed.
self-reported income, meaningless debt ratios and fraud
These two I decided to lump together. I agree with this that there is some measure of risk that people will lie, cheat, and steal and underscores once again that diversification is key as well as developing some sense of what interest rate to choose from that incorporates these aspects.
irrational interest rates
Borrowers dont seem to understand that they need to offer a significant spread over the risk-free 3-year rate, and lenders dont seem to realize that chasing double-digit returns by courting high-risk borrowers is like taking a tiger by the tailGood point, but the market will figure this out. As more sophisticated participants take part in the auction, the interest rates will become more rational over time. That's what free markets are for, right :)
lender of last resort
Maybe, maybe not. I think time will tell whether or not this idea will fly in the credit market-place. Perhaps a site like zopa will prove to be more successful. Who knows.
MB has made some very good points if you're thinking about or have been lending on prosper.com. Diversify as much as possible and have some sense of how to price a loan. My formula is a good starting point, but IT IS NOT PERFECT. You need to tweak it for your assumptions and risk-tolerance. Better yet, come up with your own formula, just try to do it somewhat intelligently. Also, rates have been rising, so your prosper rate need to be rising as well.
Excellent points MB. Looking forward to your sound rebuttal of all of my responses :).
Tags: personal finance prosper prosper.com
4:07 PM
|
|
This entry was posted on 4:07 PM
You can follow any responses to this entry through
the RSS 2.0 feed.
You can leave a response,
or trackback from your own site.
Subscribe to:
Post Comments (Atom)
12 comments:
Credit markets are too efficient for a company like Prosper to exist.
The best proof I can point to for why Prosper is an inefficient marketplace that mis-prices risk-reward, is to point out some of the more ridiculous auctions currently underway. For example:
- the E-Rated borrower fully funded for $2k at 16.45%
or
- the D-Rated borrower halfway funded for a loan of $5800 at 15%
Why on earth would lenders loan to such high risk borrowers without a commensurate interest rate premiums? Before you say that these guys are offering high rates, consider that these borrowers, if they were going through traditional credit channels, would probably qualify for 18-21% interest rates (ie, credit cards).
Prosper lenders are systematically underpricing the risk... and these auctions might even close lower.
Think about it this way. Credit markets in the US are incredibly efficient. There is a surplus of capital, and indeed, an entire industry of sub-prime lending exists to service high risk borrowers.
Enter Prosper.com, where marginal borrowers think they can get lower interest credit. If guys with D, E or NR FICO scores are coming to Prosper and getting funded, it is only because the lenders are taking risk for which they aren't being compensated. My theory is that these lenders (most of whom are unsophisticated, I'd bet), are simply chasing nominal return.
The only counter argument is that the credit industry is reaping irrationally high profits, and Prosper is just disintermediating them. Ther might be some truth to this, but even still, I am skeptical that Prosper lenders can price the risk correctly.
Or consider the flip side, the AA and A rated borrowers. The 3-year risk free rate is around 5.5% today, based on what you could get in a CD. Yet on Prosper, you routinely see A and AA rated borrowers offering only 6-7% interest rates. Now subtract the 25 basis points annually that Prosper charges, and your real rate of return is closer to 5.75 - 6.75%.
Do you think you're being properly compensated for the specific credit risk you face?
Ebay works because the transactions costs to put fragmented sellers & buyers together is very high. Particularly with things like collectibles, Ebay is the perfect market because the search costs for buyers and sellers is insurmountable. This is why Ebay can justify such high fees. (All in, the transaction costs on Ebay are probably as high or higher than those for Christies or Sothebys.)
Credit markets are different. They are already very efficient. Low risk borrowers have great access to capital, and can get teaser rates, low permanent rates, etc. that lenders on Prosper can never match. Conversely, high risk borrowers are priced according to their risk profile and granted credit terms that reflects economic reality--think paycheck advances, pawnshops, secured credit cards, subprime lending, etc.
Prosper is a marketplace where risk will never be efficiently priced in comparison to the liquidity in the existing credit markets. Somebody is always going to get screwed, and my bet is that 9 times out of 10, it will be an unsophisticated lender who thinks he is getting a "great return" when in fact, he has just purchased a time bomb.
One final point. While the Dutch Auction is a neat device, there has been a fair amount of behavioral finance research about the problem with outcry auction markets resulting in mispriced outcomes--Google "curse of the winning bidder" if you're curious. I dont know whether Dutch Auction is also subject to the curse of the winning bidder, but I suspect that many lenders compete irrationally for loans as the deadline closes.
miserly bastard, you rock!
On a high level, I agree with the Miserly Bastard. I don't believe the rewards of higher interest rates are worth the risk. That, and I dislike the rigidity of only 3-year loans. Finally, any kind of financial system that relies on completely unverified sob stories and debt ratios to choose between borrowers of equal credit rating is a bit fishy to me. At least ModestNeeds.org, a charity I support that relies on personal stories, verifies each claim to be true.
I do look forward to Zopa, however, which I believe spreads the risk better and does not involve personal stories.
Hi - Dave from Zopa here.
I (we) think this is a very important debate - we believe that there are inefficiencies in the credit market (although not at the prime end), and over time, we want to move Zopa away from our current prime focus and into that more interesting segment of mis priced borrowers.
We chose to start with credit worthy borrowers and a full 'managed service' credit model (not just relying on a FICO score or UK equivalent) because we believe that an idea as radical as Zopa (or Prosper) needs to prove itself as a safe place to invest. If the first 6 months worth of lenders all end up losing some percentage of their capital, (which our very experienced credit director tells us they would have done if we had relyed solely on, in Zopa's case, an Equifax credit score) then there is no future for the business.
We think that the future of these marketplaces is neither todays Zopa, or todays Prosper, but probably some combination of both. The majority of lenders would choose to use a 'managed service' (think mutual fund) model, and some particularly sophisticated individuals will go for the 'do it yourself' (think individual equity picks) version - where the potential returns (and hence risks) will be higher.
And I agree with the winners curse argument of miserly bastard - damn good point.
Dave--
You've hit the nail on the head.
The problem with peer-to-peer lending at the prime+ end of the credit spectrum is that it is already efficiently priced. Prime+ borrowers can get good deals. Large financial institutions can borrow cheap money, and lend it out at a slight spread, then bundle up these receivables and sell them off their B/S. It's a super sophisticated, efficient market. P2P has no value add here.
On the sub-prime side, there might be juice left to disintermediate the near-predatory lenders. I mean, you really are talking about competing for the business from folks who use pawnshops, paycheck advance, tax return advance, secured credit cards, Providian, etc. P2P could add something here b/c these markets are much less efficient.
The problem is how do you manage the risk when dealing with this end of the credit spectrum. The std. deviation of return on this end is undoubtedly much higher. Some lenders at sub-prime do this by secured lending--i.e., tax return advance, pawnshops, etc. On the unsecured lending side, guys like Providian make their money through unethical business practices, hidden fees, exorbidant interest rates, etc. Plus, they have the ability to get true portfolio diversification across a large group of borrowers (no specific risk), and the probably hold mostly market risk (e.g., recession, unemployment, etc.) They can also collateralize these receivables and sell them.
What tools does a P2P lender have? Even under a "managed" lending program? And what's the cost of the management fee? Can Prosper/Zopa manage subprime loan portfolios for 25 or 50 basis points?
At the end of the day, the problem is that if you want to play in the subprime side of the spectrum, you cant go with razor thin margins. The standard deviation is simply too high at this end, so you need to bake in higher interest rates.
When P2P lenders start giving up this juice, they're quite literally skating on progressively thinner ice.
PS--A more interesting auction method would be the sealed, second-bid price auction. However, I doubt that unsophisticated lenders are capable of using this method, and you'd still end up with them underpricing risk.
Interesting arguments about the wisdom of lending via Prosper.com. Any thoughts on the wisdom of borrowing through Prosper?
I'm only half a step above "payday loans and pawn shops".
Borrowing on Prosper is a freeroll for you. Go for the lowest possible interest rate you can get. See if somebody will fund you despite your various negative factors. You have no downside, so you should try to get the cheapest possible credit you can get, after all other lenders have rejected you. Do not overcommit to high interest rate, unless you don't fear the consequences of default, in which case you should borrow as much money as possible. Im being serious, by the way.
Shhhh! I hope you don't mind but this is too good to keep quiet about and I need to tell someone. I have found a website that is giving out FREE information about credit repair
Efficient? How many times are typical loans bought and sold with profit taken at each transaction? How much money is spent by lenders on marketing, spam mail, credit card rewards and review?
Clearly far more resources are spent on that then the basic automated services prosper.com performs to enable its p2p loans and clearly there is room for prosper in the marketplace.
How much room and who is going to benefit the most, lenders or borrowers we will have to see.
Efficient? How many times are typical loans bought and sold with profit taken at each transaction? How much money is spent by lenders on marketing, spam mail, credit card rewards and review?
Clearly far more resources are spent on that then the basic automated services prosper.com performs to enable its p2p loans and clearly there is room for prosper in the marketplace.
How much room and who is going to benefit the most, lenders or borrowers we will have to see.
As a Prosper lender, there are many ways to mitigate risk without sacrificing too much reward. My median borrower is a "D" credit risk with no more then 2 delinquencies in the past 7 years, 0 defaults and a 20% borrowing rate. Many are groupmembers and I have a few HR. I am spread ($50-$100 per) over 40 loans and reinvest all payments, though it takes 5-7 days to reinvest. Though only 26 of my loans are over 2 months old, I have no late payments. Call prosper inefficient all you want, but at least for me it is paying off, at least initially. Those who are less careful about who they lend to, or invest at rates less then 10% (where same money can be made with less risk in other investments) will have unfavorable returns.
Irrational (or mostly irrational) risk pricing is now, a year and a half later, on the way out, with the introduction of historical performance based bidding guidance by Prosper at the end of October, 2007. IE, the unsophisticated can now "fake it", at least to an ever-increasing degree... because there's a pop-up that smacks their hand if they bid too low (at least in the context of past performance of actual prosper loans).
Two months out from this change, and the data shows significant trends towards more efficient risk pricing (ie, the rates are higher on the loans that fund, and the loans that do fund are of higher credit quality). It appears to be slowly taking shape...
Further analysis (and charts of marketplace data which support above statements) on this at: Prosper Lenders.
Post a Comment