The deal, and the catch

By Felix Salmon
February 6, 2011
Ron Lieber's column on Prosper and Lending Club misses both the upside and the downside of these peer-to-peer lending platforms, I think.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Ron Lieber’s column on Prosper and Lending Club misses both the upside and the downside of these peer-to-peer lending platforms, I think. Lieber calls these platforms a “gamble”, in his headline, and lists a number of issues with them. For instance, here’s Lieber on Lending Club:

From April through the end of November 2010, the company verified income or employment data on about 60 percent of borrower applications. For the period ending in September, just 65 percent of the borrowers from those files provided it with satisfactory responses. The others ignored the inquiries, withdrew their applications or sent along data that didn’t match the original posting. In the end, fully one-third of the applications didn’t pass muster.

Frightening, right? Mr. Laplanche noted that Lending Club flagged those loans for specific reasons, which would suggest that there were probably fewer errors or lies in the 40 percent of his overall portfolio that he did not double-check. Still, it doesn’t smell quite right. And if you can’t necessarily trust some portion of the borrowers, and the still-young companies don’t have much data on completed three-year loans, which are the most popular ones, this sure seems less a bond purchase than a new type of casino game in Las Vegas.

I’m much less frightened by these numbers than Lieber is. What’s happening here is, simply put, underwriting. Lending Club has an expensive and sophisticated automated underwriting system, on which its entire business model is based. If it works — and there’s every indication, according to performance numbers, that it has worked so far — then borrowers get an attractive interest rate on their loans, and lenders get an attractive interest rate on their investment.

Any underwriting system will have ways of checking whether people are telling the truth on their loan applications, and ways of raising flags if they might be lying. It seems that Lending Club’s system raises some kind of flag about 60% of the time — that’s a high number, not a low one. And in about a third of those cases, the loans are, properly, rejected.

Lieber’s worry here is that there are liars in the 40% of loan applications where the system doesn’t seek extra information on income or employment. It’s a reasonable question, but I’m willing to give Lending Club the benefit of the doubt, since Lending Club itself has every incentive to minimize defaults and delinquencies. So here’s a question for Lending Club: from Lieber’s numbers, it seems that roughly half of your loans are based on verified income or employment data, and in the other half you didn’t ask for such verification. Can you publish separate performance data on those two sets? That should help settle Lieber’s worry, I think.

The fact is that the percentage of borrowers that “you can’t necessarily trust” is 100%, it’s not just the ones who haven’t verified their income or employment data. Lending Club is not based on trust, it’s based on underwriting, and on the law of large numbers. If the underwriting system works, then Lending Club is a good investment. If it doesn’t, then it isn’t. (The reason that Prosper failed, initially, was precisely because of weaknesses in its underwriting.)

Lieber might not fully trust Lending Club’s underwriting system, but that doesn’t make Lending Club “a new type of casino game in Las Vegas” (where the punters are statistically certain to lose). It just means that Lieber wants to wait until the first tranche of loans has started maturing before he trusts in Lending Club’s system. That’s entirely reasonable, but it doesn’t mean that people investing in Lending Club today are gamblers.

I do agree with Lieber, however, that the only people who should invest in Lending Club are people with safer investments elsewhere. And in fact even Lending Club agrees with that. Check out its criteria for being able to invest:

Investors who are residents of states other than California or Kentucky must have (a) an annual gross income of at least $70,000 and a net worth (exclusive of home, home furnishings and automobile) of at least $70,000; or (b) have a net worth of at least $250,000 (determined with the same exclusions).

The restrictions on investors from California and Kentucky, meanwhile, are even tighter. Lending Club is explicitly an investment for the rich, not for the masses: I doubt that 10% of Americans would be eligible to invest under these criteria.

There’s a very good reason why these investments should be only for the rich, and it has nothing to do with them being a high-risk gamble. Instead, it’s all about liquidity. If you lend someone money for three years, your money is essentially out of reach for three years. If you reinvest your interest payments, as Lending Club assumes you do in its return calculations, then your money is out of reach indefinitely.

Lending Club does have a secondary market, but it runs into market-for-lemons problems which Lending Club hasn’t really managed to address. This isn’t a CD, where you can take your money out just by foregoing interest payments: it’s a highly illiquid investment which you should assume essentially zero resale value and will only pay off in cash terms three (or more) years after you decide to stop reinvesting your money. That’s fine, if you know that you’re not going to need that money for the next three years. But only rich people can say that with certainty.

Meanwhile, I’ve been taking a look at another area where online deals look significantly better than anything you’re likely to find in the bricks-and-mortar world: wine. Sites like Wines Til Sold Out or Lot 18 have sprung up of late offering wines at very substantial discounts for a very limited amount of time. This is clearly a great deal if you’re familiar with the exact wine and vintage being sold — but you need to be a serious wine wonk for that, given how obscure most of the wines being sold are. These sites are aimed at a much broader population than that, and so most of us, if we buy wine from them, will be buying a wine we’ve never drunk before.

That’s fine — a very large proportion of the wine I buy at retail or in restaurants falls into the same category. But in those cases, if I like the wine, I know how to get it again. With the new discount wine sites, once you’ve bought the wine you’ll either like it or you won’t. If you don’t like it, you’ll be sad because you wasted your money on subpar juice. And if you do like it, you’ll be sad because you won’t be able to buy any more of it — certainly not at the price you just paid.

More generally, it seems to me, deals on the internet often come with a kind of catch that most people aren’t really familiar with. Lieber, looking at Lending Club and Prosper, was on the lookout for risk factors surrounding total return — but the biggest risk actually surrounds the liquidity of the investment. And wine buyers who love bargains similarly forget the importance of being able to repeat the transaction if it turns out well.

So while the internet does throw up intriguing new deals and business models on a regular basis, it also throws up unexpected catches, in areas where people often aren’t looking.

Update: It was a bit rude of me to link to Lot 18 and send y’all into an opaque registration wall. But that’s fixed now! Go here, and you’ll not only get registered immediately, they’ll even throw in a $10 coupon.

Update 2: Lending Club has provided me with the numbers I asked for.

Comments
2 comments so far

“I doubt that 10% of Americans would be eligible to invest under these criteria.”

In 2007, more than 20% of income tax returns had an AGI in excess of $75k. And surely most of these households have $70k+ of financial assets.

There are also a significant number of retirees with less income but greater than $250k of assets.

I would guess that roughly 25% of households meet these criteria.

Posted by TFF | Report as abusive

The secondary market on Lending Club is not a market-for-lemons problem at all. The sellers in the secondary market see the same performance data as you and are no “closer” to the borrower than a prospective buyer. You don’t have a market-for-lemons without information asymmetry.

Posted by drewr | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/