For a long time now, I’ve been interested in the potential for Peer-to-Peer lending (P2P). Rather than people depositing money with a bank, which then turns around and lends the money to borrowers while keeping most of the profit, in P2P lending some service matches borrowers and lenders directly. Ideally, this allows borrowers to get loans at a cheaper rate, while still giving lenders a better rate of return. It also allows nontraditional borrowers to get funded, if they can tell a good enough story. But there are problems with P2P, to which I would like to propose some solutions.

Right now, the biggest names in American P2P lending are Lending Club, and Prosper.com, where I was a small lender back around 2007 or so. When I participated on Prosper, I was able to see instances where the model worked perfectly, and other instances where the model caused problems. Among the advantages were that groups of people who knew each other (whether In Real Life or on the internet) could provide each other with funds at below-market rates, as an expression of their fellowship. This is something I liked a lot, since Judaism tends to frown on lending money at interest anyway. (More on that topic in another post…) Additionally, some borrowers whose credit score was relatively low were able to convince borrowers that they were a good risk anyway, because of factors that their credit score did not reflect.

The bad news came from several sources. First of all, because all money had to be provided by the individual lenders, as opposed to a centralized source of funds like a bank, many otherwise-attractive borrowers who didn’t know how to market themselves went unfunded. Second, the need to market yourself in the first place can be a turn-off. Borrowing from people you know can lead to tension and a loss of privacy. One of the biggest advantages of the modern system of bank or credit-card lending (as opposed to borrowing from friends and family, as the “unbanked” tend to do) is that banks can lend functionally unlimited amounts of money if they choose to, and your financial circumstances are strictly between yourself and your banker. With Prosper as it stands now, neither of these two factors are at work.

The next problem was due to greed. In the early years, lenders were attracted to the high rates paid by low-quality borrowers, particularly “HR” or high-risk borrowers who could be made to pay up to 30% interest per year. The problem was that in most cases, these borrowers were staggeringly bad risks; those of us who jumped in with both feet ended up losing quite a bit of money when the loans were not repaid. Ultimately, we lacked the knowledge and temperament to tell a good borrower from a bad one, and to resist the lure of fat profits for taking unacceptable risks. There is a reason that (until the era of government bailouts) bankers had a reputation for sobriety, prudence, and conservatism. Bankers who lacked these qualities soon ceased to be bankers.

Note that Prosper responded, ultimately, by excluding the lowest-quality borrowers from the market. This managed to improve default rates; but the fundamental problem remains, that lending decisions are being made by amateurs, many of whom do not understand the risks well enough.

Another institution that claims to be P2P, the British institution Zopa, avoids this second problem by soliciting investor capital and simply making all the lending decisions itself, as a traditional lender would. But in this case, there is no actual interaction between borrower and individual “lenders”; Zopa is actually something like a mutual fund for lending with investor funds, rather than bank capital. It is not really a true peer-to-peer vehicle.

Is there a way to mitigate the problems of P2P without going to the other extreme and quashing the social aspect altogether?

Suppose that you have a Lender institution, a depositor (David) who wants to earn some money, and several borrowers (Brad, Ben, and Betty). David puts a chunk of money (say $5000) into his bank account with the Lender. The Lender then asks David if he is willing to lend money to particular people, at any point in the future, and at what minimum interest rate. David knows Brad, Ben and Betty; he decides that he would be willing to spot Brad up to $500 at a time, but only at the market interest rate—he’s not a close friend or anything. Betty, on the other hand, is someone that David likes, so while he’d only trust her with $200, he’s willing to lend the money to her at a nominal 1% interest to cover fees. He knows that Ben is irresponsible with his money, so he’d rather not risk his own money on Ben.

Note that during this process, David does not know if Brad or Betty are actually trying to borrow money right now. That information is kept from him. So Brad and Betty get to keep their privacy. If Brad or Betty should choose to lobby David for more money, that’s their choice. In the meanwhile, money that’s not specifically earmarked for particular borrowers can be treated like a normal bank deposit, earning some amount of money for David and having some level of guarantees.

Now suppose that Betty finally decides to borrow $20,000 for a new car. The Lender can see her credit score, and other such raw numbers; but it also can see that David, and several other depositors, are willing to trust Betty with money. This does two things: first of all, it shows Lender that people trust her, which can be an additional factor in the Lender’s decision to lend or not. Second, it means that they have to risk less of their own money for the same profit, because they can use money from the depositors who know Betty while still earning servicing fees.

The final decision is the Lender’s; if Betty remains a bad risk, Lender can protect the depositors from her. But on the flip side, if people who know her only pony up $5000, the Lender would be able to make up the difference with its own funds if they like Betty as a borrower, instead of letting her languish in social-lending purgatory. The portion of the money coming from depositors directly could be at below-market interest rates if they like, keeping the social aspect of P2P lending. The balance, coming from Lender, would be at market rates.

Obviously, such a system would not be for everyone. Some people would do well enough on Prosper.com or similar sites, even without such a system. Others might prefer Zopa, or even a traditional bank. Still, I think that the proposed system would make P2P borrowing and lending more attractive for a lot of people, mitigating some of the problems while keeping most of the advantages.

(Anyone who likes can implement this system. I only ask that you drop me a note or a comment letting me know, if you got the idea from this post.)