It’s a Wonderful Loan: Economics of P2P Lending

The Financial Times wonders why big banks are going after P2P lending. Why do banks need companies like Aztec Money and Lending Club, which have negligible credit portfolios and messy business model? Well, banks themselves might say about their motivation in this case (so far they didn’t), but I can think of a good economic reason why they should pay attention to P2P lending.

This reason is older than the Internet, computers, and banks themselves. It’s information about the borrower. In between conspiracies against the public, banks do a very useful thing: they take off the lender’s headache about the borrower’s payback. Banks have to know their borrower well. And typically, they do and keep the net interest spread low. Here’s the rates for banks and credit unions:


Credit unions have been in the industry like forever. They would fit what the FT names “democratizing finance” and have much in common with the ideology behind P2P technologies. Credit unions have higher deposit rates and lower interest in the table because they know more about borrowers. Unions lend only to trusted folks and the number of individual defaults decreases, so you see better rates. Better rates also mean an even lower probability of default, so it’s reinforcing.

The Grameen Bank (and Nobel laureate Yunus) played this idea brilliantly. They radically reduced the market interest rates in poor countries, where high rates coupled with high default rates had been strangling the economy. The Grameen Bank entered very much like a credit union. Borrowers had to provide references from local peers to get access to money. The interest rates have been reduced from 50–100% annually to a single-digit number.

The Grameen-type firms and credit unions are limited in geography and expertise. You could back only your neighbor and only in a very simple business. If he tells you he’ll buy a cow to sell milk, you’re okay. But if a guy on the other coast needs a credit line to build “radar detectors that have both huge military and civilian applications,” you want to know the risks better. That’s why in a complex economy, Grameen is no longer relevant. Each loan application requires more information about the borrower, his credit history, and, most importantly, the purpose of the loan.

The purpose is vital for business loans. Banks learned to dig information about the borrower and to come up with the individual probability of default (you can try to predict yourself). But they’re getting worse in knowing the client’s business. First, businesses are getting more complex. Second, banks reduce their human workforce and local branches, while local branches provided a lot of soft information on borrowers and their performance. Jimmy Stewart’s banking was about observing his little town’s economy and deciding what would be creditworthy there. Without this source, banks pool risks and set higher interest rates, deterring borrowers.

Here comes online P2P lending. When a nuclear physicist from CERN lends money to a nuclear physicist from NASA via P2P system, it tells something about the borrower’s project. The guy from CERN is the right guy to judge. He also throws his own money into this. And that solves both the complexity (you can always find a lender-investor with the right expertise) and neighborhood problems (an expert comes from anywhere). Plus it’s technically free. The CERN physicist has already done the job banks couldn’t do: he found the borrower’s project, evaluated, and approved it. It looks like an investor’s job, and it is. P2P lending platforms like Kiva do mix investing and lending. Users do informal research before lending money.

This info allows banks do P2P loan matching (like some VC and foundations do), buy individually-backed loans, securitize them, and so on. This is a rare example when new technologies are not eating someone else’s pie (like YouTube does to mass media) but create their own. Without this easy expert-loan matching, businesses face higher interest rates, often above their breakeven point, which means no business at all.

Still, P2P platforms themselves seem distracted from this advantage. Most reasoning behind them mentions phantom problems like “predatory interest,” much paperwork, and refused applications in traditional banking. These are not the problems. The financial industry is highly competitive even after the series of post-80s M&A. It evaluates the risks with huge volumes of data, hires good quants, and saves a great deal on scale. In fact, the low market capitalization of major banks indicates that they have no means to “exploit” customers (Google and Amazon do, though in a delicate manner, as here and here). So net interest margin declines:


The bank’s paperwork and rejections are just the costs of low interest rates. It makes no sense for startups to “fix” banking in this direction because it’ll increase the rates—sort of getting the industry back into prehistoric times. The information flows between lenders and borrowers is the real thing to focus on.