The kindness of strangers

Start-ups are seeking to take a bite out of the consumer lending market by arranging person-to-person loans online. But whether borrowers will keep up the payments is an open question.

In May, Briana Stewart’s boss came to her with a tempting offer: How would she like to open a gourmet coffee bar inside the Book Worm, the suburban Atlanta bookstore where she worked as a part-time store manager? Stewart was eager to start her own business, but she lacked money to get going and didn’t want to battle red tape to get a bank loan or max out her credit cards at high interest rates. So she turned to Prosper Marketplace, an online platform that helps individuals lend money to one another.

On Prosper’s Web site Stewart posted a request for $15,000 and indicated she’d be willing to pay up to 14 percent interest. The posting, as is typical, also included her photograph, her credit rating and a personal message to prospective lenders. Prosper put the loan up for auction in early June. By the end of the month, 97 strangers, most contributing just $50 or $100 each, had combined to fund Stewart’s three-year loan at an interest rate of 10.4 percent.

“It ended up being a piece of cake,” says Stewart, now the proud proprietor of Barking Dog Coffee, which opened July 31 in the old brick bookstore on Marietta Street in Powder Springs, Georgia. “I would do it again.”

Prosper, run by Christian Larsen, the co-founder and former CEO of online mortgage lender E-Loan, is one of several upstart Web companies that have sprung up recently to offer what they call person-to-person lending. Others include Zopa, a London-based firm founded in 2004, and CircleLending, a Waltham, Massachusetts, company that helps administer private loans between family and friends.

For borrowers these online marketplaces are do-it-yourself alternatives to banks, credit cards and other traditional sources of loans. For investors they provide a chance to diversify stock, bond, real estate and cash holdings by gaining exposure to consumer credit.

The upstart companies also offer a certain populist appeal — especially for younger, Web-savvy consumers — and hope to steal business from online lending platforms like E-Loan and LendingTree, as well as from credit card and bank lenders. One of the marketing slogans on Zopa’s slickly designed home page, for instance, is “People are better than banks.” The firm’s founders are veterans of U.K. online bank Egg, which challenged the European banking establishment in the late 1990s. Prosper encourages like-minded users to form affinity groups — from Freemasons to Apple Macintosh enthusiasts — in an attempt to foster trust and communal ties among borrowers and lenders. The company’s listings page urges potential lenders to “help out a fellow human being” by lending him money.

“Most borrowers put pictures and stories up,” says Larsen, 45, an alumnus of Stanford’s Graduate School of Business and a generous Democratic Party donor who founded Prosper late last year after selling E-Loan to Banco Popular for $300 million.

The stick-it-to-the-man factor may help person-to-person lending platforms carve out a profitable niche with some consumers, but there are plenty of potential pitfalls. Most notably, the services are so young that there’s no good read on how many online borrowers will default — and no guarantee that investors who get burned as a result will come back. Consequently, the do-it-yourself firms aren’t likely to pose a huge threat to the financial establishment anytime soon. James Bruene, editor of the newsletter “Online Banking Report,” estimates that online “P2P” lending could generate $600 million in loan volume and $65 million in revenues annually by 2009. Although that’s not bad for a handful of start-ups, it would constitute barely a sliver of the U.S. consumer credit market, which today stands at $2.1 trillion in outstanding obligations and grew by $57 billion last year alone, according to the Federal Reserve Board. Even E-Loan, though successful, originated just $5.4 billion in loans in 2005.

“The concept has a chance to be at least a moderate success,” says Bruene.

Each of the three main companies in the P2P arena — Prosper, Zopa and CircleLending — takes a slightly different approach to the business. Prosper is perhaps the most ambitious. The San Francisco–based company operates an open marketplace. Any individual can apply for a loan of up to $25,000 simply by submitting personal information, including Social Security and bank account numbers and other identifying data. Prosper feeds these inputs through credit bureau Experian to generate a credit score for every borrower, ranging from the rock-solid AA to the highly speculative NC (short for “no credit”). Lenders consider these scores and other factors — including the borrower’s stated purpose for the money but not his Social Security number, home address or other specific identifying information — when deciding whether to bid on a particular loan. (The company advises lenders to sprinkle their cash among all categories to properly balance risk and returns.)

All Prosper loans carry three-year terms — a happy medium between short-term and long-term investment for lenders — with interest rates ranging from 7.7 percent to 24 percent, depending mostly on credit scores. Individual loans average about $5,000. Prosper makes money by charging a one-time 1 percent fee to borrowers and a 0.5 percent annual fee to lenders.

The company’s auction system allows individual lenders to break their investments into small pieces by bidding as little as $50 on individual loans. Prosper’s technology then pools the individual amounts together, so that loans can be funded by multiple lenders using the industry standard Automatic Clearing House network for transferring funds electronically.

One of Prosper’s key features is the personal nature of the lending. Larsen himself has loaned $120,000 to 162 different users. Serial lenders talk of spending hours surfing the site, reading the sometimes-sad loan requests. “I’m up till 3:00 in the morning dropping $50 here and there on interesting stories,” says Daniel Foster, an information-technology business analyst in San Francisco who invests through Prosper. (Two of the system’s lenders have so far committed more than $300,000 each.) Adds Larsen: “It’s not just about dollars and cents. In some ways it connects people.”

Larsen often likens Prosper to EBay — rather than auctioning goods, it auctions money. He’s not the only one who sees it that way. Venture capitalist Robert Kagle of Benchmark Capital, an early backer of both EBay and E-Loan, is also an investor in Prosper. So is EBay founder Pierre Omidyar, through his Omidyar Network venture fund. The company has raised $20 million in two rounds of funding. Since launching in February it has arranged more than 2,400 loans worth more than $11 million, according to Eric’s Credit Community, an independent Web site that analyzes Prosper data. More than 20,000 people have registered as Prosper members and created more than 1,100 affinity groups.

Foster, who since March has spread $23,000 across 149 Prosper loans of varying credit ratings, feels he’s diversified enough that defaults will slightly limit his returns, not wipe them out; only three of his loans are more than 30 days past due.

“I’m expecting at least a 10 percent return and hopefully 12 to 15 percent in all,” he says. That’s pretty good, considering five-year certificates of deposit in the U.S. are averaging rates under 5.50 percent. One-year CDs yield 5.04 percent, on average; money market savings accounts offer just 3.43 percent. “I see this as an asset class I can be comfortable with.”

Larsen too is thinking big. His target market for Prosper is the refinancing of credit card debt — debt that is being racked up at a rate of more than $10 billion per year in the U.S., according to Some analysts estimate that $100 billion in credit card debt is paid down every year by consumers tapping second mortgages and home equity loans. Larsen is also publicly releasing the site’s software building blocks, so that users of other Web sites can interact more directly with Prosper.

By using an online marketplace, he argues, consumers can secure lower rates than banks are willing to give them on these obligations. Prosper borrowers can secure rates of 8 or 9 percent, he says. That compares quite favorably with the average rate of 13 percent on credit cards and personal loans, according to the Fed. But Prosper is also seeking to tap marginal borrowers with troubled credit histories — hence the presence of affinity groups with names like “Starting Over” — as well as customers of high-interest payday loan companies, which typically serve workers living from paycheck to paycheck who occasionally have liquidity problems. These customers, along with recent immigrants and students, are part of a segment generally known in the industry as the unbanked.

The model is a little less personal and freewheeling at Zopa, which has been operating in the U.K. since March 2005 and plans to go live in the U.S. soon. The company takes its name from a business theory term, “zone of possible agreement,” which signifies the intersection between a seller’s minimum price and a buyer’s maximum. Unlike Prosper, Zopa uses credit bureau data and other information to screen out the riskiest borrowers. It divides applicants into just three relatively low-risk credit categories and requires them to have a minimum annual income of £25,000. Zopa loan rates are therefore lower than Prosper’s, ranging from 4.8 percent to 10.4 percent.

“People are using Zopa as an alternative to credit card debt and personal bank loans,” says the company’s CEO, Richard Duvall, formerly Egg’s chief marketing officer.

Whereas Prosper arranges loans through an EBay-style auction, Zopa functions more like a matching engine. Rather than sort through personalized loan requests, lenders simply input the amount they want to loan and what rate they’d like to earn. Zopa then uses its computer programs to divide the sums committed by investors among 50 or more loans. That divvying-up process, like Prosper’s system, provides for diversification across the credit-risk spectrum. For investors it’s akin to blindly buying syndicated, unsecured loans. The company charges slightly lower fees than does Prosper: 0.5 percent for each side of a transaction.

At the end of July, Zopa had 85,000 members (lenders and borrowers), up from 75,000 in June. It completes 30 loans a day and typically has more than £500,000 on offer. The draw? Zopa lenders earn an average of 7 percent annually (before defaults) on their loan portfolios, Duvall says, whereas the best U.K. savings accounts pay between 4 and 5 percent interest.

Earlier this year, Zopa raised $20 million in financing — from Bessemer Venture Partners, Benchmark, Wellington Partners and prominent venture capitalist Timothy Draper — to support its move into the U.S. But it has a long way to go to catch up with Prosper. The reason: Each state requires separate regulatory approval for the business. Prosper is already cleared to operate in all but four states; Zopa has approval only in California.

Both Prosper and Zopa face plenty of hurdles, the biggest being the likelihood of defaults. This is where differences in strategy may prove critical. Duvall, for instance, claims that Zopa currently has a default rate of just 0.5 percent. That’s significantly lower than the 2.7 percent rate on U.S. credit card debt, according to card giant Capital One Financial. He expects that rate to rise as the site grows, but considers defaults “under control” — in no small part because of Zopa’s exclusion of borrowers with low incomes and spotty or no credit history.

But Zopa, which expects to be profitable in the U.K. within 12 months, does plan to slowly expand down the credit scale toward subprime loans. Duvall can envision Zopa markets explicitly for small-business loans, car loans and other types of debt, he says.

Prosper says its current default rate is 0.04 percent. It aims to ensure that borrowers repay loans in part by encouraging users to form affinity groups. These serve a dual purpose: Designated group leaders help vet borrowers, with the goal of protecting the group’s credibility and ensuring that members can take out loans at low rates; the groups also seek to discourage defaults by fostering links between members who are total strangers.

Not everyone believes that such an approach will work in stranger-to-stranger lending. CircleLending, for instance, touts the benefit of the close ties between most of its borrowers and lenders. Rather than introducing the two sides in a transaction, the company processes and services what otherwise would be informal loans between family and friends. If an uncle wants to lend a nephew $3,000 to buy a used car, for instance, CircleLending will set up a contract and a payment schedule and administer the loan online.

“Focusing on individuals who have a less close relationship with each other tends to create higher default rates,” says CEO Asheesh Advani. Still, CircleLending’s loans default at a rate of 3.5 percent. That suggests that Prosper and Zopa, once they have been up and running long enough, may experience higher levels of nonpayment.

Advani estimates that family and friends in the U.S. loan $89 billion to one another every year. CircleLending has captured only $120 million of that flow, though its volume has more than doubled from $50 million just one year ago.

The other big challenge facing P2P lending is its quirky nature: Will anyone other than the young and Web-savvy take it seriously? Fewer than 30 percent of loan requests get funded on Prosper; coffee shop entrepreneur Stewart had to relist her request with more information to draw enough bidders. The very concept is so new and unproven that banks and other established lenders have paid it only passing notice.

“This model does have a niche, but time will tell if it has any legs with the consumers,” says Anthony Hsieh, president of, a site that matches borrowers with banks.

One way the business might evolve, especially if it doesn’t attract enough individual users, is to one day include — gasp — banks. “We see banks potentially as partners,” admits Larsen. “This is a platform in which they could participate as lenders as well.”

And at least one expert sees Zopa and Prosper as potentially offering ways to one day provide a full suite of banking services to the unbanked. “Both of these firms seem to have the risk piece down,” says Theodore Iacobuzio, an analyst at financial-technology research firm Tower Group. He’d now like to see them evolve into more-banklike entities, offering mortgages and other products.

None of the P2P players have plans to do anything like that in the near future. They’re more concerned with gaining a foothold in their chosen markets. But in the meantime, they could teach banks and other lenders a thing or two about online audiences, reaching new customers and the risks that some Web users are willing to take for higher returns. i