VISITING the homes of poor Africans who want to borrow money helps Finca International, an American microfinance company, to break down likely dead ends. If an applicant has an indoor toilet or household gifts from a relative working overseas, it is a good sign. Interviewing the neighbors also helps, says Mike Gama-Lobo, who runs Finca’s operations in Congo, Malawi, Tanzania, Zambia and Uganda. These visits work so well that only 1.5% of loans are in arrears each year, but they come at a cost: the Finca employs more than 1,200 traveling loan officers in these countries.
Hence Mr. Gama-Lobo’s interest in using other data sources to calculate solvency. Nine in ten loan seekers use a mobile phone. With the permission of potential borrowers, analyzing usage patterns can help reveal those most likely to default. Frequent calls to or from a wealthy country are a good sign. The same goes for calls on weekdays to a neighboring town: this suggests a commercial activity.
Entering all the data you can makes sense in emerging markets where credit bureaus are underdeveloped. But it also works in the rich world, where young people and immigrants often have no credit history. The offices themselves are now using everything from court records and rent payments to utility and telephone bills. And a slew of start-ups are also actively exploring alternative data.
Some companies collect scores by analyzing applicants’ online social networks. Professional contacts on LinkedIn are particularly revealing of a candidate’s “character and ability” to repay, explains Navin Bathija, the founder of Neo, a start-up that assesses the creditworthiness of applicants for automobile loans. Neo’s software helps determine if candidates’ claimed jobs are real by examining, with permission, the number and nature of LinkedIn connections with colleagues. It also assesses how quickly laid-off employees will land a new job by assessing their contacts with other employers.
As statistics accumulate, algorithms improve their detection of correlations in the data. Applicants who type only lowercase, or all caps, are less likely to repay their loans, all other things being equal, says Douglas Merrill, founder of ZestFinance, an American online lender with a default rate of around 40% less than that. of a typical payday lender. Neo’s efforts to improve accuracy include recording borrowers’ Facebook data: Mr. Bathija believes that within a year there will be enough evidence to determine whether racist comments on Facebook correlate with a lack of solvency.
Facebook data is already informing lending decisions at Kreditech, a Hamburg-based start-up that provides small loans online in Germany, Poland and Spain. Applicants are invited to provide access for a limited time to their account on Facebook or another social network. A lot is revealed by your friends, says Alexander Graubner-Müller, one of the founders of the company. A candidate whose friends seem to have high paying jobs and live in nice neighborhoods is more likely to get a loan. An applicant with a friend who has defaulted on a Kreditech loan is more likely to be rejected.
An online bank that opens this month in the United States will use data from Facebook to adjust interest rates on account holders’ credit cards. New York-based Movenbank will monitor Facebook posts and cut interest rates for those who tell friends about the bank. In the event of membership, the referent’s interest rate will fall further. Rates and fees will also drop if account holders spend wisely. Efforts to define customers “in a richer and deeper way” could potentially include higher tariffs for big players, says Brett King, founder of Movenbank.
No company has perhaps gone as far as Lenddo, a Hong Kong start-up that has online lenders in Colombia and the Philippines. Loan seekers ask their Facebook friends to vouch for them. To determine if those who say “yes” are real friends rather than just Facebook contacts, Lenddo’s software checks posts for shared slang or wording suggesting an affinity. In addition, the credit scores of those who have vouched for a borrower are damaged if the borrower does not repay. Spread the word about this “social enforcement mechanism” and “boom, the money is coming,” says Jeff Stewart, Lenddo’s boss.
Tweet and sour
Many “digital natives” who have come of age with the Web will care less about these loan terms than older people. Andrea Higuera, 21, a three-time borrower from Lenddo in Colombia, says that merging her Facebook world and her financial life “really doesn’t bother me at all.” A recent graduate and now an office worker in Bogotá, she was more hampered by the bureaucratic rigor that preceded a normal bank branch’s denial of a credit card and approval of a very small loan.
The big banks will nevertheless be cautious. Although they monitor social media for marketing purposes, using the data to assess loan applicants is “a dangerous game” that the big banks are dodging for now, says Frank Eliason, head of social media at Citibank. He notes that over the past six months, some people have ditched Facebook for paid networks like App.net that offer greater privacy. Schufa, a German credit bureau, has scrapped plans to operate Facebook, Twitter and LinkedIn after a public backlash last year.
Employees of small banks often search for candidates’ names on social media or the web, says Jack Vonder Heide of Technology Briefing Centers, a consulting firm. A loan officer may turn down a loan upon learning that, for example, an applicant is in the process of getting a divorce. But if this process were automated and industrialized, it could turn a big bank into “very juicy fodder” for the press. As a result, a bank could violate privacy laws, says Vonder Heide. And that would certainly be vilified on the very social media sites he used to make decisions.
This article appeared in the Finance & Economics section of the print edition under the title “Stat oil”