PARIS — Imagine securing a loan in just minutes — without having to visit a bank, fill out complicated paperwork, sit down with a loan officer. What if it could all be done with a few clicks on a smartphone?
For some consumers and small businesses in the United States, China, the Philippines, Mexico and Colombia, that fantasy is already a reality thanks to companies like Affirm, a San Francisco startup created in 2012 by Max Levchin, co-founder of PayPal.
Affirm is currently available to users of some 100 U.S.-based e-commerce websites. The sites offer customers the option — just as they’re about to pay — of taking out a loan with Affirm, which uses a predictive algorithm that brings together Big Data and machine learning to immediately assess the buyer’s repayment capacities.
“Algorithms make it possible to reduce costs, because they remove all human intervention,” explains Guillaume Ponsard, founder of Nordpay, a European leader in online payments.
The process has become amazingly fast, and the interest rates tend to be lower than the market standard. “Affirm offers Millennials, who are used to doing everything digitally, affordable loans and therefore more buying power,” says Jeremy Liew, a partner at Lightspeed Venture Partners, a venture capital firm that has invested in Affirm.
The next big thing
Benoit Legrand, former president of ING France who currrently heads the Dutch retail bank’s financial technology department, sees online loans as a “technological breakpoint” similar to the introduction, in the late 1990s, of online banking.
“We’re betting on it to shake up the banking sector,” he says.
The signs are everywhere, as young startups that specialize in automated underwriting mushroom in the United States (Biz2credit and Zest Finance, to name a few), Europe and in Asia.
To assess the risk posed by a borrower, these startups look through hundreds of “weak signals” such as how long it takes the customer to type his email address (to detect copy-pasting, which is often a sign that the address was created just for that purpose), the time of the request (nighttime applications arouse distrust), his social network identity, his job title on LinkedIn, the number of Facebook friends, the hour when he usually sends emails, and how many he sends.
“For instance, we check whether a potential borrower has, among his social network friends, people we already know and if they are good payers — which of course is a good thing — or bad payers,” explains Florentin Lenoir, marketing and business development director at Lenddo, a Hong Kong-based startup.
Kathryn Petralia, cofounder of Kabbage, notes that clients who authorize her company to access an active company profile on Facebook are 20% less likely to be bad payers than those who don’t.
Defenders of these algorithms believe they’re helping borrowers by giving them easier access to credit. “Big Data allows us to grant loans to people who otherwise wouldn’t have been able to borrow because they don’t have a bank account or because they once defaulted on their payment,” says Lenoir.
For the money-lenders, algorithms can reduce the number of defaults. “That’s because there’s more truth on the Internet than in what some people declare when applying for a loan,” says Damien Guermonprez, CEO of Lemon Way and former CEO of consumer credit companies.
Too impersonal?
Critics fear a further intrusion of Big Data into people’s lives. But they also point out concerns that these algorithms may reject applications for obscure reasons, and that the victims of this arbitrariness have no means to defend themselves.
Bernard-Louis Roques, cofounder and managing director of Truffle Capital, a private equity firm, says that ultimately there is little to stop the wave. “Big Data and predictive algorithms will impose themselves no matter what because they can offer the best customer service,” he says.
The next step? “Peer-to-peer” loans, predicts Erik Engellau-Nilsson of Klarna, a Swedish company that developed its own predictive algorithms for the payment of online purchases. The Payment Services Directive 2, adopted by the European Parliament last October, anticipates that third parties will be able to provide services related to bank accounts.
Put simply, a company specialized in automated underwriting could offer private individuals who have money on their own savings account to lend that money to people selected by its algorithms. Despite the inherent risk that comes with all loans, this could provide more long-term profit than just having your money sitting in your bank account.
