Researchers looking to reduce lending bottlenecks abroad say the technique can predict repayment potential.
No credit? No problem — just take a test.
That’s the message being delivered to more than 70,000 small-business owners in developing countries where credit ratings are rare and many potential entrepreneurs keep their money in cash rather than bank accounts.
Banks in 16 countries are using a psychometric test to predict behavior — specifically, whether someone will pay back a loan. Originally a Harvard doctoral project, the Entrepreneurial Finance Lab’s test has increasingly won the confidence of risk-averse bankers in places where, many economists believe, credit bottlenecks are severely stunting growth.
Now, a new partnership with MasterCard has potential to speed the model’s proliferation.
In the United States and other mature economies, assessments by multiple credit agencies based on a lifetime of bill payments and account balances help determine with relative confidence whether to give an individual or business a loan.
But the lack of such data in much of the rest of the world creates a “massive inefficiency in emerging markets,” said Bailey Klinger, 34, chief executive of the Entrepreneurial Finance Lab. Banks have money to lend, but even profitable small businesses often cannot access it, choking growth.
In 2006, Klinger was studying this problem, known as the “missing middle,” with professor Asim Khwaja at Harvard’s Kennedy School of Government. They struck upon a technique some companies have long used to screen potential employees.
For Jhonathan Darwin Montes Mendoza, a 40-minute test led to a $1,500 loan last year to buy inventory ahead of the holiday rush for his market stall in Lima, Peru. Montes’ score gave Banco Interamericano de Finanzas confidence he would pay back the loan — even though he had been in business for less than a year, with no credit history.
While similar to tools used by psychologists to assess IQ, define personalities or screen for addictions, the bank’s test was intended to measure the traits at the core of entrepreneurship: fluid intelligence, business skill, integrity and attitudes.
After paying back the first loan, Montes, 23, is on a second round, paying down a $2,500 debt. The finance lab calibrates the test for each country where it is introduced.
The lab’s model asks questions that do not necessarily have a right answer. Using an algorithm, it aims to predict whether an individual is likely to default based on how the answers relate to one another.
For example, to assess their sense of personal control over outcomes — which tends to correlate with loan repayment — respondents might be asked to rate how much they agree or disagree with the statement: “I believe in the power of fate.”
There are some unexpected findings: Optimism and self-confidence are good signs among seasoned entrepreneurs, but high levels in younger business owners do not bode well, statistically.
And the math and reasoning questions meant to measure fluid intelligence can also assess integrity — of the loan officer. Too many correct answers can reveal that an applicant was coached.
Since Standard Bank, Africa’s largest, first adopted EFL’s model in Kenya in 2008, bankers around the world have used it to lend more than $200 million, in average amounts of $7,500, to entrepreneurs who would not have otherwise qualified, the finance lab’s founders say. Now an independent company based in Lima with about 30 employees, the Entrepreneurial Finance Lab has grown with the help of grants, including a $3.6 million prize for being among the winners of a G-20 challenge for small-business financing in 2011.
Still, the program’s early successes are “by no means final validation that the model will really work,” said Peer Stein, a director at the World Bank’s International Finance Corp., which administers the G-20 challenge grant. The loan test, he said, would “probably have to go through still other iterations.”