How personality testing could help financial inclusion.

The following article, published by the Economist on October 1,  features an interview with Beedie School of Business alumnus Bailey Klinger, and profiles his Entrepreneurial Finance Lab technology.

HOW would you feel if you were invited to the moon? If you found a gold coin, would you save it, give it to charity or splurge on a holiday? Personality quizzes of this kind—“psychometrics”, in the jargon—are already the bane of many a jobseeker. Now, it is being applied to the oldest problem in finance: will a borrower repay?

In rich countries, lenders use credit scores to weigh risk. But just 7% of Africans and 13% of South Asians are covered by private credit bureaus. Bailey Klinger of the Entrepreneurial Finance Lab (EFL), which explores new kinds of credit data, argues that psychometrics could scoop many more people into the financial system. Everyone has a personality, after all.

Judging character is not new. Psychometrics attempts to make it a science. EFL began life as a research initiative at Harvard. The model used by Creditinfo, a rival firm, was developed with help from Cambridge. Their online quizzes are road-tested and tweaked for different cultures. Sifting the data reveals telling patterns: for instance, EFL found that young optimists are risky, but old ones are a safe bet.

Clever design cuts cheating. There are no obvious “right” answers; responses are cross-checked for consistency. The software monitors mouse movements for signs of indecision or distraction. And when the unscrupulous lie to get a loan, they often do so in predictable ways. In the Creditinfo test, people are shown pictures of five drinks and asked which one they would be. Choosing water over something fizzier may be a sign of cheating, says Clare McCaffery, its managing director in Britain. (Still or sparkling, you might ask.)

This all sounds fanciful, but there is evidence that it works. EFL has honed its model through trials on three continents. In one Indonesian bank, combining psychometrics with existing customer data cut default rates for small businesses by 45%. A study by the World Bank and Inter-American Development Bank found that EFL’s model boosted lending to those without a credit history (a bad record betokened problems whatever the psychometrics said).

Some lenders are convinced. Grupo Monge, a retailer, uses psychometrics to sell household goods on credit to low-income Peruvians. “Most of the time we are the first company to give them credit,” says Gabriel Trelles, its boss in Peru. The biggest market for psychometrics is for such consumer loans. But microlenders and banks are catching on. EFL’s software has been used in 690,000 loan decisions in 27 countries. Creditinfo will use its psychometrics unit, recently acquired from a marketing firm, to expand in emerging markets.

Psychometrics has so far been merged into existing loan processes. Richer data could change that. Jared Miller, CEO of EFL, describes a future in which lending is almost entirely digitised, combining psychometrics with social media and mobile-phone records. Startups are rushing to make use of these “alternative data”. One example is First Access, in Tanzania, which uses data such as mobile records to gauge the strength of borrowers’ networks, and thus how likely they are to repay.

The technique is still in its infancy and will not replace credit bureaus, says Miriam Bruhn of the World Bank. The best way to tell if somebody will repay a loan in future is to see if they have repaid one in the past. But bureaus improve more slowly than technology. Lenders, looking for an edge, will find ever more ways to peer into their customers’ souls.

Read the original article on the Economist website.

Read more about Klinger’s work from his interview at the Beedie School of Business alumni reunion earlier this year.