In developing markets, credit assessment of individuals unknown to a lender is often subjective, time-consuming and expensive, potentially involving home visits by loan officers to interview applicants and their neighbours. Credit bureau coverage may be patchy or non-existent, reflective of the fact that many consumers in these markets have little or no history with financial institutions.
Lending to lower-income households and small and informal enterprises is challenging. Many of these customers have limited familiarity with formal financial services, which inhibits their ability to make good decisions about the responsible and appropriate use of credit. And lenders often have little to none of the data they might traditionally use to make sound lending decisions e.g. official proof of income and a credit history.
In such environments, many lenders instead focus on cross-selling to existing customers or catering to those for whom credit history information is more readily accessible (typically, the more affluent or the salaried people). As a result, non-customers may be shut off from credit or face higher than necessary borrowing costs as lender costs are passed along to borrowers in the form of higher interest rates and fees. The main alternative for most in such circumstances is to borrow from friends and family or from private money-lenders at an exorbitant rate.
The underwriting processes in both high income and developing markets rely heavily on financial data inputs for the consumer. For many individuals globally, these “traditional data” inputs are simply unavailable for underwriting purposes, serving as an obstacle to consumer credit and one that impedes financial inclusion in both high income and developing markets. However, promising substitutes for traditional measures of credit risk are emerging. So-called “alternative data” from mobile-phone information, social media activity and records of online transactions offer new approaches to risk assessment.
Today, there are more than 2.5 billion people in the world without access to formal financial services, and there are hundreds of millions of micro, small, and midsize enterprises with unmet financing needs. Ensuring that these people and groups have access to a range of quality, affordable, and appropriate financial services is on the agenda of governments and businesses worldwide. New uses of data and information move us closer to this vision by enabling a more complete understanding of households’ financial needs.
Many lending companies are mining Facebook, Twitter and other social-media data to help determine a borrower’s creditworthiness. Lending companies are looking at potential problems such as whether applicants put the same job information on their loan application as they posted on LinkedIn, or if they shared on Facebook that they had been let go by an employer. Companies are tapping into other sources of data, including PayPal and eBay accounts, to determine not just whether a borrower should get a loan but whether their credit line should be increased.
There is a growing trend of fintech start-ups in emerging markets such as Africa, India and China that are taking business away from traditional banks in the areas of credit, savings and insurance. They use mobiles and reduce expenses by going directly to the consumer for loans and deposits, without branches.
Managing Director Johan Bosini of a Cape Town based micro financing startup called JUMO said to Bloomberg that the information it collects has helped it increase lending to almost 1 million loans a month of less than $200 across eight African countries. The company builds a profile of its customers from thousands of data points in calling records, ranging from who they’ve phoned to airtime and data purchases and other mobile transactions.
These new approaches have their own challenges. The practice of lending to needy borrowers based on alternative data is not free from risks. The risk modellers and code writers are working on the algorithms for cutting the loan loss-rate, but all risks cannot be eliminated. A scam artist once studied JUMO’s loan-approval patterns for several months, using 30 different sim-cards to generate data sets and deciphering the lender’s algorithms. He fleeced the firm of $30,000 in one day and then vanished.
These are exciting areas that are ripe for further innovation. The new alternative data models have cut credit losses in experimental forays into lower-income segments by 20 to 50 percent and doubled their application approval rates.