Profit motive and competition funnels funding to the unfunded better than government 'schemes', as this example of a finance company shows
The Reserve Bank of India’s (RBI’s) College of Agricultural Banking has announced a case study competition for bankers to find innovative ways of lending to the MSME (micro, small and medium enterprises) segment. The objective of sensitising bankers and getting them to think innovatively is laudable; but RBI officials would probably learn more about innovative lending by speaking to their own regulated entities in the private sector.
Last week, I was talking to the head of a rapidly growing finance company with a massive year-on-year growth in loans to first-time borrowers. He told me about the big increase in loans to micro-borrowers that included hawkers, small shopkeepers and non-salaried individuals, who were shut out of the formal borrowing system. Typically, these were loans of Rs30,000 to Rs1 lakh with tenure of about 12 months. He called it ‘algo lending’.
His lending is based on evolving a proprietary model that minimises the risk of default, having identified the common characteristics of borrowers trying to fudge while taking a loan, or more likely to default. Simply put, he arrived at these characteristics by first giving loans to a bunch of people, profiling them under numerous parameters, and then studying their repayment patterns. This model, based on the law of probability, works better when more data of borrowing and repayment continues to be added.
Some of the findings are material for further research by those keen on financial inclusion and its cost. For instance, the company discovered that borrowing through equated-monthly instalments (EMIs) for high-end, large-screen televisions are a cover for un-banked persons to get a formal loan. The shopkeeper colludes with the borrower to create loan records for EMI payments, but no TV is sold. Instead, the borrower gets the cash and the shopkeeper a nice commission. While many borrowers diligently pay back their loans, some default. The trick is to minimise the risk on such loans. By matching income and social profiles of the borrowers, the company found that, often, when the customer was trying to buy an unusually large television, it represented a financing deal and could be weeded out.
Another tiny scam that the company discovered was that borrowers tried to use the same know your customer (KYC) data with two different identities, and a minor misspelling of the name (Rajiv or Rajeev). This would, often, fly under the radar of the lending agency as a typographical error, allowing the person to avail two separate loans. A de-duplication effort, as well as a scan for past record flags such cases.
A trick by those who have a previous loan default to avoid detection by credit bureau records is to apply for a new PAN card and submit it with fresh loan applications. The company had to develop a way to flag such cases through its algorithm. Their data analysis showed that a fresh PAN, obtained by people who were 45+, was a red flag.
Yet another finding, that is being analysed, is the distance between the borrower and his place of work with that of the lending organisation. When a borrower sought out a lender, whose office was far away, it raised a red flag. The work being done by this finance company, and its success at keeping bad loans to the minimum, only proves that profit is the best driver of high-quality research, innovation, as well as risk-taking—something that will not be possible through case study contests or government-mandated programmes.