Fintech’s capital constraints can affect credit profiles

you might also like

iPhone sales in India are heading for another record

IT companies likely to cut campus hiring in FY24

What’s next for India’s faltering export history

Marico’s Q2 update fails to lift spirits

The risk to consumer credit: About one in six fintech firms is a lender. Some of these fintech lenders (FLs) may cause a deterioration in the credit profile of their borrowers. Let’s see how this can happen. Most FLs focus on consumer loans. Most such loans are in an area from ₹3000 more ₹NOK 50,000 with terms from 1 month to 12 months. These are small-ticket loans with a short term (STST). Buy-now-pay-later (BNPL) is a subcategory of such loans, with even lower ticket sizes and terms. Technically, these are unsecured personal loans (PLs). Such loans drive financial inclusion by including new-to-credit (NTC) borrowers and those with no income security or low income. The PL portfolios of FLs show higher default rates than banks, as the latter target borrowers with lower risk profiles than FLs, who expect higher interest costs to compensate for the additional risk they carry.

Well-established FLs have significant expertise in using advanced analytics to facilitate credit decisions. However, some FLs may have weakened credit policy to chase growth. Certain worrying trends are emerging which are neither new nor unique to India.

Exaggerate the theme of “repeat customers”: Leveraging data from existing customers to improve credit decisions is a global best practice. However, it can be too much of a good thing. Some FLs can come close to the sector’s equivalent of evergreen. Say the borrower is expected to repay the loan via cheque. Even before the check is cashed, FL can extend a new STST loan for a higher amount within 36 hours. The borrower benefits from a credit transfer and does not have to repay the principal. The borrower thus does not get a chance to default. Other lenders view this borrower from a credit bureau lens as someone who is not delinquent and has serviced loans of increasingly higher ticket sizes. Basically a good credit profile!

Laddering and loan stacking: Adding the competitive dimension, it is possible for a borrower to get a STST loan from one lender and repay another. Meanwhile, the previous lender will use analytics on data that shows a deceptively improving credit profile. This lender will be ready to disburse the next STST loan. Here, the borrower almost climbs up the ticket size eligibility ladder, not because his income has improved, but because of suboptimal and questionable credit practices. Loan stacking is just one step away, where a borrower whose distress may not have been disclosed in the loan performance applies for a large number of loans from various borrowers and gets most of them.

The process in which such pockets of influence are built up continues until liquidity conditions change and FL itself faces financing challenges. Then it will focus on improving the quality of the loan book and try to improve the collection of cash and also underwrite new loans in a more prudent manner. Borrowers, some of whom were new to credit and thus credit immature, would then be seriously surprised to see their on-loss credit dry up. Such borrowers whose credit scores to the agency may have improved all this will suddenly exhibit “jump-to-default” behavior.

Shock transmission to the banking sector: Borrowers who had used credit flows from FL loans to service bank loans are likely to start defaulting on their banks. These banks will then go into risk-off mode and restrict credit further. However, such a situation can be avoided if measures are taken in time. The banks must look at their credit models and lending policies again. Typical loan lending rules, such as instances of 30+ or ​​60+ days past due payments in the past 6 months, cannot capture the inherent risk of such borrowers, since they avoided default. Risk amplifiers that rise can also be missed. Ever since the introduction of STST loans, banks have relaxed or done away with leverage-based limits such as “two loans in the last three months” because many loan applicants were flagged. But such rules may have to come back. Risk management is an art form, one where science precedes art. If data doesn’t capture all risks, machine learning won’t help. This is where a good assessment of risk management must come in.

Deep Mukherjee is a quantitative risk management professional and on the visiting faculty of risk management at IIM Calcutta.

Elsewhere in Mint

In Opinion Nouriel Roubini says that stagflationary debt crisis he had prophesied is here. Vivek Kaul writes about the price we pay for overprinting of money. Long Story is unveiled the promise and perils of Ameerpet.