Retail Lending: What got us here, will no longer take us too far ahead

1. Credit penetration used to be much lower. New-to-credit (NTC) borrowers formed a substantial portion (25% to 40%) of retail loans. NTC are not high risk borrowers, and typically have delinquency levels much better than sub-prime borrowers.  However, NTC as a proportion of new loans has fallen from 18% (September 2022) to 10% (September 2024).

2. In the past consumer leverage was comfortably low. Currently, the leverage level of a segment of borrowers, that is, borrowers below 35 years of age has spiked. (Source: BCG’s Report titled Banking for Viksit Bharat). This segment has been a major driver of unsecured loan growth in last 3-4 years.

3. Between 2019 and 2023, credit application fraud have spiked by four times ( Source: BCG’s Report titled Banking for Viksit Bharat).

Lenders: New Day, New Game

The battle for profit pools: Even without RBI’s higher ask for capital on unsecured lending, growth would have been moderated by prudent lenders. In the medium-term it is likely to remain lower than long-term average growth. Further, the profit pool that retail lending offered may grow at even more moderate rates. Select players may have disproportionate share of the retail profit pool. These need not be among the successful players of the past decades. The game has opened up .Players of some scale but more integrated digital underwriting , supported with well-designed and best in class risk analytics will corner much higher share of profit pool than suggested by the proportion of their loan book. However digital laggards and haphazard digital adopters may observe asset growth but their profitability from retail lending may dwindle further.

Welcome to ‘analytical & information arbitrage’:  Most players have some acceptable, if not outstanding credit risk score cards typically based out of credit bureau data, application data and bank statements. So similar set of customers will be identified as high risk or low risk. But lenders with better designed credit models and underwriting journey and who are adept at using alternate data may ‘arbitrage out’ lenders of moderate analytical capabilities. So these borrowers will be able to identify and reject high risk borrowers who may be tagged as low to moderate risk by an average underwriting setup and vice versa. So lenders with competitively better models and credit decision framework will grow more profitably than others in a scenario where high risk-adjusted growth is not a given.

Further differentiator would be development of advanced fraud detection systems. Since digital fraud is surging, lenders with digital lending capability but manual or purely rule-based fraud check mechanism may be the worst casualty.

Investors: Seek more disclosures and sharper due diligence

Investors have in the past over indexed on growth of portfolio and credit cost was secondary. The reason being the credit cost was very low to start with. At least some of the hubris in the current retail landscape may be attributable investor over-enthusiasm and naïveté with respect to fintech and new age lenders.

Risk Adjusted Returns: For starters the target return metrics should also have risk adjusted return measures (RAPM) such as Return on Economic Capital (ROEC) and Return on Regulatory Capital (RORC). Popular measures such as ROE and ROA per se do not fully consider the risk of the portfolio nor are they a direct measure of shareholder value.

Differentiated insights: The portfolio itself need to be assessed differently. For example, the conventional coincidental NPA rate, hides early signal of deterioration because even a moderate growth deflates the gross NPA rate. However lagged NPA (NPA of current period divided by Assets of previous period) captures those signals without fail.

Transparency: More disclosures may be requested with respect to collection efficiency. Further, numbers related to credit loan upsell to high-risk customers- for that may be an indicator of soft evergreening. Additionally, early delinquency rate may be looked at to track possible instances of credit fraud.

Regulators: Play a more enabling role

Better Disclosure: RBI and SEBI may help investors and other stakeholders take better decisions by pushing lenders to disclose more metrics about their portfolio performance on the lines detailed above in this article.

Database on Fraud: Further RBI may play a critical role in fraud risk management by creating a comprehensive database of frauds furnished by banks. The process of identifying and reporting fraud and suspected fraud may also be made smoother.

Enabling Risk Innovation: At times, even in certain developed nations excess scrutiny and restrictions on risk models tended to stifle innovations. Given India has among the richest data ecosystems in the world, RBI can play a critical role in supporting innovations in credit modelling. Here a principle-based approach focusing on the end outcome of the model may be useful for model risk management.

Financial inclusion remains an unfinished agenda. How the lenders, investors and regulators support each other to adapt to the new reality will determine the future of retail lending.

(Deep Narayan Mukherjee is a Risk Consultant and Visiting Faculty of Finance at IIM A & IIM C.)

Views are personal and do not represent the stand of this publication.