Loan Pricing – The Holy Grail for Establishing Just Impacts from Credit Inclusion

Retail credit is essentially any form of formal credit accessed by individuals, households, associations of persons, unincorporated sole proprietorships and any other unsophisticated legal persons in a country. A key objective of inclusive financial systems relates to how its credit ecosystem prices credit, and in doing so, whether it is able to achieve risk ordinality, and how efficiently risk ordinality might be manifesting. This is very salient to the success of retail credit markets development. We look at this for India.

Risk ordinality means loan pricing should preserve the correct ranking of risk across borrowers, that is, there must exist a monotonic relationship between risk and price of credit, atleast theoretically. The relationship is likely to be monotonic when considering the same product, tenor, collateral, ticket size, competitive conditions, regulatory treatment and in the absence of any interest rate caps or subsidies. At an aggregate level, high-risk borrowers should be charged a higher interest rate than low-risk borrowers.

The Report of the Reserve Bank of India's (RBI) Committee on Comprehensive Financial Services for Small Businesses and Low-Income Households¹ (Chair: Dr. Nachiket Mor, hereafter called Mor Committee) in 2014, stated the desired outcome for its vision of 'Sufficient Access to Affordable Formal Credit'² is for affordable rates (the rate here is the price at which borrowers can avail credit) to be "ordinal by risk level in the long-run after adjusting for reasonable transactions charges". This means that there must be an ordering of the risk-interest rate combinations, and if the ordering breaks down, risky borrowers get cheaper loans than safer ones, and credit ends up flowing to the wrong borrowers, crowding out productive but lower-risk investments. If pricing does not reflect the inherent risk order, then safer borrowers may exit the market because they are overcharged, leaving mostly high-risk borrowers (case of adverse selection). This weakens the entire lending system. The Mor Committee had found very high levels of violations of ordinality within the formal system and in the economy as a whole then (see Tables 2.2.4 a and b in the report).

Interest rate charged to a borrower is a sum of cost of capital for the lender (A), the operating expense (B), expected losses or loan loss provisions (C), unexpected losses or cost of equity, and a profit margin on top of all these. Cost of capital is cheaper for banks than NBFCs as they have access to low-cost retail deposits. Operating costs increase for serving difficult-to-reach borrowers and for monitoring them, and decrease with adoption of technology if used well. Expected losses (indicated by past performance of a particular asset class) are typically arrived at using credit bureau scores and credit ratings for individuals and corporates respectively. The credit information company (CIC) regulations under RBI and the credit rating agency regulations under SEBI govern the production of credit scores and ratings respectively, and are meant to uphold the ordering of risk needed to maintain the risk ordinality principle.

It is widely accepted that the more difficult-to-access borrowers, the vast under-served segments of India's population that matter for delivering just impacts, present larger informational asymmetries than the easy-to-access borrowers; they are the more difficult and more expensive to underwrite and monitor; and monitoring itself can shape credit positive behaviors. Better monitoring can in turn reduce expected losses and improve credit pricing over time.

There are specific pools of data that capture the borrowing costs for various categories of borrowers. RBI's data on interest rate ranges charged for loans to 'household sector -individuals' and 'household sector – others', and data from national surveys such as the All India Debt and Investment Surveys (AIDIS) and India Human Development Surveys (IHDS) are known to have these variables. However, they are ill-equipped to directly inform whether –

1. Pricing of loans follows the risk ordinality principle for under-served borrowers in the last 10 years since the Mor Committee's analysis
2. Borrowers who have demonstrated credible history of timely repayments over a period of time have experienced commensurate improvements in their credit scores and thus reductions in loan pricing
3. Mispricing of credit risk is being discovered efficiently and removed or is accumulating at various levels

Since these datasets are not panel datasets following the same set of respondent households, it is difficult to ascertain whether a borrower has experienced improved loan pricing over his/ her lifetime of engaging with the lending ecosystem. Further, for new-to-credit borrowers, analysing their score journey across time and across different CICs can indicate how smoothly and how uniform the score-building approaches are across different CICs.

RBI could construct a household panel and require CICs to provide data on loan amounts, loan types, sources of borrowing, pricing, and repayment behaviours across time for such a panel of borrowers, and score transitions. RBI can test whether pricing sharply jumps after certain risk thresholds, whether certain borrower classes are excluded entirely, and whether the presence of collateral dramatically changes pricing (indicating the bias for lenders towards reducing loss given default rather than ensuring cashflow-based appropriateness assessments). A subset of borrowers can be surveyed or interviewed to collect valuable qualitative information on them, which, when combined with the CIC data, can help to answer the question of whether risk ordinality is being upheld in India's lending ecosystem for retail borrowers.