By Deep Narayan Mukherjee & Abhinav Bansal, respectively partner and associate director, & managing director and partner, BCG

Microfinance institutions (MFIs) are critical for facilitating credit access and thereby income generation for some of the economically weakest sections in India. For two decades, they have been fulfilling their mission while expanding their footprint. Pre-2022, there were periods when delinquency spiked in MFI portfolios, triggered by exogenous shocks. These shocks such as a sharp economic downturn, Covid-19, or local political issues exposed the MFI portfolio to systemic risks. The portfolios bounced back to health once the exogenous shocks subsided.

However, the gradual deterioration of MFI portfolios from 2022 till date is different. There are no external shocks. Arguably, it was specific risks driven by borrowers who ended up having or being given more debt than they can pay back. The current deterioration in MFI portfolio could be primarily attributed to a lapse in risk management practices on the part of at least some entities. As collection efforts by MFIs intensified, sociopolitical interferences crept in over debt collections. But these political moves were not the trigger of the current MFI stress. In the run-up to the current situation, the Reserve Bank of India (RBI) has also raised concerns about the underwriting practices and risk appetite of MFIs.

Despite pockets of excellence, if the typical MFI lender does not improve risk management capability, the portfolio quality is unlikely to improve.

Complacency of 99% collection

Historically, the typical MFI lender would collect 99% or more of the amount outstanding with borrowers. This level of collection efficiency may be the envy of large retail lenders. MFIs in normal times exhibited this performance due to the joint liability group (JLG) mechanism. Here, five to 10 borrowers form a group. This group becomes jointly responsible to ensure that the MFI loans to all members of the JLG are paid back. The JLG tended to minimise borrower-specific risks, while the peer support and pressure ensured low delinquency. However, when systemic risk events impacted all members of a JLG, they would default. The success of the business model in the past as well as spikes in MFI defaults only during black swan events possibly created some hubris.

Lessons not learnt

The MFIs have possibly missed the lessons of the retail credit crisis of 2007-08. Back then, few lenders provided small-ticket personal loans (STPLs) with limited background check of the borrowers’ credit history. Further, basic credit scorecards with low predictive power, predominantly judgemental credit policy, and reactive portfolio management aggravated the credit blow-up. Worryingly, a lot of the erstwhile weaknesses of retail lenders are present in MFI lenders. However, after the STPL blow-up in 2007-08, successful retail lenders changed their lending approach. Their decision-making became more date-driven, improved analytical rigour, and raised the standards of credit policy and governance. It has served retail lenders well until of late. MFIs need to travel on that path.

Regulatory guidelines

The RBI has been justifiably worried about some of the lending practices, and thus tightened certain guardrails. This included classifying MFI borrowers as belonging to households with income below `3 lakh per annum. It capped debt service requirement at 50% on monthly income. However, constrained techno-analytical and data infrastructure, at least for a section of MFIs, has prevented comprehensive implementation of such risk guidelines.

It is not uncommon to find MFIs who use judgemental or quasi-analytical rule of thumb to assess household incomes. Discretion has remained in the assessment of debt servicing ability, which made some lending decisions suboptimal, ultimately creating over-leveraged borrowers. Further, certain lenders adopted aggressive loan upsell strategies. These loans were often given outside the JLG structure without rigorous assessment of debt servicing abilities. The JLG-imposed credit discipline was breached. Thus, it is not surprising that MFI default rates spiked even where there was no economic shock. Clearly, the sector needs to reinvent itself in terms of its risk management capabilities.

The MFI sector needs to work on the following dimensions:

More and better data use: For existing customers, the behavioural risk model (B-Score), which is a staple for most retail borrowers, is often missing in MFIs. They need to tap into internal behaviour data better for predicting individual and JLG debt servicing patterns. In India, non-traditional data sources are maturing faster than anticipated. Such sources like locational profiles, payment app behaviours, and satellite data, when used for predicting credit risk, are seen to have predictive power over and above credit bureau data-based models.

Analytics need to leapfrog: Seasoned MFI underwriters have astute knowledge of borrower behaviour and locational dynamics. However, this is not institutionalised. Building credit models which marry this institutional insight with advanced data analytical capabilities will provide powerful and explainable credit score. Such transparent risk models may address questions raised by regulators and stakeholders about underwriting quality and risk-based pricing.

Portfolio monitoring: For a business exposed to macro risks, MFIs typically do not have stress testing models to simulate the loss rates under different scenarios. A stress testing capability can forecast losses at least four to six quarters ahead for alternative macro-economic scenarios. This will enable MFIs to adjust credit policy to minimise losses.

Risk culture and governance: The lack of digitalisation and weak data analytical capability sometimes breeds poor risk culture and governance. In such set-ups, tracking compliance with regulations and governance norms becomes challenging. The problem is aggravated if lending decisions are predominantly judgemental, supported by broadly defined credit policies. Discretion and opportunistic “cutting the corners” creeps in. Digitalisation of the underwriting process will enhance governance and support risk culture.

MFIs must make a course correction quickly. An impaired MFI sector will be detrimental to the credit access of the weakest sectors in India.