India's MSME ABS market has moved through a rapid maturation phase over the past five years. Issuance volumes that stood at roughly ₹8,000 crore in financial year 2019–20 have exceeded ₹22,000 crore in more recent periods, driven by the expansion of NBFC and fintech origination into the small and micro-enterprise segment and by growing appetite from domestic mutual funds, insurance companies, and banks seeking PSL-qualifying assets. That growth has brought with it a broader investor base and tighter pricing — but it has also brought a cohort of investors into MSME pools whose analytical frameworks were built for more predictable asset classes and have not always been recalibrated for the specific characteristics of MSME credit.
This note covers the structural features of MSME ABS that differ from other retail asset classes, what the rating agency methodologies are measuring and where they diverge, and what an analytically sound view of an MSME pool looks like in practice.
From bank books to structured pools: the originator shift
The MSME securitisation market was, for most of its history, dominated by pools originated by established NBFCs with multi-year track records — STFC, Mahindra Finance, and similar lenders whose underlying books were more vehicle-secured or equipment-secured than true working-capital MSME. The expansion in volume over the past five years reflects the entry of a second cohort: fintech-originated pools, where the lending model relies on GST data, bank account statement analysis, and bureau scores rather than physical asset security or relationship-based credit assessment.
The distinction matters for securitisation analytics because the two originator types produce pools with different cash flow characteristics. Bank and established NBFC-originated MSME pools tend to show more stable monthly collections, lower prepayment rates, and default curves that follow established seasonal patterns. Fintech-originated pools often show higher prepayment rates (reflecting the short-tenor, revolving nature of much fintech MSME lending), less predictable seasonal patterns, and default curves that are younger in the vintage sense — meaning there is limited stress-cycle performance data on which to base stress scenarios.
Investors who apply the same base-case loss rates and stress multipliers to both originator types are not conducting equivalent analysis. A 3% base-case net credit loss assumption may be well-supported by historical data for a seasoned NBFC originator with a 12-year pool performance track record. The same assumption applied to a fintech-originated pool with three years of origination history and no through-cycle data carries a fundamentally different confidence interval.
Cash flow volatility: why MSME is not like residential mortgage
The benchmark securitisation asset class in most markets — and the one around which investor credit frameworks are typically calibrated — is residential mortgage. Residential mortgage pools have predictable amortisation schedules, low intra-period volatility, and defaults that follow observable housing cycle patterns with a long lead time. MSME pools are none of these things.
MSME borrowers are, by definition, exposed to business-cycle volatility in a direct way that salaried borrowers are not. A manufacturing MSME's ability to service debt in a given month depends on whether its receivables collection was on time, whether its key customer paid, and whether input costs moved in a way that compressed margins. This creates a lumpy, episodic pattern of defaults and delinquencies that does not lend itself to the smooth loss-rate curves used in residential mortgage pool analysis.
The practical implication is that MSME pool modelling requires dynamic cash flow scenarios that account for correlated stress — the scenario where one economic shock generates simultaneous delinquency across a geographically or sectorally concentrated portion of the pool — rather than independent loss distributions. MSME pools that look well-diversified at the borrower level may still carry significant concentration risk at the sector or geography level. A pool of 1,200 MSME loans spread across Maharashtra and Gujarat, where 400 of those borrowers are textile manufacturers, is not as diversified as the borrower count implies.
Rating methodology differences: CRISIL versus ICRA
Both CRISIL and ICRA publish methodology frameworks for MSME securitisation ratings, and both apply stress scenarios to base-case pool assumptions to arrive at credit enhancement sizing. The methodologies have converged over time but retain meaningful differences in how they treat originator-specific factors, bureau reliance, and geographic concentration.
CRISIL's MSME pool methodology places significant weight on the originator's servicing capability and collection infrastructure. For established NBFCs with dedicated collection teams and field staff, CRISIL's originator assessment provides credit in the form of lower stress multipliers on base-case loss rates. For fintech originators who rely predominantly on digital collections and have limited physical recovery capability, the originator assessment carries a more adverse adjustment. The practical consequence is that identical pool-level characteristics — same loan tenor, same geographic distribution, same credit score distribution — attract different CE requirements depending on originator type.
ICRA's methodology places comparatively heavier emphasis on pool-level bureau data quality and the completeness of borrower-level data disclosure. ICRA has been more explicit in its criteria about the consequences of missing or stale bureau data within a pool — a feature that reflects the Indian MSME lending environment, where a meaningful proportion of MSME borrowers across fintech lenders have partial or inconsistent bureau histories. Where CRISIL may accept an originator's internal credit model outputs as supplementary support for pool quality, ICRA typically requires bureau-validated data for a higher proportion of the pool.
These methodology differences are not simply academic. A transaction rated by CRISIL may attract a materially different CE sizing than the same pool rated by ICRA, and the gap can be wide enough to change the economics of a transaction from viable to unviable. Structured credit investors who rely on the rating alone, without understanding the methodology that produced it, are not equipped to assess whether the CE is sized at the right level for their own risk tolerance.
Geographic concentration: the underweighted risk factor
Geographic concentration is consistently the most underweighted structural risk in Indian MSME securitisation. Most rating agency frameworks include geographic concentration in their analysis, but the way concentration risk is typically haircut — through incremental stress on loss rate assumptions — does not adequately capture the tail risk of localised economic stress events.
The AP/Telangana over-lending episodes (which the MFI market experienced acutely and which have a parallel in MSME lending in those states), the impact of state-level tax or GST enforcement actions on specific industrial clusters, and the effect of infrastructure disruption on corridor-dependent MSME logistics businesses are all examples of localised stress that can produce correlated defaults across a geographically concentrated pool that a rating stress scenario does not fully capture.
Pools where more than 30–35% of the outstanding portfolio is concentrated in a single state, or where more than 15–20% is in a single industrial cluster, require explicit scenario analysis for localised stress — not simply a haircut to the pool-level loss rate. The CE structures of many MSME ABS transactions currently in the market were sized before this risk was fully priced; investors in seasoned tranches from programmes originated between 2021 and 2023 should consider whether the CE available today, after amortisation of the pool, is still adequate against a concentration stress.
Bureau data reliance and what it does not capture
The expansion of MSME credit bureau coverage — through CRIF High Mark, Experian, and CIBIL — has improved the quality of underwriting data available to MSME lenders materially over the past decade. Bureau penetration in the formal MSME lending segment is now high enough that bureau-based origination decisions are the norm rather than the exception for NBFC and fintech originators.
What bureau data does not capture is the true indebtedness of an MSME borrower across informal or unreported credit channels. Bureau data reflects reported institutional lending. It does not reflect outstanding dues to trade creditors, informal moneylender borrowings, or intercompany financing within business groups. For MSME borrowers — particularly micro-enterprises in semi-urban and rural markets — these unreported obligations can represent a significant portion of total financial leverage. The consequence is that bureau-based underwriting systematically understates the debt-service burden of the most informally-financed segment of the MSME market.
Pools with a higher proportion of micro-enterprise exposures, or pools originated in markets with lower formal financial inclusion, carry a bureau data adequacy risk that larger-ticket MSME pools do not. This is rarely quantified explicitly in transaction documentation or rating reports, but it is material in stress scenarios.
CE sizing challenges and what good structure looks like
Credit enhancement sizing for MSME ABS is more difficult than for vehicle or housing loan pools because the historical loss data is shorter, the loss volatility is higher, and the concentration risks are harder to capture in a single-factor stress model. The consequence is that CE levels for MSME ABS have historically been set with less precision than for more established asset classes, with rating agencies applying larger stress multiples to compensate for data uncertainty.
A well-structured MSME ABS transaction has several features that distinguish it analytically from a transaction that simply passes a rating committee minimum. The pool is large enough — typically at least 500 borrowers and ideally above 1,000 — to ensure that individual-loan defaults do not drive disproportionate pool-level loss variability. Geographic and sectoral concentration is explicitly disclosed and capped, with the CE analysis including a specific scenario for maximum permitted concentration. The originator's collection performance is disclosed at the pool level with sufficient vintage detail to construct a loss curve, and that curve is updated on a periodic basis throughout the transaction's life.
Overcollateralisation and cash reserves should be sized to cover at least 1.5 times the rating agency's base-case loss estimate for transactions rated at AA and above, with additional stress scenarios carried at the transaction review stage rather than deferred to a rating review trigger. Excess spread — the difference between the pool yield and the sum of senior costs and tranche interest — is the first line of CE absorption and should be modelled explicitly across the life of the pool, including in scenarios where prepayment rates are significantly higher than base case, which compresses excess spread duration.
Investor appetite for MSME ABS has proven resilient even through periods of MSME credit stress, but that appetite is not uniformly sophisticated. The pricing convergence between well-structured and poorly-structured MSME ABS transactions that has characterised the past two years of a favourable credit environment is a warning sign rather than a comfort. When the next MSME credit cycle turns, the structural differences will matter more than the spread differential currently implies.
For pool-level analytics, CE sizing review, or originator assessment for MSME ABS transactions, get in touch.
