The Firm Investment Approach Insights Newsroom Leadership Contact LP Portal
47th Floor, Kohinoor Square
Dadar West, Mumbai 400028
investments@areion.in

How large India’s stressed debt market is, and who holds it now

The gross non-performing asset ratio of Indian banks fell from 11.2 per cent in March 2018 to 2.8 per cent in March 2024, but much of the debt behind that fall was written off or sold and remains legally recoverable. Counted across every holder, the stock is far larger, and the domestic capital committed to it is small.

October 202610 min read

Contents
  1. How the 2015 review made the stock visible
  2. Where the stock went: holders outside the bank ratio
  3. How large the stock is
  4. Stock, transactable volume and capacity
  5. Why the stock replenishes
  6. Why committed domestic capital is still small

The figure most often quoted for Indian credit stress is the gross non-performing asset (NPA) ratio of scheduled commercial banks: loans overdue for more than ninety days, as a share of total bank advances. It fell from 11.2 per cent in March 2018 to 2.8 per cent in March 2024. Read on its own, that describes a banking system that has dealt with its bad loans.

A bank can bring the ratio down by recovering a loan, writing it off or selling it. Only recovery extinguishes the claim. A write-off removes the loan from the numerator while the debt, and any security behind it, survives in law. A sale moves the loan to another holder, usually an asset reconstruction company (ARC), which continues to pursue it. The fall between 2018 and 2024 drew on recoveries, write-offs and sales, supported by provisioning and the recapitalisation of banks.

The ratio therefore measures one holder at one point in time. It leaves out debt that has already left bank balance sheets and stress forming at lenders outside the banking system, which makes it the narrowest of the available measures of the market.

11.2% to 2.8%

Gross non-performing asset ratio of scheduled commercial banks, March 2018 to March 2024. The decline combines recovery, write-off and sale.

How the 2015 review made the stock visible

Before the ratio could fall, the stress had to be recorded. For most of the decade before 2015, Indian banks kept failing loans out of the non-performing category through repeated restructuring. The bank extended new terms, the borrower made a token payment and the account was re-classified as standard. Each side gained in the short term: the bank avoided the provisioning charge that classification would have triggered, and the borrower avoided enforcement.

The cost fell on the wider economy. Capital stayed locked in enterprises that were no longer productive, banks carrying such loans had less capacity to lend to growing businesses, and a borrower who defaulted strategically faced years of negotiation before any real threat to the collateral.

The Reserve Bank of India (RBI) forced recognition through its Asset Quality Review, begun in 2015. Banks were directed to examine their restructured portfolios against objective criteria and to classify accounts as non-performing wherever the evidence required it. The gross NPA ratio of scheduled commercial banks rose from 4.3 per cent in March 2015 to 11.2 per cent in March 2018. The increase recorded stress that had accumulated behind accounting adjustments for a decade or more; very little of it was new.

Recognition changed the incentive to sell. Before the review, selling a loan to an ARC meant crystallising a loss that restructuring allowed a bank to defer. Once a loan had been classified and provided for, that reason disappeared. A sale, even at a steep discount, released capital, improved the reported ratio and freed management time for performing business. The review turned a hidden supply of stressed loans into a visible one that banks were motivated to sell.

A market for distressed assets cannot exist until the assets are acknowledged to be distressed. Recognition was the precondition for pricing, and pricing was the precondition for transfer.

Where the stock went: holders outside the bank ratio

The fall in the bank ratio after 2018 moved a large part of the stock to holders that the ratio does not count. Three matter for an investor, and each behaves differently as a seller.

Written-off bank books

Figures cited in early 2026 put cumulative write-offs by Indian banks at approximately ₹19 lakh crore over eleven years. A write-off is an accounting entry with no legal effect on the debt. The claim and the security interest both survive, and the bank keeps its right to recover. The only change is that the exposure is carried at nil or near-nil book value.

That produces an unusual kind of seller. A bank holding a written-off claim has no capital tied up in it, little reason to run a long enforcement process on a small account and a plain reason to sell the claim for cash. Book value plays almost no part in the price, which is set by negotiation between what the bank will accept and what a buyer expects to recover.

Reconstruction trusts

ARCs are licensed by the RBI under the Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interest Act, 2002 (SARFAESI) to acquire non-performing loans and pursue their recovery. Loans sold to them leave the bank balance sheet entirely and sit in trusts that issue security receipts against them. The aggregate held this way is substantial and its quality varies widely. Some trusts hold well-secured loans bought at sensible prices; others hold older pools acquired when competition among buyers was heavier and diligence lighter. Positions in existing trusts can occasionally be bought from holders who no longer wish to wait for resolution (see “How stressed loans change hands in India: reconstruction companies, security receipts and special situation funds”).

Non-bank balance sheets

The third holder is the non-bank lending sector: non-banking financial companies (NBFCs), housing finance companies and, in small-ticket retail credit, fintech lenders. Their balance sheets fall outside the bank ratio altogether, and much of the stress now forming in Indian credit sits there (see “Where India’s next stressed assets are forming: non-bank loan books, stalled housing projects and renewable energy plants”). Loans to micro, small and medium enterprises (MSMEs) cut across all of these holders, spread between banks and non-banks alike.

How large the stock is

There is no single figure for the size of the Indian stressed debt market, because there is no agreed definition of what belongs in it. The most useful estimate divides the stock by segment, since each segment is priced and resolved differently. The table below gives estimates for each segment as at April 2026.

SegmentEstimated stressed poolTypical acquisition discountMain recovery route
Bank NPAs (legacy)₹4 to 5 lakh crore40 to 60 per centSARFAESI enforcement, insolvency, settlement
NBFC and housing finance portfolios₹1.5 to 2 lakh crore50 to 70 per centGranular portfolio servicing
Mid-market real estate₹1 to 1.5 lakh crore40 to 65 per centProject completion, asset sale
Renewable energy project companies₹30,000 to 50,000 crore35 to 55 per centOperational turnaround
MSME credit₹2 to 3 lakh crore60 to 80 per centSettlement, collateral enforcement
Total₹8.8 to 12 lakh crore
Estimates as at April 2026, drawn from RBI Financial Stability Reports, Insolvency and Bankruptcy Board of India bulletins, Securities and Exchange Board of India data on Alternative Investment Funds and public transaction data. Discounts are against outstanding loan value. The total is the sum of the rows.

The rows sum to ₹8.8 to 12 lakh crore. Legacy bank NPAs remain the largest single segment, but more than half of the estimated pool lies in the other four. MSME credit is the second largest segment and the one bought at the deepest discounts. Renewable energy is the smallest and is bought at the narrowest discount; its assets are operating plants with contracted revenue (see “Where India’s next stressed assets are forming: non-bank loan books, stalled housing projects and renewable energy plants”).

Higher totals are sometimes quoted for India’s stressed pool. A total well above the sum of these rows would have to include categories the table leaves out, of which written-off exposure is the largest, and the written-off figure cannot simply be added. The ₹19 lakh crore records what banks wrote off over eleven years. It does not record how much of that remains outstanding and enforceable today.

Stock, transactable volume and capacity

Three different quantities are often presented as the size of the market, and they answer different questions.

MeasureWhat it countsFigure and dateWhat it is useful for
Stock of stressed debtStressed loans across the five segments, wherever they are held₹8.8 to 12 lakh crore, April 2026The scale of what the system has to resolve
Transactable volumeStressed assets likely to change handsAbout USD 7 billion in 2021, about USD 14 billion in 2025, a projected USD 17 billion in 2027The amount buyers can realistically deploy against
Annual addition from new bank lendingStress expected from one year’s new bank credit within five years₹40,000 to 60,000 crore a year, April 2026A floor for how fast the stock is replenished
Stock and annual addition: estimates as at April 2026. Transactable volume: third-party estimates for 2021 and 2025 and a projection for 2027.

Transactable volume is far smaller than the stock. At any plausible exchange rate, USD 14 billion is a small fraction of ₹8.8 lakh crore. The gap is expected, because part of the stock is resolved by the original lender through restructuring, settlement or its own enforcement and never reaches a buyer. Transactable volume is also the measure most sensitive to regulation, since each widening of the channels through which lenders may sell stressed loans, such as direct sales to special situation funds, raises it.

For an investor, none of these totals is the operative number. What matters is how much capital a given buyer can deploy at prices that leave an adequate margin for recovery. That depends on the buyer’s sourcing, its ability to service and enforce in the regions where the collateral sits, and the segments it can actually resolve. A buyer with strong origination in one segment can reach its capacity limit far below the size of the market, and capital raised beyond what its pipeline can absorb tends to be pushed toward assets it should not buy.

Why the stock replenishes

The annual addition in the table rests on two inputs. Indian banks originate approximately ₹20 lakh crore of new credit a year (April 2026), and non-bank lenders add a significant further volume. If two to three per cent of a year’s new bank lending turns non-performing within five years, each year adds ₹40,000 to 60,000 crore to the stressed pool from bank credit alone. Historically the rate has been higher. The figure is a floor, because it counts only one year’s new lending and leaves out fresh defaults on loans already on the books. A broader measure that applies an annual formation rate to the whole credit stock is set out separately (see “Where India’s next stressed assets are forming: non-bank loan books, stalled housing projects and renewable energy plants”).

A stock that is resolved and not replaced shrinks toward nothing. India’s does not, because new stressed debt is generated by the ordinary working of a large and growing credit market. Every period of credit growth produces a cohort of loans that will not perform, and the stock is replenished as each cohort is recognised. Its size rises and falls with the cycle, and the market for it is a lasting part of the financial system.

In the United States an institutional market for stressed debt has operated for decades, sustained by the same regeneration of credit stress. India’s market is at an earlier stage, and its legal framework is newer and national in scope: the Insolvency and Bankruptcy Code of 2016 applies uniformly across the country, and enforcement of security interests follows one statute everywhere, even though its speed varies from state to state.

Why committed domestic capital is still small

Against a stressed pool of ₹8.8 to 12 lakh crore, the domestic capital committed to buying it is modest. Total commitments to Alternative Investment Funds (AIFs), the pooled vehicles regulated by the Securities and Exchange Board of India (SEBI), were approximately ₹9 lakh crore as at April 2026, up from negligible levels a decade earlier. Less than five per cent of that was deployed in distressed credit, which puts AIF capital in the strategy below ₹45,000 crore, at most about five per cent of the lower estimate of the pool. ARCs also buy through their own trusts, but the gap between the stock and the specialist capital available to buy it remains wide.

The reasons are largely institutional and psychological. The economics of the asset class changed with the reforms of 2002 and 2016; investor perceptions have moved more slowly.

  • Stigma and institutional memory. Non-performing loans are associated with failure, and domestic allocators remember the years when a valid secured claim could take a decade to enforce. An investor comfortable with a growth equity fund, where the whole investment can be lost, may still avoid a loan secured against tangible collateral because of the word “distressed”.
  • Unfamiliarity among advisers. Most wealth managers and advisers in India have no training in distressed credit, and the distribution network able to present the strategy accurately is still small.
  • A short record. Allocators look for realised outcomes across a full cycle. The insolvency code dates from 2016, so the framework under which such a record can be built is young.
  • Operational intensity. The strategy needs legal, forensic and asset management capability that most credit managers lack and cannot quickly acquire.
  • Illiquidity. Investors who built wealth in listed equities are used to daily prices and daily exit. Capital in a stressed debt strategy is committed for several years with no interim exit, which sits outside many mandates.
  • Headline risk. Stressed investing draws scrutiny, and some allocators prefer to forgo the strategy than to explain the position to their own boards or clients.

What is changing

Several of these barriers are weakening. The special situation fund, a sub-category of Category I AIFs under SEBI’s rules, gives the strategy a regulated domestic vehicle designed to invest in stressed loans (see “How stressed loans change hands in India: reconstruction companies, security receipts and special situation funds”). Wealth managers have begun to include distressed credit in their advisory frameworks. The growth of AIF commitments over the past decade has also made investors more familiar with closed-ended, illiquid structures, which lowers resistance to long commitments.

The scarcity of committed capital has a direct effect on price. The discount compensates a buyer for an uncertain timeline and the cost of enforcement, and with few buyers competing for each portfolio, sellers accept wider discounts than those costs alone would justify. As committed capital grows, competition should increase and discounts should narrow, which makes pricing discipline and access to bilateral transactions more important to each buyer.

Prepared for educational purposes only. Not an offer to sell securities, investment advice or a recommendation. Views are as at the date of publication and are subject to change.