Non-Performing Loans: A crisis too long ignored

For years, Bangladesh's approach to bad loans — repeated rescheduling, periodic circulars, and rarely hard enforcement — treated the problem as containable. That approach has now visibly failed. The numbers tell a different story: one of a banking system that allowed its rot to accumulate for so long that the reckoning, when it finally came, arrived all at once.

At the end of March 2026, the gross non-performing loan (NPL) ratio in Bangladesh's banking sector stood at 32.26%. That means nearly one in every three taka lent by a bank in Bangladesh has gone bad. The trajectory is also telling: the ratio rose from 20.2% in December 2024 to 24.6% in March 2025, before peaking at a sharp 36.73% in September 2025. Although the ratio briefly moderated to 31.2% by December 2025, it deteriorated once again. Much of that spike was not new lending going bad in real time; rather, it was old rot finally being counted.

The scale of the crisis did not emerge from bad luck or a single economic shock; it is the product of a specific, well-documented set of institutional failures. Governance breakdowns, political interference, fabricated financial statements, and policy leniency towards wilful defaulters have been the primary drivers of the crisis, with its roots being fundamentally psychopolitical rather than purely economic.

State-owned commercial banks have consistently posted some of the worst asset quality in the sector, with their NPL ratio remaining above 41% over the past year, rising from 41.4% in March 2025 to 41.9% in December 2025 and 42% by March 2026. This pattern is tied to directed, politically motivated lending insulated from market discipline.

More alarming still is the trajectory at full-fledged Islamic banks, where the NPL ratio has more than doubled in a single year, surging from 29.2% in March 2025 to 56.5% by December 2025 and 58.4% by March 2026.

Legal enforcement has been just as weak. As of May 2026, 74,679 recovery cases involving nearly Tk4 lakh crore remained tied up in the country's Money Loan Courts, with a disposal rate of less than 10% over the last couple of years. This has turned legal recourse into a further delay tactic rather than a deterrent. Since early 2021, more than 2,150 writ petitions contesting bank seizures and auctions of mortgaged properties in default cases have remained stalled in the High Court, tying up more than Tk8,500 crore.

The consequences have moved well beyond the balance sheets of individual banks. With capital buffers eroded by mounting provisioning requirements, banks have grown reluctant to extend fresh credit, a squeeze that is slowing private investment, dampening job creation and weighing on the broader economic recovery. Compounding this is a second, less-discussed dynamic: heavy government borrowing is now crowding out private credit.

The International Monetary Fund (IMF) warns that over-relying on local banks to fund the government's budget deficit risks starving the private sector of credit and pushing the banking system past its lending capacity. Government securities' share of domestic debt rose to 62.3% in FY25 from 53.6% in FY24. 

The root cause is fiscal: tax collection fell sharply from 7.4% of GDP in FY24 to just 6.8% in FY25, forcing the government to rely on domestic bank borrowing to cover the shortfall, borrowing that then displaces private-sector access to the same pool of credit. The result shows up starkly in the data: private sector credit growth stood at 6.1% in December 2025, 6.03% in January 2026, and 6.03% in February 2026, before falling further still to 4.72% by March 2026.

Facing this compounding crisis, Bangladesh Bank has set an ambitious target of reducing non-performing loans within the next 18 months, with a seven-point structural roadmap. This combines tighter supervision and asset quality reviews, recovery-linked capital and provisioning rules, restructuring reserved only for viable borrowers, faster legal action against large defaulters, liquidity support kept separate from bank capital restructuring, new resolution and deposit-protection laws, and a shift to forward-looking loss provisioning under IFRS 9 by 2027. 

The crowding-out problem sits largely outside the roadmap altogether. None of the seven points addresses the government's own borrowing. Without fiscal reform, banks will keep finding government bonds more attractive than the harder, riskier work of private lending, blunting the impact of NPL clean-up on actual credit flows to the real economy.

Alongside it, the central bank has rolled out a one-time "Golden Exit" settlement scheme, a shorter write-off timeline, mandatory dispute resolution targets, and explicit NPL-reduction goals, 10% for state-owned banks, under 5% for private ones.

The honest answer sits between the two extremes: meaningful progress, but not a quick or complete fix. The roadmap is among the most structurally serious efforts Bangladesh Bank has produced. It moves away from the perpetual-rescheduling model that allowed the crisis to fester, towards genuine legal consequences, better data and international-standard provisioning. But several structural constraints limit how far it can go within the announced timeline.

Legal bottlenecks are the load-bearing constraint: no supervisory framework can recover assets faster than the courts can process them. Fast-tracking proceedings is necessary, but it will take years to clear the huge backlog. Political will remains untested. Every diagnosis of the crisis identifies political interference and connected lending as root causes. While the roadmap addresses supervision and legal processes, whether Bangladesh Bank will be allowed genuine operational independence to act against powerful, politically connected defaulters remains an open question.

The crowding-out problem sits largely outside the roadmap altogether. None of the seven points addresses the government's own borrowing. Without fiscal reform, banks will keep finding government bonds more attractive than the harder, riskier work of private lending, blunting the impact of NPL clean-up on actual credit flows to the real economy.

The Expected Credit Loss (ECL) transition is itself a multi-year project, with full IFRS 9 implementation not expected until 2027. This means the more rigorous, forward-looking loss-recognition framework that could identify emerging problem loans earlier will not be fully operational during the initial push.

Bangladesh has announced NPL-reduction schemes before, many of which were undermined by loopholes that allowed defaulters to delay repayment indefinitely. The credibility of this round rests on whether the Golden Exit and legal fast-tracking are genuinely one-time measures and are properly enforced, or become yet another avenue for preferential treatment.

Rebuilding genuine confidence in Bangladesh's banking system will likely take five to 10 years of sustained reform, not 18 months. Bangladesh Bank has, for once, produced a plan that treats the disease rather than the symptoms. Whether the political system allows it to work is the harder question and, based on the evidence of the last 20 years, the one that remains unanswered.

Sudeepto Roy is a research associate at South Asian Network on Economic Modeling (SANEM). Email: sudeeptoroy232@gmail.com 

Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the opinions and views of The Business Standard.