ILLUSTRATION: MATT KENYON
ON PAPER Nepal’s financial system looks enviable. Foreign-exchange reserves have touched $23bn, enough to finance more than 18 months of imports. The current account is in surplus to the tune of Rs553bn ($3.7bn). Inflation has fallen below 4%. Workers abroad sent home Rs1.45trn in the first eight months of the current fiscal year, a flow equivalent to nearly a quarter of GDP. By the usual metrics of emerging-market vulnerability, the land of Everest should be sleeping soundly.
Yet inside its banks, something has gone badly wrong. Non-performing loans have reached 5.42% of total advances, the highest in a generation. For finance companies, the smallest and weakest tier of the formal system, the ratio stands at 11.86%. Loan-loss provisions have eaten up Rs340bn, a sum bigger than the government’s entire annual capital budget. And the interest suspense account, where banks park income they have accrued but will never collect, has grown 165% in a year to Rs54bn. That last number is the one to watch. It means part of the profits banks report is not real cash, but figures created by accounting rules.
The standard diagnosis of such troubles blames a credit cycle. Lending booms, bad loans accumulate, banks tighten standards, the economy slows, and eventually the system cleans itself out. What makes Nepal different is that the cleaning mechanism has seized up. The country’s banks are trapped in a vicious circle, not a cycle. And the trap has six interlocking parts.
First, the NPL problem feeds on itself. As loans go bad, regulations force banks to set aside more provisions. Those provisions consume capital. Weaker capital leaves less room to lend. Less lending slows the economy. A slower economy generates more bad loans. The loop closes and tightens. Aggregate provisions have risen 43% for development banks year on year, slamming their capacity to lend.
Second, income is evaporating while costs are not. Interest income grew 13.4% in the first eight months of the fiscal year. That sounds healthy until you notice that staff expenses rose 14% and office operating costs rose 13.6%. The incremental revenue is being absorbed before it reaches retained earnings. Multiple banks now carry negative retained earnings, meaning cumulative losses exceed accumulated profits. Karnali Development Bank is technically insolvent, with negative capital of Rs3.5bn. It was placed under the control of Nepal Rastra Bank in 2024.
Third, banks are drowning in cash they refuse to use. Total deposits have crossed Rs7.8trn, up 12.4% from a year ago. Yet monthly credit growth was just 0.31% in the most recent data. Instead of lending, banks have parked Rs1.17trn in government securities, or 15% of all deposits. The credit-to-deposit ratio stands at 74%, well below the regulatory ceiling of 90% and far from the 80-86% range that prevailed before the crisis. Banks have some Rs2trn of headroom before hitting the limit. They are choosing not to use it.
Fourth, monetary policy has lost its power to persuade. The Nepal Rastra Bank, the central bank, has cut rates repeatedly. The weighted average deposit rate has fallen from above 8.5% to 3.45%. The base lending rate has dropped in lockstep. But credit growth has barely budged. This is not a puzzle. Rate cuts work when the constraint on lending is the cost of funds. In Nepal the constraint is fear. Banks do not want new NPL, and no reduction in the repo rate changes their capital position or their memory of the last wave of defaults. The central bank is pushing on a string.
Fifth, capital ratios lie. Commercial banks report core capital of 9.61% of risk-weighted assets and total capital of 12.64%, both above regulatory minimums. Those numbers assume that reported income is real. It is not. The interest suspense account alone, if properly provisioned, would wipe out a significant slice of stated capital. And the risk weights themselves may understate danger if loan classification is lagging reality.
Sixth, there is no way out. Nepal lacks any credible mechanism for disposing of bad loans. India has asset reconstruction companies. Bangladesh has a dedicated NPL tribunal. Sri Lanka built a post-crisis resolution framework. Nepal has voluntary restructuring, which turns bad loans into bad loans with longer maturities and lower interest rates. The merger drive pushed by the central bank has reduced the number of financial institutions from more than 200 to 106, but merging two stressed balance sheets produces one larger stressed balance sheet. Resolution requires a buyer. There is none.
The macroeconomic context makes this trap all the more frustrating. The external sector is stronger than it has ever been. Remittances alone have grown 37.7% year on year. The government’s fiscal deficit is manageable. Inflation is so low that it signals weak demand rather than price stability. Real GDP growth is expected to slow this year, according to the World Bank. Gross fixed capital formation has fallen from 33.8% of GDP in 2019 to 24.1% now. In other words the economy is idling.
Who breaks the circle? The central bank could tighten provisioning rules, forcing banks to recognise losses faster. But that would push some institutions below capital thresholds, triggering runs or bailouts. The government could create a state-backed asset reconstruction company, capitalised at, say, Rs20bn, to buy bad loans at market discounts and work them out over half a decade. That would clear balance sheets without destroying them overnight. But such a vehicle would need political will that is conspicuously absent, not least because the banking sector is intertwined with business and political networks that benefit from forbearance.
The most likely outcome is protracted stagnation. NPL will stabilise in the 5-6% range. Credit growth will stay below 8%. GDP will chug along at 4-4.5%. A finance company or two will fail quietly. No single event will trigger a crisis. But the economy will remain dependent on remittances, unable to generate the kind of productive investment that raises wages and builds infrastructure. The window for non-disruptive reform is probably two or three years. After that, the odds of a remittance shock or a geopolitical crisis large enough to break confidence rise materially.
Banks are flush with cash and paralysed by fear. They are not about to collapse. But they are not about to lend, either. And until someone builds a mechanism to clear the deadwood from their balance sheets, they will stay what they have become: a drag on the economy rather than an engine of it. ■
