There is a number making the rounds this Bhadra that will please shareholders and mislead almost everyone else. Nepal's 20 commercial banks booked a combined net profit of NPR 69.86 billion in FY 2025/26, up a striking 32.28 percent from NPR 52.81 billion the year before. On paper, the banking sector has staged a handsome recovery.

Read the fine print, though, and a different story emerges. This is not the sound of an engine roaring back to life. It is the sound of a household paying its bills by selling furniture and skipping the mechanic. The profit is real. The health it is supposed to signal is not.

Let me explain why, because the distinction matters far beyond bank boardrooms.

The core engine has stalled

The heart of banking is simple: take deposits, lend them out, earn the spread. That spread, aggregated across the system, is Net Interest Income (NII), and it is the truest measure of whether banks are doing their actual job of moving money from savers to borrowers.

Last year, that engine barely turned. NII crept up just 1.27 percent, from NPR 189.95 billion to NPR 192.36 billion. In an economy with any inflation at all, standing still by 1.27 percent is going backwards.

Why? Because deposits are flooding in while credit refuses to move. Deposits grew 14.93 percent, an addition of roughly NPR 962 billion. Credit grew a limp 6.55 percent, or NPR 303.81 billion. Put plainly: for every three rupees Nepalis deposited, banks could lend out barely one. The rest piles up as idle liquidity, parked back at Nepal Rastra Bank for a pittance while banks keep paying interest on the deposits that funded it.

The individual numbers are starker still. Rastriya Banijya Bank pulled in deposits up 37.19 percent against credit growth of 13.60 percent. Standard Chartered grew deposits 17 percent while its loan book actually shrank 9 percent. Citizens Bank grew deposits nearly 10 percent and lent out less than 1 percent more. And at NIMB, NIC Asia, and Prabhu, credit went outright negative year-on-year.

A bank whose loan book is shrinking is not being prudent. It is being starved. When credit contracts across the country's largest lenders simultaneously, that is not a banking story. It is a demand story, and it is telling us the real economy has run out of appetite.

So where did the profit come from?

Three places, and none of them is lending.

First, non-interest income. Banks leaned hard on fees, LC commissions, remittance charges, and digital banking, lifting fee and commission income 9.91 percent to NPR 39.49 billion. More dramatically, treasury desks feasted: net trading income surged 39.97 percent to NPR 13.11 billion, helped by NRB relaxing the ceiling on Non-Deliverable Forward positions. Kumari's trading income jumped 94.22 percent; Laxmi Sunrise's rose 94 percent. Impressive. Also entirely disconnected from whether a single factory got financed.

Second, and most importantly, the provisioning writeback. This is the quiet lever that did the heaviest lifting. Loan-loss provisioning across the sector fell 27.65 percent, from NPR 51.45 billion to NPR 37.37 billion. The share of NII being swallowed by provisions dropped from over 25 percent to under 20 percent.

Some of this reflects genuine recoveries. Much of it is an accounting release. When a bank writes back a provision, no new money enters the building; a cushion set aside earlier is simply reclassified as profit. Kumari Bank is the poster child: it swung from a NPR 3.45 billion provisioning charge to a NPR 1.44 billion writeback and watched operating profit leap 305.96 percent. Nepal Bank did the same, flipping a NPR 2.25 billion charge into a NPR 1.01 billion writeback for a 61.68 percent operating profit gain. Both cleared accumulated losses and can pay dividends again.

I do not begrudge them the turnaround. But we should be honest about its nature. A one-time writeback flatters this year's income statement and cannot be repeated next year. It is a sugar rush, not a balanced meal.

This is why the aggregate conversion ratio looks so good: banks turned 42.50 percent of total operating income into operating profit, up from 36.15 percent. A genuine improvement in efficiency, yes, but one substantially borrowed from provisions that will not be there to release again.

The four banks that should keep us awake

Here is where a central banker's eye stops moving to the averages and fixes on the outliers, because systemic risk never lives in the average.

Four institutions are flashing red. NIC Asia, once the sector's most aggressive balance-sheet builder, saw net profit fall 39.21 percent, its NPL ratio climb to 9.26 percent, and its core capital ratio sink to 6.27 percent, far below the 8.50 percent regulatory floor, on accumulated losses near NPR 15 billion. Prabhu's NPLs have reached a jarring 15.55 percent, its core capital at 8.17 percent, and it posted an outright loss of NPR 240 million. NIMB's NPLs hit 8.66 percent with distributable profit negative to the tune of roughly NPR 5.09 billion. And Himalayan, despite a nominal profit uptick, is nursing accumulated losses approaching NPR 10 billion with core capital of just 8.19 percent.

Look at operating-profit conversion and the same four sit at the bottom of the table: NIC Asia at 10.18 percent, Prabhu at 12.57 percent, Himalayan at 23.53 percent, NIMB at 25.54 percent, each bleeding a third or more of operating income straight into provisions. When capital buffers this thin meet asset quality this weak, the arithmetic leaves little margin for a bad surprise.

What the headline hides

None of this is a crisis narrative. Nabil and Everest posted EPS of NPR 28.36 and NPR 36.95 on disciplined, credit-driven performance and deserve their applause. Prime Commercial runs the sector's leanest shop, spending just NPR 25.31 to earn NPR 100 of income. The strong are genuinely strong.

But the sector average is doing what averages always do: hiding the dispersion that actually matters. A 32 percent profit jump built on treasury gains and provisioning releases, while credit stagnates and four banks slide toward capital distress, is not a recovery. It is a holding pattern dressed up as a rebound.

The real test comes when the writebacks run dry and the only lever left is lending. For that, banks need borrowers, and borrowers need an economy that wants to grow. Until real-sector credit demand returns, we should read these results for what they are: a well-managed pause, not a turning of the tide. The number is 32 percent. The story is stall.