A central bank speaks in two registers. One is the language of reassurance, the steady voice that tells a country its banks are sound and its savings are safe. The other is the language of scrutiny, quieter and less often heard in public, in which the supervisor records what it actually found when it opened the books. The Bank Supervision Report for 2024-25, released recently by Nepal Rastra Bank, is notable because it lets the second voice be heard clearly. That candour is worth appreciating before anything else.
On the surface, Nepal’s commercial banks are doing what banks are supposed to do. Their combined assets grew by more than twelve percent over the year, crossing Rs 7,756bn. Deposits rose by nearly fourteen percent. Capital adequacy across the system sat above twelve and a half percent, comfortably over the regulatory floor. Anyone reading only those figures would conclude that the sector is in healthy condition, and in important respects it is.
The report’s contribution is to ask a sharper question. Not how large the capital ratio is, but how it was arrived at. A capital ratio, after all, is only a fraction. The capital a bank holds sits above the line, and the risk it carries sits below. Make the lower number smaller than it ought to be, and the ratio rises on its own, without a single rupee of fresh capital being raised. The supervisor’s reviewers went looking for exactly this effect, and they describe what they saw without euphemism.
In the department’s words, capital adequacy ratios in some banks were not reflected accurately because risk-weighted exposures had been misstated and provisions for likely loan losses fell short. Translated into ordinary terms, a few institutions were carrying loans on terms that made their balance sheets look stronger than the underlying reality. The report sets out how. Loans that had already slipped into the watch-list, substandard, doubtful, or loss categories were sometimes given gentler risk weights than their condition deserved. Unpaid claims on guarantees the bank had issued were quietly left out of the risk calculation altogether. Each adjustment, taken alone, looks small. Together they lift a ratio off the page.
There was a second route to the same destination. Banks are allowed a lighter capital charge on loans that fall within what is called the regulatory retail portfolio, on the logic that many small, scattered loans are safer as a group than a few large ones. The review found cases where loans were swept into this favorable bracket on the strength of a single qualifying test rather than the full set, easing the capital they consumed. None of this is exotic. It is the everyday arithmetic of regulatory arbitrage, and the value of the report is that it names the practice plainly enough to act on.
Why should anyone outside a bank’s finance department care about the mechanics of risk weighting? Because the capital ratio is the single number on which a great deal of trust rests. The fund that insures small depositors prices its exposure on the assumption that reported capital is real. Investors value bank shares on it. Banks lend to one another overnight on the basis of it. When the figure flatters, every one of those judgments is made on slightly false information, and the error does not stay contained. It travels. A supervisor that catches the distortion early, and says so in print, is protecting far more than the bank in question.
The report is equally frank about a quieter weakness, one of process rather than accounting. Every bank is meant to run an internal capital adequacy assessment, a yearly exercise in which it judges for itself how much capital its own risks demand. In a number of institutions this had withered into a formality. Some did not complete it on time. Others filled it in by simply copying across the regulatory minimum, as if the floor set by the central bank were also the ceiling of their ambitions. A few had no genuine capital plan at all, neither for the coming months nor for the years beyond. The report warns, in measured language, that an institution without such a plan will find a future crisis far harder to weather. The caution is sensible, and it is offered in time.
Set the question of measurement aside, and the plain numbers still ask for attention. Bad loans across the commercial banks rose by more than a fifth during the year, reaching some Rs 220bn, and the share of loans that have gone sour climbed to 4.44 percent from 3.76 percent a year earlier. At the state-owned banks, the deterioration was steeper still. The assets banks repossess when borrowers default, the foreclosed property that sits on their books earning nothing, swelled by more than forty percent. These are not figures the report conceals. They are figures it publishes, and the publishing is itself a mark of confidence.
There is an unusual backdrop to all of this. Nepal’s banks are not short of money. Deposits have flowed into the system faster than loans have flowed out, leaving the sector liquid to an almost awkward degree. Deposits now exceed the size of the economy itself, while credit has slipped back relative to it. That imbalance has its own costs, thinner margins chief among them, but it also hands the regulator something valuable: room to manage a rise in bad loans from a position of strength rather than scarcity. Profits, meanwhile, held up at the headline level, though almost entirely on the back of the state-owned banks, while private lenders saw earnings dip. The resilience is real but unevenly spread.
What lifts this report above a routine annual filing is its honesty about the road ahead. The supervisor states openly that the task is to move past tick-box inspection toward a model that can tell when the economic substance of a loan quietly contradicts the form in which it has been reported. That is a demanding ambition. It leans on judgment rather than checklists, on experienced supervisors willing to look past a clean-looking return to the reality beneath it. Naming that direction out loud is a sign that Nepal’s risk-based supervision is growing up.
The report does not leave matters at diagnosis. It records that corrective directives have already been issued, and the direction they point is clear enough. Reported risk weights should be reconciled honestly against what the supervisor independently calculates. Loans should be classified for what they are and provisioned accordingly. The internal capital assessment should reflect a bank’s own risks rather than echo the regulator’s minimum. The functions that exist to say no inside a bank, its risk and audit teams, should answer to the board rather than to the executives whose decisions they are meant to check. And capital should be built while the weather is fair, not scrambled for once the storm has arrived.
None of this requires new legislation or dramatic intervention. It asks the supervisor to keep using the powers it already holds, and to use them with conviction. The most reassuring thing about the 2024-25 report is that it exists in this candid form. A central bank willing to write down, for anyone to read, where its banks have fallen short is a central bank doing the unglamorous core of its work. The findings are on the table and the directives are in force. The measure of the year ahead will be how faithfully they are carried out, by banks and supervisors alike, in the same constructive spirit in which they were set down.
The author writes on financial regulation and fintech policy. The views expressed are personal and based on a public document
