The Banking Sector Playbook

First published 23 Aug 2026 · Last verified 29 Aug 2026

Menuka Rai still remembers the afternoon she bought her first bank stock. It was a Thursday in Ashadh, the market was thin because everyone was busy closing their fiscal-year accounts, and she had just discovered that a certain commercial bank was trading at a price-to-book ratio that looked, on the surface, absurdly cheap next to its peers. She bought two hundred shares before lunch. By the following Baisakh, the bank had been quietly absorbed into a larger institution after Nepal Rastra Bank flagged it for capital shortfalls, her shares were converted at a swap ratio she had no say in, and the "cheap" stock she thought she had bought turned out to be cheap for a reason she had never bothered to check. She had looked at one number — price to book — and ignored the six or seven numbers that would have told her the real story.

That mistake is the reason this chapter exists. Banks and other bank-like financial institutions — the commercial banks, development banks, and finance companies that NRB collectively calls BFIs — make up one of the largest and most heavily weighted segments of NEPSE. On many trading days, the banking sub-index moves the whole market, simply because banks carry so much combined market capitalisation and so much trading float. An investor who does not know how to read a bank is, in effect, unable to read a third or more of the exchange. And banks are genuinely different animals from manufacturing companies, hydropower developers, or trading houses. A bank's balance sheet is not a description of a business — it is the business. Its main "product," a loan, sits directly next to its main "raw material," a deposit, and the spread between what it pays for money and what it earns on money is the entire engine of its profitability. Get the spread wrong, or misjudge the quality of what sits behind that spread, and everything else — dividends, book value, share price — eventually corrects itself.

This chapter builds Menuka's replacement discipline: a complete, repeatable playbook for evaluating any NEPSE-listed commercial bank or BFI, grounded in the specific regulatory architecture Nepal Rastra Bank has built around them.

Lesson 71.1 — Why Banks Are Not Like Other Companies

Start with the balance sheet itself. A manufacturing company's balance sheet is a supporting document; the income statement tells you whether the business works. For a bank, the balance sheet is the business. Assets are loans and investments; liabilities are deposits and borrowings; the gap between what the bank earns on the asset side and what it pays on the liability side, scaled across a balance sheet that is often ten to fifteen times the size of the bank's own equity, is what produces earnings. That leverage is normal and necessary for banking — it is also why a small deterioration in asset quality can wipe out a large share of equity in a way that would be unthinkable for a hydropower company or a manufacturer with the same percentage revenue shock.

This leverage is precisely why banks are regulated the way ordinary companies are not. Nepal Rastra Bank does not merely supervise BFIs the way the Securities Board supervises listed companies generally — it sets binding rules on how much capital a bank must hold against its risk-weighted assets, how much it may lend relative to what it collects in deposits, how it must classify and provide for loans that stop performing, and, periodically, how large its paid-up capital must be to keep operating as a commercial bank at all. Every one of these rules shows directly in the numbers a bank reports each quarter, and every one of them has, at some point in Nepal's market history, driven a multi-year cycle in bank share prices. An investor who treats a bank stock like any other cyclical industrial — cheap when earnings are down, expensive when earnings are up — will miss the fact that a bank's earnings cycle is substantially manufactured by regulatory cycles, not by end-market demand alone.

The second reason banks deserve a dedicated playbook is homogeneity. Unlike manufacturing or trading companies, which differ wildly in product, market, and cost structure, Nepal's roughly twenty listed commercial banks are, on paper, doing nearly the same thing: taking deposits, underwriting loans, managing a securities portfolio, and generating fee income from trade finance, remittance, and cards. This homogeneity is a gift to the analyst, because it means cross-sectional comparison is unusually reliable — a bank's net interest margin, cost-to-income ratio, or non-performing loan ratio can be benchmarked directly against a dozen near-identical peers reporting on an identical quarterly schedule in an identical NRB-mandated format. Few other NEPSE sectors offer this quality of apples-to-apples comparison. The playbook in this chapter is built to exploit exactly that advantage.

Lesson 71.2 — The Six Numbers That Matter

Every serious evaluation of a Nepali bank stock should begin with six figures, in this order: net interest margin, the credit-to-deposit ratio measured against its regulatory ceiling, the capital adequacy ratio measured against its regulatory floor, the non-performing loan ratio together with provisioning coverage, the cost-to-income ratio, and return on equity. Each tells a different part of the story, and together they tell you whether a bank is safe, efficient, and capable of compounding shareholder value, or whether it is one bad quarter away from a rights issue.

Net interest margin, or NIM, is net interest income — interest earned on loans and investments minus interest paid on deposits and borrowings — expressed as a percentage of average interest-earning assets. It is the single best proxy for a bank's core profitability engine, because unlike gross interest income alone, it nets out the cost side and normalises for balance sheet size. A closely related figure that NRB itself watches and periodically constrains through directive is the interest rate spread — average lending rate minus average deposit rate — which has hovered in the mid-single digits for large commercial banks in recent cycles and has been explicitly capped by NRB at various points to prevent banks from profiteering off the spread between what depositors earn and what borrowers pay. When you see a bank's NIM compress sharply quarter over quarter, the first question is not "did loan demand fall" but "did NRB tighten the spread ceiling, or did deposit competition force the bank to pay up for funds faster than it could reprice loans."

The credit-to-deposit ratio, universally shortened to CD ratio, measures total loans and advances against total deposits (with some adjustments for core capital and certain qualifying borrowings depending on the exact NRB formula in force). This is arguably the single most Nepal-specific metric in this entire playbook, because NRB has, for over a decade, imposed a binding regulatory ceiling on it — historically 90 percent — that functions as a hard credit-growth speed limit across the entire banking system. A bank sitting near the ceiling cannot grow its loan book meaningfully without first growing deposits, which forces a scramble for deposit mobilization (and often pushes deposit rates up, compressing NIM) every time credit demand outpaces deposit growth. A bank sitting comfortably below the ceiling has room to grow loans without a deposit chase, which is a genuine competitive advantage worth paying for.

Capital adequacy ratio, or CAR, measures a bank's own capital — core (Tier 1) capital plus supplementary (Tier 2) capital — against its risk-weighted assets. Under NRB's Capital Adequacy Framework, aligned with Basel III, commercial banks are required to hold minimum common equity, minimum Tier 1 capital, and minimum total capital, each with an additional capital conservation buffer layered on top. In practice this pushes effective working minimums to roughly seven percent common equity, eight and a half percent Tier 1, and around twelve to thirteen percent total capital once the conservation buffer is included — though an investor should always check the currently circulated unified directive rather than assume last year's figure still applies, since NRB has adjusted specific components (loan-loss provision treatment, countercyclical buffers, and what counts as supplementary capital) more than once in recent years. A bank running its CAR only marginally above the floor has essentially no capacity to absorb a shock without either raising fresh capital or shrinking its loan book.

Non-performing loan ratio and provisioning coverage go together and should never be read separately. The NPL ratio is simply non-performing loans divided by total loans, using NRB's loan classification categories (pass, watchlist, substandard, doubtful, and loss). Provisioning coverage measures how much of that non-performing exposure the bank has already set aside as loan-loss provision — a bank with a high NPL ratio but very high provisioning coverage has already absorbed most of the pain into past earnings, while a bank with a lower NPL ratio but thin provisioning coverage is deferring pain into future quarters. As of the most recent Nepal Rastra Bank financial stability reporting, sector-wide NPLs had climbed to roughly 5.6 percent, with considerable dispersion across individual banks — some names running close to nine percent, others closer to five percent — which is exactly the kind of dispersion that makes this metric worth checking bank by bank rather than trusting a sector average.

Cost-to-income ratio measures operating expense against operating income and is the clearest available signal of managerial discipline. A well-run Nepali commercial bank typically keeps this figure in the low-to-mid thirties or low forties; a bank that has expanded its branch network aggressively, hired ahead of revenue, or simply runs an inefficient back office will often show cost-to-income ratios well north of forty-five to fifty percent, quietly eating into the margin the other five metrics work so hard to protect.

Return on equity, finally, is the metric that converts all of the above into a single per-share compounding number: net profit divided by average shareholders' equity. Nepali banking ROE ran comfortably in the high teens through parts of the last decade; more recently, margin compression, elevated provisioning against rising NPLs, and periodic rights issues that expand the equity base faster than earnings have compressed sector ROE into the single digits to low teens for a meaningful share of listed banks, even as a policy-easing cycle beginning in 2025/26 has started to lift reported profits again for several names. ROE is useful, but only after you have checked the five metrics that feed it — a high ROE built on a thin CAR cushion or under-provisioned NPLs is not a durable ROE.

Lesson 71.3 — Reading the NRB-Mandated Quarterly Disclosure Format

Every NEPSE-listed bank publishes its quarterly results in a standardised format that NRB itself prescribes, which is a considerable gift to the analyst once you know how to navigate it, because every bank's disclosure lands in the same sequence of schedules. Menuka's early mistake was reading only the headline profit figure and the balance sheet total; the format rewards reading in a specific order instead.

Begin with the capital fund schedule. This is where you find core capital, supplementary capital, total capital fund, total risk-weighted exposure, and the resulting CAR — broken into the common equity, Tier 1, and total capital ratios discussed above. This schedule alone tells you whether the bank has genuine headroom to grow its balance sheet or whether a rights issue or debenture raise is likely on the horizon. A bank whose CAR has been drifting down toward the regulatory floor over consecutive quarters, even while reporting rising profit, is signalling that its risk-weighted asset growth is outrunning its internal capital generation — profitable growth that is nonetheless capital-hungry.

Next, move to the loan classification and provisioning schedule. This breaks total loans into the five NRB categories — pass, watchlist, substandard, doubtful, and loss — with the specific provisioning percentage NRB mandates for each category (rising steeply as loans move down the classification ladder). Sum substandard through loss to get gross NPL, divide by total loans for the NPL ratio, and compare total loan-loss provision held against that NPL figure for provisioning coverage. Watch particularly for growth in the watchlist category quarter over quarter — loans that are still technically "pass" or borderline but have been flagged for early warning signs — because this is often the leading indicator of NPL deterioration one or two quarters before it shows up in the headline ratio.

Then read the interest income and expense schedule, which lets you reconstruct NIM and the interest rate spread directly, along with the average yield on loans and average cost of deposits and borrowings separately — useful because a bank can defend its NIM either by repricing loans upward (which its borrowers feel) or by keeping deposit costs down through a favourable low-cost current and savings account (CASA) mix, and the two paths have very different sustainability profiles.

Finally, the CD ratio and liquidity schedule shows loans and advances against deposit mobilization, along with the statutory liquidity ratio and, increasingly in recent NRB guidance, liquidity coverage and net stable funding metrics as the central bank gradually shifts emphasis toward those internationally standardised measures alongside the traditional CD ratio ceiling.

A subtlety worth flagging here: distributable profit, the figure that ultimately determines dividend capacity, is not simply net profit. NRB requires banks to route a portion of profit into regulatory reserves — for items like deferred tax assets, unrealized investment gains, and certain non-banking assets acquired through loan recovery — before what remains becomes distributable to shareholders. A bank can report a healthy net profit while showing a much thinner distributable profit line, and the gap between the two is itself informative: a persistently wide gap often signals a bank whose reported earnings quality is lower than the headline suggests, propped up by items regulators require it to reserve against rather than pay out.

Lesson 71.4 — How NRB Monetary Policy Drives the Bank Profit Cycle

If there is one lesson that separates an investor who merely reads bank financial statements from one who can actually forecast a bank's next few quarters, it is this: Nepali bank earnings move in cycles that NRB itself substantially creates, through three levers — the interest rate corridor, the CD ratio (and its emerging liquidity-ratio successors), and capital requirements — each announced or adjusted through the annual monetary policy and periodic directives.

The interest rate corridor is the most direct lever. NRB sets a policy rate, a standing liquidity facility rate (sometimes called the bank rate, the ceiling of the corridor) at which banks can borrow overnight from the central bank, and a standing deposit facility rate (the floor) at which they can park excess liquidity. In the monetary policy for fiscal year 2025/26, NRB cut the policy rate from 5 percent to 4.5 percent, lowered the standing liquidity facility rate by half a percentage point to 6 percent, and reduced the standing deposit facility rate from 3 percent to 2.75 percent — an unmistakably accommodative stance intended to ease borrowing costs and support credit growth after a stretch of sluggish private-sector lending. When the corridor narrows and policy rates fall, banks' cost of short-term funds falls quickly, but their loan books reprice more slowly (many loans carry contractual reset lags), which typically produces a temporary NIM boost in the easing phase before competitive deposit pricing catches up. Watch for this lag: a rate-cutting cycle tends to lift bank earnings for several quarters before the benefit fades as deposit competition intensifies again.

The CD ratio has historically been the second, and arguably more Nepal-specific, lever. Since NRB replaced its earlier core-capital-cum-deposit ratio framework with a straightforward CD ratio ceiling — set at 90 percent starting in fiscal year 2078/79 — the ratio has functioned as a hard cap on system-wide credit growth relative to deposit mobilization. When credit demand outpaces deposit growth, banks bump against the 90 percent ceiling, a scramble for deposits ensues (often through promotional fixed deposit rates), deposit costs rise faster than loan yields can be repriced, and NIM compresses even as loan books are frozen against growth. This exact dynamic played out visibly in the deposit-scramble years, when a stretch of high fixed deposit rates squeezed margins across the sector simultaneously — a sector-wide compression, not a company-specific failure, which is an important distinction for an investor deciding whether a given quarter's weak NIM reflects mismanagement or simply the CD-ratio cycle biting every bank at once. More recently, the 2025/26 monetary policy has signalled a gradual shift away from the blunt CD ratio ceiling toward internationally standard liquidity metrics — the liquidity coverage ratio and net stable funding ratio — though the 90 percent CD ceiling remains the figure most banks and analysts still quote and monitor in practice during this transition.

The third lever, capital requirements, moves less often but reshapes the sector more dramatically when it does. The defining example remains NRB's 2015/16 monetary policy directive quadrupling the minimum paid-up capital for commercial banks from roughly two billion to eight billion rupees, with a multi-year window to comply. That single directive triggered the largest consolidation wave in Nepali banking history, forcing dozens of banks and finance companies into mergers between 2016 and 2020 simply to meet the new capital floor, and it permanently reshaped the competitive landscape of the sector NEPSE investors trade today. More recent capital-side easing has moved in the opposite direction: the 2025/26 monetary policy allows banks to count certain regulatory reserves created from non-banking assets as supplementary capital for up to two years, and eases the path for banks to raise additional capital — including rights issues — with central bank approval, effectively giving capital-constrained banks a release valve rather than forcing consolidation. The lesson for an investor is that capital directives can compress a bank's ROE for years by forcing dilutive rights issues, or can suddenly loosen and free up lending capacity — and both directions are policy decisions made in Kathmandu, not decisions the bank's own management team controls.

A fourth, softer lever worth watching is NRB's directive-level intervention on provisioning and asset quality — most recently, proposals to establish asset management companies to absorb distressed loans off bank balance sheets, and greater flexibility around loan restructuring and write-offs for "genuine" cases. These interventions matter because they change how quickly a rising NPL trend actually hits reported earnings; a policy environment that permits more generous restructuring can flatter near-term profit at the cost of deferring recognition of real credit losses, which loops back directly to the importance of tracking watchlist loan growth discussed in Lesson 71.3 rather than trusting the headline NPL figure alone.

Lesson 71.5 — Mergers, Acquisitions, and What Consolidation Means for Shareholders

Bank mergers in Nepal are not a rare corporate event — they are close to routine. Since NRB introduced its formal merger and acquisition bylaw in 2068 BS (2011), the sector has recorded more than sixty merger or acquisition transactions among commercial banks, development banks, and finance companies. For an investor, this means the possibility of a merger, forced or voluntary, needs to sit permanently in the evaluation checklist for any BFI below the top tier of the sector — not as a tail risk, but as a base-rate event.

It helps to separate two distinct flavors of consolidation. Voluntary mergers are typically driven by competitive logic: two mid-sized banks combine to gain scale, branch network reach, or a stronger deposit franchise, often initiated by the institutions' own boards and negotiated on a share-swap ratio basis, subject to NRB and shareholder approval. The merger of Nepal Investment Bank and Mega Bank into what became Nepal Investment Mega Bank, with final regulatory approval secured and joint operations commencing in early 2023, is the clearest recent example of this voluntary category — two established banks combining by choice to create one of the largest banks in the country by asset size, rather than being pushed together by a capital shortfall. Forced or regulatory-encouraged mergers, by contrast, are what happened en masse after the 2015/16 capital hike: banks and finance companies that could not independently raise fresh capital to meet the new Rs 8 billion floor had little choice but to merge with a stronger partner or exit the industry, and NRB actively encouraged this consolidation as a matter of financial stability policy rather than leaving it purely to market forces.

For shareholders, the mechanics that matter most in any merger are the swap ratio (how many shares of the merged entity each existing shareholder receives per share held, typically set by relative book value and sometimes adjusted by an independent due diligence valuation), the resulting dilution or accretion to per-share metrics, and — critically — what happens to the weaker partner's problem assets. A merger genuinely strengthens a shareholder's position when the combined entity has a stronger CAR, better NPL coverage, and lower funding costs than either predecessor bank alone; it merely delays a reckoning when a chronically undercapitalized or NPL-heavy institution is absorbed into a stronger partner mainly to avoid an outright regulatory intervention, in which case the acquiring bank's own ratios often show visible strain for several quarters after the merger closes while it works through the inherited loan book.

It is also worth noting which banks have never been part of a merger — among them Agricultural Development Bank, Everest Bank, Nepal Bank Limited, Nepal SBI Bank, and Standard Chartered Bank Nepal. Several of these are foreign-joint-venture banks or state-linked institutions with capital structures and shareholder bases that made them either naturally compliant with capital floors or institutionally resistant to consolidation pressure — a useful reminder that ownership structure, not just balance sheet health, shapes merger probability.

Lesson 71.6 — Well-Run Bank versus Growth-at-Any-Cost Bank: The Checklist

Two banks can show similar headline loan growth and similar reported profit in a given quarter while being fundamentally different investments — one compounding shareholder value safely, the other borrowing against its own future stability to produce this quarter's number. Consider two illustrative composites, Bank Alpha and Bank Beta, both mid-tier commercial banks reporting roughly 18 percent year-on-year loan growth and comparable ROE for the quarter Menuka was comparing them. Bank Alpha funded that growth mostly through low-cost current and savings deposits, kept its CD ratio well below the 90 percent ceiling with genuine headroom, held its NPL ratio below sector average with provisioning coverage comfortably above 100 percent of non-performing exposure, and kept its cost-to-income ratio in the mid-thirties. Bank Beta funded similar growth mostly through expensive promotional fixed deposits, ran its CD ratio right at the regulatory edge, showed a fast-growing watchlist category behind a still-modest headline NPL figure, and had let cost-to-income creep toward fifty percent while opening branches faster than its deposit base could support them. Both reported similar growth; only one of them was building something durable.

The qualitative signals that separate these two banks rarely appear as a single dramatic red flag — they accumulate. Deposit mix is one of the most reliable: a bank with a high proportion of current and savings account deposits (low-cost, sticky funding) has structurally cheaper and more stable funding than one dependent on fixed deposits chased with promotional rates, and that difference shows up directly in NIM resilience during a CD-ratio-driven deposit scramble. Loan book concentration is another: heavy exposure to a single sector (real estate, margin lending against shares, or a handful of large corporate borrowers) creates correlated risk that a diversified retail and SME loan book does not carry. Related-party and connected lending, disclosed in the notes to financial statements and in NRB's periodic supervisory findings, is a persistent governance concern in Nepali banking specifically, and a bank with recurring findings here deserves a discount regardless of how clean its headline ratios look. Management and board stability matters too — frequent CEO turnover or repeated NRB directive actions against a specific institution are both observable, low-effort signals available well before the next quarterly disclosure. Finally, dividend history and capital-raising pattern tell you whether a bank has historically grown its book value organically through retained earnings or has repeatedly needed dilutive rights issues to stay above its capital floor — the latter pattern erodes per-share value even when the underlying institution survives and grows in absolute terms.

Menuka now runs every bank she considers through the same sequence, and it is worth setting it out explicitly as the playbook to apply to any NEPSE-listed bank. First, pull the last eight quarters of the capital fund schedule and chart CAR against the regulatory minimum — look for a stable or rising trend, not one drifting toward the floor. Second, do the same for the CD ratio against the 90 percent ceiling, checking headroom for future loan growth. Third, build the NPL and provisioning coverage trend across the same eight quarters, paying particular attention to watchlist category growth as an early warning signal rather than waiting for the headline NPL number to move. Fourth, reconstruct NIM and the interest rate spread, and separately track average yield on loans against average cost of deposits, to see whether margin resilience is coming from loan repricing power or from a genuinely low-cost deposit franchise. Fifth, compute cost-to-income ratio and compare it against at least four to five peer banks reporting the same quarter, since this is one of the most directly comparable efficiency metrics on NEPSE. Sixth, compute ROE, but only after completing the first five steps, and discount any ROE figure that rests on a thin CAR cushion, weak provisioning, or a widening yield-cost gap that looks unsustainable. Seventh, check ownership and merger-probability signals specifically — promoter shareholding stability, any history of NRB directive action, and whether the bank sits near a capital or CD-ratio threshold that would make it a plausible merger candidate. And eighth, read at least the headline of the current fiscal year's monetary policy and any recent NRB directive affecting BFIs, since — as this chapter has tried to establish — a bank's next four quarters are shaped as much by decisions made at NRB's Baluwatar headquarters as by decisions made in the bank's own boardroom.