Case Study 1 — A Commercial Bank

First published 26 Aug 2026 · Last verified 29 Aug 2026

Case Study 1 — A Commercial Bank

Sushmita Karki had read fifteen Parts of this book before she opened her first spreadsheet to actually use any of it. That gap — between understanding a framework and running it, cold, on a real company with real numbers that refuse to line up neatly — is the gap this entire Part exists to close. Parts XIII through XV gave you the machinery: the Canon Score itself, its seven dimensions, the sector adjustments that keep a bank from being graded like a trekking company, the backtesting mindset, the calibration habit, the exit-rule discipline, and the drift-management routine that keeps all of it honest as the market changes underneath you. Part XVI is where that machinery meets eight real companies, one per chapter, worked start to finish the way you would actually do it at your own desk on a Saturday morning with tea going cold beside your laptop.

Sushmita chose to start with a commercial bank for a reason that will sound almost too simple: banks are the most "textbook" sector on the Nepal Stock Exchange (NEPSE). Every commercial bank in the country reports to the same regulator, Nepal Rastra Bank (NRB), on the same quarterly schedule, using largely the same handful of ratios. A hydropower company's value depends on rainfall and river flow. A hotel's value depends on tourist arrivals nobody can fully predict. A bank's value depends on numbers that show up, every quarter, in a disclosure format NRB itself prescribes. If you are going to practice the Canon Score for the first time on a real company, a bank is the fairest test — the data exists, it is comparable across the whole peer set, and there is nowhere to hide behind a story.

The bank Sushmita picked was Nabil Bank Limited, traded on NEPSE under the ticker NABIL. This chapter follows her through all six lessons of the case study, using Nabil's real, publicly reported figures as of mid-2025 and the first half of fiscal year 2082/83. Some of those figures move a little from one disclosure to the next — audited versus unaudited, one quarter versus the next — and rather than smoothing that mess away, this chapter treats it as part of the lesson. Real companies do not hand you a single clean number. They hand you a range, and your job is to read the range honestly.

Lesson 83.1 — Setting Up the Case: A Bank Called Nabil

Nabil Bank has a claim to being the single most historically important private bank on NEPSE. It was incorporated in 1984 as Nepal Arab Bank Limited, the country's first joint-venture commercial bank — meaning a domestic bank formed in partnership with a foreign banking group, in this case Emirates-linked capital that was later replaced by Bangladesh's IFIC Bank as the principal foreign promoter. It was later renamed Nabil Bank Limited, and it has traded on NEPSE for decades under the scrip code NABIL, long enough that several generations of Nepali retail investors have owned it at one point or another, often as a first bank stock.

In July 2022, Nabil completed a merger with Nepal Bangladesh Bank Limited (NB Bank), swapping 43 Nabil shares for every 100 NB Bank shares held. Mergers of this size reshape a bank's numbers for years afterward — loan books get blended, deposit bases get combined, and ratios that looked clean before the merger often look messier for several reporting cycles afterward as the two institutions' loan quality, systems, and staff get folded into one. Keep that merger date in mind; it explains almost every awkward number in this chapter.

As of its most recent disclosures, Nabil operates 268 branches and around 316 ATMs across Nepal, with a paid-up capital of roughly Rs 32.06 arba (Rs 32.06 billion) — about four times NRB's old minimum capital mandate, a detail Lesson 83.4 will return to in detail. Its shares outstanding sit at about 27.06 crore (270.6 million), promoter shareholders hold roughly 58.44 percent of the company, and the general public holds the remaining 41.56 percent. Market capitalisation has recently sat around Rs 146 arba (Rs 146 billion), making Nabil one of the larger banks on the exchange by market value, though no longer automatically the largest — a decade of mergers among its peers has produced several banks of comparable or greater size.

Sushmita's first move, before touching a single Canon Score dimension, was to pull up Chapter 71's banking-sector playbook and re-read the six numbers it insists any competent bank analysis must include. She wrote them on an index card and taped it above her desk, because — as she put it in her notes — "if I can't explain these six things to my mother in one sentence each, I don't actually understand the bank yet."

Sushmita's working assumption going in — one she wrote at the top of her notebook page — was that because Nabil is old, large, and closely watched, its Canon Score would probably land somewhere comfortably good, and the exercise would mostly be about confirming that instinct with numbers. By the end of Lesson 83.5, the process would confirm the broad instinct but sharply revise the reasoning behind it — Nabil does score comfortably well, but not for the reason a reputation-based guess would assume, and with one specific weak spot the gut-feel version of this exercise would never have isolated. That is, in fact, exactly what a first full case study is supposed to do to you: not just tell you yes or no, but tell you why, in pieces you can individually check.

Lesson 83.2 — Hemisphere 1: Governance, Tradability, and Durability

The Canon Score splits into two broad hemispheres, grouping Chapter 64's seven dimensions by the kind of question each one asks. Hemisphere 1 covers the qualitative dimensions — Governance & Promoter Behaviour (15 points), Liquidity & Tradability (10 points), and Sector & Business Model Durability (15 points), 40 points in total — is this stock actually tradeable, who controls the company, and will this kind of business still matter in ten years. Hemisphere 2 covers the quantitative dimensions built from the financial statements themselves. Sushmita worked Hemisphere 1 first, on the theory that if a stock fails the basic "can I actually buy and sell this at a real price" test, the fundamentals barely matter.

Liquidity & Tradability (10 points). Average daily volume (ADV) — the average number of shares that actually change hands each trading day — is the practical measure of whether a stock can absorb a meaningful order without moving the price against you. Nabil's ADV has recently run around 46,000 shares over a one-week window and about 52,000 shares over a one-month window. At a share price near Rs 540, a single day's typical volume represents roughly Rs 2.5 to 2.8 crore (NPR 25–28 million) of turnover — comfortably above Chapter 64's NPR 5 million top band, so the volume sub-component scores a full 5 out of 5. Free float — the portion of shares actually available to trade, as opposed to locked up with promoters — sits at roughly 41.56 percent of the company, since promoters hold the remaining 58.44 percent. That clears Chapter 64's 40 percent free-float threshold for full marks, another 5 out of 5. Liquidity & Tradability: 10 out of 10 — Nabil is, straightforwardly, one of the more easily tradeable names on the exchange.

Governance & Promoter Behaviour (15 points). Nabil's promoter group has historically centred on IFIC Bank of Bangladesh as the principal foreign joint-venture partner, alongside Nepali business interests including the Chaudhary Group. In 2025, news emerged of a planned transfer of roughly 17.7 million promoter shares from IFIC Bank to the Chaudhary Group — a transaction that, at the time Sushmita did her research, was sitting under a court stay order rather than proceeding cleanly. Chapter 64's first governance check asks specifically about pledging — promoter shares used as loan collateral, which can trigger a forced sale with no warning to ordinary shareholders. Sushmita's research turned up no evidence that Nabil's promoter shares are pledged; the live issue is a contested ownership transfer under litigation, which is a different risk from pledging but still real uncertainty at the top of the shareholding table. She scored this sub-component 5 out of 6 — full marks for holding stability and the absence of pledging, minus one point for the unresolved transfer.

The third governance check — disclosure timeliness and board independence (4 points) — was the hardest to verify from the outside. Nabil's quarterly figures throughout this chapter arrived on the schedule NRB requires, which is itself a form of on-time disclosure; Sushmita did not, however, separately pull Nabil's board composition to confirm its independent-director count against the regulatory minimum. She scored this sub-component 3 out of 4, noting in her worksheet that board independence specifically was assumed rather than verified, and belongs on her list of things to actually check rather than infer next time.

Governance & Promoter Behaviour: 5 + 3 + 3 = 11 out of 15. On the positive side of the ledger, Nabil is the oldest continuously operating private commercial bank in Nepal, has survived four decades of political and economic cycles, completed a complex merger without a depositor-facing crisis, and in July 2025 commissioned a fresh issuer rating from CARE Ratings Nepal — the kind of voluntary external scrutiny that a bank hiding something usually avoids inviting. Balanced against that is the audited-versus-unaudited revision pattern, the unresolved promoter-share dispute, and the unverified board-independence detail. Sushmita's honest note in her worksheet read: "Old and mostly clean, but not spotless, and not finished changing hands at the top." That sentence, more than any single number, is what this dimension is trying to capture.

Sector & Business Model Durability (15 points). Chapter 64's moat check asks whether the business sits in a licensed, regulated sector with high entry barriers — Nabil, as an NRB-licensed commercial bank operating under a regulatory regime that makes new full banking licenses rare, scores a full 8 out of 8 here. The revenue-concentration check (7 points) asks whether the business depends too heavily on any single customer, segment, or counterparty. Sushmita did not pull Nabil's loan book down to individual-segment detail, but as a 268-branch, nationally diversified retail-and-corporate lender — not a single-sector specialist the way a hydropower company depends on one buyer — Nabil is directionally diversified rather than concentrated. She scored this sub-component 5 out of 7, full marks withheld only because she had not verified the exact segment breakdown herself. Sector & Business Model Durability: 8 + 5 = 13 out of 15.

Lesson 83.3 — Hemisphere 2: What the Numbers Actually Say

With Hemisphere 1 assessed, Sushmita turned to the financial statements themselves — Hemisphere 2, the quantitative half of the Canon Score, covering Financial Strength & Profitability (20 points), Valuation Reasonableness (15 points), Growth Trajectory (15 points), and Dividend & Capital Return Discipline (10 points), 60 points in total.

Financial Strength & Profitability. For fiscal year 2081/82 (roughly mid-2024 to mid-2025), Nabil reported customer deposits of about Rs 5.24 kharba (Rs 524 billion), up 13.53 percent year-on-year, and loans and advances of about Rs 4.12 kharba (Rs 412 billion), up 10.50 percent. Its CASA ratio — the share of deposits sitting in low-cost current and savings accounts rather than more expensive fixed deposits — stood at about 44.35 percent, recovering from a post-merger trough near 35 percent. Its cost of funds fell from about 5.86 percent to 4.41 percent over the same stretch, and continued falling toward roughly 3.25 percent in the following quarters. Earnings per share (EPS) rose from about Rs 22.90 to about Rs 26.34 for the fiscal year. Reported net profit for the year, using the unaudited figure, was about Rs 7.13 arba; the audited figure, as already noted, came in lower at about Rs 5.92 arba.

Chapter 64's sub-component A scores a 3-year average ROE. Nabil's ROE has come in at roughly 10 to 11 percent for fiscal year 2081/82 on the most commonly cited basis — a sharp fall from a pre-merger run rate closer to 19 to 20 percent, with the intervening post-merger years sitting somewhere in between as the integration worked through the numbers. (Different data providers compute this differently — some use average shareholders' equity, some use period-end equity, some annualize a single strong quarter — which is why you will see figures for Nabil's ROE ranging from under 10 percent to over 18 percent depending on the source. Sushmita's practical fix was to compute it herself, net profit divided by average equity for the full fiscal year, and treat any other figure as a cross-check rather than a substitute.) A 3-year average that starts near 19–20 percent and ends near 10–11 percent lands close to Chapter 64's "near current system average" band rather than its top band. Sushmita scored ROE 4 out of 8.

Sub-component B scores capital adequacy. This is where a bank's numbers earn their keep, and where NRB's regulatory floors matter most directly.

Nabil's capital adequacy ratio has recently sat close to 11.94 percent against that 11 percent floor — comfortably within Chapter 64's 11.5–13 percent band, which is worth 5 out of 7, but well short of the 13 percent-or-higher band that would earn full marks. Lesson 83.4 looks at this number in more depth, because a bank's precise position relative to its peers, not just relative to the regulatory floor, turns out to matter.

Sub-component C scores 5-year earnings quality and consistency. Nabil's non-performing loan ratio has moved from a pristine 0.84 percent in fiscal year 2078/79, up to a post-merger peak near 4.78 percent in 2080/81 as NB Bank's loan book was absorbed and reclassified under Nabil's own underwriting standards, and back down to a range of roughly 4.3 to 4.5 percent in the most recent quarters — a large move from the bank's historic norm that has not yet fully round-tripped back. Combined with the ROE compression and the unaudited-to-audited profit revision already discussed, that reads as at least one meaningfully "flat or declining" stretch inside the last five years rather than a smooth, uninterrupted profit climb. Sushmita scored this sub-component 3 out of 5. Financial Strength & Profitability: 4 + 5 + 3 = 12 out of 20.

On the liquidity side, separate from the Liquidity & Tradability dimension scored in Lesson 83.2, Nabil's loan-to-deposit ratio has recently sat around 82.5 percent, comfortably under the 90 percent regulatory ceiling, implying roughly Rs 39 arba of additional lending capacity before that ceiling would bind.

Valuation Reasonableness (15 points). Chapter 64 is explicit that NEPSE valuation must be judged against sector and historical medians, never a fixed textbook number — this book has already noted that NEPSE's banking sector has typically traded closer to 15–16 times earnings and around 1.5 times book value, well below the exchange's broader index. At a recent share price near Rs 541 to Rs 548, against a book value per share of roughly Rs 235 to Rs 247, Nabil trades at a price-to-book (P/B) ratio of about 2.19 times — against a sector median near 1.5 times, that works out to roughly 1.46 times the sector median, which falls in Chapter 64's 1.1x–1.5x band rather than its top band. Its price-to-earnings (P/E) ratio sits around 19.5 to 20.5 times trailing earnings against a sector median near 15.5 times — also roughly 1.29 times the sector median, the same 1.1x–1.5x band. Sushmita scored P/E 3 out of 8 and P/B 3 out of 7. Valuation Reasonableness: 3 + 3 = 6 out of 15 — the single weakest dimension in Nabil's entire scorecard, and the clearest read on the whole case study: Nabil is not an expensive bank in absolute terms, but it is a bank priced above its own sector on both major valuation yardsticks, at exactly the moment its return on equity has compressed the most.

Growth Trajectory (15 points). Nabil's loan book grew 10.50 percent and its deposits 13.53 percent for the fiscal year — a single year's figure rather than a verified 5-year CAGR, but one that sits inside Chapter 64's 8–15 percent band. Sushmita scored the growth sub-component 6 out of 8, flagging that a full 5-year CAGR would sharpen this if she pulled it later. EPS grew from Rs 22.90 to Rs 26.34 this year, but set against the sharp post-merger ROE swing already discussed, that reads as one strong year inside a choppier multi-year pattern rather than a clean, low-volatility uptrend — Chapter 64's "positive in 3 of 5 years, moderate swings" band, worth 4 out of 7. Growth Trajectory: 6 + 4 = 10 out of 15.

Dividend & Capital Return Discipline (10 points). Nabil has paid a dividend every year for at least the last five fiscal years Sushmita could find a public record for: 38 percent (mostly bonus shares) for FY2077/78, 30 percent for FY2078/79, then a shift to cash-only payouts as the sector moved away from bonus-heavy distributions — 11 percent cash for FY2079/80, 10 percent cash for FY2080/81, and 12.5 percent cash for FY2081/82. On an EPS of Rs 26.34, a 12.5 percent cash dividend (Rs 12.5 per Rs 100 of face value) works out to a payout ratio near 47 percent of profit — squarely inside Chapter 64's 30–70 percent "sensible" band. Paid every year, consistently in that band: 6 out of 6. Sushmita found no evidence the dividends were funded by anything other than ordinary distributable profit, though she had not separately confirmed reserves were intact from the balance sheet itself, so she scored the funding-source sub-component 3 out of 4 rather than a full 4. Dividend & Capital Return Discipline: 6 + 3 = 9 out of 10 — comfortably Nabil's strongest dimension after Liquidity.

Put together, Nabil's Hemisphere 2 picture is a bank still working through the tail end of a large merger: deposit and loan growth are healthy, funding costs are falling, and the dividend has been paid reliably every year — but return on equity has compressed sharply, asset quality has not yet fully normalised, and the stock's valuation multiples have not obviously priced any of that compression in.

Lesson 83.4 — Applying the Drift Adjustment: Capital Adequacy, Relatively Judged

Chapter 82 flagged a specific, well-documented problem with an older, cruder way of scoring bank capital: an approach built years ago around NRB's landmark paid-up capital mandate — the rule, dating back roughly a decade, that forced every commercial bank to hold at least Rs 8 billion in paid-up capital or merge with one that did. That single policy triggered the wave of bank mergers (including, indirectly, the one that created today's larger Nabil) that consolidated dozens of small banks into the current field of survivors. At the time, "does this bank clear Rs 8 billion in paid-up capital" was a genuinely useful, discriminating question — plenty of banks did not clear it, and it correctly separated the institutions that would survive from those that would be absorbed.

A decade later, that question has stopped discriminating between anything. Nabil's paid-up capital sits at roughly Rs 32 arba — four times the old mandate — and so does essentially every other surviving commercial bank on NEPSE, because the ones that could not get there were merged out of existence years ago. A scoring rule built around that absolute rupee threshold would now give every single remaining bank full marks regardless of how well-capitalised it actually is today. That is model drift in its purest form: the rule is still technically true, and still completely useless for telling Nabil apart from its peers.

Chapter 64's capital-adequacy sub-component already avoids the worst of this by scoring the CAR ratio itself against NRB's current thresholds (11 percent, 11.5 percent, 13 percent) rather than an absolute paid-up-capital rupee figure — which is why Nabil's 11.94 percent lands in the 11.5–13 percent band for 5 out of 7, not automatic full marks. But Chapter 82's deeper point still applies one layer down: a bank can clear its regulatory floor comfortably and still be thinly capitalised relative to its actual peer group, which the rubric's fixed bands alone will not show you. That is exactly what Sushmita checked next, using a mid-2025 sector snapshot of commercial bank CAR figures.

Nabil's 11.94 percent CAR sits below the sector average of 13.08 percent, ranks ninth out of the twelve major banks in this snapshot, and carries a thinner capital buffer above the regulatory floor than eight of its listed peers. That does not make Nabil unsafe — it remains comfortably compliant, and its 5-out-of-7 score under Chapter 64's own bands already reflects "comfortable but not exceptional" — but the peer table adds something the point band alone cannot: Nabil is one of the less well-capitalised large banks on the exchange relative to its own peer set, not merely "above some old, now-meaningless absolute rule." That distinction — a fixed rulebook band giving you a number, and a peer table telling you what that number actually means in context — is the more durable lesson Chapter 82 is teaching here.

Sushmita's note here was blunt: "The old rule said Nabil's capital position was basically perfect. The relative rule says it's fine, but it's the ninth-best-capitalised bank out of twelve where nobody's actually bad — and there's a new systemic-bank rule and a stalled asset-quality review both headed this way." That is precisely the kind of adjustment the drift-management discipline of Chapter 82 is meant to produce: not a dramatic reversal, but a materially more honest, more discriminating number than the one the un-updated model would have handed her.

Lesson 83.5 — The Full Worked Canon Score

With both hemispheres assessed and the drift adjustment applied, Sushmita assembled her full Canon Score tally, using exactly the seven dimensions and point weights Chapter 64 defines: Financial Strength & Profitability (20), Governance & Promoter Behaviour (15), Liquidity & Tradability (10), Valuation Reasonableness (15), Sector & Business Model Durability (15), Growth Trajectory (15), and Dividend & Capital Return Discipline (10).

Summed, Nabil Bank's Canon Score comes to 71 out of 100 — inside Chapter 64's Strong band (70–84), a rung above Adequate and a rung below Exceptional. No dimension triggers Chapter 64's governance override, since Governance & Promoter Behaviour scored 11 out of 15, well above the 5-point floor that would cap the whole score in the Weak/Avoid band regardless of everything else.

Sushmita was honest with herself about which dimensions were hardest to judge, and a fair-minded analyst working the same facts could reasonably land a few points higher or lower on any of them. Three stood out.

Valuation Reasonableness was the clearest, and the harshest, call in the whole scorecard. A 2.19x book multiple and a 20x earnings multiple, on a bank earning roughly 10 to 11 percent on that book right now, is not obviously cheap by the plain arithmetic of return on equity versus valuation — which is exactly why it is Nabil's lowest-scoring dimension at 6 out of 15. But Nabil's ROE compression is plausibly temporary — a direct byproduct of digesting a large merger rather than a structural decline in the underlying franchise — and a market willing to pay up for eventual recovery is not automatically an irrational market. An analyst more confident in a swift recovery could reasonably score this dimension several points higher; one who suspects the compression reflects a permanently larger, slower-growing, harder-to-manage post-merger institution could reasonably score it even lower still.

Financial Strength & Profitability's capital-adequacy sub-score was the second hardest call. Nabil is unambiguously compliant with every NRB capital and liquidity requirement in force today. The question is how much the 11.5–13 percent band should reflect a bank sitting in the lower half of its peer group specifically, especially with a systemic-bank framework and an asset-quality review both pending. Score it too harshly and you punish a bank for a relative ranking the rubric's fixed bands do not ask about directly; score it too gently and you ignore exactly the peer-context Lesson 83.4 surfaced. Sushmita's 5 out of 7 reflects a deliberate middle path, and she flagged it as the sub-score most likely to move at the next data refresh.

Governance & Promoter Behaviour was the third genuine judgment call. A recurring pattern of unaudited results running ahead of the audited restatement is a real signal, but it is also common enough across the sector that treating it as disqualifying would be disproportionate; reasonable analysts could argue for anywhere between 2 and 4 out of 5 on that particular sub-check depending on how much weight they put on disclosure discipline versus the bank's much longer clean operating history. Similarly, the unresolved IFIC-to-Chaudhary Group promoter-share transfer is real uncertainty but not evidence of pledging or expropriation, which is why it cost one point rather than several.

Lesson 83.6 — The Decision, and What Would Change It

A score of 71 out of 100 sits inside Chapter 64's Strong band — a solid long-term holding candidate worth owning with normal monitoring, a rung above merely Adequate but short of Exceptional. Reading the seven sub-scores rather than stopping at the total tells the real story: Nabil is comfortably strong on Liquidity, Durability, and Dividend discipline, respectably placed on Governance and Growth, and dragged down almost entirely by one dimension — Valuation Reasonableness, at 6 out of 15, its weakest score by a wide margin. Sushmita's conclusion, working purely through the process and holding no existing position, was this: Nabil Bank is a fundamentally sound, well-run institution still absorbing the tail end of a large merger, priced by the market as though that absorption were mostly complete, when the numbers say it is not quite there yet. That combination argues for a bank worth owning as a long-term holding, but not one worth chasing at the current price — an existing holder has little reason to sell a Strong-scoring bank, while a new buyer has good reason to wait for the valuation to catch down to the current, still-compressed return on equity, or for the return on equity to catch up to the valuation.

The more durable value of the exercise is the watch list it produces, because a Canon Score is never a one-time verdict — it is the current reading on an instrument you keep checking, exactly as Chapter 82 insists. Four specific developments would move Nabil's score, and Sushmita wrote all four into a recurring calendar reminder rather than trusting herself to remember them unprompted.

First, a sustained recovery in return on equity back toward the mid-teens, sourced from Nabil's own audited annual report rather than a single flattering quarter, would lift both the Financial Strength and Valuation Reasonableness dimensions meaningfully — it would be the clearest sign that the merger's integration costs are genuinely behind the bank rather than still working through the numbers, and it is the single change most likely to move Nabil from Strong toward Exceptional. Second, a capital adequacy ratio that climbs back above the sector average, rather than sitting in the bottom third of the peer table, would lift the Financial Strength score under the same relative logic Lesson 83.4 applied — and conversely, a further slide would be a real warning sign, not a rounding error. Third, resolution of the IFIC-to-Chaudhary Group promoter-share transfer, whichever way the pending court matter concludes, would remove a live piece of governance uncertainty from the Governance & Promoter Behaviour dimension. Fourth, the outcomes of NRB's planned asset quality review and its forthcoming Domestic Systemically Important Bank framework — both explicitly flagged in the 2025/26 monetary policy — could reshape Nabil's effective capital and provisioning requirements independent of anything the bank does on its own, and deserve a fresh look the moment either is published in final form.

Sushmita closed her notebook on this exercise with a line that captures the whole point of a first case study better than any score could: "I went in assuming the answer was obviously yes. The process mostly agreed — Strong, not Exceptional — but it also told me exactly which one number is doing all the damage, and that's a sharper answer than my gut ever gave me." That is what a disciplined framework is supposed to do to a beginner's instinct — not override it with false precision, but sharpen it into something checkable.