Case Study 3 — A Microfinance Institution

First published 26 Aug 2026 · Last verified 29 Aug 2026

Anjana Koirala opened her notebook to a fresh page and wrote two words at the top: "small loans." She had spent Chapter 83 inside the balance sheet of a commercial bank, learning to read capital adequacy ratios and net interest margins the way a mechanic reads an engine's vital signs. Chapter 84 had taken her up a hillside penstock, learning to price a hydropower company's water year by water year, discounting a river the way you would discount a bond. Now she was looking at a business built on loans of twenty, forty, sixty thousand rupees each, made to women running vegetable stalls, buffalo sheds, and tailoring shops in villages she had never visited. Her bank case study had had a market capitalisation measured in tens of arba (one arba equals one billion rupees, roughly the scale at which Nepal habitually counts its largest listed companies). Her hydropower case study had had a single, physical, inspectable asset: a dam, a tunnel, a powerhouse. This company had neither a single site to visit nor a single number that fully captured its risk. It had, instead, hundreds of thousands of tiny, unsecured promises, scattered across dozens of districts, each one no bigger than a mid-sized grocery bill back home in Kathmandu.

That, she reminded herself, was the whole point of doing this as her third case study. The Canon Score — the 100-point, seven-dimension scoring framework this book has built chapter by chapter — is only as good as an investor's willingness to apply it honestly to businesses that do not look like each other. A bank borrows short and lends long against collateral and a regulatory capital cushion. A hydropower company sells a single physical commodity, electricity, under a long-term contract, and lives or dies by rainfall and interest rates. A microfinance institution (an MFI — a company whose entire business is making small, typically collateral-free loans to low-income borrowers, often organised in groups) does something different again: it manufactures trust, at scale, in places the formal banking system finds too expensive to reach. Getting the Canon Score right for an MFI meant testing whether the framework's discipline survived contact with a genuinely different kind of risk. This chapter follows Anjana through that test, using a real, currently NEPSE-listed microfinance institution: Chhimek Laghubitta Bittiya Sanstha Limited, trading under the ticker CBBL.

Lesson 85.1 — Setting Up the Case: A Different Kind of Bank

Chhimek Laghubitta was established in 2058 BS (2001 AD) as one of Nepal's earliest Grameen-model microfinance replicators — meaning it borrowed its founding method from the Grameen Bank of Bangladesh, the institution credited with popularizing group-based microlending to poor, largely rural women. Over two and a half decades it grew from a single-district pilot into one of Nepal's largest laghubitta ("small finance," the Nepali term used for microfinance institutions) networks, with a branch footprint spanning well over a hundred offices across the country and a borrower base numbering in the hundreds of thousands, concentrated among women in rural and semi-urban households who use small loans to finance livestock, retail trade, agriculture, and cottage-industry work. It is one of the older, larger, and more closely watched names in a sector that, unusually for Nepal, has an enormous number of separately listed companies: roughly fifty microfinance institutions trade on NEPSE, a level of sector fragmentation with no real parallel among Nepali banks or hydropower companies.

Anjana's first task, before she touched a single number, was to understand why an MFI needs its own scoring lens rather than a copy of the bank playbook from Chapter 65. Three structural facts drove this.

The first is the joint-liability-group (JLG) lending model itself. Rather than lending to individuals against collateral, as a commercial bank does, most Nepali MFIs organise borrowers into small groups — typically five to seven women — who meet regularly, often weekly, and who each guarantee the others' loans informally through peer pressure and shared reputation rather than through a mortgage or a fixed deposit.

This is a powerful mechanism when it works, because it turns a village's own social fabric into an underwriting tool a bank could never replicate. But it is also a mechanism that can fail all at once: when a whole village's cash flow is hit by the same shock — a bad monsoon, a remittance slowdown, a local market collapse — the very thing that made repayment reliable (shared local conditions) turns into a common point of failure. A bank's thousands of individually underwritten mortgage borrowers are far less correlated with each other than a microfinance centre's thirty group members farming the same watershed.

The second structural fact is the dual mandate. An MFI is chartered, in part, on a social mission — expanding financial access to households the formal banking system has historically ignored — alongside its obligation, as a NEPSE-listed company, to deliver a commercial return to shareholders. These two mandates usually pull in the same direction, but not always: a purely profit-maximising lender might raise loan sizes and interest rates faster than borrower repayment capacity can bear, while a purely mission-driven one might resist raising loan-loss provisions during a stress cycle for fear of looking like it is retreating from its social purpose. An investor reading Chhimek's numbers has to keep both mandates in view.

The third is regulatory. Nepal Rastra Bank (NRB, the central bank), which regulates banks, hydropower financing, and microfinance institutions all under one roof but with very different rulebooks for each, has historically applied tools to microfinance that it does not apply to commercial banks at all.

This single regulatory change matters enormously for an MFI's future profitability, because it determines how much room an institution has to reprice its loan book as its own funding costs move — a lever a commercial bank takes almost for granted but that microfinance institutions in Nepal have only recently been given in a more flexible form.

Anjana's note to herself at the end of this lesson: everything from here on has to be read through these three lenses — social collateral that can fail correlated, not independent; a dual mandate that can quietly trade off against itself in a downturn; and a regulatory regime for pricing that has just changed shape under the sector's feet.

Lesson 85.2 — Hemisphere 1: Liquidity, Governance, and Durability

Anjana followed the same method Sushmita and Sabina had used before her: split Chapter 64's seven dimensions into Hemisphere 1 (Liquidity & Tradability, Governance & Promoter Behaviour, and Sector & Business Model Durability, 40 points total) and Hemisphere 2 (the four dimensions built from the financial statements). She worked Hemisphere 1 first.

Liquidity & Tradability (10 points). Chhimek's 30-day average trading volume runs close to 9,700 shares a day; at a recent price near Rs 930, that works out to roughly Rs 90 lakh (about NPR 9 million) of daily turnover — comfortably above Chapter 64's NPR 5 million top band, worth 5 out of 5. Its 52-week range of roughly Rs 882 to Rs 1,093 is a genuine two-way band of about 24 percent, not a stock pinned at a circuit limit. Free float is stronger than Anjana expected going in: with promoters holding 51 percent and the public holding the remaining 49 percent, Chhimek clears Chapter 64's 40 percent free-float threshold comfortably, worth another 5 out of 5. Liquidity & Tradability: 5 + 5 = 10 out of 10 — a genuinely liquid microfinance name, not the thin, hard-to-trade stock the sector's reputation might suggest.

Governance & Promoter Behaviour (15 points). Chapter 65 flags this as the sector's heaviest-rewrite dimension, supplementing the generic checklist with two microfinance-specific red flags: loan recycling (issuing a fresh loan to a struggling borrower specifically to repay an old one about to go overdue, which keeps reported NPL artificially low while the borrower's real debt burden worsens) and multiple borrowing, or overlap, across competing lenders in the same village. Promoter shareholding and pledging (6 points) scores cleanly: the 51 percent promoter stake shows no evidence of pledging or recent decline, worth a full 6 out of 6. The recycling-and-overlap check (5 points) is where Chapter 65's own warning matters most: a clean NPL ratio alone is not proof of clean lending, since recycling can suppress the headline number without fixing the underlying borrower distress. But Chhimek's non-performing loan ratio held near 2.3 percent through the entire 2021-2026 sector stress period — the lowest in the sector by a wide margin, against a sector average that climbed to roughly 11.35 percent and eighteen listed peers crossing 10 percent NPL — and a multi-year, full-cycle result is a materially harder thing to fake through recycling than one clean quarter would be. Anjana scored this sub-component 4 out of 5, crediting the durability of the result while still not awarding an automatic full mark for a metric Chapter 65 explicitly warns can mislead. Disclosure timeliness and board independence (4 points) — Chhimek files regular quarterly results and is one of the more closely watched names in its sector, supporting on-time disclosure, but board independence was not separately verified, so Anjana scored this sub-component 3 out of 4. Governance & Promoter Behaviour: 6 + 4 + 3 = 13 out of 15.

Sector & Business Model Durability (15 points). Chapter 65 flags this as a heavy adjustment for microfinance too, for a different reason than governance: regulatory concentration risk. Because interest rates, provisioning rules, and lending practice across the entire sector are tightly directed by NRB policy — including the periodic rate caps and the 2025 shift to a base-rate-plus-premium regime described in Lesson 85.1 — a microfinance institution's durability is unusually exposed to regulatory shifts compared with a less directly regulated business. The moat sub-component (8 points) reflects a licensed, NRB-regulated activity with real entry barriers (branch buildout capital, a lending license, years of borrower-trust-building), but the sector is also unusually fragmented — roughly fifty separately listed microfinance institutions compete on NEPSE, a level of fragmentation with no real parallel among Nepali banks or hydropower companies — which caps this at a moderate rather than dominant position: 5 out of 8. The revenue-concentration sub-component (7 points), read here as geographic and borrower-livelihood diversification rather than customer concentration, favours Chhimek clearly: its decades-long, multi-district branch network across the Tarai, hills, and mountain belts is real diversification against the correlated local shocks Lesson 85.1 described, a genuine and unglamorous edge over smaller, more regionally concentrated laghubitta peers. Anjana scored this 6 out of 7. Sector & Business Model Durability: 5 + 6 = 11 out of 15.

Lesson 85.3 — Hemisphere 2: What the Numbers Actually Say

With Hemisphere 1 assessed, Anjana turned to the quantitative half of the Canon Score: Financial Strength & Profitability (20 points), Valuation Reasonableness (15 points), Growth Trajectory (15 points), and Dividend & Capital Return Discipline (10 points).

Financial Strength & Profitability. Chapter 65 says this dimension needs real surgery for microfinance: loan-loss provisioning coverage and portfolio quality should weigh more heavily than they would for a commercial bank, since unsecured group lending carries structurally higher default risk than a typical bank's collateralized book. Chhimek's net profit came to roughly Rs 100 crore for the most recent fiscal year, down 5.83 percent from the year before, against equity of roughly Rs 9 arba (net worth per share near Rs 249, against 3.62 crore shares outstanding) — a return on equity near 11 percent, in Chapter 64's 10-15 percent band, worth 6 out of 8 on the ROE sub-component. Capital adequacy or leverage discipline (7 points) was the hardest sub-component to pin down precisely: Anjana could not find a single disclosed capital-adequacy figure for Chhimek the way NRB requires banks to publish one, so she used net worth per share (roughly 2.5 times the Rs 100 face value, reflecting two decades of retained earnings) as a rough proxy for a well-capitalised institution, scoring this conservatively at 5 out of 7 and flagging the data gap rather than pretending a precise ratio existed. Earnings quality and consistency (5 points) is where the recent asset-quality evidence matters most: profit dipped in one recent year, but with the sector's provisioning-and-margin story in mind — a rate-cap transition compressing spreads sector-wide, not a credit-quality collapse — and set against Chhimek's NPL ratio holding near 2.3 percent through the entire multi-year sector stress cycle (versus a sector average that climbed to roughly 11.35 percent), Anjana scored this generously at 4 out of 5 rather than the default "one flat year" band, on the reasoning Chapter 65 specifically asks for: weight provisioning-quality evidence more heavily than a bank's smoother profit-history norm would suggest. Financial Strength & Profitability: 6 + 5 + 4 = 15 out of 20.

Valuation Reasonableness. Chhimek trades around Rs 930 per share against trailing earnings near Rs 34-36 per share and book value near Rs 250 per share — a price-to-earnings ratio near 26-28 times and a price-to-book ratio near 3.7-3.9 times. Nepal's microfinance sector has recently carried one of the richest average valuations on the whole exchange — a sector-average P/E cited around 47 times, second-highest among NEPSE's eleven sectors — which makes Chhimek's own P/E, at roughly 0.6 times that sector average, land in Chapter 64's cheapest band: 0.8x sector median or below, worth a full 8 out of 8. That is a genuinely interesting, current finding: the microfinance sector as a whole trades rich, but the market is pricing Chhimek — arguably the cleanest name in the sector through its worst stress cycle in years — at a real discount to its own peers. P/B is a harder comparison to make cleanly: a microfinance-specific sector P/B median was not something Anjana could pin down with confidence, so she checked Chhimek's 3.7-3.9x against NEPSE's whole-exchange average of roughly 2.8x, a rougher but honestly-labelled anchor, landing at about 1.4 times that broader benchmark — Chapter 64's 1.1x-1.5x band, worth 3 out of 7. Valuation Reasonableness: 8 + 3 = 11 out of 15.

Growth Trajectory. Loan book growth has slowed deliberately across the sector since the 2021-22 stress episode, and Chhimek's own recent figures show the same pattern — profit essentially flat-to-declining in the most recent full fiscal year before a partial recovery in more recent quarters. Anjana scored the revenue/loan CAGR sub-component 3 out of 8, in Chapter 64's 0-8 percent band, reading the deceleration as a discipline choice consistent with the sector's post-stress caution rather than a demand collapse. EPS consistency showed a similar mixed-but-recovering pattern — one down year, followed by EPS climbing back above its prior level in the most recent quarter — fitting Chapter 64's "positive in 3 of 5 years, moderate swings" band, worth 4 out of 7. Growth Trajectory: 3 + 4 = 7 out of 15 — Chhimek's weakest dimension, and a real one: deliberate post-stress deceleration is a defensible choice, but it is still a growth story that has genuinely slowed, not merely a market misperception to correct.

Dividend & Capital Return Discipline. For the most recent fiscal year, Chhimek distributed a 12.5 percent bonus share issue plus a 12.5 percent cash dividend. The cash component alone, against EPS near Rs 34, works out to a payout ratio near 37 percent of earnings — squarely inside Chapter 64's 30-70 percent sensible band. Combined with two decades of retained earnings visible in its net worth per share, this reads as a company with a consistent distribution record, worth 6 out of 6 for consistency. Anjana found no evidence the distribution was funded by anything other than genuine profit or that reserves were being drawn down, worth 4 out of 4. Dividend & Capital Return Discipline: 6 + 4 = 10 out of 10 — tied with Liquidity as Chhimek's strongest dimension.

Lesson 85.4 — Reading a Clean NPL Ratio Honestly, and What Chapter 80 Confirms

Before assembling the full score, Anjana paused on the single most important piece of evidence in this whole case study: Chhimek's non-performing loan ratio holding near 2.3 percent, the best in its sector, through a multi-year stress cycle that pushed the sector average above 11 percent and put eighteen listed peers over the 10 percent threshold. This is exactly the kind of sector-specific evidence Chapter 65 says to weight heavily for microfinance — and exactly the kind of number Chapter 65 also warns not to trust blindly, since loan recycling can manufacture a clean-looking NPL ratio that does not reflect real borrower health.

Two things convinced Anjana this particular number deserved real trust rather than suspicion. First, duration: recycling can flatter a single quarter, but sustaining the sector's best asset-quality figure across a multi-year cycle that genuinely broke many of its peers is a far harder thing to fake. Second, corroboration: Chhimek's governance profile — the multi-district diversification from Lesson 85.2, the absence of any disclosed pledging or ownership instability — is consistent with an institution that built supervisory capacity alongside its branch network, rather than one manufacturing a clean number while its underlying book quietly deteriorated. Neither point makes recycling impossible to rule out with total certainty from the outside — Anjana was honest that a retail investor working from public disclosures alone cannot fully verify this the way an on-site auditor could — but together they were enough to treat the 2.3 percent figure as real evidence rather than a number to distrust reflexively.

This is also the right place to connect Chhimek's case to something else in this book. Chapter 80, on calibrating the scoring model, follows an NRB veteran named Bimal Sharma discovering — through his own, independently chosen test cases — that the costliest gap in how he had actually been applying the Canon Score was exactly this one: failing to ask Chapter 65's two microfinance-specific Governance questions (whether loan growth or NPL trends show signs of recycling, and whether borrowers show evidence of overlapping debt with other lenders in the same district) before finalising a score. Bimal's fix was a mandatory checklist forcing those two questions to actually be asked every time — a process change, not any change to Chapter 64's point weights. Anjana had not read Chapter 80 before starting this case study, and arrived at the same place from a different direction: she asked exactly those two questions of Chhimek's NPL record before trusting it, rather than crediting a clean number on sight. Two analysts, two different real companies, landing on the same sector-specific check Chapter 65 had already written down — that is what a well-calibrated framework is supposed to produce.

Lesson 85.5 — The Full Worked Canon Score

With both hemispheres assessed, Anjana assembled her full tally using exactly the seven dimensions and point weights Chapter 64 defines.

Summed, Chhimek's Canon Score comes to 77 out of 100 — inside Chapter 64's Strong band, a rung above Adequate and within real reach of Exceptional. No dimension triggers the governance override, since Governance & Promoter Behaviour scored 13 out of 15, comfortably above the 5-point floor.

Anjana was honest about the closest calls. The capital-adequacy sub-score was scored conservatively specifically because she could not find a disclosed figure — a reader with access to Chhimek's full regulatory filings could reasonably move this up or down once a real number is in hand. The NPL-driven governance and earnings-quality credit was a genuine judgment call, not something the rubric handed her cleanly — she leaned on duration and corroboration to justify trusting the number, but a more skeptical analyst could reasonably score both sub-components a point or two lower. And the sector-fragmentation discount inside Sector & Business Model Durability reflects a real, structural feature of Nepali microfinance — fifty-odd separately listed competitors — that a reader focused only on Chhimek's own individual strength might be tempted to score higher.

Lesson 85.6 — The Decision, and What to Watch

A score of 77 out of 100 sits inside Chapter 64's Strong band — a solid long-term holding candidate worth owning with normal monitoring. Reading the seven sub-scores rather than stopping at the total tells the real story: Chhimek is comfortably strong on Liquidity, Dividend discipline, and Valuation, respectable on Governance and Financial Strength, and dragged down by two related weaknesses — Growth Trajectory at 7 out of 15 and, more moderately, Sector & Business Model Durability at 11 out of 15 — both rooted in the same underlying reality: a mature institution in a fragmented, post-stress sector that has deliberately chosen discipline over speed. Anjana's honest conclusion: Chhimek looks like one of the stronger names in a sector the broader market is still pricing with real caution, trading at a discount to its own rich sector despite the best demonstrated asset quality through a genuine multi-year stress test. That combination argues for a real, properly sized position rather than either enthusiasm or avoidance — a name whose price has not yet caught up to what its loan-book discipline has already proven.

Chapter 82's drift-monitoring discipline applies here as much as it did to the bank and the hydropower plant. Three items matter most for a microfinance holding. First, the quarterly NPL and provisioning trend, read against the sector average rather than in isolation — a modest uptick that merely tracks the whole sector's cycle is a different signal than one that outpaces peers facing the same regulatory and macro environment. Second, any further change to NRB's base-rate-plus-premium regime, since it is still new enough that its full effect on sector-wide portfolio yields has not fully shown up in the numbers yet. Third, whether loan growth resumes at a pace that outruns the institution's own supervisory capacity — the same growth-versus-oversight discipline that separated strong from weak microfinance institutions during the 2021-2026 stress cycle in the first place.

Anjana closed her notebook on this third case study having learned something the bank and hydropower cases had not fully prepared her for: that a clean-looking number like a 2.3 percent NPL ratio is only trustworthy once you have actually asked the sector-specific questions Chapter 65 insists on, rather than crediting it on sight — and that this book's own Chapter 80 had arrived at the same requirement independently, from a completely different direction. A framework is only as trustworthy as an analyst's willingness to keep checking its cleanest-looking numbers against the questions that could unmask them, not just against the company being scored.