For years, laghubitta was the trade that never seemed to lose. Microfinance stocks paid fat dividends, split their capital again and again, and rewarded anyone who held through the noise. Retail investors learned the pattern and kept buying the past. That is the mistake worth naming up front. The microfinance business model in Nepal made real money for a real reason, but the conditions that produced those returns have changed, and buying laghubitta today on the strength of a five-year-old dividend record is looking straight into the rearview mirror.
To see why, you have to understand how these companies actually earn. It is not complicated. It is a spread business, and the spread is under attack from several directions at once.
What a Class D institution is
Nepal Rastra Bank (NRB), the central bank, sorts licensed banks and financial institutions into classes. Commercial banks are Class A, development banks Class B, finance companies Class C, and microfinance institutions (MFIs, locally laghubitta bittiya sanstha) are Class D. If you want the fuller taxonomy, we cover it in how Nepal’s Class A, B, and C banks differ. Class D sits at the bottom of that ladder by size and at the frontier by reach.
The core idea is old and well proven. Formal banks do not find it worth their while to chase a woman in a village who wants NPR 40,000 to buy two goats. The paperwork, the collateral checks, the branch visit, none of it pencils out on a loan that small. Microfinance exists to serve exactly that borrower, and it does so through a model borrowed from Bangladesh: group lending. Borrowers, usually women, form a group. The group meets regularly, and members effectively guarantee one another. If one member misses a payment, the others feel the pressure, socially and sometimes financially. That peer structure replaces the collateral a bank would demand, and it is the reason recovery rates were historically very high without any physical security behind the loan.
The sector splits into two layers. Wholesale MFIs lend to retail MFIs and cooperatives rather than to individuals. Retail MFIs, the ones you actually meet in the field and the bulk of what trades on the exchange, lend to the end borrower. As of April 2025, NRB counted 49 retail MFIs operating across all 77 districts and reaching roughly 6 million households (NRB Key Financial Indicators of Microfinance Institutions, mid-April 2025, cited by FinDev Gateway, July 31, 2025). That is a genuine national footprint, and it is the good the sector did. Account ownership in Nepal roughly doubled, from 25 percent in 2011 to 60 percent in 2024 (World Bank Global Findex, via FinDev Gateway). Microfinance was not the only driver, but it was one of them.
The engine: borrow wholesale, lend retail, turn it fast
Here is the model in one sentence. A microfinance company borrows money relatively cheaply and lends it out at a much higher rate, and the gap between the two is where the profit lives.
Where does the cheap money come from? Two main sources. The first is wholesale borrowing from commercial banks. Class A banks in Nepal have long been required to direct a share of their lending toward the “deprived sector,” and one easy way to meet that obligation is to lend the money to a Class D microfinance company, which then does the hard work of finding thousands of tiny borrowers. So the big banks are both shareholders in and lenders to the MFIs. The second source is customer deposits. MFIs are allowed to take savings from their own members, and those deposits are cheaper and stickier than bank borrowing. NRB data puts deposits at around 40 percent of MFI funding (FinDev Gateway, July 31, 2025). Deposits are cheap, but they are short to medium-term, which becomes a problem we will return to.
Now the lending side. The end borrower pays far more than the MFI pays for its funds. Historically, effective rates that borrowers faced ran as high as 24 to 30 percent once service charges were added (Nepal Rastra Bank directives, reported by Fiscal Nepal, July 2024). The MFI’s cost of funds was a fraction of that. The difference is the spread, and NRB caps it. The regulator has limited the interest rate spread for MFIs to. A spread cap sounds punishing, but on a large, fast-turning book it is still very profitable, which brings us to the second half of the engine.
Microfinance loans are small and short. A borrower takes a working-capital loan, repays it in weekly or fortnightly installments over a year, and takes another. Capital does not sit idle. It cycles. High turnover means the same rupee of lending capacity earns its spread multiple times relative to a bank’s multi-year term loan, and it means fees get charged again on each fresh cycle. Combine a protected spread, high turnover, and near-total recovery through group pressure, and you get the returns on equity that made laghubitta a dividend darling. That is the whole story of why these stocks paid so well.
If you want to see how spread, cost of funds and recovery quality show up in the numbers, our explainer on NPL, CD and CAR banking ratios walks through the same metrics that apply here.
A worked example, plainly
Say a retail MFI raises NPR 1 crore. It borrows part of that from a commercial bank at, for argument’s sake, 9 percent, and funds the rest with member deposits at 8 percent, so its blended cost of funds sits around 8.5 percent. It lends that money to groups at 15 percent. On the pure interest, that is a spread of about 6.5 percentage points. Add loan-processing fees and the compulsory savings members must keep with the institution, and the effective margin is wider than the headline spread suggests. Now cycle that NPR 1 crore book more than once a year as short loans repay and reissue. The margin compounds. That is why an MFI could earn a return on equity that a commercial bank, lending long at thinner spreads, could not touch.
Note the fragility hiding in that example. The whole thing depends on three assumptions holding at once: the spread stays open, the borrowers keep repaying, and the cheap funding keeps flowing. All three are now in question.
Why the model is under pressure
The spread is capped, and the cap has been squeezed
The first pressure is regulatory, and it is direct. For years NRB held microfinance loan rates under a hard ceiling of 15 percent, having earlier pushed the spread cap down toward 6 to 7 percent (Kathmandu Post, July 15, 2016; Investopaper). Then it changed the mechanism. From Shrawan (mid-July) 2025, NRB moved MFIs onto a base-rate system, the same logic banks already use, where an institution calculates its own base rate from cost of funds, administrative cost, and loan risk and may add a premium of up to 3 percent on new loans (NRB directive, reported by NEPSE Trading and Investopaper, 2025). The exact numbers are worth checking.
Whichever mechanism applies, the effect on profitability is the same. It squeezes. The World Bank-affiliated FinDev Gateway, working from NRB data, found that under the 15 percent cap only three MFIs reported a return on assets above 2 percent, and the sector average return on assets was just 0.7 percent as of mid-April 2025. That is thin. It is far below the profitability that built the dividend reputation, and it leaves little cushion to absorb losses.
Over-lending and multiple borrowing
The second pressure the sector brought on itself. In the growth years, MFIs crowded into the same accessible towns and villages and competed for the same borrowers. Households that should never have qualified for a second loan took a third and a fourth, one from each institution, often to repay the last one. This is multiple borrowing, and it is the classic way a microfinance market overheats. An NRB survey back in 2017 already found that 37 percent of borrowers had loans from at least two MFIs (reported via The Himalayan Times). Field observers put the real figure higher today.
NRB has responded with hard limits: capping the number of MFIs a single borrower can owe and capping total microfinance debt per borrower, alongside a rise in the group-loan ceiling to NPR 7 lakh from NPR 5 lakh (NRB directive, 2024-2025, via Investopaper and myRepublica). Those rules protect borrowers. They also cut off the growth. A model that expanded by lending more to the same people cannot keep expanding once each person is capped.
Rising defaults
The third pressure is the one that shows in the accounts. As the multiple-borrowing bubble deflated and an anti-microfinance movement disrupted repayment, defaults climbed sharply. Sector non-performing loans (NPLs, the share of the book that has stopped paying) rose from 2.6 percent in mid-July 2022 to 7.2 percent by mid-April 2025 (NRB data via FinDev Gateway). That is not a wobble. That is the recovery engine, the thing that made the model work, seizing up. FinDev found that 17 MFIs, together holding about 40 percent of the sector’s portfolio, carried NPLs above 7 percent. With average capital adequacy at only 10.3 percent and the average NPL at 7.2 percent, the sector’s cushion against bad loans is uncomfortably thin.
Political and regulatory backlash
The fourth pressure is political, and it is the one that turned a credit-quality problem into a solvency risk. Beginning in 2023, an organized borrowers’ movement, at times calling itself the Microfinance Victims Struggle Committee, accused MFIs of predatory rates, abusive recovery, and driving families into debt traps. The movement demanded debt relief, in some cases outright loan waivers. Repayment discipline, the social glue of group lending, broke down in affected districts as borrowers were encouraged to stop paying.
The state has since pushed back. In a ruling reported in May 2026, Nepal’s Supreme Court held that the agitation posed a serious threat to the financial system and the rural economy while also acknowledging that some institutions had over-lent, chased profit too hard, and used abusive collection tactics, and directing NRB to enforce its rules strictly (Clickmandu, May 2026). Read that carefully. Both sides were told they were partly right. The court did not clear the sector’s conscience. It confined the movement’s methods.
The dividend question, honestly
Here is where an investor has to be clear-eyed. NRB now ties dividend payouts to health metrics: broadly, an MFI needs capital adequacy above 12 percent and NPLs under 5 percent to distribute up to 25 percent, with lower bands for weaker institutions (NRB policy, reported by ShareSansar, August 25, 2025). Read against the sector averages above, most MFIs no longer clear the top bar. The generous, uniform payouts of the past are gone by regulation, not by choice.
When that dividend policy was announced, laghubitta stocks jumped and led NEPSE’s weekly gainers (Khabarhub, 2025). That reaction tells you how much of the sector’s market value still rests on the dividend story rather than on the fundamentals underneath it. Microfinance is among the most volatile corners of the exchange for exactly this reason, and we look at that trading behavior separately in why microfinance stocks are so volatile on NEPSE. Dividend policy across the wider market is worth understanding too; our piece on commercial bank dividends in Nepal sets out the same rules from the other end of the ladder.
The verdict
Give the sector its due. Microfinance in Nepal put credit into hands that formal banks would not touch, doubled account ownership, and financed millions of tiny enterprises. That was real financial inclusion, not marketing. The spread model, borrowing wholesale and lending retail at high turnover with group recovery, worked because all its assumptions held.
They no longer hold. The spread is capped and being managed down. The growth trick of lending ever more to the same borrowers is fenced off. Defaults have tripled off their base. And the political ground has shifted from expansion to protection and relief. None of this means the sector dies. Consolidation, individual and collateral-based lending, and stronger balance sheets can produce a smaller, sounder industry. But it will not be the dividend machine of 2018.
So the honest position is this. The easy-growth era is over. An investor buying laghubitta because it paid handsomely five years ago is pricing a business that no longer exists. Judge these stocks on today’s spread, today’s NPLs, and today’s capital, or do not judge them at all.
This is analysis, not financial advice.
