Walk into any IPO discussion in Kathmandu and the first number anyone quotes is megawatts. “It’s a 25 MW project.” “This one is 86 MW, much bigger.” The logic feels obvious. Bigger plant, bigger company, bigger payday. It is also mostly wrong, and it is the single most expensive mistake retail investors make when they try to value a hydropower stock on NEPSE.
Hydropower is the largest and noisiest IPO pipeline on the exchange. Dozens of promoters chase generation licenses, sign a power purchase agreement, build for four or five years, and then float shares to a public that has been trained to subscribe on faith. The marketing leans on installed capacity and a one-line dividend hope. Almost nobody reads the contract that actually decides whether the company makes money. That contract is the PPA, and once you understand it, you stop counting megawatts and start counting rupees.
This is a framework piece, not a stock tip. By the end you should be able to look at a hydropower prospectus and ask the three questions that matter: What does the PPA actually pay? How much electricity will this plant really produce across a full year? And how heavy is the debt that has to be serviced before you see a single paisa of dividend?
Why megawatts mislead you
Installed capacity is the nameplate rating of the turbines. It tells you the maximum the plant can produce at any instant, under ideal flow. It does not tell you how often the plant hits that maximum, and in Nepal the honest answer is not often, and almost never in winter.
Most projects listed on NEPSE are run-of-river plants. A run-of-river plant has little or no reservoir. It generates from whatever the river is carrying that day, which means its output tracks the hydrology, not your hopes. According to research on Nepal’s river basins, generation is high during the monsoon and falls sharply in the dry winter months (ScienceDirect / PMC study on climatic variation and hydropower generation in Nepal). The Nepal Electricity Authority has reported dry-winter output falling roughly 20% below the winter average in a year with little rain and weak snowmelt, with the country importing power to cover a large share of peak winter demand (NepalEnergyForum).
So a 25 MW run-of-river plant is a 25 MW plant for a few monsoon weeks and something far smaller for months at a stretch. The number that matters is the capacity factor: actual annual energy produced divided by the theoretical maximum if the plant ran flat out all year. Two plants can share the same nameplate and have very different capacity factors depending on the river, the catchment, and the design. When you read a prospectus, find the design energy or the projected annual generation in gigawatt-hours, not just the MW headline. That is the number that turns into revenue.
Read the PPA before you read anything else
The power purchase agreement is the contract between the developer and NEA, which for almost every domestic project is the only buyer. There is no open market to sell to. If you do not understand the PPA, you do not understand the company. Four things inside it drive the whole valuation.
The tariff and the wet/dry split. NEA does not pay a single flat rate. It pays one rate for wet-season energy and a higher rate for dry-season energy, precisely because dry-season power is scarce and valuable. Reported run-of-river rates have sat around (Hydropower & Dams International). Hold that structure in your head, because it creates a cruel mismatch. A run-of-river plant produces the most energy in the wet season, when the price is lowest, and the least in the dry season, when the price is highest. The high winter tariff barely helps you if the river has shrunk to a trickle. Reservoir and storage projects, which can hold water and release it in winter, are built to capture exactly that premium, which is part of why policy now offers them better terms.
Escalation. A rupee in year fifteen is worth far less than a rupee today, and a fixed tariff would be slowly eaten by inflation. NEA’s standard run-of-river PPAs have included an annual escalation, commonly cited as around (Hydropower & Dams International). Note the limit. Escalation that stops after a handful of years means the tariff is effectively flat for most of the project’s life. That is a quiet erosion that the MW headline will never show you.
Take-or-pay versus take-and-pay. This is the most underpriced clause in the entire sector, and most retail buyers have never heard of it. Under take-or-pay, NEA is obligated to pay for the contracted energy whether or not it actually uses it. That guarantee is what gave banks the confidence to lend, because the revenue was contractually certain. Under take-and-pay, NEA pays only for the electricity it chooses to take when it chooses to take it. In June 2025 the government’s budget moved new PPAs toward take-and-pay, set off a sector-wide alarm, and was then walked back later that month after fierce pushback from developers and lenders (Kathmandu Post; Fiscal Nepal). The episode told you everything about where the real risk sits. When NEA already spills surplus power in the wet months and worries about oversupply, the question is no longer just “Will the plant generate?” but “Will NEA actually buy what it generates?” Read the specific PPA terms in the prospectus. Do not assume the old guarantee applies.
The term and the handover. A hydropower license is not forever. Nepal’s framework grants generation licenses for a defined period, commonly cited as 35 years for domestic projects under the Hydropower Development Policy 2001, after which most projects are built on a Build-Own-Operate-Transfer (BOOT) basis and handed to the government (Clickmandu; Unihome). Read that twice. The company you are buying has a finite life. At the end of the license, the asset goes to the state, and the equity that remains is whatever cash was distributed along the way, not a perpetual annuity. Some of the oldest BOOT projects are now approaching their handover dates, which has already created real uncertainty about the residual value of those shares (clickmandu). If you are buying a hydro stock as a “forever” dividend payer, the contract disagrees with you.
Cash flow, not capacity, is the company
Put the pieces together and you can build the only model that matters. Revenue is annual energy times the seasonal tariff. From that you subtract operating and maintenance costs, which for hydro are relatively low, and then the big one: debt service.
Hydropower in Nepal is built heavily on borrowed money. Debt-to-equity ratios are high by design, and high debt means a heavy interest burden that has to be paid before any dividend (Bajarko Chirfar English). There are two debt traps worth naming.
The first is interest during construction. A plant earns nothing for the four or five years it is being built, but the loan clock is already running. That interest gets capitalized into the project cost, which inflates the total capital the company must eventually earn a return on. A project that runs over schedule, and many do, simply piles on more interest before the first unit is sold.
The second is the repayment hump in the early operating years. In the first years after commissioning, a large slice of cash flow goes to repaying principal and interest. This is exactly when new retail shareholders expect their first fat dividend and exactly when the company can least afford one. A plant can be operating perfectly, generating as designed, and still pay a thin dividend because the bank is first in line. That is not failure. That is the structure. But it is invisible if you only looked at the megawatts.
Here is a deliberately simplified worked example to make the shape clear. The numbers are illustrative, not a forecast.
Take a 20 MW run-of-river plant. Assume it produces around 105 gigawatt-hours a year across wet and dry seasons combined (a capacity factor in the rough region of 60%, which is generous and depends entirely on the river). Blend the wet and dry tariffs, and suppose the average realised price works out to roughly NPR 6 per unit. Revenue is then about NPR 63 crore in a year. Now suppose the project cost NPR 360 crore, financed 70% by debt, so about NPR 252 crore of loans. Interest and principal in the early years can easily consume a very large portion of that revenue, leaving thin distributable profit for the first several years even though the plant is running well.
Change one input and watch it swing. If the river underdelivers and annual energy comes in at 90 GWh instead of 105, revenue drops by roughly NPR 9 crore straight off the top, and because debt service is fixed, almost all of that loss lands on the equity holders’ share. That sensitivity, the way hydrology risk flows directly to your dividend, is the whole game. Megawatts cannot show it to you. A cash-flow model can.
The risk retail is mispricing
If you force me to name the single most mispriced risk in NEPSE hydropower, it is not construction delay, and it is not the headline tariff. It is the combination of dry-season revenue weakness and the slow shift in offtake certainty.
Retail buyers price these stocks as if every unit generated will be bought at a healthy blended rate forever. The reality is narrower. Run-of-river plants are structurally weakest in winter, the season when power is most valuable, so they capture the lowest-value energy in volume and the highest-value energy only in trickles. On top of that, NEA’s appetite to buy is no longer unconditional. The 2025 take-and-pay scare was reversed, but it revealed the direction of travel: a single buyer that already spills surplus in the wet season and is increasingly unwilling to guarantee offtake. For new projects, especially larger run-of-river ones, the risk that some wet-season generation gets curtailed or goes unpurchased is real and rising (Kathmandu Post).
My view, stated plainly: the market is paying for installed capacity and IPO scarcity and underpaying attention to offtake certainty and seasonal revenue quality. A storage or peaking project with a strong, escalating, take-or-pay PPA can be worth meaningfully more per megawatt than a run-of-river plant twice its size on a weak contract. Until retail starts pricing the PPA instead of the nameplate, the cheap-looking big-MW IPO will keep disappointing on dividends, and the boring, well-contracted project will keep being underrated.
A practical checklist before you subscribe
When the next hydropower IPO opens, skip the WhatsApp hype and open the prospectus. Find these in this order. Is it run-of-river or storage, and what is the projected annual energy in GWh, not just the MW? What are the wet and dry tariffs? What is the escalation rate, and how many times does it apply? Is the PPA take-or-pay or take-and-pay, and is there any curtailment or limited-energy clause? What are the debt-to-equity ratio and the repayment schedule in the first five operating years? And how many years are left on the license before BOOT handover? If a prospectus is vague on any of these, that vagueness is itself information.
If you want to understand the wider IPO mechanics behind these floats, our guide to how IPO allotment works in Nepal covers the subscription and allotment side, and the difference between an IPO, an FPO, and a rights share matters because hydro companies often come back to shareholders for more capital mid-construction. Once the stock is listed, learning to read the NEPSE floor sheet will tell you whether the post-IPO trading is genuine interest or a thin pump.
Hydropower can be a sound long-term holding. Some Nepali projects throw off steady cash and respectable dividends for years. But the ones that do are not chosen by megawatts. They are chosen by the contract, the river, and the balance sheet. Count the cash flow, read the PPA, and respect the dry season, and you will value these stocks the way the lenders already do, which is the only group in the room that consistently gets paid.
This is analysis, not financial advice.
