The Hydropower Sector Playbook

First published 23 Aug 2026 · Last verified 29 Aug 2026

Anjali Basnet keeps two folders on her desk in her flat above a stationery shop in New Baneshwor. One is thick, full of prospectuses, audited statements, and photocopied newspaper clippings about landslides and transmission lines. The other is thin, holding only her share certificates and a dividend record book. Both folders are about hydropower. She has learned, the hard way, over eleven years of applying for hydropower IPOs and holding a rotating cast of run-of-river and storage project shares, that the thick folder is what protects the thin one. In 2019 she bought into a pre-construction hydropower IPO on the strength of a glossy prospectus cover showing a turbine hall and a projected internal rate of return that seemed too polished to argue with. The project's commercial operation date slipped by nineteen months after a monsoon debris flow destroyed a section of the headrace tunnel access road. By the time the plant finally synchronized to the national grid, Anjali had held a non-dividend-paying, book-value-only stock for the better part of four years while bank fixed deposits paid her neighbours 9 percent annually. She did not lose her principal. She lost four years of opportunity cost, and she never wants to make that mistake again.

Hydropower is the closest thing NEPSE has to a national identity sector. It is also the sector where investors most reliably confuse two entirely different kinds of businesses because they trade under the same sub-index and wear the same green turbine-and-mountain logo language in their annual reports. This chapter is the playbook Anjali eventually built for herself, expanded into the systematic framework every NEPSE investor needs before putting capital into a hydropower ticker, whether that ticker represents a hole in the ground with a construction schedule or a fully commissioned plant with eleven years of dividend history.

Lesson 72.1 — Two Species Under One Ticker: Pre-COD and Operating Hydropower

The single most important distinction in hydropower investing is not which river a project sits on, nor how many megawatts its nameplate capacity claims. It is whether the company has reached its commercial operation date, universally abbreviated COD, the day the Nepal Electricity Authority formally accepts the plant's output onto the national grid and the power purchase agreement's revenue clock starts running. Before COD, a hydropower company is a construction project wrapped in a public company's legal form. After COD, it is a cash-generating infrastructure asset with a predictable, contractually defined revenue stream. These are not two stages of the same investment. They are two different asset classes that happen to share a listing category.

A pre-COD company's balance sheet is dominated by capital work in progress, an accounting line that grows every quarter as the developer draws down loan tranches and equity calls to pay the EPC contractor, the transmission line builder, and the land acquisition compensation committee. There is no revenue line, or a token one from a small owned diesel genset or advance energy sales that barely registers. Earnings per share is negative, zero, or a rounding artifact of interest income earned on undisbursed loan proceeds sitting in a bank account. Book value per share is essentially a statement of how much capital has been spent so far, not a statement of what the asset is worth once finished. None of the ratios retail investors reach for reflexively, price-to-earnings first among them, mean anything applied to a pre-COD company, because there are no earnings to divide the price by. Valuing a pre-COD hydropower stock is an exercise in project appraisal: estimated total project cost per installed megawatt, financing structure, expected COD, and a discounted cash flow built on the PPA tariff schedule the company has actually signed, not the tariff schedule of some comparable operating plant on a different river.

An operating company is the opposite. It has a stabilised, or maturing, revenue stream governed by its PPA, actual generation data measured against design estimates, a dividend track record that can be examined the way you would examine any income-generating equity, and a debt schedule that is now amortizing rather than being drawn. Its risks are operational and regulatory rather than construction risks: plant availability, hydrology in a given fiscal year, royalty and tax step-ups, and the health of its relationship with the single buyer that is legally obligated to purchase everything it produces. Table 72.1 lays out the contrast Anjali eventually drew on a single sheet of paper and taped inside her thick folder.

Anjali's rule now, stated in her own handwriting on the inside cover of the thick folder, is this: before reading a single ratio, determine which species of company you are looking at. If the annual report's income statement has more footnotes about capitalised borrowing costs than it does about units sold to NEA, you are still in construction-phase territory, and the entire analytical toolkit shifts.

Lesson 72.2 — The Physical Risk Layer: Hydrology, River Type, and Evacuation

Hydropower's defining risk is not financial in origin. It is hydrological, and it sits upstream of every other number in the annual report. Nepal's rivers fall broadly into two behaviours that matter enormously for how a project generates and earns. A run-of-river, or ROR, project has no meaningful reservoir. It diverts a portion of a river's flow through a headrace tunnel or canal to a powerhouse and returns the water downstream, generating power more or less in proportion to whatever the river happens to be carrying at that moment. A storage project, by contrast, impounds water behind a dam, allowing the operator to release it when it is needed rather than when the river happens to deliver it, which means storage plants can generate through the dry season and can be dispatched during peak demand hours regardless of instantaneous river flow. A peaking run-of-river, or PROR, project sits in between, with a small pondage that allows a few hours of daily flow-shifting to align generation with the hours NEA's grid needs it most, typically the evening peak.

This distinction matters because Nepal's rivers are overwhelmingly monsoon-fed rather than glacier-fed in their seasonal amplitude, meaning roughly 80 percent of annual flow arrives in the four monsoon months of June through September, and dry-season flow in many rivers, particularly in the middle hills, can fall to a fraction of wet-season flow. A pure ROR plant sized to its river's monsoon discharge will run near full capacity for a third of the year and at a much lower plant load factor for the rest of it. This is precisely why NEA's power purchase agreements pay a materially higher per-unit tariff for dry-season energy than for wet-season energy: dry-season electrons are scarce and valuable to the grid, wet-season electrons are, in aggregate across dozens of ROR projects commissioning in the same river basins, increasingly abundant and at the margin close to worthless if the grid cannot absorb them. An investor evaluating an ROR-heavy hydropower stock must model revenue on a seasonally weighted basis, never simply annual generation multiplied by an average tariff, because the mix between wet and dry season units generated determines earnings far more than total annual kilowatt-hours does.

Hydrology risk compounds this seasonality with year-to-year variability that no engineering study fully eliminates. A below-average monsoon, a glacial lake outburst flood upstream, an unusually early or late withdrawal of the monsoon, or accelerated sedimentation silting up an intake all reduce actual generation below the design energy figure quoted in the prospectus. Sedimentation deserves particular attention in Nepal's young, geologically active Himalayan and Chure hill catchments, where rivers carry heavy sediment loads during the monsoon that erode turbine runners and can, over years, reduce reservoir storage capacity in storage-type projects faster than initially modelled.

The final physical risk layer, and one Nepali retail investors historically underweight, is transmission and grid evacuation. A hydropower plant that generates power it cannot deliver to the grid earns nothing for that unserved energy, and Nepal's transmission network has repeatedly lagged the pace of generation capacity addition in specific river corridors, most visibly where multiple ROR projects on the same tributary system commission around the same period and compete for headroom on a single evacuation line. During peak monsoon flow, when many ROR plants across a basin are simultaneously near full output, NEA's system has at times been unable to absorb all available generation, resulting in curtailment or spillage where developers are instructed to reduce output regardless of available river flow. Whether a given company's PPA and grid interconnection agreement shield it financially from curtailment, through deemed generation payments, or expose it directly to the lost revenue, is a contractual detail every investor must verify rather than assume.

Lesson 72.3 — Reading the PPA: Tariff Structure, Escalation, and Take-or-Pay

The power purchase agreement is the single document that matters most for an operating hydropower company, and yet it is the document most NEPSE retail investors have never actually read. Every unit of revenue an operating hydropower company earns flows from this one contract with a single buyer, the Nepal Electricity Authority, and the contract's structure is knowable, standardised within categories, and worth understanding line by line rather than trusting a summarised tariff figure quoted in a broker note.

For a typical small to mid-size run-of-river project, NEA's standard PPA structure sets two separate per-unit tariffs, one for energy delivered during the wet season months and a materially higher one for energy delivered during the dry season, reflecting the relative scarcity of dry-season generation across the system. An illustrative real example is instructive: a run-of-river project with a required commercial operation date in the late 2020s carries a wet-season tariff of roughly Rs 4.80 per kilowatt-hour and a dry-season tariff of roughly Rs 8.40 per kilowatt-hour, close to a 1-to-1.75 ratio between the two seasons. Both figures escalate at a fixed 3 percent annually, but only for the first eight years from COD; from year nine through the remainder of the PPA's typically thirty-year tenure, the tariff is frozen in nominal terms. This detail is easy to miss and expensive to misunderstand. An investor who extrapolates the escalating early-year tariff growth rate forward across the full life of the PPA will overstate long-run revenue growth substantially, because in real, inflation-adjusted terms, a frozen nominal tariff after year eight means the plant's revenue per unit actually declines in purchasing power for the following two decades, even as its physical output stays constant.

Table 72.2 shows this schedule in illustrative form for a standard run-of-river PPA of the kind that governs the majority of NEPSE-listed hydropower companies.

Storage and peaking projects are priced differently, and the regulatory framework governing them has been evolving. Nepal's Electricity Regulatory Commission has floated tariff structures for storage hydropower that separate an energy charge, again split by wet and dry season, from a capacity charge paid per kilowatt of available capacity per month regardless of how much energy is actually dispatched, compensating the plant for standing ready to deliver power during system peaks rather than only for the energy it generates. This capacity-plus-energy structure is meant to reflect the genuinely different economic value storage and peaking plants provide to a grid still dominated by seasonal, weather-dependent generation, and tariffs proposed under this framework run considerably higher than standard run-of-river rates precisely because they are designed to make otherwise more expensive dam and reservoir infrastructure financeable. An investor comparing a storage project's per-unit tariff against a run-of-river project's tariff without adjusting for the capacity payment component and the different risk and cost structure behind a dam project is comparing two different products as though they were interchangeable commodities.

Every operating hydropower company's annual report and every pre-COD company's prospectus discloses its specific PPA tariff schedule, escalation terms, and tenure. There is no substitute for pulling this schedule directly rather than relying on a single blended average tariff figure quoted secondhand, because the wet-dry season split, the escalation cutoff year, and the remaining tenure of the agreement are together what determine the shape of the company's revenue for the next two or three decades.

Lesson 72.4 — Royalty, Taxation, and the Year 10/15 Value Cliff

Even a hydropower company generating exactly its design energy and selling every unit at its full contracted tariff does not keep all of that revenue. Two government levies, royalty and income tax, both step up sharply at specific points in a project's life, and both step-ups tend to cluster around the same window, roughly ten to fifteen years after COD, creating what is best understood as a value cliff that every long-horizon hydropower investor needs to model explicitly rather than discover in an annual report a decade after buying the stock.

The royalty regime charges hydropower generators two separate royalties, a capacity royalty assessed per kilowatt of installed capacity per year and an energy royalty assessed as a percentage of total revenue, and for smaller plants in the roughly one to ten megawatt range, both royalties are set on a schedule that rises steeply after year fifteen. The capacity royalty runs at approximately Rs 100 per kilowatt per year for the first fifteen years of operation and then jumps to approximately Rs 1,000 per kilowatt per year, a tenfold increase, from year sixteen through year thirty. The energy royalty runs at approximately 2 percent of total revenue for the first fifteen years and then rises to approximately 10 percent of total revenue for years sixteen through thirty, a fivefold increase. Larger projects sit on their own government-set royalty schedules that are generally steeper still, reflecting their scale, and any investor holding a large storage or peaking project's shares should confirm that project's specific royalty bracket directly from its PPA and licensing documents rather than assuming the small-project schedule above applies.

Income taxation follows a parallel, deliberately front-loaded incentive structure meant to help projects survive their highest-leverage early years. A hydropower company enjoys a full, 100 percent income tax holiday for its first ten years of commercial operation, followed by a 50 percent rebate on the standard 20 percent corporate income tax rate for years eleven through fifteen, meaning an effective 10 percent tax rate during that window, before paying the full 20 percent rate from year sixteen onward. This benefit has historically been tied to a COD cutoff date in government notices, so an investor should verify that a specific company's COD falls within whatever eligibility window currently applies rather than assuming every listed hydropower stock automatically qualifies for the full schedule.

Table 72.3 lays out the combined effect for a typical small to mid-size run-of-river company, using the illustrative royalty and tax brackets above.

Stack this against the tariff freeze from Lesson 72.3, where nominal per-unit revenue stops growing after PPA year eight, and a pattern emerges that is easy to miss if an investor simply extrapolates a company's current dividend per share forward. A hydropower stock in years four through eight after COD typically shows its most attractive-looking headline numbers: revenue still escalating, zero income tax, low royalty. Somewhere between years ten and sixteen, three things happen close together: the tax holiday ends and gives way first to a partial then a full tax charge, the energy royalty quintuples, and the capacity royalty rises tenfold, while the top-line tariff has already been frozen in nominal terms for several years. None of this necessarily means the stock becomes a bad investment at that point; by then debt is typically well into amortisation or fully repaid, which is its own offsetting improvement in free cash flow, a dynamic covered in Chapter 68's discussion of how dividend capacity shifts across a company's life cycle. But it does mean that a naive five-year-forward projection built by simply compounding a company's most recent dividend growth rate will be wrong, sometimes badly wrong, for any company approaching this window.

Anjali now builds a simple year-tracker for every operating hydropower stock she holds: COD date, current year of operation, tax bracket, royalty bracket, and years remaining on the PPA. It takes five minutes to fill in from an annual report and it has saved her from over-extrapolating a company's best years at least twice.

Lesson 72.5 — Debt Structure, Covenants, and the Moratorium Question

Hydropower is a leveraged business by design. Nepali project financing for hydropower typically runs with a debt-to-equity ratio in the range of 70:30 to 75:25, meaning for every rupee of promoter and public equity in the project, roughly two and a half to three rupees are borrowed, usually from a syndicate of Nepali commercial banks since single-bank exposure limits and the combined lending capacity of the domestic banking sector constrain how much any one institution, or even the sector collectively, can lend to hydropower without syndication. This leverage is precisely what makes hydropower returns attractive to equity holders when things go according to plan, since a fixed-tariff, contracted revenue stream financed mostly with debt magnifies the equity return, and precisely what makes construction delays and cost overruns so punishing when things do not go according to plan, since additional debt drawn to cover an overrun dilutes the equity return on a project whose revenue ceiling is fixed by the PPA regardless of final cost.

Loan tenure for Nepali hydropower project debt has typically run eight to twelve years, a period that usually begins with a moratorium, a grace interval, generally coinciding with the construction period, during which the borrower pays interest only, often capitalising part of it into the loan principal, before principal repayment begins in earnest once the plant is commissioned and generating revenue. The length of this moratorium matters directly to an equity holder because it determines how much of the loan principal remains outstanding at COD and therefore how steep the amortisation schedule needs to be to retire that debt within the remaining loan tenure, which in turn determines how much of the plant's early operating cash flow is absorbed by debt service rather than being available for dividends. A project that emerges from a longer, more generous moratorium with a larger outstanding principal balance will typically show weaker dividend capacity in its first several operating years than an otherwise identical project that entered construction with a smaller loan or a shorter moratorium, even though both may show similar operating profit.

Lenders protect themselves through covenants, chief among them a minimum debt service coverage ratio, or DSCR, defined as cash available for debt service divided by the debt service due in a period. Nepali hydropower project loans commonly carry minimum average DSCR covenants, and a breach, or even a DSCR trending toward the covenant floor, typically triggers consequences that flow directly through to equity holders: restricted or prohibited dividend distributions until a debt service reserve account is topped up, mandatory cash sweeps that divert surplus cash to accelerated principal repayment instead of dividends, and in persistent breach scenarios, renegotiation or acceleration rights for the lender. This is the single most important reason an operating hydropower company with genuinely positive net profit can still pay no dividend, or a token one, in a given year: the loan agreement, not the income statement, governs how much cash the board is legally free to distribute.

Construction-phase cost overruns deserve a final, specific warning, because they interact with both the debt structure and the PPA in ways that compound rather than simply add. When a project's actual cost exceeds its appraised cost, the shortfall is typically financed with additional debt drawn against the same fixed future revenue stream, which both raises the debt-to-equity ratio beyond what was originally underwritten and lengthens the period before DSCR comfortably clears its covenant. Separately, and just as importantly, missing the required commercial operation date can cost a project its escalation entitlement.

Lesson 72.6 — The Nine-Step Evaluation Checklist

Anjali's thick folder now opens with a single page, a nine-step checklist she runs through for any hydropower stock, pre-COD or operating, before she commits capital. It is written to be worked through in order, because early steps determine which later steps even apply.

The first step is to establish life-cycle stage precisely: for a pre-COD company, the physical construction progress percentage against the required commercial operation date; for an operating company, the exact number of years elapsed since actual COD. Everything else in the checklist is read differently depending on the answer to this first question.

The second step is to classify the project's physical type, run-of-river, peaking run-of-river, or storage, and to locate the hydrological study or, for an operating plant, at least three to five years of actual generation data measured against the original design energy estimate. A persistent shortfall against design energy is a flag that must be explained, not waved past.

The third step is to obtain and read the actual PPA: the wet and dry season tariffs, the escalation rate and the year it stops, the total tenure and years remaining, and whether the agreement is take-or-pay or take-and-pay. This single document, more than any ratio in the annual report, determines the revenue ceiling of the business for the next two to three decades.

The fourth step is to check transmission and grid evacuation readiness for the specific corridor the project sits in, looking for any history of curtailment, spillage, or interconnection delay affecting that river basin or transmission line, since a plant that cannot deliver its power earns nothing for the units it cannot evacuate regardless of how well it generates them.

The fifth step is to pull the debt structure: debt-to-equity ratio at financial close, loan tenure, moratorium period, and, where disclosed in a credit rating report or bond prospectus, the minimum DSCR covenant and whether a debt service reserve account is fully funded. This step tells the investor how much of near-term operating cash flow is legally available to reach shareholders at all.

The sixth step is to place the company on the royalty and tax bracket timeline: which year bracket of the capacity and energy royalty schedule it currently sits in, and which year bracket of the income tax holiday and rebate schedule, so that current-year margins are not mistaken for a permanent steady state.

The seventh step, applicable only to operating companies, is to examine actual plant load factor against design capacity factor, and to build a simple dividend history table showing payout ratio trend and whether dividends have tracked, lagged, or outpaced reported profit, since a rising profit with a flat or declining dividend is usually a debt covenant story rather than a governance story, though both are worth ruling out.

The eighth step, applicable only to pre-COD companies, is to research the EPC contractor's track record on other Nepali hydropower projects, cross-check the current physical progress percentage against the required commercial operation date, and identify what cost overrun and delay penalty provisions exist in the construction contract and the PPA respectively.

The ninth and final step is valuation, applied with the correct tool for the life-cycle stage established in step one: for a pre-COD company, cost per installed megawatt against comparable recently commissioned projects and a discounted cash flow built on the actual signed PPA tariff schedule rather than a generic sector average tariff; for an operating company, a price-to-earnings multiple used only with the year-bracket caveats from step six firmly in mind, supplemented by an EV/EBITDA comparison and, where the investor has the patience for it, a dividend discount model that explicitly steps the royalty and tax assumptions at the correct future years rather than holding them constant.

Running this checklist against the pre-COD stock that cost her four years of opportunity cost, Anjali now sees clearly what she missed the first time: an EPC contractor with a public record of delays on two prior projects, and a construction contract with no meaningful delay-penalty clause protecting the developer. She would not buy that stock today at any price without first pricing in that specific contractor risk. On her more recent purchase, an operating run-of-river company nine years past COD with a clean three-year generation record against design energy, a DSCR comfortably above covenant according to its published rating rationale, and four more years remaining before its royalty and tax step-up, she bought with the year-bracket clock already running in her notebook, so that when the step-up arrives, it will be an expected event she has already modelled into her return expectations rather than a surprise that shows up one year in a dividend that is smaller than she hoped.