Dividend Discount Model (DDM)

First published 23 Aug 2026 · Last verified 29 Aug 2026

Imagine you own a small rental house in Baneshwor. You don't plan to sell it next year. What it's "worth" to you, in a very real sense, is the rent it throws off, year after year, for as long as you hold it — discounted back to today because a rupee promised five years from now is worth less than a rupee in your hand today. If the tenant pays reliably and the rent grows a little each year as you renovate and as market rents rise, the house is worth more. If the tenant is erratic, or the neighbourhood is declining, the house is worth less, even if the paint is fresh and the plot is large.

A share of common stock in a NEPSE-listed company is not so different. You do not own the factory, the bank branch, or the powerhouse. You own a claim on the cash the company chooses to pay out to you, its shareholder, over time — the dividend. The Dividend Discount Model, or DDM, values a share exactly the way you would value that rental house: as the present value of the cash it is expected to pay you, forever, discounted at a rate that reflects how risky that cash stream is.

This chapter builds the DDM from its simplest form to a more realistic multi-stage version, and — because this is a book about Nepal — spends real time on why this particular model, more than almost any other valuation tool in Part IX, fits NEPSE like a glove, and where it can also mislead an investor who applies it carelessly to the wrong kind of company at the wrong stage of its life.

Lesson 48.1 — The Rental House Logic: Gordon Growth Model Mechanics

Start with the simplest possible version of a dividend stream: a company that pays a dividend today, and that dividend is expected to grow at a constant rate, forever. This is the world of the Gordon Growth Model, named after the economist Myron Gordon, who formalised it in the 1950s and 1960s. It is the workhorse version of DDM, and nearly every more complex DDM you will ever build is just several Gordon Growth calculations stitched together.

The formula is:

Value per share = D1 / (r − g)

Where D1 is the dividend expected next year (not the one just paid — the one coming), r is the cost of equity (the return shareholders require for holding this stock, given its risk — covered in depth in Chapter 44), and g is the constant expected growth rate of the dividend, forever.

Go back to the rental house. D1 is next year's rent. r is the return you need to be compensated for tying your money up in this property rather than a government bond or a fixed deposit — it bakes in the risk that the tenant stops paying, the building needs repairs, or the neighbourhood declines. And g is how fast you expect the rent to grow, year after year, forever — perhaps 4-5% a year as the area develops and market rents drift upward.

If you expect Rs 100,000 in rent next year, need a 12% return to compensate you for the risk of being a landlord in that particular neighbourhood, and expect rent to grow 4% a year forever, the property should be worth:

Rs 100,000 / (0.12 − 0.04) = Rs 100,000 / 0.08 = Rs 1,250,000

Notice something important: the model requires r to be strictly greater than g. If growth ever equals or exceeds your required return, the formula breaks — it produces an infinite or negative value, which is a mathematical way of telling you that no company can grow its dividend faster than the market's required return, forever. A company can grow explosively for five years, or ten, but not forever — eventually competition, market saturation, or the sheer scale of the business drags growth back down toward something close to the growth rate of the overall economy. This is a crucial constraint you will lean on again in Lesson 48.5.

Now put a NEPSE bank into the same frame. Suppose Himalayan Ridge Bank Ltd. (a hypothetical commercial bank, used for illustration throughout this chapter) is expected to pay a cash dividend of Rs 15 per share next year. Investors require a 14% return on bank equities of this risk profile (built up, as Chapter 44 showed, from a risk-free rate plus an equity risk premium adjusted for the bank's beta and Nepal-specific risk factors). The bank's dividend is expected to grow at 6% a year indefinitely, roughly in line with nominal GDP growth and the banking sector's steady expansion of its loan book.

Value per share = 15 / (0.14 − 0.06) = 15 / 0.08 = Rs 187.50

If Himalayan Ridge Bank trades at Rs 150 on NEPSE, the Gordon Growth Model says it is undervalued relative to its dividend-paying capacity, holding the growth and discount rate assumptions fixed. If it trades at Rs 260, the model says the market is pricing in either faster growth, a lower required return, or both, than your assumptions capture — a signal to revisit your inputs, not necessarily proof the market is wrong.

The two inputs that make or break this model — r and g — are exactly the two places where an analyst's judgment, not a formula, does the real work. Chapter 44 covered how to build r (the cost of equity) carefully. The rest of this chapter is substantially about g: how to estimate it responsibly for a Nepali company, and why Nepal's own dividend culture makes this both easier and, in specific ways, trickier than in markets where dividends are smaller and less central to how companies return value to shareholders.

Lesson 48.2 — Why DDM Is Nepal's Natural Valuation Tool

In many developed equity markets, dividends are almost a rounding error in total shareholder return. American technology companies often pay no dividend at all for decades, returning cash instead through share buybacks, and an analyst who tried to value Amazon or a similar growth company using a simple DDM would get a value of approximately zero — an obviously wrong answer, because it ignores nearly all the ways in which that company actually creates value for its owners.

NEPSE is a different animal entirely. Nepali listed companies — above all, commercial banks, development banks, finance companies, microfinance institutions, insurance companies, and hydropower companies — have a deeply entrenched culture of paying out a large share of annual profit as dividends, in cash, in bonus shares, or a mix of both. Several structural features of the Nepali market explain why.

First, share buybacks are rare and, for most sectors, effectively unavailable or heavily restricted as a mechanism for returning cash to shareholders. A Nepali company that generates a profit and wants to reward shareholders has essentially one channel available at scale: declare a dividend. Second, retained earnings for regulated financial institutions are subject to capital and reserve requirements (discussed in Lesson 48.4), which push a portion of profit into reserves rather than reinvestment in new ventures, but the remainder is customarily distributed rather than hoarded on the balance sheet, partly because minority shareholders — a large share of the register for most NEPSE banks — expect and demand it. Third, listed hydropower companies, once a plant reaches commercial operation and starts generating predictable cash flow under a long-term Power Purchase Agreement (PPA) with the Nepal Electricity Authority, tend to have limited further capital expenditure needs for that specific plant, so free cash flow converts into dividends rather than fresh reinvestment, at least until the company undertakes a new project.

Because dividends are the primary channel through which Nepali listed companies return value to shareholders, and because that dividend stream is what most retail and institutional NEPSE investors actually watch, budget for, and reinvest, the Dividend Discount Model is not an academic curiosity for this market — it is close to the most natural lens available. Where a Silicon Valley analyst might reasonably reach first for a discounted cash flow model built around free cash flow to equity or to the firm (the subject of the next chapter), a NEPSE analyst covering a bank or a seasoned hydropower stock can often go straight to a well-built DDM and get a defensible answer, precisely because the dividend is not a fragment of the return story — it is most of the story.

This is not to say DDM is the only tool an analyst needs for NEPSE — Chapter 49 and Chapter 50 will build discounted cash flow and relative valuation approaches that remain essential, especially for companies without a long dividend history, or where you suspect the dividend policy itself is not a reliable signal of underlying value creation. But for the core of the NEPSE investable universe — the banks, the insurers, the seasoned hydropower names — DDM deserves to be the first model you reach for, not the last.

Lesson 48.3 — Cash Dividends, Bonus Shares, and the Dilution Trap

Here is where Nepal's dividend culture becomes genuinely tricky, and where an analyst who is careless will produce a badly wrong valuation even while following the DDM formula correctly.

Nepali companies routinely declare dividends in two forms simultaneously: a cash dividend (an actual cash payment per share, subject to a 5% dividend tax withheld at source for individual investors, as covered in the tax chapters) and a bonus share dividend (additional shares issued to existing shareholders, in some fixed proportion to their current holding, at no cost to them). A company might, for instance, declare a "20% dividend" consisting of 15% bonus shares and 5% cash — meaning a shareholder holding 100 shares receives 15 new shares plus Rs 500 in cash (5% of a Rs 100 par value, before tax).

This is the single most important mechanical point in this lesson, and it is worth stating as plainly as possible: a bonus share does not create wealth. It is not a real cash return to the shareholder in the way a cash dividend is. When a company issues bonus shares, it moves an amount from its reserves (retained earnings) to its paid-up capital account on the balance sheet, and it issues new share certificates to match. The company's total equity value is unchanged. But now that same total value is divided among a larger number of shares — so the per-share price mechanically adjusts downward on the ex-bonus date, roughly in proportion to the bonus percentage. A shareholder holding shares worth Rs 100,000 before a 15% bonus issue still holds shares worth approximately Rs 100,000 immediately after (100,000 shares become 115, and the per-share price falls by close to the same ratio) — they simply now hold more certificates, each worth less.

So how should bonus shares be treated in a DDM framework? There are two defensible approaches, and a good NEPSE analyst should be comfortable with both.

The first, and simpler, approach is to value only the cash dividend stream in the DDM, and to treat the bonus share issuance as a non-event for valuation purposes — a cosmetic increase in share count that is not part of "the dividend" in the economic sense DDM is trying to capture. Under this approach, D1 in your Gordon Growth calculation is strictly the expected cash dividend per share next year, and your growth rate g is the expected growth of that cash dividend per share, on an already-adjusted, post-bonus share count basis. This is the cleaner, more rigorous approach, because it never confuses a reserve-to-capital bookkeeping entry with a genuine cash return, and it is the one this book recommends as the default.

The second approach — used by some practitioners as a rough shorthand — treats bonus shares as a proxy for the growth rate itself: the logic being that a company retains earnings (rather than paying them out in cash) specifically to reinvest in growing the business, and bonus shares are one visible marker of that retention. Under this reading, a company with a large bonus component alongside a smaller cash dividend is signalling that a larger share of its profit is being retained to fund future growth, which should, if that reinvestment is productive, translate into a higher future g for the cash dividend stream. This is not wrong as an intuition, but it is dangerous to apply mechanically, because a bonus share connected to genuinely value-accretive reinvestment (funding a bank's loan book growth without diluting existing shareholders through a rights issue, for example) is very different from a bonus share issued simply to make an existing shareholder base feel better about a "big dividend number," with no corresponding improvement in the underlying earnings power per share.

This connects directly to the cost-basis and WACC material from the tax chapters (Chapter 46 and Chapter 47). Recall that when you compute a shareholder's effective return, or a company's effective cost of equity capital, bonus shares change the denominator (number of shares, and therefore the shareholder's cost basis per share for capital gains tax purposes) without changing the numerator (total value or total cash invested). A shareholder who receives bonus shares does not pay tax on receipt (bonus shares are not treated as taxable income at issuance in Nepal's current framework, unlike a cash dividend, which is taxed immediately at the 5% withholding rate) — but their original cost basis is spread across more shares, which matters enormously when they eventually sell and calculate capital gains. In a very real sense, a bonus share is a *timing* device: it defers the taxable event and changes its character (from dividend income to capital gain, taxed differently) rather than creating new value. A DDM analyst who forgets this and treats a bonus-heavy "dividend yield" as equivalent to a cash-heavy one from another company is comparing two things that are not alike — one is real current income, taxed now; the other is a deferred, differently-taxed claim on a company's reserves.

Lesson 48.4 — NRB's Regulatory Leash on Bank Dividends

If DDM is unusually well suited to NEPSE banks because of their high payout culture, it is equally important to understand that Nepal Rastra Bank (NRB), the central bank and banking regulator, does not let banks and financial institutions distribute dividends however they please. NRB's dividend distribution framework directly constrains a bank's dividend *capacity*, which means it constrains what a sensible analyst should ever project as achievable D1 and g for a bank stock, no matter how strong the bank's raw profit looks on paper.

The core logic of NRB's approach is that a bank's capital is not simply its owners' money to distribute at will — it is also the cushion that protects depositors and the financial system from losses. NRB accordingly links a bank's maximum permissible dividend distribution to its capital adequacy position (how much regulatory capital it holds relative to its risk-weighted assets, discussed in earlier chapters on the banking sector) and to the quality of its loan book (typically proxied by its non-performing loan, or NPL, ratio). A bank sitting exactly at or barely above its minimum required capital adequacy ratio, or carrying a rising NPL ratio, will be permitted to distribute a smaller share of its profit — sometimes none at all in cash — because the regulator wants that capital retained inside the institution as a buffer, not paid out to shareholders.

This has a direct, mechanical consequence for a DDM built on a Nepali bank. A naive analyst forecasts D1 as "expected EPS times last year's payout ratio" and calls it done. A careful analyst instead asks: given this bank's current and projected capital adequacy ratio, and its NPL trajectory, what dividend capacity will NRB actually permit next year, and for the several years after that? A bank rapidly growing its loan book (which consumes capital, since more risk-weighted assets require more capital to support them under the same capital adequacy ratio) may be earning healthy profit yet be constrained by the regulator to plough most of that profit back into capital rather than pay it out — meaning its true sustainable D1 is lower than a simple extrapolation of past payout ratios would suggest, especially in a period of aggressive branch or loan-book expansion. Conversely, a mature bank with slower loan growth and ample capital headroom may be able to sustain a genuinely high payout ratio for an extended period, exactly the kind of company where DDM will earn its keep.

This regulatory dimension is one reason DDM for Nepali banks benefits from being paired with a capital-adequacy forecast: project the bank's risk-weighted asset growth, its expected retained-earnings contribution to capital, and back into a realistic maximum payout ratio consistent with staying safely above NRB's minimums, rather than simply trending the historical payout percentage forward. It is also a reminder that "regulated industry" cuts both ways in valuation — the same regulatory apparatus that gives Nepali banks a relatively stable, licensed, oligopolistic operating environment (a source of durable competitive advantage discussed in earlier chapters) is the same apparatus that can, in a given year, cap how much of that durable earnings power actually reaches shareholders as cash.

Lesson 48.5 — When One Growth Rate Isn't Enough: The Multi-Stage DDM

The Gordon Growth Model in Lesson 48.1 assumed a single, constant growth rate forever. That assumption is fine for a mature, stable business — a well-established bank growing roughly in line with the economy, for instance — but it breaks down badly for a company going through a distinct phase of unusually fast (or unusually slow, or negative) growth that will not persist indefinitely.

The solution is the multi-stage DDM: instead of one growth rate applied forever, you model an explicit early period with one (or several) elevated or depressed growth rate, and then assume the company settles into a stable, sustainable long-run growth rate from some terminal year onward — at which point you apply the Gordon Growth formula to that terminal, stable dividend stream, and discount that terminal value back to today alongside the explicit near-term dividends.

The two-stage version works like this:

Step 1: Forecast dividends explicitly, year by year, for the high (or transitional) growth period — say, five years. Step 2: At the end of that explicit period, calculate a terminal value using the Gordon Growth Model, applied to the first "stable" year's dividend and the long-run growth rate expected from that point forward. Step 3: Discount each of the explicit-period dividends, plus the terminal value, back to the present at the cost of equity, and sum them.

Let's build a worked example for a hypothetical hydropower company, Trishuli Ridge Hydropower Ltd., that has recently begun commercial operations and is still in the early years of its PPA tariff structure (more on this in Lesson 48.6). Assume the company is expected to grow its dividend per share at 18% annually for the first three years as its tariff escalates and it works through initial capital structure deleveraging, slowing to 10% in years four and five as growth moderates, before settling into a stable 6% long-run growth rate from year six onward, roughly matching the sector's long-run nominal growth. The cost of equity for a hydropower stock of this risk profile is estimated at 13%. Current dividend per share (D0) is Rs 8.00.

Summing the present values of years 1 through 5 (8.35 + 8.72 + 9.11 + 8.87 + 8.64 = 43.69) and adding the present value of the terminal value (130.79) gives an estimated intrinsic value per share of approximately Rs 174.48.

This worked example already hints at the danger explored fully in the next lesson: notice that the entire calculation depended on correctly identifying *when* the 18% growth phase ends and the stable 6% phase begins. Get that transition wrong — assume the 18% growth persists for eight years instead of three — and the valuation changes dramatically, because you are compounding a high growth rate over a much longer explicit period, and you are pushing back a large terminal value to a later date without properly capturing what actually stops the growth from being that fast in the interim.

Lesson 48.6 — The Tax-Holiday and Tariff-Escalation Trap

This final lesson addresses the single most dangerous way DDM goes wrong on NEPSE: applying a single, extrapolated growth rate to a company that is currently in a temporary, structurally elevated phase of its life, and mistaking that temporary phase for the company's normal, sustainable state.

Chapter 43 examined how Nepali hydropower companies benefit from income tax holidays and concessional tax rates in their early years of commercial operation (a defined number of years at 0% tax, followed by a period at a reduced rate, before reverting to the standard corporate tax rate), and how many hydropower PPAs with the Nepal Electricity Authority are structured with an escalating tariff schedule in their early years — a dry-season energy rate that rises by a fixed annual percentage for a set number of years before flattening out at a fixed rate for the remainder of the PPA term. Both features are deliberate policy design, intended to help project economics work during the early, most financially fragile years of a hydropower asset's life, when debt service is heaviest and the project is least seasoned. Both features are also, from a valuation standpoint, temporary — and temporary is exactly the word an analyst applying DDM needs to take seriously.

Here is the mechanism in full. During a hydropower company's early operating years, three tailwinds often combine to produce dividend per share growth that looks spectacular on a trailing basis: the PPA's built-in tariff escalation is raising the average realised tariff per unit of electricity sold each year; the company is paying little or no income tax under its holiday provisions, so a larger share of revenue converts to net profit; and the plant, having only recently reached full commercial operation, may still be ramping up toward its full expected generation capacity as any initial teething issues are resolved. Stack these three effects together and a hydropower company can post 15%, 20%, even 25%+ annual growth in dividend per share for several consecutive years — not because the underlying business is compounding at that rate sustainably, but because three separate, finite tailwinds are all blowing in the same direction at once.

Then, in a specific year (or a short window of years), all three tailwinds can end. The PPA tariff escalation schedule flattens at its ceiling rate. The tax holiday period expires and the statutory or concessional rate steps up, immediately compressing the net margin on the same revenue. And full-capacity generation, if not already reached, plateaus. The result is that dividend per share growth does not gently decelerate toward a stable long-run rate — it can fall off a cliff, or even see the dividend per share decline in absolute terms in the transition year, even while the underlying physical asset (the powerhouse, the water rights, the PPA itself) is completely unchanged and arguably just as valuable a long-run cash-generating asset as it was the year before.

The discipline this demands of a NEPSE analyst is straightforward to state, if not always easy to execute: before building any DDM on a hydropower company (or any other company benefiting from a temporary tax concession or a contractually scheduled, time-limited cash flow enhancement), go directly to the company's PPA and its tax status. Find the specific number of years remaining in any tariff escalation schedule. Find the specific number of years remaining in the tax holiday or concessional tax period, and the tax rate the company will revert to afterward. Build your explicit-forecast stage in the multi-stage DDM to match those actual, contractually or statutorily defined timelines — not a generic five-year template — and only transition to a stable terminal growth rate once both the tariff schedule has flattened and the full statutory tax rate has taken effect. If the tax holiday ends in year four but the tariff escalation continues through year seven, your model needs at least two distinct transitional stages before it reaches a genuinely stable terminal phase, not one.

The payoff for doing this work carefully is that DDM, applied with this level of diligence, becomes one of the most reliable tools available for valuing precisely the kind of company that dominates NEPSE's investable universe — and one of the most dangerous tools available when applied carelessly to that same company. The formula never changes. What changes, entirely, is whether the analyst using it has actually looked underneath the dividend history to understand what is driving it, and for how much longer that driver will last.