KARACHI: Strip away the charts, the ratios and the market noise, and a share is a claim on a stream of cash that its company will one day hand to its owners. For a dividend-paying company, that stream is visible: a dividend this year, a bigger one next year, and so on into a future nobody can fully see. The dividend discount model (DDM) takes that simple idea with complete seriousness. It says the value of a share today is the present value of all the dividends it is expected to pay, forever. Elegant, austere and brutally sensitive to its inputs, the DDM is less a pricing machine than a discipline: it forces whoever uses it to state, in numbers, what they believe about a company's payouts, its growth and the return they demand for waiting.
The core idea: why future dividends are worth less today
A rupee next year is worth less than a rupee today, because today's rupee can be invested — in a bank deposit, a government security, or another share — and because waiting involves risk. Discounting converts future payments into today's money by shrinking each one at a rate that reflects both time and risk: divide next year's dividend by (1 + r), the dividend in two years by (1 + r)², and so on, where r is the investor's required rate of return, often called the cost of equity. Add up the discounted values of every expected dividend and you have the model's estimate of intrinsic value. That is all the DDM is. Everything else — the famous formulas, the arguments in analyst reports — is technique for estimating an infinite stream of dividends without writing down an infinite number of terms. The compounding arithmetic behind the shrinking is the mirror image of the growth arithmetic in our guide to the power of compounding: discounting is compounding run backwards.
The Gordon growth model: one formula, three beliefs
The workhorse version, associated with Myron Gordon, assumes dividends grow at a constant rate g forever. The infinite sum then collapses into a single line: value = next year's dividend ÷ (r − g). Suppose a fictional company, Chenab Foods Limited — invented for illustration — is expected to pay a dividend of Rs 6 per share next year, its dividends are expected to grow at 5 per cent a year indefinitely, and investors require a 13 per cent return for holding a share of its risk. The model values the share at 6 ÷ (0.13 − 0.05) = 6 ÷ 0.08 = Rs 75. Stare at that arithmetic and you see the model's character. The denominator is a sliver — the gap between required return and growth — so small changes in either belief swing the answer violently. Raise expected growth to 6 per cent and the value jumps to about Rs 86; lower the required return to 12 per cent and it leaps to Rs 100; let growth approach the required return and the formula explodes toward infinity, which is the model's way of telling you the assumptions have become impossible, since no company outgrows the economy forever.
Where the inputs come from — and why they fight back
Each input is an estimate wearing a suit. Next year's dividend is usually anchored to the latest declared dividend, adjusted for any announced change in policy — the mechanics of declaration, book closure and payment are covered in our guide to how PSX dividends work. The required return r is commonly estimated with the capital asset pricing model — a risk-free rate, usually from government securities, plus a premium for equity risk scaled by the share's volatility — the subject of our explainer on the capital asset pricing model. Growth g is the troublemaker. One honourable way to estimate it is the sustainable growth identity: growth equals return on equity multiplied by the retention ratio — the share of profit the company keeps rather than pays out. That identity quietly ties the DDM to the rest of financial analysis: a company paying out 80 per cent of earnings cannot also grow at 15 per cent unless its ROE is extraordinary, and our guide to DuPont analysis shows where the return figures come from. In practice, analysts working in high-inflation economies like Pakistan's must also keep nominal and real terms consistent — nominal dividends, nominal growth and a nominal required return, or all three in real terms, but never a mixture.
Multi-stage models: growth in chapters
Few real companies grow at one rate forever, so practitioners let the story have chapters. A two-stage model projects dividends explicitly for a high-growth phase — say five or ten years, year by year, from an earnings and payout forecast — then applies the Gordon formula at the end of that phase to capture all subsequent years as a single terminal value, and discounts everything back to today. A three-stage or "H" model smooths the transition between fast growth and maturity. These variants are more realistic and more dangerous in equal measure: realistic because they can honour a visible expansion — a new plant, a market being entered; dangerous because the terminal value typically dominates the answer, often contributing three-quarters or more of the total, so a model that looks richly detailed is still mostly one assumption about the distant future wearing a spreadsheet's clothing. The honest response is sensitivity analysis: value the share across a grid of plausible r and g values, and treat the output as a range with a story attached, not a point estimate with false precision.
When the model fits — and when it cannot be used at all
The DDM fits companies whose dividends are the genuine, principal channel through which shareholders receive the business's cash: mature, profitable, steadily growing firms with a long, stable payout record and a policy minority shareholders can rely on — in Pakistani conditions, think of the character (not the current data) of established banks, fertiliser producers and consumer companies that have paid through cycles. It fits poorly, or not at all, where dividends are absent, erratic or set by a controlling shareholder's cash needs rather than by policy; where buybacks have replaced dividends as the payout channel; and where the company's value lies in assets or growth options rather than in distributions — an explorer, a young technology firm, a turnaround. A zero-dividend company is not worthless under the model's logic — the model simply insists its value rests on dividends expected eventually, which pushes the forecaster so far into speculation that other tools, from free cash flow models to asset values, usually serve better. Note also what the model values from the minority shareholder's seat: the dividends you will receive, which is exactly right for a portfolio investor, and different from the cash a controlling buyer could extract — one reason takeover offers can exceed DDM values.
Using the DDM backwards: what is the market assuming?
Perhaps the model's most practical use inverts it. Take the current price as given and solve for the growth rate the market must be assuming, holding your required return fixed. If Chenab Foods traded at Rs 120 rather than Rs 75, the implied perpetual growth, at a 13 per cent required return and a Rs 6 next dividend, would be 0.13 − (6 ÷ 120) = 8 per cent — a demanding but not absurd expectation for a strong consumer franchise, and a far more useful sentence than "the stock looks expensive." Repeat the exercise across a sector and you have a map of where optimism is concentrated. This reverse-engineering habit — asking what must be true for today's price to be right — is the mature way to use every valuation model, and it pairs naturally with the margin of safety idea: you buy not when a model pronounces a share cheap to the second decimal, but when the price requires so little of the future that even your mistakes leave you whole. Pakistani investors in income shares should also remember the tax and purification dimensions that sit alongside any valuation when they build a payout-focused portfolio.
Interest rates, inflation and the model's Pakistani weather
The DDM's denominator — r minus g — makes it a hostage to macroeconomic weather, and Pakistan's weather is dramatic. When the State Bank's policy rate swings through the wide ranges Pakistani investors have lived with, the risk-free leg of the required return swings with it, and dividend-stock valuations swing mechanically: high-rate eras compress the value of distant dividends and punish long-duration payers, while falling-rate eras do the reverse with flattering speed. This is not a defect of the model so much as a true statement about the assets: a dividend stream genuinely is worth less when a depositor can earn high-teens returns at the bank next door. Inflation cuts the other way through the g term, since companies with pricing power grow their nominal dividends with the price level. The careful user's habits follow directly: build the model in consistent nominal terms, use a required return that reflects today's opportunity cost rather than a fond memory of cheaper money, and never compare a DDM value computed in one rate regime with a market price set in another. Pakistani income investors who watched solid payers de-rate as rates rose — and re-rate as rates fell, without the companies changing at all — have seen the denominator at work in their own portfolios.
Why the model survives its critics
Academics have been announcing the DDM's inadequacy for decades — it ignores buybacks, flounders on non-payers, and demands clairvoyance about growth — yet it refuses to die, because its failures are the price of its virtues. It is the only mainstream model that values exactly what a minority shareholder actually receives, in cash, in hand. Its inputs are few enough to argue about honestly: a debate over a DDM value is a debate about payout, growth and required return, conducted in the open, rather than a fog of multiples chosen for convenience. And its very fragility is informative — a valuation that cannot survive a one-point change in assumptions is a valuation the market should not lean on, whatever the model says. The sensible modern synthesis treats the DDM as one instrument on the panel: decisive for mature, committed payers; a cross-check elsewhere; and everywhere a device for converting a share price into the expectations that price implies. A model that tells you what you must believe is never useless, even — especially — when what you must believe turns out to be hard to believe.
There is one more discipline the model quietly teaches, about the relationship between dividends and price that beginners often reverse. In the DDM's world, the dividend is not a bonus sprinkled on top of a share's "real" value; for the long-term holder, the dividends and the eventual resale — itself just the next buyer's claim on later dividends — are all the value there is. A company that will never pay, never be bought and never liquidate is, under this lens, a collectible rather than an investment, whatever its story. Holding that thought while reading promotional material for non-paying companies is a cheap and effective scepticism exercise, and it costs nothing but the willingness to ask of every share the model's one blunt question: where, eventually, is my cash? Ask it of every holding, every year, without exception.
This article is educational content explaining a classic valuation model and its limitations; it is not financial advice, and the fictional figures used are illustrations, not data about any real company.
Key takeaway: the dividend discount model values a share as the present value of all its future dividends — in its Gordon form, next year's dividend divided by the gap between your required return and perpetual growth — a formula whose simplicity exposes exactly which beliefs about payout, growth and risk a price depends on.
