Dividend Discount Model (DDM)

The dividend discount model values shares from expected future dividends, with results sensitive to dividend timing, growth, and the required equity return.

The dividend discount model (DDM) estimates a share’s value today by discounting its expected future dividends to the present. It is an equity valuation model: the cash flows belong to shareholders, and the discount rate reflects their required return.

DDM helps investors and analysts test whether a dividend forecast could support a share price. Its result is an estimate conditional on the forecast, not a promised selling price or a guarantee that dividends will be paid.

Key Takeaways

  • The Gordon growth model uses the next dividend, not the dividend just paid.
  • Constant perpetual growth must be below the required return for the standard positive-dividend formula to produce a finite value.
  • A multistage model values near-term dividends separately from the later dividend stream.
  • Dividend policy, funding needs, and long-run growth assumptions can matter more than calculation precision.

General Dividend Discount Formula

With annual end-of-year dividends and a constant annual required return:

$$ V_0=\sum_{t=1}^{\infty}\frac{E[D_t]}{(1+r)^t} $$

V0 is today’s estimated value per share; E[Dt] is the expected dividend per share at the end of year t; and r is the required equity return. The expectation matters: future common dividends are not contractual interest payments.

Use a cost of equity consistent with the dividend risk and currency, not a blended debt-and-equity discount rate applied indiscriminately. Annual nominal dividends require a compatible annual nominal rate. Quarterly or irregular payment dates need a corresponding discounting schedule.

CFA Institute’s discounted dividend valuation overview explains DDM and how it differs from free-cash-flow and residual-income approaches.

Worked Example: Constant Dividend Growth

The Gordon growth model, also called the constant-growth DDM, assumes dividend per share grows at a constant rate indefinitely:

$$ D_1=D_0(1+g),\qquad V_0=\frac{D_1}{r-g},\qquad r>g $$

Here, D0 is the annual dividend just paid, D1 is next year’s expected dividend, and g is the perpetual annual dividend-growth assumption.

Suppose a fictional company has these inputs. Dollar amounts are U.S. dollars per share; future payments occur at year-end, and investor-specific taxes and fees are excluded.

InputAssumption
Dividend just paid, D0$2.00
Perpetual dividend growth, g4%
Required equity return, r9%

Next year’s dividend is $2.00 multiplied by 1.04, or $2.08. The model value is:

$$ V_0=\frac{2.08}{0.09-0.04}=\$41.60 $$

Putting the just-paid $2.00 in the numerator instead would give $40.00. That mixes the dividend dates: the formula values payments still to come.

The $41.60 result depends on both the 4% growth assumption and the 9% required return. It does not establish that a market price below $41.60 must rise. The analyst’s assumptions may be wrong.

Worked Example: Two Growth Stages

Now change the same company’s forecast: start from the $2.00 just paid, assume 10% dividend growth for two years, then 4% growth from year 3 onward. Keep the required return at 9% throughout.

The expected dividends are $2.20 in year 1, $2.42 in year 2, and $2.5168 in year 3.

At the end of year 2, immediately after the year-2 dividend, the value of all subsequent dividends is:

$$ V_2=\frac{D_3}{r-g} =\frac{2.5168}{0.09-0.04} =\$50.336 $$

This is a year-2 value, not today’s value. Discount it for two years, alongside the first two dividends:

ComponentTimingFuture amountPresent value at 9%
First dividendEnd of year 1$2.20$2.0183
Second dividendEnd of year 2$2.42$2.0369
Value of dividends from year 3 onwardEnd of year 2$50.336$42.3668
Total estimated value todayTime 0Not applicable$46.4220
$$ V_0=\frac{2.20}{1.09} +\frac{2.42+50.336}{1.09^2} \approx \$46.42 $$

Calculations use unrounded inputs; the table rounds present values to four decimals. The first two dividends add about $4.06 of present value. Approximately 91.3% of the estimate comes from the discounted later dividend stream, making the mature-growth assumption especially important in this example.

Do not add the year-3 dividend separately to $50.336: that terminal value already includes it. Do not discount the terminal value for three years merely because its numerator is the year-3 dividend.

Damodaran’s NYU dividend discount model notes illustrate the separation between explicit dividends, terminal value, and discounting back to the valuation date.

Sensitivity to Growth and Required Return

Return to the constant-growth case with a $2.00 just-paid dividend. Each cell recalculates next year’s dividend at the column’s growth rate and then divides by required return minus growth.

Required return3% perpetual growth4% perpetual growth5% perpetual growth
8%$41.20$52.00$70.00
9%$34.33$41.60$52.50
10%$29.43$34.67$42.00

At 4% growth, raising the required return from 9% to 10% lowers estimated value from $41.60 to about $34.67. At a 9% required return, raising growth from 4% to 5% increases estimated value to $52.50.

These are sensitivity cases, not equally probable outcomes or a recommended valuation range. A small positive gap between required return and growth makes the estimate highly sensitive. Growth equal to or above the required return invalidates this perpetual-growth calculation; a negative denominator is not evidence of a negative stock price.

Zero Growth and Gradually Changing Growth

A zero-growth DDM assumes the same dividend continues indefinitely. Its value is the constant dividend divided by the required return. With a $2 annual dividend and 9% required return, that is approximately $22.22. This is a model assumption, not a promise that a common-stock dividend is fixed.

The H-model approximates a gradual, linear transition from an initial dividend-growth rate to a stable long-run rate:

$$ V_0\approx \frac{D_0(1+g_L)+D_0H(g_S-g_L)} {r-g_L} $$

gS is initial growth, gL is stable growth, and H is half the total transition length in years. A six-year transition means H equals 3, not 6. Because H already represents half the transition, the growth adjustment uses H, not H divided by 2.

The approximation assumes a constant required return above stable growth. It is not an exact year-by-year forecast. See the H-model discussion in Damodaran’s valuation methods survey.

When DDM Is Useful, and When It Is Weak

DDM is most practical when a company’s dividend policy has a reasonably understandable relationship to its earning and distribution capacity. A forecast should account for reinvestment, financing obligations, and payout policy rather than extend the latest dividend increase forever.

A company paying no dividend today is not automatically impossible to value with DDM. A multistage forecast can include initial zero payments and later distributions. However, the result then depends heavily on when payments begin and how large they become. Damodaran addresses this distinction in his valuation questions on non-dividend-paying stocks.

Other valuation approaches may be more useful when dividends poorly represent cash available to equity investors. Buybacks also require care: they can change future share counts and per-share dividends, but a continuing shareholder does not personally receive the cash paid to selling shareholders. Adding buyback cash to dividend per share without adjusting the model can double count benefits.

This article is educational, not personalized investment advice. Common dividends can be cut or suspended, and even a carefully constructed valuation can differ substantially from future market prices.

  • Dividend per Share (DPS): The distribution attached to each share, used to build the dividend forecast.
  • Dividend Growth Rate: The change in dividend per share; historical growth is not a perpetual-growth forecast.
  • Dividend Payout Ratio: Relates dividends to earnings and helps assess whether payout assumptions are plausible.
  • Terminal Value: Represents cash flows after the explicit forecast and must be discounted from the correct date.
  • Price-Dividend Ratio: Compares a market price with a specified annual dividend rather than forecasting an entire dividend stream.

Knowledge Check

Loading quiz…

FAQs

Can the dividend-growth model estimate the cost of equity?

Yes, conditionally. If the observed market price is assumed to equal model value, rearranging the Gordon formula gives required return equal to next year’s dividend divided by current price, plus perpetual growth. At $41.60, a $2.08 next-year dividend and 4% growth imply 5% plus 4%, or 9%. This is a model-implied rate, not a guaranteed realized return.

Why discount dividends beyond the period I expect to own the stock?

In a dividend model, the expected selling price reflects the dividends later owners are expected to receive. A finite holding-period calculation includes dividends during ownership and an estimated sale value at the end. It should not add both that sale value and the full stream of dividends it already represents.
Browse Valuation and Analysis