The Dividend Discount Model in Excel
The dividend discount model values a share as the present value of every dividend it will ever pay. The simplest form, the Gordon growth model, assumes dividends grow at a constant rate forever and discounts them at the cost of equity. It works best for mature, stable payers like utilities and large banks, where dividends are predictable and the growth rate is sensible.
The formula and its parts
The Gordon growth model is Value = D1 / (r - g), where D1 is next year's dividend, r is the cost of equity, and g is the constant growth rate. D1 equals this year's dividend grown one period: D1 = D0 * (1 + g).
The discount rate r is the cost of equity, because dividends are cash flows to equity holders. Estimate it with CAPM: r = Rf + beta*(Rm-Rf). The model only holds when r > g; if growth meets or exceeds the discount rate, the denominator collapses and the value is meaningless.
When growth is not constant, use a two-stage model: project explicit dividends through a high-growth phase, then apply Gordon growth to a terminal year and discount everything back.
Worked example
Last year's dividend D0 was 2.00, the cost of equity r is 9 percent, and the long-run growth rate g is 4 percent. First grow the dividend one year, then apply the formula.
- Next year's dividend:
=2.00*(1+0.04)returns 2.08. - Spread:
=0.09-0.04returns 0.05. - Value per share:
=2.08/0.05returns 41.60. - Sensitivity: cut
gto 3 percent and=2.06/0.06returns 34.33, showing how sensitive the model is to the growth input.
| Input | Value |
|---|---|
| This year's dividend D0 | 2.00 |
| Cost of equity r | 9.0% |
| Growth g | 4.0% |
| Next dividend D1 | 2.08 |
| Value per share | 41.60 |
=2.00*(1+0.04)/(0.09-0.04) returns 41.60.
Laying it out in a model
Keep the model auditable by separating the three drivers into their own input cells and computing D1 and value in their own steps.
- Put
D0,r, andgin three labeled input cells, never inside the value formula. - Compute
D1in its own cell so the grown dividend is visible. - Reference a CAPM block for
rrather than typing a rate, so the discount rate is traceable. - Add a small data table flexing
randgto show the valuation range.
Pitfalls
The most common error is setting g too close to r. A growth rate of 8 percent against a 9 percent cost of equity gives a 1 percent denominator and an absurd value; the long-run growth rate should not exceed the economy's nominal growth.
Applying the model to a company that pays no dividend, or whose payout is erratic, produces a number with no meaning. Use the two-stage version for firms still ramping their payout, and reserve the single-stage model for mature, steady payers. Confusing D0 with D1 and forgetting to grow the dividend one period understates value by exactly one growth factor.
Formula Trace
Formula Trace follows the value cell back through D1, r, and g so you can confirm the discount rate really comes from your CAPM block.
Get ModelMint See how it worksFAQ
What discount rate does the dividend discount model use?
The cost of equity, because dividends are cash flows that belong to shareholders. Estimate it with CAPM as Rf + beta*(Rm-Rf). Do not use WACC; that blended rate is for unlevered cash flows in a DCF, not for equity cash flows.
When does the Gordon growth model break down?
When growth meets or exceeds the discount rate. If g is greater than or equal to r, the denominator r - g is zero or negative and the value is infinite or nonsensical. Keep g below r and below long-run nominal economic growth.
What is the two-stage dividend discount model?
It projects explicit dividends through a high-growth phase, then values the remaining stream with Gordon growth at a stable rate. You discount the explicit dividends and the terminal value back at the cost of equity, which suits firms not yet in steady state.