Assumptions Behind Constant-Growth Valuation

Why the Assumptions Matter

The Gordon growth model is useful because it compresses an entire stream of expected dividends into a simple expression. That simplicity comes from strong assumptions: dividends grow at a constant rate indefinitely, investors require a constant return, and the growth rate is lower than that required return. These conditions are not just technical details. If a company is still changing rapidly, its payout policy is unstable, or its forecast growth cannot persist, the model’s central description of future cash flows may be a poor fit.

This tutorial focuses on deciding whether those assumptions are plausible. The question is not whether a constant-growth model can produce a number, but whether its inputs describe the business and its expected dividends with reasonable discipline. This assessment is aligned with the SOA VEE Accounting and Finance syllabus.

Definition: In the constant-growth dividend model, $D_0$ is the most recently paid annual dividend per share, $D_t$ is the dividend expected at the end of year $t$, $g$ is the constant annual dividend growth rate, and $r$ is the constant annual required return on equity. The model assumes $D_t=D_0(1+g)^t$ for every future year $t$, and, when $r>g$, its value at time 0 is $P_0=D_1/(r-g)$, where $D_1=D_0(1+g)$.

The standard interpretation assumes annual dividends paid at each year-end, a current price measured at time 0, and rates expressed on the same annual basis. The required return is the return investors require for bearing the stock’s risk; the growth rate is the expected long-run rate of growth in dividends per share. Neither rate should be treated as a short-term forecast simply because recent observations are available.

What Constant Growth Requires

The model describes a company whose expected dividends follow one stable pattern indefinitely. It therefore requires more than a history of dividend increases. The firm should be sufficiently mature that its business and distribution policy are not expected to undergo a major transition. Its earnings and cash-generating capacity should be capable of supporting dividends that rise at a persistent rate. The payout decision should not be expected to shift sharply between retaining cash and distributing it.

These are economic plausibility checks, not guarantees. A mature company can still face competition, regulation, shocks, or strategic change. A history of steady dividends can support a forecast, but it cannot prove that the future will repeat the past. Likewise, a young or rapidly expanding firm may have attractive growth prospects but still be a poor candidate for a single growth rate that applies forever.

Assumption	Evidence that may support it	Warning sign
Mature, relatively stable business	Established operations and no evident near-term structural transition	Rapid expansion, major restructuring, or a business model still changing
Stable payout policy	Dividends that move in a reasonably consistent relationship to earnings and capacity to pay	Large payout swings, irregular special dividends, or repeated cuts
Persistent long-run growth	Growth supported by durable business conditions rather than one unusual year	Recent unusually high growth treated as permanent without a supporting reason
Required return greater than growth	A required return that exceeds the perpetual growth assumption	Growth equal to or above the required return

Why Growth Must Be Below the Required Return

The formula is the present value of a perpetuity whose payments grow at a constant rate. At the end of year $t$, the dividend is $D_1(1+g)^{t-1}$. Discounting it to time 0 at the required return gives the present value of that dividend. Summing all such present values gives a geometric series.

P_0=\sum_{t=1}^{\infty}\frac{D_1(1+g)^{t-1}}{(1+r)^t} =\frac{D_1}{1+r}\sum_{t=1}^{\infty}\left(\frac{1+g}{1+r}\right)^{t-1}

A geometric series with first term $a$ and common ratio $q$ has a finite sum $a/(1-q)$ only when $|q|<1$. Here, with ordinary nonnegative annual growth and return rates, the ratio is $q=(1+g)/(1+r)$. It is below 1 exactly when $g<r$. Substituting the first term and simplifying gives the Gordon formula:

P_0=\frac{D_1}{1+r}\cdot\frac{1}{1-\frac{1+g}{1+r}} =\frac{D_1}{r-g}

If $g=r$, the discounted dividend terms do not shrink to zero; if $g>r$, they grow rather than shrink. In either case the infinite sum does not converge to a finite value under this model. This is a mathematical failure, not a signal that the stock has an extraordinarily high finite value. Even when $g<r$, a very small difference between the two rates makes the model highly dependent on the assumptions used.

Condition: For a finite value in the perpetual constant-growth model, the long-run dividend growth rate must be strictly below the required return, with both rates measured on the same basis. A forecast that violates this condition cannot be used as a valid perpetual-growth assumption in the formula.

Assessing Business Maturity and Payout Stability

A mature business is not simply one with a long operating history. For this purpose, maturity means that the company’s future dividend pattern can reasonably be summarized by a stable long-run rate rather than by a major period of transition. A firm may be old but facing a disruptive change; another firm may be relatively young but already have stable operations and distributions. The assessment should connect the business outlook to dividends, not rely on a label.

Payout stability is also different from an unchanging dividend. A company can raise its dividend each year and still have an unstable payout policy if its dividend is growing much faster or slower than the earnings that support it. Conversely, a temporarily unchanged dividend does not by itself establish that long-run growth is zero. Examine several years of dividends alongside earnings or other relevant capacity-to-pay measures. Look for whether dividend changes appear deliberate and supportable, and whether unusual payments or cuts make the record misleading.

Dividend growth per share should also be distinguished from growth in total company earnings or total dividends. Share issuance, repurchases, and changes in payout policy can make those measures move differently. The model concerns expected dividends per share to the stockholder. A business’s growth story is relevant only insofar as it supports a plausible long-run pattern in those per-share dividends.

Worked Example: Is a Mature Dividend Record Consistent?

A hypothetical utility has paid annual dividends per share of $0.80, $0.88, $0.97, $1.05, and $1.13 over the last five years. Its earnings per share over those years were $2.00, $2.20, $2.45, $2.55, and $2.75. Calculate the four annual dividend growth rates and payout ratios, then assess whether the record supports—but proves—a constant-growth assumption.

Method. Compute each dividend growth rate as the current year’s dividend divided by the prior year’s dividend, less 1. Compute each payout ratio as dividend per share divided by earnings per share. These are historical diagnostics; they do not establish future growth.

The annual dividend growth rates are:

\frac{0.88}{0.80}-1=0.1000=10.00\%,\quad \frac{0.97}{0.88}-1=0.10227=10.23\%

\frac{1.05}{0.97}-1=0.08247=8.25\%,\quad \frac{1.13}{1.05}-1=0.07619=7.62\%

The four growth rates are approximately 10.00%, 10.23%, 8.25%, and 7.62%. The compound annual growth rate from $0.80 to $1.13 over four intervals is $(1.13/0.80)^{1/4}-1$, approximately 9.02% when rounded to two decimal places. The decline in annual growth rates indicates that the single historical average should not be mistaken for a guaranteed perpetual rate.

The payout ratios are $0.80/2.00=40.00\%$, $0.88/2.20=40.00\%$, $0.97/2.45=39.59\%$, $1.05/2.55=41.18\%$, and $1.13/2.75=41.09\%$. Rounded to two decimal places, the ratios are 40.00%, 40.00%, 39.59%, 41.18%, and 41.09%.

Assessment. The hypothetical record has a fairly stable payout ratio and a dividend that has increased in each observed year. Those facts are consistent with a mature, steady-payout interpretation. However, dividend growth has eased from about 10% to about 7.6%, and five observations cannot demonstrate that one rate will persist indefinitely. The evidence supports further consideration of constant growth, but does not by itself justify carrying the historical average forward forever.

Check. The payout ratios use dividend divided by earnings, not earnings divided by dividend; reversing the ratio would not measure the share of earnings distributed.

Growth Persistence and the Required Return

A forecast must distinguish a temporary growth phase from a perpetual one. A few years of unusually strong dividend increases may follow recovery, expansion, or a change in payout policy. Even if that near-term pattern is credible, the same rate may not be credible indefinitely. A constant-growth model uses one rate for every future year, so its growth input must represent a defensible long-run expectation rather than simply the most recent annual increase.

The required return and growth rate also have different roles. Growth describes the expected path of dividends; the required return reflects the return investors demand for the risk of those dividends. The condition $r>g$ is necessary for convergence, but it is not sufficient to establish that either input is well estimated. A rate pair can satisfy the inequality and still be economically implausible if, for example, perpetual growth is based on a short-lived surge.

Use consistent units and timing. An annual dividend growth rate should not be compared with a monthly required return, or a nominal rate with a real rate, without converting to a consistent basis. A dividend forecast stated as a dollar amount also needs to refer to the correct year: the model’s numerator is the next expected dividend $D_1$, not the already-paid $D_0$. These distinctions matter before any numerical valuation is attempted.

Worked Example: Separating Recent Growth From Perpetual Growth

A hypothetical medical-device company increased its annual dividend from $1.20 to $1.50 over three years. Management expects a required return of 10% for the stock, while an analyst proposes using the observed three-year growth rate as the perpetual dividend growth rate. Determine the observed rate and evaluate whether directly extending it is a sound assumption.

Method. Use the compound annual growth rate over three annual intervals, not the total percentage increase as though it occurred in one year:

g_{\text{observed}}=\left(\frac{1.50}{1.20}\right)^{1/3}-1 =(1.25)^{1/3}-1\approx 0.0772=7.72\%

The calculation is consistent with the data: compounding a 7.72% annual rate for three years gives approximately $1.20(1.0772)^3=\$1.50$, subject to rounding. If 7.72% were assumed to persist forever, it would be below the stated 10% required return, so the mathematical convergence condition would be satisfied. The gap would be $10.00\%-7.72\%=2.28$ percentage points.

That arithmetic does not establish that 7.72% is a plausible perpetual rate. The observed interval is only three years, and the scenario supplies no evidence that the company’s growth drivers or payout policy will sustain that pace indefinitely. The appropriate conclusion is that the historical rate is a possible reference point, not a demonstrated long-run assumption. A constant-growth model would require a defensible long-run rate and a consistent required return.

Common mistake avoided. Treating the total increase of $1.50/1.20-1=25\%$ as annual growth overstates the rate. It is the cumulative change over three years; the annualized rate is about 7.72%.

Common Exam Trap: A Valid Inequality Is Not a Valid Forecast

A frequent error is to check only whether $g<r$, then conclude that the model’s assumptions are appropriate. The inequality is a mathematical requirement for a finite sum, not proof that the inputs describe the company. Another error is to carry a recent growth rate into perpetuity without considering maturity, payout support, or the possibility of a business transition.

For example, suppose a proposed perpetual growth rate is 11% and the required return is 9%. The incorrect reasoning is: “The firm has grown quickly, so use 11%.” The model’s growth-to-return ratio is $(1.11)/(1.09)$, which exceeds 1; the discounted dividend series therefore does not converge. The correction is not to insert the rates anyway, nor to interpret the result as a very large finite value. Reconsider the long-run growth assumption or use a valuation framework that can represent a changing growth pattern. The first issue is whether the assumptions fit, not how to force the formula to produce an answer.

If a proposed rate is 8% and the required return is 9%, the inequality holds, but a separate economic assessment is still needed. Ask whether dividends per share can reasonably grow at 8% indefinitely, whether the payout policy is stable, and whether the required return corresponds to the stock’s risk. A small spread between rates should prompt particular care because the formula’s denominator $r-g$ is then small, making the result especially dependent on the inputs.

Examine the business.
Decide whether operations appear mature enough for one long-run dividend growth pattern to be credible, and identify known transitions that could interrupt it.

Examine the payout record.
Compare dividends with earnings or relevant capacity to pay; distinguish a stable policy from a merely rising dividend series.

Separate recent from perpetual growth.
Annualize historical changes correctly, then ask whether the observed rate has support as an indefinite long-run expectation.

Check the model’s mathematical conditions.
Use annual rates on a consistent basis and confirm that the perpetual growth rate is strictly below the required return.

Putting the Assessment Together

A constant-growth assumption is most defensible when a company’s operating outlook is relatively mature, dividends are supported by earnings and a stable payout policy, and the chosen long-run growth rate is more than a continuation of a short exceptional period. The required return must be appropriate to the stock’s risk and greater than that growth rate. These checks should be stated explicitly: “growth is below return” alone is not a full justification.

The model remains an idealization. Its value is not that every company literally grows at one exact rate forever, but that a stable long-run pattern can sometimes summarize expected dividends usefully. When the available evidence instead points to a meaningful change in growth or payout, forcing one perpetual rate hides that feature rather than resolving it.

Key takeaway: Treat perpetual constant growth as a long-run economic assumption, not a mechanical extrapolation. Assess maturity, payout stability, and growth persistence, then verify that the annual growth rate is strictly below the annual required return.

Check Your Understanding

For each situation, identify the relevant assumption and give the requested calculation or conclusion.

A hypothetical retailer’s dividend rises from $0.90 to $1.08 over two years. Calculate the compound annual growth rate, rounded to two decimal places. (Answer: 9.54% per year.)
A company has steady dividends, but its payout ratio rises from 35% to 70% over several years. Name the assumption that deserves scrutiny and explain why. (Answer: Payout stability; the sharply higher share of earnings distributed may not be sustainable.)
An analyst proposes perpetual dividend growth of 6% and a required return of 8%, both annual. Does the mathematical convergence condition hold, and is that alone enough to validate the model? (Answer: Yes, $g<r$; no, economic plausibility must also be assessed.)
A stock’s recent dividend growth is 12%, while the required return is 10%. What is the model-condition problem if 12% is used perpetually? (Answer: $g>r$, so the infinite discounted-dividend series does not converge.)
A hypothetical firm has a stable business and payout policy, but its 15% dividend growth occurred during a one-time recovery. What is the central forecasting concern? (Answer: A recovery-period rate may not persist indefinitely and should not automatically be treated as perpetual growth.)

What You'll Learn