Stock Valuation: Multi-Stage Growth Models

Valuing a mature, stable company is challenging enough, but what about a fast-growing tech startup, a pharmaceutical firm on the cusp of a blockbuster drug launch, or a cyclical company emerging from a downturn? The constant-growth dividend discount model (DDM) fails here because it assumes a perpetual, unchanging growth rate. Multi-stage growth models are the essential tool for valuing companies whose growth trajectories are expected to change dramatically over time. By breaking a company's future into distinct phases—like high growth, transition, and stable maturity—these models provide a more realistic and nuanced picture of intrinsic value, which is critical for informed investment and corporate finance decisions.

The Rationale for a Multi-Stage Approach

The fundamental premise of stock valuation is that a share's worth equals the present value of all its future cash flows to the owner, typically dividends. The standard Gordon Growth Model (GGM), with its formula $P_0=D_1/(r-g)$, requires a constant growth rate $g$ that is less than the cost of equity $r$. This is unrealistic for most companies in the real world. A young company may reinvest all earnings to fuel expansion, paying no dividends initially, before eventually initiating and rapidly growing a payout. An established firm might enjoy a period of superior growth from a new product line before settling back to industry norms.

This is where multi-stage models excel. They acknowledge that competitive advantages are often temporary. High growth rates inevitably attract competition and become unsustainable as a company matures and its market saturates. By modeling separate stages, you can explicitly forecast the duration of exceptional growth, the path of its decline, and the eventual point at which the company converges to a long-run, stable growth rate comparable to the broader economy. This structured approach forces you to justify your assumptions about the length and intensity of each growth phase, leading to a more defensible valuation.

Constructing a Two-Stage Dividend Discount Model

The most common implementation is the two-stage DDM. It assumes an initial period of high, often unstable growth, followed by a perpetual stable growth phase. The model's engine is the calculation of the terminal value (TV), which represents the value of all dividends from the start of the stable phase onward, discounted back to a single point in time—the horizon date.

Here is the standard formula for a two-stage model, where high growth lasts for $n$ years:

$$P_0=\sum _{t=1}^n\frac{D_0(1+g_h{)}^t}{(1+r{)}^t}+\frac{TV}{(1+r{)}^n}$$

Where the Terminal Value $TV$ is calculated using the Gordon Growth Model at the horizon date:

$$TV=\frac{D_{n+1}}{r-g_s}=\frac{D_0(1+g_h{)}^n(1+g_s)}{r-g_s}$$

Let's define the components:

• $P_0$ = Intrinsic value per share today.
• $D_0$ = Current dividend per share.
• $g_h$ = High growth rate for the first $n$ years.
• $g_s$ = Stable growth rate in perpetuity after year $n$.
• $r$ = Required rate of return (cost of equity).
• $n$ = Length of the high-growth period.

Example Calculation: Assume a tech company pays a current dividend $D_0$ of $\mathrm{1.00.}Itisexpectedtogrowat$gh = 15\%MATHINLINE17n=5MATHINLINE18gs = 4\%$forever.Thecostofequity$r$ is 10%.

1. Calculate dividends for the high-growth phase:

• Year 1: $D_1=1.00\ast (1.15)=$1.15$
• Year 2: $D_2=1.15\ast (1.15)=$1.32$
• ... and so on until Year 5: $D_5=1.00\ast (1.15{)}^5=$2.01$

1. Calculate the Terminal Value at the end of Year 5:

• First, need $D_6$, the first dividend of the stable period: $D_6=D_5\ast (1.04)=2.01\ast 1.04=$2.09$
• $T{V}_5=D_6/(r-g_s)=2.09/(0.10-0.04)=2.09/0.06=$34.83$

1. Discount all cash flows to present value:

• $PV(D_1)=1.15/(1.10{)}^1=$1.05$
• $PV(D_2)=1.32/(1.10{)}^2=$1.09$
• $PV(D_3)=1.52/(1.10{)}^3=$1.14$
• $PV(D_4)=1.75/(1.10{)}^4=$1.19$
• $PV(D_5)=2.01/(1.10{)}^5=$1.25$
• $PV(T{V}_5)=34.83/(1.10{)}^5=$21.62$

1. Sum the present values: $P_0=1.05+1.09+1.14+1.19+1.25+21.62=$27.34$

This step-by-step process shows how the terminal value often constitutes the majority of the total present value, especially for companies with long high-growth periods.

Estimating and Justifying Growth Rate Assumptions

The output of the model is only as good as its inputs. Arbitrary growth rates yield meaningless valuations. Each growth rate must be grounded in fundamental analysis.

• High Growth Rate ($g_h$): This should be linked to the company's sustainable competitive advantage. Ask: What is the source of this superior growth (e.g., patented technology, regulatory barrier, powerful brand)? How long can it be defended? The rate should also be consistent with the company's reinvestment policy. A useful fundamental growth formula is $g=\text{Return on Equity (ROE)}\times \text{Retention Ratio (b)}$. For a high-growth company not paying dividends ($b=1$), $g_h$ should be close to its expected future ROE.
• Length of High-Growth Period ($n$): This is not a guess. It should reflect your estimate of the competitive advantage period. For a software company with network effects, it might be 10 years. For a trendy retailer, it might be 3. Industry analysis and Porter's Five Forces are key tools here.
• Stable Growth Rate ($g_s$): This is the most critical and often most abused assumption. It must be conservative and perpetual. A rational rule is that $g_s$ cannot exceed the long-term growth rate of the economy in which the company operates (often 2-4% for developed markets). Furthermore, when a company reaches stable growth, its characteristics should reflect that maturity: its risk (beta) should converge towards 1, and its return on equity should trend toward the industry/cost of equity. A common error is assuming a high $g_s$ while also assuming high ROE and low risk—an impossible trifecta for a mature company.

Common Pitfalls

1. The Unjustifiable High Stable Growth Rate: Assuming a stable growth rate of 6% or 7% in perpetuity is a major red flag. It implies the company will forever outgrow the economy, eventually becoming larger than the GDP itself. This inflates the terminal value and the entire valuation. Correction: Cap $g_s$ at or below the nominal GDP growth rate. For a US-focused company, a range of 2-4% is prudent.

1. Inconsistent Reinvestment Assumptions: A model might project a drop from high growth ($g_h=12\%$) to stable growth ($g_s=4\%$) without adjusting the fundamental drivers. If high growth was fueled by a 40% reinvestment rate and a 30% ROE ($g=0.40\ast 0.30=12\%$), the stable phase must be consistent. If ROE falls to 10% in maturity, to achieve a 4% growth, the reinvestment rate must be 40% ($0.10\ast 0.40=4\%$). If the model assumes dividend payout increases instead, this must be explicitly modeled. Correction: Use the fundamental growth equation as a sanity check to ensure your phase transitions are internally consistent.

1. Ignoring the Cost of Equity Transition: Risk changes over a company's lifecycle. A volatile startup has a high beta and cost of equity. A stable utility has a low beta. Using a single discount rate $r$ across both phases misprices risk. Correction: For a more advanced valuation, use a higher discount rate for the high-growth phase (reflecting higher risk) and a lower rate for the stable phase. This often involves adjusting the beta in the CAPM formula over time.

1. Over-reliance on the Terminal Value: In our example, the terminal value made up nearly 80% of the total value. When this proportion is excessively high, the valuation is extremely sensitive to your $g_s$ and $r$ assumptions, making it speculative. Correction: If the terminal value exceeds 70-75% of total value, scrutinize your high-growth assumptions. Perhaps the high-growth period is too short, or the high-growth rate is too low. Extending the forecast period or adding a third, transitional stage can help.

Summary

• Multi-stage DDMs are necessary for valuing companies with non-constant growth, allowing you to model distinct high-growth, transitional, and stable-maturity phases.
• The terminal value, calculated at the horizon date using the Gordon Growth Model, represents the present value at that time of all subsequent dividends in the perpetual stable phase and is a dominant component of total value.
• All growth rate assumptions must be justified fundamentally. The high-growth rate should be tied to competitive advantage and ROE, while the stable growth rate must be conservative, perpetual, and not exceed the economy's long-term growth rate.
• Avoid common errors like assuming unrealistic perpetual stable growth, using inconsistent reinvestment rates across stages, or applying a single discount rate to dissimilar risk profiles. The model's output demands rigorous sensitivity analysis on these key drivers.