Key Concepts of Dividend Discount Models

Why This Matters

Dividend Discount Models (DDMs) form the backbone of equity valuation in financial mathematics. These models test your understanding of time value of money, growth rate assumptions, and present value calculations, all of which connect to broader portfolio theory and investment analysis. The central idea is straightforward: a stock's intrinsic value equals the present value of all its future dividends.

Different companies have different growth trajectories, so different models apply. Don't just memorize formulas. Know when to apply each model and why certain growth assumptions matter. You need to match real-world company characteristics (stable blue chips, high-growth startups, transitioning firms) to the appropriate valuation framework.

---

Constant Growth Models

These models assume dividends grow at a single, unchanging rate forever. The mathematical simplification works because a perpetuity with constant growth converges to a clean formula, but only when the growth rate is less than the required return.

Gordon Growth Model

• Core formula: $P_0 = \frac{D_1}{r - g}$, where $P_0$ is the current stock price, $D_1$ is next year's expected dividend, $r$ is the required rate of return, and $g$ is the constant dividend growth rate
• Best applied to mature, stable companies like utilities or consumer staples with predictable dividend policies and minimal earnings volatility
• Critical constraint: $g < r$. If growth equals or exceeds the required return, the denominator hits zero or goes negative, producing an infinite or negative price. That's the model telling you the constant-growth assumption doesn't hold for this company.
• Note that $D_1 = D_0(1 + g)$, so if you're given the most recent dividend ($D_0$) rather than next year's dividend, you need to grow it forward one period before plugging into the formula.

Constant Perpetuity Model

• Simplest DDM formula: $P_0 = \frac{D}{r}$. This assumes zero growth, treating dividends as a level perpetuity.
• Ideal for preferred stock valuation. Preferred shares typically pay fixed dividends with no growth component, making this model a natural fit.
• Limitation: It ignores growth entirely, so it's inappropriate for common stock of growth-oriented firms where reinvestment drives value.

---

Two-Phase Growth Models

These models recognize that companies often experience distinct growth periods, typically high growth followed by stable maturity. The valuation splits into two parts: calculate the present value of dividends during the initial period, then calculate a terminal value using the Gordon Growth Model at the point where stable growth begins. Sum those present values together.

Two-Stage Dividend Discount Model

Here's how to work through a two-stage valuation:

1. Forecast dividends during the high-growth phase. Using the initial growth rate ($g_1$), project each dividend from $D_1$ through $D_n$, where $n$ is the number of high-growth years.
2. Calculate the terminal value at year $n$. Apply the Gordon Growth Model using the stable growth rate ($g_2$): $TV_n = \frac{D_{n+1}}{r - g_2}$.
3. Discount everything back to today. Find the present value of each individual high-growth dividend and the terminal value, all discounted at the required return $r$.
4. Sum all present values to get $P_0$.

This model captures company lifecycle transitions and is useful when a firm's current growth rate is unsustainable but will eventually normalize. Accuracy depends heavily on forecasting when and at what rate the transition occurs.

H-Model (Half-Life Model)

• Assumes a gradual linear decline in growth rather than an abrupt shift. Growth smoothly transitions from the short-term rate ($g_S$) to the long-term rate ($g_L$).
• Formula: $P_0 = \frac{D_0(1 + g_L)}{r - g_L} + \frac{D_0 \cdot H \cdot (g_S - g_L)}{r - g_L}$, where $H$ is the half-life of the high-growth period (half the number of years over which the decline occurs). So if growth declines linearly over 10 years, $H = 5$.
• More realistic than an abrupt two-stage shift because it captures the gradual competitive erosion that typically slows growth over time.

---

Multi-Phase Growth Models

For companies with complex growth trajectories, these models add additional phases. Each phase requires separate dividend forecasting and present value calculations, increasing both accuracy and complexity.

Three-Stage Dividend Discount Model

• Adds an intermediate transitional phase between high growth and stable maturity. The three stages are: high growth → gradual slowdown → stable maturity.
• Most comprehensive standard model. It provides flexibility for companies moving through multiple lifecycle stages, such as a tech firm that's growing fast now, will slow as the market matures, and will eventually settle into a steady state.
• Requires three separate growth rate estimates. Forecasting errors compound across phases, so sensitivity analysis (testing how the valuation changes when you adjust your growth assumptions) becomes especially important.

Multi-Stage Dividend Discount Model

• Accommodates any number of growth phases for companies with highly irregular growth patterns.
• Maximum flexibility, maximum complexity. Each additional phase adds forecasting requirements and calculation steps.
• Best for unique situations where standard models don't capture a company's specific growth trajectory. In practice, you'd forecast dividends individually for each non-standard phase, then apply a terminal value once growth stabilizes.

---

Variable Growth Models

These models handle situations where growth rates fluctuate unpredictably before eventually stabilizing. The approach involves forecasting individual dividends year-by-year during the variable period, then applying a terminal value calculation once growth normalizes.

Non-Constant Growth Dividend Discount Model

• Projects specific dividends for each year. Rather than applying a single growth formula, you forecast $D_1, D_2, D_3, ...$ individually, discount each back to the present, and then add a terminal value once constant growth begins.
• Handles cyclical or unpredictable industries where earnings and dividends vary with economic cycles (e.g., commodity producers, airlines).
• Computationally intensive because it requires detailed year-by-year analysis, but more accurate for firms with volatile dividend patterns.

Supernormal Growth Model

• Specifically designed for exceptionally high initial growth. This targets startups or firms in explosive growth phases where early dividends grow at rates far above what's sustainable long-term.
• Calculation approach: Discount each high-growth-period dividend individually back to the present, then calculate a terminal value at the point where growth normalizes and discount that back as well.
• Duration estimate is critical. The valuation is highly sensitive to how long you assume supernormal growth continues. Even shifting the supernormal period by one or two years can meaningfully change $P_0$.

---

Quick Reference Table

---

Self-Check Questions

1. Model selection: A mature utility company has paid steadily increasing dividends for 20 years. Which DDM would you use, and why would a Three-Stage Model be inappropriate?

2. Compare and contrast: How does the H-Model's treatment of growth transition differ from the Two-Stage Model, and in what real-world scenario would this difference significantly affect valuation?

3. Formula application: If $g > r$ in the Gordon Growth Model, what happens mathematically, and what does this indicate about the model's applicability?

4. Matching models to companies: You're valuing a biotech startup expecting 25% dividend growth for 5 years before normalizing to 4%. Which two models could handle this, and what's the key difference in their approach?

5. FRQ-style prompt: Explain why the Constant Perpetuity Model is appropriate for preferred stock but not for common stock of a growth company. What fundamental assumption makes this distinction?