Skip to content

Dividend Discount Model (DDM)

One-line definition

Values a share as the present value of all future dividends it will pay — the equity analog of pricing a bond as the present value of its cash flows.

Formula

Gordon growth (constant growth forever): V = D1 / (r - g)

Two-stage: discount an explicit growth rate for N years, then a Gordon-growth terminal value at year N, discounted back to today.

Why PMs care

Every multiples-based shortcut (P/E, PEG) is a compressed version of this — relative_valuation_multiples.md shows the algebra connecting them. When a stock's price doesn't match a DDM value at a defensible (r, g), that gap is either a real growth/risk difference the market is pricing correctly, or a genuine mispricing — DDM is how you tell which.

Worked example

See src/pm/equity/valuation.py (gordon_growth_value, two_stage_ddm_value) and notebooks/equity/28_equity_valuation_and_capm.ipynb.

Common mistakes

  • using this year's dividend (D0) in the Gordon-growth numerator instead of next year's (D1) — forgetting to grow it one period first
  • picking g >= r, which makes the model diverge (infinite value) — a real constraint on the assumption, not a numerical accident
  • applying constant-growth Gordon growth to a young, high-growth company where growth clearly isn't constant forever (use the two-stage version instead)

Limitations

Real companies don't grow at a single rate forever, and the required return r is itself an assumption — usually sourced from CAPM (see capm_and_beta.md), not directly observable.