What If Growth Is Higher Than Your Discount Rate?

Owner earnings, Buffett’s input

From his 1986 letter: owner earnings = reported earnings + depreciation/amortization − maintenance capital expenditures (± working-capital changes). It’s the cash an owner could pocket without hurting the business — the honest number to discount.

The formula that breaks

The Gordon growth shortcut values a perpetuity:

As g approaches r, the denominator approaches zero and V rockets toward infinity. At g ≥ r the formula is simply invalid — not a bargain signal, a model failure.

Buffett’s practical answer

No company grows faster than the economy forever — it would eventually become the economy. High growth is a temporary phase. So the pros use a two-stage model: explicit high growth for 5–15 years, then a terminal rate below r (typically 2–3%, near long-run GDP growth).

Worked example

1. Y1 owner earnings: $10/share, g = 12%, r = 10%. Gordon: invalid (g > r).
2. Two-stage instead: 12% for 10 years → Y10 OE = $27.73.
3. Terminal at g=2.5%: V₁₀ = 27.73×1.025/(0.10−0.025) = $379.
4. Discount everything to today: ≈ $265/share — finite, defensible.

What the sliders teach

• With r − g = 5pt, distant bars shrink fast: the far future barely matters.
• At r − g = 1pt, green bars stay tall for decades — value lives in the far future, which is why “growth” stocks whipsaw when rates move.
• At g ≥ r, green bars stop shrinking at all: each future year is worth as much or more than the last, and the sum never converges.

The margin-of-safety point

Buffett’s deeper lesson: if your valuation only works because g is within a point of r, you don’t have a valuation — you have a hope. Demand a wide r − g spread, or a price so far below your estimate that the model’s fragility can’t hurt you.