Lesson 5 of 8
Fade, and what no fade asserts
A constant growth rate for ten years claims competition never arrives. Across 2,198 US companies the median return on capital sits below its cost.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- A single growth rate held for ten years asserts that nothing competes the return away over that decade.
- Across 2,198 US-listed companies carrying both figures, the median return on capital sits 3.00 points below its estimated cost, and only 841 sit above it.
- Fading one real company's growth from 5 percent to 2.5 percent over the forecast cut the answer by 9.2 percent.
Fade is the assumption that a company's excess return shrinks towards the cost of capital as competition arrives. Most models leave it out, not by argument but because the input box holds one number.
Return on capital less its estimated cost, US-listed companies
Cost of capital: between the median (−3.00pp) and the 75th percentile (3.47pp).As a floor it admits between a quarter and a half of the cohort.
- 10th percentile: −59.96pp
- 25th percentile: −15.94pp
- median: −3.00pp
- 75th percentile: 3.47pp
- 90th percentile: 12.44pp
What the distribution says
The median listed company earns three points less on its capital than its capital is estimated to cost. A quarter of them are nearly sixteen points short. Earning more than the cost of capital is the minority case, not the default.
Procter and Gamble sits at plus 13.57 points, a return on invested capital of 19.76 percent against an estimated cost of 6.19 percent. That is above the ninetieth percentile of the same pool.
A model that grows this company's cash flow at a constant rate for ten years and then forever is asserting that a top-decile spread survives a decade of everyone else noticing it. That assertion may be right. It should at least be made on purpose.
What a fade costs, in dollars
Fading the rate cost 12.06 a share, 9.2 percent of the answer. That is the price of an assumption most models never state.
Fade is about the return, not the growth
Growth is what a fade is usually applied to, because growth is the input the model exposes. The thing that actually fades is the return on new capital.
A business earning 20 percent on capital and reinvesting can grow profitably. When competitors arrive, the return on the next dollar falls first: reported return on capital can still look high while the return on the capital most recently added is already at the cost of capital. The growth rate keeps working for a while after the economics have stopped.
Doing it when the model has one box
ValueScreener's workbench takes one stage-one growth rate. It has no fade input, which is a real limit and worth naming.
Two honest ways round it. Use the equivalent constant rate, as the example above does, and write down that it stands for a fading path. Or build the forecast rows outside the model, fade them explicitly, and bring in the terminal value separately. What is not honest is holding the high rate flat because the box only takes one number.
Check yourself
What does a constant stage-one growth rate for ten years assert about competition?
- Nothing, because growth and competition are unrelated
- That the company's excess return survives the whole forecast without being competed down
- That competitors will arrive in year eleven
Show the answer
Answer: B.
Holding the rate flat is a claim that nothing erodes the economics over the forecast, which is the strongest claim in most models and usually an unexamined one.Across the 2,198 US-listed companies carrying both figures, where does the median return on capital sit against its estimated cost?
- About ten points above it
- About three points below it
- Exactly at it, by construction
Show the answer
Answer: B.
The median is minus 3.00 points and 841 of the 2,198 sit above zero. A positive spread is the minority case.Which figure falls first when competition arrives?
- Reported return on capital
- The return on the capital most recently added
- The growth rate
Show the answer
Answer: B.
The average return on the whole capital base moves slowly because old capital is still earning. The return on new capital moves first, which is why incremental returns are the early signal.Sources
- Procter and Gamble, Form 10-K, fiscal year ended 30 June 2026, filed 2026-08-04: free cash flow, debt, cash and share count.
- Damodaran, Investment Valuation: the excess return period and the fade of returns towards the cost of capital as growth companies mature.
- Mauboussin and Callahan, 'The Math of Value and Growth', 2020: growth creates value only above the cost of capital, and the market-implied duration of that spread.
What this leaves out
- The spread distribution mixes a stored return on capital with an estimated cost of capital, and a spread near zero sits inside the error of the estimate.
- Cost of capital is not computed for every company: lenders are excluded by design, and a company whose filed interest is not a cost of its debt is refused rather than estimated.
- The fade path above is linear because that is the simplest defensible shape, not because businesses decay in equal annual steps.
- Nothing here is advice about any security.