Learn
How long a high multiple takes to come down
A high multiple does not need the share price to fall to come down. Hold the price perfectly still, let the earnings keep compounding underneath it, and the multiple falls on its own — at a rate you can work out in your head. The only question worth asking is how many years it takes.
Where this follows from
Its companion page, why a great business can still lose you money, settles the identity everything here rests on:
What a share price actually is
Share price=earnings per share ×the multiple
That page shows the multiple can fall faster than earnings rise, which is why a price can drop on good results. This one asks the follow-up: if the earnings simply keep growing and the price stands still, how long before the multiple is somewhere you would call normal? Nothing has to go wrong for it to happen. Nobody has to sell. The earnings only have to show up.
Move the sliders
Start at a multiple, name one you would be comfortable with, and read off the wait at each growth rate. The default is a round forty times becoming a round twenty.
The ladder worth memorising
The time it takes to halve a multiple depends only on the growth rate. Not on where the multiple started — 40× and 12× halve on exactly the same schedule. That makes it six numbers, and six numbers will fit in anyone's head.
Seven, five, four, three, two-and-a-half — and then about two, whatever else happens.
The rungs are fine at the bottom and coarse at the top on purpose. Five points of earnings growth is worth two whole years down at 10%, and about ten weeks up at 40% — so everything above 35% collapses into one band, because the difference up there is not worth carrying around. That asymmetry is the useful part: being precise about the growth rate matters enormously for a steady compounder, and barely at all for a fast one.
The shortcut
72 ÷ growth % ≈ years to halve
The rule of 72, run backwards. It always lands a little short, and shorter the faster the growth — so treat what it gives you as the optimistic edge of the range rather than the middle of it.
shortcut vs actual 10% → 7.2 vs 7.2715% → 4.8 vs 4.96 20% → 3.6 vs 3.8025% → 2.9 vs 3.11 30% → 2.4 vs 2.6440% → 1.8 vs 2.06
The same thing as a table — 40× after n years
| Earnings growth | Yr 1 | Yr 2 | Yr 3 | Yr 4 | Yr 5 | Yr 6 | Yr 7 | Yr 8 | Yr 9 | Yr 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| 5% a year | 38.1 | 36.3 | 34.6 | 32.9 | 31.3 | 29.8 | 28.4 | 27.1 | 25.8 | 24.6 |
| 8% a year | 37.0 | 34.3 | 31.8 | 29.4 | 27.2 | 25.2 | 23.3 | 21.6 | 20.0 | 18.5 |
| 10% a year | 36.4 | 33.1 | 30.1 | 27.3 | 24.8 | 22.6 | 20.5 | 18.7 | 17.0 | 15.4 |
| 12% a year | 35.7 | 31.9 | 28.5 | 25.4 | 22.7 | 20.3 | 18.1 | 16.2 | 14.4 | 12.9 |
| 15% a year | 34.8 | 30.2 | 26.3 | 22.9 | 19.9 | 17.3 | 15.0 | 13.1 | 11.4 | 9.9 |
| 20% a year | 33.3 | 27.8 | 23.1 | 19.3 | 16.1 | 13.4 | 11.2 | 9.3 | 7.8 | 6.5 |
| 25% a year | 32.0 | 25.6 | 20.5 | 16.4 | 13.1 | 10.5 | 8.4 | 6.7 | 5.4 | 4.3 |
| 30% a year | 30.8 | 23.7 | 18.2 | 14.0 | 10.8 | 8.3 | 6.4 | 4.9 | 3.8 | 2.9 |
| 40% a year | 28.6 | 20.4 | 14.6 | 10.4 | 7.4 | 5.3 | 3.8 | 2.7 | 1.9 | 1.4 |
| 50% a year | 26.7 | 17.8 | 11.9 | 7.9 | 5.3 | 3.5 | 2.3 | 1.6 | 1.0 | 0.7 |
The marked cell is the first year that growth rate has reached 20×. Share price held flat throughout.
What the curve is not telling you
Assumption 01
Growth arrives in steps, not as a curve
Earnings land four times a year, in lumps, with revisions behind them. A smooth line is an average path through a decade, not a path any company walks. A rate that holds for three years and breaks in the fourth does not de-rate a share slowly — it re-rates it downwards in an afternoon.
Assumption 02
The price is doing something too
With the price held flat, the whole of earnings growth is spent compressing the multiple and the holder banks nothing. Let the price drift up and only the gap between the two rates compresses: 20% growth against 10% drift de-rates at 8.3% a year, not 16.7%, so the wait roughly doubles.
Assumption 03
Arriving is not the same as cheap
Half of 90× is 45×. The clock says how fast a multiple unwinds; it never says the destination is a multiple worth paying. That is settled by how durable the growth rate actually is, and by what the business earns on the capital it keeps back to produce it — which is what the series spends its time on.
Terms used here are defined in Key terms. Every figure on this page is arithmetic on constant rates — multiple(t) = multiple now × ((1 + drift) ÷ (1 + growth)) raised to t — and no company data is involved. General explanation, not personal advice, and not a recommendation to buy or sell anything. Capital is at risk.