# Valuing a Compounder
## Three inputs: reinvestment rate, return on capital, and duration
Everything else is noise. REQ's framing (REQ, Jul 2025, p.267, p.269):
> [!important] The requisite blend
> Durability in reinvestment opportunities, a high return on capital, and **substantial certainty that the reinvestment continues for years**. Combine the three and the justifiable price for the stock usually exceeds initial expectations. Investors commonly overvalue growth while overlooking durability.
The worked example (p.131–132), all companies earning $100:
| | Reinvest | At return | Justified P/E |
| --- | --- | --- | --- |
| Company A | 100% | 10% (= cost of capital) | ~10 |
| Company B | 100% | 14% for 20 years | ~20 |
| Company A′ | 35% at 20%, rest paid out | 20% | ~15 |
| Company B′ | **75% at 20%**, long private-market runway | 20% | **~30** |
A′ and B′ earn *identical* incremental returns. B′ is worth double, purely because it has somewhere to put the money. The 65% "leak" is what kills A′ — shareholders have to find comparable returns themselves in public markets.
And the premium table, with cost of capital set at 9.5% (p.280): growth at a return equal to the cost of capital creates nothing. At **16% incremental ROIC and 5% growth**, a 46% premium to a no-growth company is justified. At **22% and 8%**, the company is worth **three times** a no-growth peer.
> [!tip] The implication for management
> "The CEO of a company generating 18% return on capital should spend his or her time finding growth opportunities instead of trying to further increase profitability."
## Why these are systematically underpriced
REQ's most interesting argument is behavioural. Standard DCF uses **exponential discounting** — time-insensitive, the discount factor falling at a constant rate, so distant cash flows are worth very little. Humans do not behave that way. We use **hyperbolic discounting**: steep discounts in the near term (we want the $100 today over $110 in a year) and much shallower ones far out (we happily wait year 11 for $110) (p.270–272).
> [!info] The size of the effect
> In REQ's 20-year illustration, the present value under a hyperbolic function is **almost 40% higher** than under exponential discounting. They are not advocating you rewrite your DCF — the point is that the market, in aggregate, appears to price highly predictable long-duration businesses closer to hyperbolic than exponential. Using textbook discounting on very long-term cash flows may significantly **undervalue** these compounders.
Mark Leonard's version of the same idea: "a company that is growing quickly, that the market expects to stop growing within the next 5–7 years, but that actually keeps growing quickly for much longer... may appear expensive on a PE basis, but actually be an attractive long-term investment on a value investing basis."
## The uncomfortable holding rule
> [!warning] These are almost never cheap on near-term multiples
> "Our businesses are often overvalued in the short term and significantly undervalued over the long term." REQ's discipline is to be careful about the price paid **and** to be comfortable holding businesses at multiples where they would not buy. A stock should not be sold on price alone; there must be something more than a temporarily high multiple.
>
> "Making an error in estimating the *price* of a business is preferable to underestimating the *quality* of a business" (p.269).
They are, however, careful about companies reinvesting **100%** of profits — those are vulnerable if growth disappoints. The sweet spot named explicitly: **80% of profits reinvested at more than 20% return on equity, for a very long time** — a combination the market almost never prices correctly (p.267).
## What the market actually pays for
- **Not hyper growth.** Regressing EV/EBITA and EV/Sales against 21–23E sales CAGR gives R² of 0.06 and 0.003 — no relationship (p.277).
- **Not margins per se.** EBITA margin vs EV/EBITA: R² 0.13. What earns a premium is **stable or increasing** margins, and size (p.278, p.113).
- **Durability and proof.** Larger Nordic compounders (Addnode, Addtech, Beijer Ref, Indutrade, Lagercrantz, Lifco, Vitec) re-rated away from smaller ones (AQ Group, Beijer Alma, Ependion, Bergman & Beving, Medcap, OEM, Xano) from around **2017** — because they had shown higher reinvestment rates, higher growth and higher returns on capital, and because confidence in duration rises with track record (p.274).
## Why it matters
This is the note that changes behaviour. It says the mistake to avoid is not overpaying by 20% — it is selling a genuine compounder because the multiple got uncomfortable.
## Related
- [[Acquisition-Driven Compounders MOC]]
- [[Quality Compounders]]
- [[Scaling M&A and the Runway]]
- [[Buffet Indicator]]
- [[Time in the Market]]