The Multiple Is the Size of Your Bet on the Future

A multiple is not just a price. It is how much of the future you pay for upfront.

It does not measure how much capital you are risking. It measures how much of the valuation depends on the future resembling your assumptions.

If you buy at 10 times next year's expected UFCF, that cash flow equals 10% of the price. At 30 times, it is barely 3.3%. Cash arriving later has to justify everything else.

And that cash does not exist yet. It depends on margins, reinvestment, growth, and discount rates that can change long before the money arrives.

A high multiple is neither good nor bad by itself. It means that more of the price lives in distant cash flows and that the valuation will react more sharply when expectations change.

Before we continue, this post is the next part of a short series. It is worth reading the earlier pieces first:

In the previous articles, we saw that the value of a perpetuity depends mainly on growth and the discount rate. Here we will measure how much a valuation changes when those expectations are wrong, and why the same mistake hurts more at higher multiples.

But first we need an intermediate step: translating any operating multiple into cash. If someone talks about EBITDA or sales, we still need to know how much of that metric ends up as UFCF. Only then can we see the true size of the bet.

All operating multiples are EV/UFCF in disguise

Imagine an analyst tells you:

The company trades at only 9x EBITDA.

It might sound cheap. But the same business could also trade at 12 times EBIT, 16 times NOPAT, or 20 times UFCF. The price has not changed. Only the denominator has.

EV/EBITDA or EV/Sales can be useful when cash flow is negative, erratic, or unrepresentative. The mistake is treating them as final answers. To value the business, we need to know how much of that metric becomes unlevered free cash flow.

The conversion can be summarized in a single identity:

EV/EBITDA, EV/EBIT, EV/NOPAT, and EV/Sales are simply EV/UFCF multiplied by a conversion ratio. Return to the lemonade stand: with an EV of €1,800, €1,000 of sales ultimately becomes €90 of UFCF.

It is the same business. Risk, growth, and value have not changed, but the denominator turns 1.8x sales into 9x EBITDA and, eventually, 20x UFCF.

A company trading at 20x UFCF does not become a bargain because it also trades at 9x EBITDA. The problem is not using EBITDA. The problem is forgetting the costs, taxes, and reinvestment still standing between EBITDA and cash.

Before interpreting a multiple, convert it into UFCF. That is the economic price we care about, and the one that lets us ask how much future the valuation requires.

The multiple is the size of your bet

A high multiple forces you to recover the price through a longer chain of future cash flows. You are not paying only for next year's cash. You are betting that the business can sustain or improve that cash flow for a long time.

To isolate the mechanism, we will use a stable-growth perpetuity. We assume normalized UFCF expected next year, a constant WACC, and perpetual growth below WACC. This is not the right model for every company, nor does it replace a multi-stage DCF. It is a simplification designed to measure one specific sensitivity while holding everything else constant.

Return to the lemonade stand. We expect it to generate €90 of normalized UFCF next year. We valued it at €1,800, or 20 times that UFCF, because we required a 9% return and expected it to grow steadily at 4%:

Look at the denominator. The entire valuation rests on a subtraction that equals five percentage points.

Now imagine you are wrong by a single point. The business does not grow at 4%, but at 3%. The spread widens from five to six points:

The stand is no longer worth €1,800. It is worth €1,500.

A one-point mistake has erased 17% of your valuation.

And how much would that same point have hurt in more expensive businesses?

Let us repeat the mistake in three companies that differ only in the multiple paid:

The same one-point mistake destroys 9% of a valuation at 10x and 25% of one at 33x.

For very small changes, the local sensitivity can be approximated by multiplying the change by the initial multiple:

The multiple you pay approximates your valuation's sensitivity to small changes in (WACC − g).

The approximation says that an adverse one-point change costs 10% at 10x and 30% at 30x. The exact calculation gives −9.1% and −23.1%, respectively. One point is no longer a particularly small change when the multiple is high. I leave the exact formula in Appendix I.

There is also a favorable side. Starting from 20x, a one-point improvement raises value by 25%, while an equivalent deterioration reduces it by 16.7%. Neither move happens merely because you bought at a high multiple. Both require a surprise relative to what the price already discounted.

If expectations are met, you earn the return embedded in the price. The multiple measures the size of your exposure to surprises, not their direction.

A high-multiple investment behaves like a long-duration bond

There is another way to see the same idea. When you pay a high multiple, the cash generated over the next few years is small relative to the price. What you are really buying is cash that will arrive far into the future.

And we can measure it. Assume a 9% WACC. Here is the share of value that depends on cash flows beyond year 10:

At 20 times cash flow, almost two-thirds of the value sits in cash flows that will not arrive for more than a decade. At 33 times, three-quarters does.

A high multiple makes the investment behave like a long-duration asset. It is not a bond because its cash flows are not fixed, but it shares a bond's high sensitivity to changes in the discount rate.

That duration helps explain the multiple compression of 2022. For many compounders, the share-price decline was far greater than the deterioration in the following year's cash flow. Rising discount rates did much of the work. The business does not have to collapse for a demanding valuation to fall.

When errors compound

The previous rule held expected cash flow constant. In real life, we can also be wrong when estimating that cash flow, especially if we start from sales or EBITDA.

Return to the lemonade stand one last time. Reality will be €1,000 of sales, €90 of UFCF, stable growth of 3%, and a value of €1,500 at a 9% WACC.

Analyst A works close to cash. A correctly estimates €90 of UFCF but expects 4% growth. That single-point mistake leads A to pay €1,800 and lose 16.7% when reality arrives.

Analyst B works down from sales. To get from €1,000 of revenue to cash, B has to move through the entire funnel, making an assumption at every step. B uses the real reinvestment figure and makes no huge mistake anywhere else, just a 10% optimistic error at each stage:

Three small mistakes multiply. B expects €120 of UFCF instead of €90. B also shares A's mistaken 4% growth assumption, so B pays €2,400 for a business worth €1,500.

The multiple rule measures only the error in (WACC − g) while holding cash flow constant. That describes A. B is also wrong about the numerator. Moving from the €120 B expected to the actual €90 is a 25% reduction that does not disappear at a lower multiple.

The two errors combine. The multiple amplifies the growth revision, while poor conversion from sales inflates cash flow before that multiple is applied. The exact decomposition and the 10x, 20x, and 33x scenarios are in Appendix III.

Read the multiple backwards

If the multiple is the size of your bet, the least you can do before placing it is know what you are betting on. And you can solve for that. Instead of calculating a justified multiple from your assumptions, observe the market multiple and ask what has to be true for the price to make sense.

Within the same stable perpetuity, using UFCF and EBITDA expected over the same period, we can combine the identity from the beginning with EV/UFCF = 1/(WACC − g):

Suppose a company's EV equals 12 times next year's expected EBITDA, with an estimated WACC of 9% and stable growth of 4%:

At that price, the model needs a stable 60% conversion from EBITDA to cash for the assumptions to fit together. Perhaps the business can achieve it. Perhaps it will grow faster, or its WACC will be lower. We are not discovering what every investor literally expects. We are making explicit the combination of assumptions that justifies the price.

The calculation does not solve the valuation, but it turns a vague statement, "it trades at 12 times EBITDA," into a concrete question:

Can this business sustainably convert around 60% of EBITDA into UFCF while growing at 4%?

The same method works with EV/Sales by solving for the margin. A company trading at 5 times next year's expected sales, with the same 9% WACC and 4% growth, implies:

The price requires a 25% free cash flow margin. That is not necessarily impossible. But we are no longer saying that "five times sales is reasonable for a quality company." We are saying something measurable.

And measurable statements have the inconvenient virtue of being able to be wrong.

Conclusion

The multiple is the size of your bet because it tells you how much future you need to defend today's price.

1. Near-term cash carries less weight. At 10x, next year's UFCF equals 10% of the price. At 30x, only 3.3%.
2. Value lives further away. The higher the multiple, the more of the valuation depends on cash flows that will take many years to arrive.
3. Assumptions matter more. Those distant cash flows depend on margins, reinvestment, growth, and discount rates that will change before you collect them. In a stable perpetuity, a high multiple makes small changes in (WACC − g) move value dramatically.

The denominator adds a second layer. EV/EBITDA or EV/Sales can hide how much UFCF you are actually buying. If you overestimate conversion, the effective multiple on real cash will be higher than you thought. You will have made an even bigger bet without seeing it.

The next time someone tells you a company is cheap because it trades at 8 times EBITDA, the answer should not be yes or no. It should be:

What multiple of actual cash flow am I paying, how much of that value lives in distant cash flows, and what must remain true for those cash flows to arrive?

That is valuation. The rest is division.

And one more difficult question remains. If, at 20 times cash flow, almost two-thirds of value lives beyond year 10, what exactly does that distant value depend on? That territory has a name: terminal value.

We will talk about it in the next chapter.

Appendix I: where the multiple rule comes from

The rule follows directly from comparing value before and after a change in the spread between WACC and growth.

1. The multiple is the inverse of the spread

In a growing perpetuity, value and the multiple on next year's expected UFCF are:

Paying 20x therefore means starting from a 5% spread: 1 / 0.05 = 20. Paying 10x means starting from a 10% spread. The higher the multiple, the narrower the spread supporting the valuation and the more any change in it matters.

2. What happens when the spread changes

Let Δ be the change in (WACC − g). If expected UFCF does not change, the ratio of new value to original value is:

Subtracting 1 from that ratio gives us the exact percentage change in value:

For small changes, 1 + M × Δ is approximately 1. This gives us the simple rule used in the article:

Δ is expressed as a decimal, so one percentage point is 0.01. For a one-point deterioration, the approximation therefore estimates a 10% decline at 10x, a 20% decline at 20x, and a 30% decline at 30x. The exact formula moderates those figures:

The asymmetry is mathematical, but it does not make high multiples better investments. An improvement creates a return only if the new spread is more favorable than the one already discounted by the price when you bought. Likewise, a deterioration destroys value only if expectations worsen after the purchase. The multiple tells you how strongly value will react to that surprise, not which direction the surprise will take.

3. Why a high multiple depends more on the future

In a growing perpetuity, the share of value that depends on cash flows after year N is:

With a 9% WACC and 4% growth, 62% of value comes from cash flows in year 11 and beyond: (1.04 / 1.09)¹⁰ ≈ 62%. That is the other side of the rule. At a given WACC, a high multiple does not merely amplify changes in assumptions. It also places a larger share of the valuation in distant, more uncertain cash flows.

Appendix II: the full EV/EBITDA derivation

For anyone who wants to see the entire bridge, nuts and bolts included. We start from the valuation formula and the definition of UFCF:

Substitute EBIT = EBITDA − D&A, expand, and collect the depreciation terms (−D&A(1 − t) + D&A = D&A × t):

Insert this into EV₀ = UFCF₁ / (WACC − g), then divide by EBITDA expected over the same period:

The numerator is literally the conversion from EBITDA to UFCF written out piece by piece: taxes, the depreciation tax shield, capex, and working capital.

Using the lemonade stand figures, t = 25%, D&A/EBITDA = 25%, capex/EBITDA = 30%, ΔWC/EBITDA = 6.25%, and WACC − g = 5%:

The 45% numerator is exactly the conversion from EBITDA to UFCF: 90 / 200 = 45%.

Nothing magical. Just cash coming in and going out.

Appendix III: when the cash-flow estimate is wrong too

The multiple rule assumes expected UFCF remains constant. If we also revise cash flow, the approximation gains an additional term:

The multiple appears only in the second term. Revising expected UFCF from €120 to an actual €90 is a 25% decline from the price paid, regardless of the multiple. For large changes, the exact decomposition is:

Let us compare the same mistakes, €120 of expected UFCF versus €90 actual and 4% growth versus 3%, at three initial multiples. A estimates cash correctly and is wrong only about growth. B suffers from both errors:

B's cash factor is always 75%: 90 / 120. What changes with the multiple is the second factor. At 10x, it retains 90.9% of the multiple; at 20x, 83.3%; and at 33x, 75%. That is why the lower multiple cushions the error in (WACC − g), but does not eliminate the earlier mistake in estimating cash.