Growth Is Not Always Good

In the previous post, we saw where shareholder-side valuation multiples come from: P/E, P/B, shareholder P/FCF, dividend yield, earnings yield, and shareholder P/S.

We were trying to answer one question: How much is the part of the business that belongs to shareholders worth?

If you have no idea what I am talking about, take a look at the previous article:

Now we change the task.

We are going to look at the entire operating business, not just the shareholders' slice. The lemonade stand itself, not the part left for you after paying the bank.

The foundation does not change: we are still discounting cash flows.

But valuing the whole business requires a few new pieces:

• another way to measure its value (the Enterprise Value),
• another discount rate (the WACC),
• and the relationship between growth, returns, and reinvestment.

That relationship is the star of this post and the reason for its title. Growth is not free: you have to reinvest cash to achieve it. And if that cash earns less than the capital requires, the company grows while becoming less valuable.

Counterintuitive. I know.

That is why understanding it matters so much.

In this post, we will assemble those three pieces. Multiples for the entire business (EV/EBITDA and company multiples) will come in the next one.

So things are about to get a little more complicated.

Enterprise Value

A business does not always belong entirely to its shareholders. Banks and bondholders also have economic claims on it when debt is involved.

This is where enterprise value, or Enterprise Value (EV), enters the picture. We mentioned it in the previous post. It measures the value of the entire operating business, regardless of whether it is financed with debt, equity, or a mix of both. Market capitalization (Market Cap), by contrast, measures only the shareholders' slice.

The simplified calculation looks like this:

EV = Market Cap + Debt − Cash

Where

EV = enterprise value

value of the entire operating business

Market Cap = market capitalization

market value of the shareholders' slice

Debt = financial debt

money the company owes its creditors

Cash = surplus cash

cash the company does not need to operate and that could reduce the purchase price

This is the back-of-the-envelope version. If a business is worth €100,000, has €40,000 of debt, and holds €10,000 of cash, the shareholders' slice is worth €70,000:

Market Cap = €100,000 − €40,000 + €10,000 = €70,000

Enterprise Value = €70,000 + €40,000 − €10,000 = €100,000

That is why two companies can have the same market capitalization and very different enterprise values. It depends on their capital structure and how they are financed.

It is also why operating metrics are usually compared with EV: they belong to the business before the pie is divided.

Revenue, EBITDA, EBIT, and UFCF are not shareholder-only figures. They belong to the whole business. And note that only UFCF is free cash flow. The others need a bridge to get there.

Free cash flow is what we like, after all.

Enterprise Value is the first piece of the puzzle. We have it.

The other two are the business's discount rate, the WACC, and the relationship between growth, returns, and reinvestment, the ROIC.

Now the fun starts.

Adapting the DCF

In the previous article, we used shareholder cash flows and a discount rate, “r,” focused only on us:

Equity value =

FCFE₁÷r_equity − g

Now we use cash flows for the entire business: unlevered free cash flow, or UFCF.

UFCF is operating cash after taxes and reinvestment but before interest, debt repayment, or financing decisions. It is the business's cash before we ask whether the company is financed with debt or equity.

And here comes an important change from the previous post: the discount rate cannot be only the cost of equity either.

It has to be the rate required by all providers of capital.

WACC: shareholders and debt

That is the WACC. Or “weighted average cost of capital” if we want to sound fancy.

As the name suggests, it is the weighted average of the rates required by shareholders and creditors.

Its formula is:

WACC = r_equity × % equity + r_debt × (1 − tax rate) × % debt

Where

r_equity = cost of equity

return required by shareholders

r_debt = cost of debt

return required by creditors

tax rate = tax rate

adjustment for the fact that interest expense usually reduces taxes

% equity = equity weight

share of the company's capital financed by shareholders, measured at market value

% debt = debt weight

share of the company's capital financed by debt, measured at market value

Two details matter here:

1. The weights are based on the market value of the capital financing the business, not its book value.

   The balance sheet records what things cost in the past; the market tells us what they are worth today. Required returns apply to today's value: someone buying shares at the market price expects a return on that price, not on an accounting entry.

   If the company's market capitalization is €900 and its debt is €100, equity has a 90% weight and debt a 10% weight, whatever the balance sheet says.

2. Interest expense is usually tax-deductible, so the cost of debt is measured after taxes.

   If debt costs 6% and the tax rate is 25%:

   6% × (1 − 25%) = 4.5%

   Part of the cost of debt is therefore offset by tax savings. That is why the tax rate appears in the formula.

   But that saving does not arrive by divine decree. The company must generate taxable income and be allowed to deduct the interest. If it accumulates losses or the deduction is limited, the tax shield may arrive late, arrive only partially, or never arrive at all.

To simplify:

• r_equity = what shareholders require for bearing residual risk.
• r_debt = what creditors require for lending to this company.
• WACC = the weighted average of both requirements.

There is nothing new about estimating r_equity: it is the same “r” from the previous post, the minimum return you require for taking the risk of this business. The difference is that it is no longer alone in the formula.

The new piece is r_debt. Contrary to what it may seem, this is not the historical coupon on a bond.

Bondholders do not lend for free. They could buy government debt or lend to another company, so they demand a return that matches this company's risk. That opportunity cost lives inside r_debt. It should therefore be the current cost of debt: the rate at which the company could borrow today, or the market yield on its bonds if they trade with reasonable liquidity.

If a Treasury bond pays 4% and this company's debt yields 7%, that 7% already includes compensation for the extra risk:

r_debt ≈ risk-free rate + credit spread

You may disagree with the market. If you think it understates the debt's risk, use a higher r_debt; if you think it overstates the risk, use a lower one. But do not mix layers. Your required return as a shareholder does not belong inside r_debt. That part must represent the economic cost of debt for this business.

And that is the summary.

Now we account not only for what shareholders require but also for what creditors require. Fairly logical.

Substituting these terms into our beloved simplified formula gives us this:

EV₀ =

UFCF₁÷WACC − g

Where

UFCF1 = expected unlevered free cash flow

operating cash available to shareholders and creditors after taxes and reinvestment

WACC = weighted average cost of capital

required return for the entire business

g = growth

long-term growth rate of operating cash flow

The same as before, except now the numbers belong to the whole company rather than only its shareholders.

All the disclaimers to keep in mind:

1. This formula is not a full DCF with ten years of forecasts. It is a stable-growth perpetuity. A simplification. From this point onward, we assume normalized UFCF₁, sustainable growth, margins, reinvestment, and capital structure.
2. WACC and g must use the same currency and the same terms: both nominal or both real.
3. A perpetual g cannot exceed the long-term growth of the economy in which the business operates. Nothing can outgrow its environment forever.
4. We use forward metrics: EV₀ is today's value, and figures with a subscript 1 are expected for next year. Future cash flows are what matter.

With that, we have a working idea of what WACC is.

Now we move to the most fundamental and interesting concept.

ROIC, reinvestment, and growth

We need to understand how companies accumulate and invest capital over the long term in order to grow.

Growth does not appear ex nihilo.

To grow, a business usually has to reinvest. And the important question is: at what return does it reinvest?

This is where Return on Invested Capital (ROIC) comes in.

ROIC is the return a company earns on every euro invested in the business. In other words, if we invest €100, how much money does that capital generate?

Measuring this is not entirely trivial.

We are talking about a metric for the entire business, not just its shareholders, so net income will not do. It already includes interest and belongs on the shareholder side. We need a measure taken before shareholders and creditors are paid.

That is why we use NOPAT (Net Operating Profit After Taxes).

NOPAT is operating profit after taxes but before interest and other financing decisions. It measures what the business produces before that value is divided between creditors and shareholders.

The formula is fairly intuitive:

NOPAT = EBIT × (1 − tax rate)

That lets us measure the money “earned.” But to understand the return, we still need the money invested: Invested Capital (IC).

Invested capital is the net operating capital the business needs to function: broadly speaking, operating assets minus operating liabilities.

The exact calculation has nuances (which assets count as operating, what should be excluded, and how to handle goodwill), so I have left the details for Appendix I.

The important point is the idea: this is money that is “working,” not money sitting around doing nothing.

We can now define ROIC:

ROIC =

NOPAT÷Invested Capital

Where

NOPAT = net operating profit after taxes

profit from the operating business after taxes but before interest and financing

Invested Capital = invested capital

operating capital the company needs to generate that profit; the average between the beginning and end of the period is used

Suppose the business has €500 of invested capital and generates €100 of NOPAT. Its ROIC is 20%.

Nothing surprising so far.

ROIC is not growth

A ROIC of 20% does not mean the business will grow at 20%.

It depends on how much money it reinvests at that return.

If it reinvests the full €100 of NOPAT and the new capital also earns a ROIC of 20%, that reinvestment will add €20 to the following year's NOPAT. It will rise from €100 to €120: growth of 20%.

If it reinvests only €50, it will add €10. NOPAT will rise from €100 to €110: growth of 10%.

The other €50 does not automatically become dividends. It forms part of the UFCF available to all providers of capital. The company may pay interest, repay debt, buy back shares, distribute dividends, or leave the cash gathering dust on the balance sheet.

Here is the heart of the matter.

There is a relationship between the money a company reinvests to grow and the operating cash left over after that reinvestment.

The relationship is expressed like this:

g = expected marginal ROIC × reinvestment rate

Where

expected marginal ROIC = return on new capital

expected after-tax operating profit generated by each additional euro reinvested

reinvestment rate = reinvestment rate

percentage of NOPAT reinvested in the business

Here, g is the sustainable growth in NOPAT generated by new investments. Under the stable assumptions we are using, we also treat it as growth in UFCF.

The relevant ROIC is the expected marginal return on new capital, not the attractive historical ROIC you see on Yahoo Finance. They are not the same thing.

The difference between historical and marginal ROIC has plenty of nuance, so I have moved it to its own Appendix II.

Even so, the idea to remember is that ROIC measures the historical return on past investment, while we are looking forward. We want to know the ROIC on the next investment. That is the expected marginal ROIC.

They are not the same.

The key ideas are:

• To grow more, you need to reinvest more.
• To grow with little reinvestment, you need a high marginal ROIC.
• Two companies can have the same g and very different economic quality.

For example, one company can grow at 5% by reinvesting 12.5% of its NOPAT at a 40% return. Another can grow at the same 5% by reinvesting 62.5% at an 8% return.

The same growth, but very different business quality.

If both are going to grow at the same rate, of course we prefer the company that can do more with less. It is a more efficient money-making machine.

Back to the DCF

Let us connect what we have learned to our favorite simplified formula.

EV₀ =

UFCF₁÷WACC − g

We have just seen that marginal ROIC is directly related to g and that not all growth is equal.

So let us look at the denominator:

WACC − g

• The lower the WACC, the more the business is worth.
• The higher the g, the more the business is worth.

It is a small subtraction, but any change here has an enormous effect on valuation.

With a WACC of 8% and a g of 3%, you divide cash flow by 5%. If g rises to 5%, you divide by 3%. Same cash flow. Sixty-six percent more value.

The interesting part is that we already know g depends on marginal ROIC and reinvestment. We can substitute it:

WACC − g = WACC − (marginal ROIC × reinvestment rate)

We are beginning to see how marginal ROIC affects valuation. But this is only the first half of the story.

If marginal ROIC is positive, more reinvestment increases g, reducing WACC − g.

That sounds fantastic. At this point, we might conclude that more reinvestment is always better. But we cannot look only at the denominator.

The cost of reinvestment

Reinvestment also reduces the cash available today.

That is the problem.

The reinvestment rate does not appear in the top half of the formula, so it is hard to see how it affects the cash available today.

Luckily, we have already seen that UFCF₁ can be calculated like this:

UFCF₁ = NOPAT₁ × (1 − reinvestment rate)

This works because UFCF₁ is the part of NOPAT₁ that does not need to be reinvested in the business's operations. It is cash available to shareholders and creditors, and we have not yet decided how to divide it.

Here is the old friend we were looking for: the reinvestment rate.

To keep things clear and avoid scaring anyone, it is simply the percentage of NOPAT being reinvested in the business.

The basic way to calculate reinvestment is:

Reinvestment = capex − depreciation + increase in operating working capital

The reinvestment rate we have been using is simply this figure divided by NOPAT.

Now that we understand the pieces, we can substitute them into the simplified formula:

EV₀ =

NOPAT₁ × (1 − reinvestment rate)

÷

WACC − marginal ROIC × reinvestment rate

Now we can see the trade-off:

• Reinvestment reduces the numerator because there is less free cash flow today.
• But it also reduces the denominator because it increases growth.

If reinvestment can be good and bad at the same time, how do I know whether I should reinvest?

Put another way:

Does the additional growth compensate for the cash you sacrifice today?

The answer is not as simple as “if you have a positive ROIC, you should invest.” Money has a cost, and so does risk.

Imagine a project with an expected return of 3% and a cost of capital of 8%. The return is positive, but it is five percentage points below what the capital requires. The project destroys value.

So it is not enough to say that more reinvestment is better.

The economic intuition is: reinvest while the expected marginal ROIC exceeds the cost of capital (WACC).

Convincing at first glance.

It means we are creating value above our cost.

And we do not have to rely on intuition. We can prove that the multiple rises or falls according to the relationship between marginal ROIC and WACC. I have left the complete proof, derivative included, in Appendix III.

What we need to understand is:

• If marginal ROIC is greater than WACC, more reinvestment increases value.
• If marginal ROIC equals WACC, more reinvestment does not change value.
• If marginal ROIC is lower than WACC, more reinvestment destroys value.

Let us put numbers on it. In every scenario, NOPAT₁ is €100, WACC is 10%, and the reinvestment rate is 25%:

Marginal ROIC	g	UFCF1	EV0	vs. no reinvestment
No reinvestment	0%	€100	€1,000	(base)
20%	5%	€75	€1,500	+50%
10%	2.5%	€75	€1,000	0%
5%	1.25%	€75	€857	−14.3%

All three reinvestment scenarios sacrifice the same cash: €25. The company grows in every scenario; only the return changes.

• At 20%, growth creates 50% more value.
• At 10%, exactly the WACC, reinvestment is running to stand still: €1,000, the same as no growth.
• At 5%, the company grows every year while being worth 14% less than if it had stayed still.

It bears repeating: an average ROIC below WACC tells us about the past: capital that has already been used poorly.

It is an autopsy.

We care about marginal ROIC. What destroys value is reinvesting new capital below the appropriate cost for its risk.

Why, then, do companies with a high marginal ROIC not reinvest all their money?

Because the best opportunities run out. Marginal ROIC tends to fall with each additional euro invested. Reinvestment makes sense while the next euro still earns more than its cost. Not one euro more.

If you find a company capable of reinvesting at high returns above its cost of capital and maintaining a high reinvestment rate for a long time, you have found a damn unicorn.

---

This relationship between reinvesting money and returning it to shareholders is useful for understanding why many companies should not distribute dividends, repay debt immediately, or use the cash for other things.

It explains the implicit pact between investors and managers: we accept less cash today if the company can reinvest it at attractive returns and turn it into more cash tomorrow.

In summary:

• WACC answers: what return does capital require?
• g answers: how fast do cash flows grow?
• Marginal ROIC answers: how much operating profit does each euro reinvested generate?
• The reinvestment rate answers: how much capital must be committed to achieve that growth?

We can already infer that a company with a higher marginal ROIC is preferable to one with a lower marginal ROIC if both grow at the same rate.

I will spare you the proof this time. You are welcome. But the intuition is simple: it can grow at the same rate using less money, leaving more free cash flow to distribute.

Conclusion

Growth is not always good. Now we know why.

Growth is not a quality of a business. It is an amplifier. It magnifies what the business already is: if the company deploys new capital above its cost, growth magnifies value. If it deploys capital below its cost, growth magnifies the hole.

Paradoxically, the fastest way to make or lose money is to reinvest heavily. The outcome depends on the relationship between marginal ROIC and WACC.

That is why “how fast does it grow?” is a bad question.

Try these instead:

• At what return does the company deploy new capital (its marginal ROIC)?
• How much cash must it sacrifice today to achieve that growth (the reinvestment rate)?

When the market pays a high or low multiple for a company, it is implicitly answering those questions. In the next post, with the engine now assembled, we will derive EV/UFCF, EV/EBIT, and EV/EBITDA from this same formula, along with their assumptions, variables, and traps.

Once again, multiples are compressed DCFs.

---

Appendix I - Invested capital in detail

We have introduced the idea of Invested Capital. Let us explain it in a little more detail.

We want to measure the return on investment using only the money and assets that are actually working.

If you want to measure your employees' productivity on a project, you look at the output and the hours worked by employees who contributed to that project. You do not count the hours of employees on leave, substitutes who did no work, or anyone else who did not participate.

For example, if I want to measure the return a company earns by opening a new factory, I look at the money spent to open that factory. Otherwise, I would be measuring the return on something else, and we have other return metrics for that beyond ROIC.

In other words, when measuring profitability with ROIC, we want to use the money working inside the operating business. That is Invested Capital.

We can calculate it in two ways that should lead to the same place.

Two paths, same destination

On the operating side, add what the business uses to operate:

Invested Capital = operating working capital + net fixed assets + operating intangible assets

Operating working capital is receivables plus inventory minus payables, excluding surplus cash and short-term debt. Those two are financing, not operations. Fixed assets and operating intangibles are measured at net book value. The business in the main body of this post would have €100 of working capital, €350 of fixed assets, and €50 of intangible assets: the €500 that generated €100 of NOPAT.

On the financing side, look at who provided the money:

Invested Capital = book equity + financial debt − surplus cash

Here, equity means book equity, not market capitalization. We are measuring the money that went into the business, not what the market thinks that money is worth.

The general rule is: if an asset does not help generate NOPAT, exclude it from invested capital. That is why surplus cash and financial investments are excluded. Their returns are not in the ROIC numerator, so they cannot be in the denominator either.

Of course, these are approximations based on the public data companies report. If you have access to better data or inside information, I recommend making good use of it.

The goodwill question

When one company buys another, the premium paid is recorded on the balance sheet as goodwill.

Does it count as invested capital? It depends on what you want to measure.

• Including goodwill measures management: what return did it earn on all the capital it deployed, including any acquisition premium?
• Excluding goodwill measures the underlying business, regardless of whether it was acquired cheaply or expensively.

To project organic growth with the g formula, the version excluding goodwill is usually more useful. To judge management's capital allocation, use the version including goodwill.

There is a nuance here. If your business model is to be a serial acquirer, you cannot exclude goodwill from invested capital while including the acquired profits in NOPAT. Acquisitions are a recurring source of growth.

If you do, you inflate ROIC and ROIIC without anyone creating anything. Either include the full cost of acquisitions or separate acquired growth from organic growth.

Appendix II - Average ROIC, marginal ROIC, and ROIIC

We have insisted that the ROIC relevant to growth is the expected marginal ROIC. The ROIC on a financial website will not do. Neither will ROIC from past years.

Let us explain why. We will also see how to find a better approximation of the “right ROIC.”

The average looks backward; the marginal looks forward

The ROIC reported by financial websites is an average return: the year's NOPAT divided by all the capital accumulated over the years. It is a useful metric because it summarizes how good past investments were. But it is a snapshot of old capital.

Marginal ROIC answers a different question: what return will the next euro reinvested generate? Future growth depends only on those new euros. In the formula g = expected marginal ROIC × reinvestment rate, the capital already invested does not appear anywhere.

The two figures can be very different. A chain of stores with an average ROIC of 25% may already have opened its best locations. The next stores (in worse locations and facing more competition) may earn 10%. To project g, that 10% is the number that matters, not the 25% shown on the statistics page.

This matters because the best opportunities are usually taken first. The future ROIC (expected marginal ROIC) is therefore quite likely to be lower than the past ROIC.

ROIIC: a more measurable approximation

The problem with expected marginal ROIC is that it does not appear in any financial statement. It is an expectation.

Fortunately, we can measure its historical cousin: ROIIC (Return on Incremental Invested Capital).

Instead of comparing one year's profit with all accumulated capital, ROIIC compares how much new profit appeared with how much new capital was required to generate it:

ROIIC =

ΔNOPAT÷Δinvested capital

Where

ROIIC = return on incremental invested capital

additional operating profit generated by each additional euro invested during the period

ΔNOPAT = change in NOPAT

NOPAT in the latest year minus NOPAT at the beginning of the period

Δinvested capital = change in invested capital

invested capital at the end of the period minus invested capital at the beginning

For example, if NOPAT rose from €100 to €130 over the last three years and invested capital increased by €150, ROIIC is €30 / €150 = 20%. Each new euro invested over that period generated 20 cents of additional annual operating profit. That 20% looks much more like marginal ROIC than any historical average.

It is still a historical metric and therefore not perfect. But it lets us see the trend in new investments. We can determine whether the latest investments are earning more or less than the historical average, which may provide a useful clue about how new investments will perform in the near term.

The fine print on ROIIC

Before you fall in love with the metric, three warnings:

• It is noisy from year to year. Today's capex takes time to produce profits, and a single period can produce absurd figures, including negative values for healthy businesses. That is why ROIIC is calculated over three- to five-year windows, preferably using several different windows.
• Not all ΔNOPAT comes from new capital. If margins improve because of efficiency, pricing, or the cycle, profit grows without additional investment and ROIIC becomes inflated. It is worth asking how much of the improvement comes from reinvestment and how much from margins. Acquisitions, divestitures, and currency movements have the same effect: they move ΔNOPAT and Δinvested capital without telling us anything about the return on new capital.
• It still looks backward. ROIIC measures the return on incremental capital that has already been invested. It is the best available clue about marginal returns, but it remains a clue. You must then ask whether future investments resemble past ones. If the company expands into worse businesses, the past is a poor prophet.

In short: use average ROIC to judge the quality of the existing business and ROIIC to estimate the return at which the company is deploying new capital. Is it sustainable? That is up to your judgment. The final figure (the expected marginal ROIC) is the one that enters the growth formula.

Appendix III - The proof with the derivative

In the main body, we said that more reinvestment creates value only if expected marginal ROIC exceeds WACC. Here is the proof.

We start with the value formula using NOPAT and marginal ROIC, then divide both sides by NOPAT₁. That removes the size of the business and leaves us with the multiple. The math works the same whether the business earns €1 or €1,000,000.

EV₀/NOPAT₁ =

1 − reinvestment rate÷WACC − marginal ROIC × reinvestment rate

The idea is simple, even if the formula is about to look complicated.

We want to observe how increasing or decreasing the reinvestment rate affects the EV₀/NOPAT₁ multiple. If the multiple rises when reinvestment rises, the business is worth more and we are creating value. If it falls, we are destroying value.

Mathematically, that means looking at the effect of a small increase in the reinvestment rate. The cleanest way to see it is with the derivative:

∂(EV₀/NOPAT₁)÷∂reinvestment rate

=

mROIC − WACC÷(WACC − mROIC × reinvestment rate)²

This formula tells us how an increase in the reinvestment rate affects the multiple while WACC and marginal ROIC remain constant.

We care less about the derivative's magnitude than its sign:

• If it is positive, the multiple rises: value is created.
• If it is negative, the multiple falls: value is destroyed.

The denominator is squared, and any squared number is positive. That means the sign of the derivative depends only on the numerator:

marginal ROIC − WACC

If marginal ROIC exceeds WACC, the derivative is positive and more reinvestment creates value. If they are equal, it is zero. If marginal ROIC is lower, the derivative is negative. Exactly the rule from the main body.