Multiples Are DCFs for Lazy People

In the previous post we valued the simplest business in the world: a lemonade stand.

We said that the value of the stand depends on the cash it can generate for its owner, adjusted for the risk we take, the time we wait, and the other opportunities we give up.

Very deep. And also very annoying.

Because in the real world, investors do not usually talk like that. They do not say:

This business is worth the present value of all its future cash flows discounted at an appropriate required return.

(only a few nerds do)

They usually say things like: "it trades at 15 times earnings", "it's cheap at 6 times EBITDA", "I would never pay 10 times sales", or "the market has compressed its multiple".

Wonderful and sophisticated. Very Wall Street.

But what does all of that actually mean? How does it relate to the valuation formula we saw in the previous post?

Today we close that circle.

The value of a stock is determined using a DCF, that is, a discounted cash flow. And, before we start, one thing needs to be made clear:

They are compressed DCFs. The same thing, but more manageable and with the assumptions hidden.

Or, to be more precise, a justified multiple is a compressed DCF.

Why do I say "justified"? Because sometimes people use multiples that are not connected to the discounted cash flow formula. They are not shortcuts, they are references. They can be more or less useful.

For example, I could say that company X trades at "30 times the square meters of its factory".

If you can connect those square meters to its cash flows, maybe there is something there. Without that connection, no. It is a "reference" multiple, not a real valuation multiple. Later we will see a very famous example: the PEG.

It is true that most of the market uses multiples only to compare prices between similar companies, without thinking about flows or rates. That use is also legitimate, but notice that it is a different method: instead of estimating flows, you assume the market is pricing the comparables correctly. This post is about something else: what the multiple means underneath.

In short, multiples are shortcuts. Sometimes very useful ones. But shortcuts nonetheless. And shortcuts are dangerous when you forget which road they are cutting short.

One last clarification before we start: in this post we will almost always be on the shareholder's side. The flows will be shareholder cash and the price will be the market cap, the market value of the shareholders' slice. There is another family of multiples that looks at the whole business using enterprise value, but we will cover that distinction calmly at the end.

Simplifying the formula

Let's start at the beginning. Recall the full discounted cash flow formula:

V=n∑

t = 1

FCF_t

(1 + r)^t

Where

V = value

What the thing we are valuing is worth today.

FCF_t = free cash flow in year t

The real money we expect to receive, or expect the business to generate, in each year.

r = discount rate

The minimum return we demand for waiting, taking risk, and giving up other opportunities.

t = specific year

The further away the cash flow is, the less it is worth today.

n = estimated life

The number of years during which we expect to receive or generate those flows.

In essence, it adds this up, with each year discounted to present value:

FCF year 1 / (1+r) + FCF year 2 / (1+r)² + FCF year 3 / (1+r)³...

But to understand where multiples come from, we do not need to project year by year. A very simplified version is enough.

Imagine a business generates $10,000 of free cash flow next year. After that, the cash grows at a stable rate forever.

The formula would be:

V = FCF1 / (r − g)

Where

V = value

the value that corresponds to the flow being used: equity value if the flow belongs to shareholders, enterprise value if the flow belongs to the whole business

FCF1 = FCF1

free cash flow expected for next year

r = required return

the minimum return we demand for taking the risk of this business

g = growth

the rate at which we expect the cash to grow over the long term

Do not panic. It looks complicated, but it is the same idea as before. The business is worth the cash it will give us in the future, adjusted for time, risk, and opportunity cost.

The formula is a simplified model. Very simplified. Instead of having to put a little number on every future year, we take next year's cash flow and assume constant growth.

But it is useful because it shows the three forces that drive value: the expected cash, the return we demand, and the growth.

It is interesting to notice that many fundamental investing strategies are different ways of obsessing over one specific part of the formula.

Pure value looks mostly at FCF1. It wants cash today relative to the price paid.

Quality or defensive looks mostly at r. If the business is more stable and less fragile, the required return should be lower.

Growth looks mostly at g. It wants duration. The question is how much the flows can grow and for how long.

Multiples compress three questions: how much cash is there?, how much can it grow?, and what return do we demand for taking the risk?

That is all. You can disguise them and you can rename them; but they are always there.

That is the skeleton.

The condition you cannot break

This formula only works if the required return is greater than the perpetual growth rate.

In symbols: r has to be greater than g.

r > g

The reason is simple: the denominator is r − g. If that denominator hits zero or goes negative, the formula stops representing a business with finite value.

Also, r and g have to speak the same language: both nominal or both real, same currency, same type of flow, and same level of risk.

• When r is greater than g, all good.

Growth pushes the cash upward, but the discounting is still stronger. Distant flows weigh less and less and the value makes sense.

• If r = g, the denominator is zero.

That implies infinite value. It makes no economic sense.

• And if g is greater than r, the denominator turns negative.

That does not mean the business is worth a negative amount. It means you are using the formula when you should not.

A company can grow faster than r for a few years. Of course it can. What you cannot assume is that it will grow faster than r forever. In a perpetuity, long-term growth has to be conservative and lower than the discount rate.

Think of r as gravity and g as the engine. The engine can push hard during a phase, but if you say it pushes harder than gravity forever, the value shoots off to infinity. And the formula breaks.

There is a second common-sense bound: over the very long run, no company can grow faster than the economy it lives in forever. If you assume a perpetual g above nominal GDP growth, you are saying the company will eventually become bigger than the economy itself. In practice, the g of a perpetuity sits near the economy's growth rate or below it.

And a warning that every multiple in this post will inherit: the denominator r − g is a small subtraction between two numbers nobody knows with precision. With r = 8% and g = 4%, the justified multiple is 25x. Raise g to 5% and it jumps to 33x. A single percentage point in an estimate moves the "fair value" by 33%. Justified multiples do not eliminate the fragility of the DCF: they concentrate it.

And let nobody laugh at the multiple from the DCF bleachers. In most discounted cash flow models, most of the value sits in the terminal value, which is calculated with this very same perpetuity or, directly, with an exit multiple. The DCF and the multiple lean on the same two numbers nobody knows with precision. The difference is how much paper you use to hide it.

Now we can take this logic from dividends toward cash flow, EPS, sales, and other numbers investors are more familiar with.

From the dividend to P/FCF

Let's start with the cash.

Cash that reaches, or could reach, the shareholder.

But first we need to remember something important: as owners, we do not care about accounting profit in the abstract. We care about collectible cash. We want to make money for ourselves.

The cleanest representation of that cash is the dividend. It is not a promise. It is money that leaves the company and reaches the shareholder.

So, if we are valuing the shareholders' slice, we can rewrite the formula using dividends per share:

P = D1 / (r − g)

Where

P = price per share

the value of the slice of the business that corresponds to one share

D1 = expected dividend

the dividend per share expected for next year

r = required return

the minimum return we demand for taking the risk of this stock

g = growth

the rate at which we expect the dividend to grow over the long term

We divide by the dividend. The first multiple appears:

P / D1 = 1 / (r − g)

This multiple is the odd cousin of the famous dividend yield. More precisely, here we are talking about the forward dividend yield, because we use next year's expected dividend:

Forward dividend yield = D1 / P = r − g

An important nuance before moving on, because it applies to every multiple in this post. These formulas say what the multiple should be given r and g. That is the justified multiple we talked about at the beginning. The multiple you see trading in the market is something else: it comes from the market price, not from your assumptions. Read through this formula, it tells you what combination of cash, growth, and risk the market is assuming at that price. Comparing the justified multiple with the traded one is, precisely, the game of valuation.

But the dividend has a problem: a company can generate cash and not distribute it as a dividend. It can return it through other channels, like share buybacks. Or it can pay down debt, accumulate cash, and reinvest for years before sending a single dollar to the shareholder.

Back to the lemonade stand. After paying for lemons, cups, taxes, and all the year's reinvestment, it has $10,000 of free cash left. But it pays only $4,000 to the owner and leaves the other $6,000 piling up in the business's account.

The dividend looks at the $4,000. But we may want to look at the $10,000, as long as those retained $6,000 still truly belong to the shareholder and do not get wasted.

Here comes the P/FCF: price over free cash flow to the shareholder.

To be precise, here I use FCF in the equity-side sense: FCFE. If we used the cash of the whole business, before debt, the correct numerator would be enterprise value, not market cap. We will cover that distinction calmly at the end of the post.

If we use FCFE per share, the formula looks like this:

P = FCFE1 / (r − g)

Where

FCFE1 = expected FCFE

free cash flow to equity expected per share for next year

We divide again. This time, by that free cash flow:

P/FCFE = P / FCFE1 = 1 / (r − g)

In that sense, P/FCF is more direct than the P/E. It is closer to what really matters: cash available to the owner.

The key assumption is that this FCFE is representative.

And there is one condition you cannot skip: for the formula to work, the FCFE has to be measured after all reinvestment, including the reinvestment that funds the growth g. If your "FCF" only subtracts maintenance capex and you still assume growth, you are assuming that growing is free. And we will see that it is not.

And a warning: FCFE can be dressed up with debt. Issuing debt inflates the cash reaching the shareholder today at the expense of tomorrow's, so a low P/FCFE can be leverage instead of a bargain.

One practical problem is that many websites call very different things FCF. It can be cash from operations minus capex, be after interest, ignore debt repayments, or not separate maintenance from growth.

That is why P/FCF can be a better option than the P/E, but it also demands that you check exactly which FCF you are using.

From P/FCF to P/E

But almost nobody talks only about price to dividend or price to free cash. The market usually talks about price to earnings: the PE Ratio (PER).

To get there we need a bridge between dividends and earnings:

D1 = EPS1 × payout

EPS1 = expected earnings per share

the accounting profit expected per share for next year

payout = payout ratio

percentage of earnings the company distributes as a dividend

Substituting the dividend with earnings and payout, we first arrive at this formula:

P = (EPS1 × payout) / (r − g)

Now we divide both sides by earnings per share. The full version of the P/E would be:

P/E = payout / (r − g)

Now we make two strong simplifications:

• Assumption 1: we assume all free cash flow is distributed as a dividend.
• Assumption 2: we treat earnings per share as an acceptable approximation of that cash.

They are shortcuts. Useful, yes. But notice the trap we just accepted: a 100% payout and, at the same time, a growth rate g. In other words, we are assuming that growing is free. It is not, and we will come back to this in a moment. For now, keep your eyes open.

With a 100% payout:

D1 ≈ EPS1 and payout = 1

And it becomes:

P/E = 1 / (r − g)

That means the P/E is not a made-up multiple. It comes from the original discounted cash flow formula, as long as you accept the two assumptions above.

One practical detail before moving on: this is a forward P/E, because it uses next year's expected earnings. The P/E you see on most websites is trailing: price over the last twelve months of earnings. The idea is the same, but do not mix the two when comparing.

Then come the two variables that move the multiple:

• Variable 1: the expected rate of return r, which summarizes risk and opportunity cost.
• Variable 2: the expected growth g, which pushes future flows upward.

This is what people forget when they do not know where the PE comes from.

• A business that will never return its cash to the shareholder (no dividends, no buybacks, no future sale) deserves a lower multiple. Even 0.
• A business whose free cash flow is lower than its accounting profit deserves a lower multiple, if that gap is not funding profitable growth.
• A safe business (lower r) deserves a higher multiple.
• A business with growth (higher g) deserves a higher multiple.

So we cannot say "a P/E of 30 is expensive". It depends on the business.

That said, for companies that are similar in cash, risk, and growth, the multiple can be very useful for comparison.

The condition is that the data be representative. You cannot use the earnings of an extraordinarily good year, because that does not represent the business over the long run.

Earnings yield: the P/E flipped upside down

Another way of looking at exactly the same idea is to flip the P/E upside down.

• Instead of asking, how many times earnings am I paying?
• We say, what percentage of earnings am I buying for each dollar invested?

That is the earnings yield:

Earnings yield = EPS1 / P = 1 / P/E

EY = earnings yield

earnings per share divided by price per share; the inverse of the P/E

If a company trades at 20x earnings:

Earnings yield = 1 / 20x = 5%

This is useful because it turns the multiple into something comparable with other returns. For example, if a government bond pays 4% and a stock has an earnings yield of 5%, at least we can start a conversation:

Earnings yield = 5% vs. bond yield = 4%

But careful! They are not the same thing.

A bond's coupon is usually a contractual obligation. A company's earnings are not. They can fall, they can be cyclical, they can fail to convert into cash, they can require reinvestment, or they can never reach the shareholder at all.

There is a more subtle trap: the bond yield is nominal, while a company's earnings tend to grow with inflation. Remember the earlier rule: to compare, same language.

The earnings yield is not a guaranteed return. It is an implied accounting return.

It is there to ask yourself "am I getting enough earnings for the risk I am taking compared to a bond?"

The comparison with bonds is useful, but incomplete. A stock needs to compensate you for more uncertainty, more volatility, and more risk. In exchange, it can also offer something the bond usually does not: growth.

Growth is not free

So far we have talked about g as if it were a simple variable: more growth, more value.

And yes, in the formula that is what happens. If g goes up and everything else stays the same, the denominator r − g gets smaller and the multiple goes up.

But in a real business, growth costs money.

It does not just appear.

Many companies can only grow by reinvesting cash into concrete things: opening stores, hiring people, buying inventory, developing product, financing customers, or acquiring other businesses.

That cash does not arrive today.

In practice, g and payout go together. Paying out more cash today can mean growing less tomorrow. Retaining more cash can help you grow, but it leaves less immediate money for the shareholder.

The simplified relationship is this:

g = ROE × retention

ROE = return on equity

profit generated per dollar of book equity

retention = retention

percentage of earnings the company does not distribute and reinvests in the business

Notice one detail. This relationship breaks the 100% payout assumption we used to simplify the P/E. If you distribute all your earnings, you retain nothing. And the formula is clear: without retention, there is no growth through reinvestment. At most, a company that pays everything out can grow more or less with inflation if it has pricing power, but not much more. That is why the earlier version was a shortcut, not an exact description of any real business.

The formula is not perfect, but the intuition is good.

This separates good growth from bad growth.

• Good growth: the company reinvests at high returns, increases future cash, and creates value.
• Bad growth: the company grows sales or earnings, but consumes so much capital that it barely leaves any cash for the shareholder.

It is not enough to say "this company grows a lot". You have to ask:

"How much capital does it need to grow, and what return does it earn on that capital?"

In reality, the topic of returns and growth is a bit more complicated, but we will leave that for another day.

The idea you need to keep is: to grow you need to invest, and the money you invest cannot be distributed to the shareholder.

And that ROE we just met gives us the perfect bridge to the next multiple.

From P/E to P/B

The P/B, or price to book value, is another of the oldest and most widely used multiples.

Book value per share tries to measure the accounting capital that belongs to each share. It is not perfect, but in some businesses (banks, insurers, and very asset-intensive companies) it can be a useful reference.

Also, book value has the appeal of fluctuating less than accounting profit, which makes the P/B a more stable multiple over time.

The bridge between earnings and book value is the ROE:

EPS1 = BVPS × ROE

BVPS = book value per share

book value of equity divided by the number of shares

ROE = return on equity

profit generated per dollar of book equity

If before we had:

P = (EPS1 × payout) / (r − g)

we can substitute EPS1 with BVPS × ROE:

P = (BVPS × ROE × payout) / (r − g)

Now we divide both sides by book value per share:

P/B = (ROE × payout) / (r − g)

The P/B is the same logic with one more piece. Instead of starting only from earnings, we connect those earnings with the accounting capital that supposedly generates them.

And watch out: the simplifications we made to get to the P/E do not disappear here. They accumulate.

The assumptions:

• Assumption 1 (repeated): the cash that matters can be approximated with dividends or payout.
• Assumption 2 (repeated): earnings per share are a good approximation of that cash.
• Assumption 3: book value reasonably represents the economic capital of the business.
• Assumption 4: the ROE is sustainable, not the result of an odd year, excess leverage, or aggressive accounting.

The variables also multiply:

• Variable 1: return on equity ROE, which connects book value with earnings.
• Variable 2: the payout, which determines how much cash reaches the shareholder.
• Variable 3 (repeated): expected rate of return r, same as in the P/E.
• Variable 4 (repeated): expected growth g, also inherited from the base formula.

There is an even more compact form if we assume growth comes from reinvesting retained earnings.

We recover the sustainable growth relationship from the previous section: g = ROE × retention.

And since payout = 1 − retention, the P/B can also be written like this:

P/B = (ROE − g) / (r − g)

The point is that a company deserves to trade above book value if it generates returns on its equity above the return we demand. Sounds logical.

If the ROE is high and sustainable, the P/B can be high. If the ROE is low, mediocre, or artificial, a high P/B should scare you.

So the P/B does not mean much on its own. A bank at 0.8x book can be cheap if it earns well on its equity and has a healthy balance sheet. Or it can just look cheap, because that book value is not worth what the balance sheet says, or because the business does not generate sufficient returns on it.

From P/E to P/S: sales are not cash

Let's go down one more step: the P/S, or price to sales.

Sales are real. Without sales there is no business. But sales are not yet cash for the owner.

Two lemonade stands can sell $100,000 a year. One earns $30,000 after lemons, cups, taxes, and maintenance. The other earns $3,000.

Same sales. Very different businesses.

Between a sale and a dollar of value for the shareholder, many things happen: margins, taxes, reinvestment, debt, dilution, working capital, and cash conversion.

P/S has no direct valuation reading. For it to make sense, we need to convert sales into earnings or cash.

The simplest bridge is the margin:

EPS1 = Sales1 × margin

Sales1 = expected sales per share

expected sales for next year divided by the number of shares

margin = net margin

percentage of sales that ends up becoming profit attributable to the shareholder

If we force the equity-side derivation, we can start from the same price formula we used for the P/E:

P = (EPS1 × payout) / (r − g)

And substitute EPS1 with Sales1 × margin:

P = (Sales1 × margin × payout) / (r − g)

Now we divide both sides by sales per share:

P/S = (margin × payout) / (r − g)

Mathematically it works. But look at what we just did: P/S inherits the payout, the risk, and the growth from the P/E, and on top of that it adds a new variable: the margin.

That is: P/S does not value sales directly. First it has to convert sales into earnings or cash. Then it asks the same questions as always: payout, risk, and growth.

The assumptions pile up again:

• Assumption 1 (repeated): the cash that matters can be approximated with dividends or payout.
• Assumption 2 (repeated): earnings per share are a good approximation of that cash.
• Assumption 3: sales convert into earnings at a sustainable margin.

The relevant variables are four:

• Variable 1: the margin, which converts sales into earnings.
• Variable 2: the payout, which connects those earnings with cash for the shareholder.
• Variable 3 (repeated): expected rate of return r, same as in the P/E.
• Variable 4 (repeated): expected growth g, also inherited from the base formula.

This is why comparing companies only on P/S can be dangerous. Two companies can sell the same and deserve completely different multiples if one converts those sales into a lot of cash and the other does not.

A business with a high margin, good cash conversion, low risk, and durable growth can deserve a high P/S. A business with a low margin, heavy reinvestment, losses, or dilution can look cheap at 1x sales and still be expensive.

This teaches us another important intuition:

The P/E already requires several assumptions. And the P/S goes one step further because it adds even more variables and assumptions.

That does not make it useless. It makes it less direct and more dangerous.

P/S can work for early-stage businesses, companies temporarily without earnings, or quick comparisons within the same sector. But it is not a tool for valuing a business with precision. It is more useful as a reference and as a comparison between similar businesses.

PEG: a shortcut on top of a shortcut

We still have the example we promised at the beginning: the PEG, the most famous "reference" multiple of them all.

The PEG is not a clean derivation like the P/E. It is a shortcut on top of a shortcut.

PEG stands for price/earnings to growth. That is: P/E divided by a growth rate.

PEG = P/E / expected near-term growth

PEG = price/earnings to growth

P/E divided by an expected growth rate, usually EPS growth for next year or the next few years

growth_near = expected near-term growth

usually expected EPS growth over the short or medium term; not the same as the perpetual growth in the formula

At first the logic seems reasonable:

"A P/E of 30x does not mean the same thing if the company grows at 3% as if it grows at 30%."

So far, so good.

But there is an important distinction here. The growth the market uses to calculate the PEG is usually expected earnings growth for next year, or for the next few years.

It is not necessarily the g from our formula.

In this article, g represents perpetual or long-term growth. That is a different thing. A company can grow EPS at 30% next year, but you cannot plug in 30% as perpetual growth without breaking the valuation.

That is why the PEG is NOT saying:

PEG = P/E / perpetual growth

But exactly what we saw in its definition: P/E divided by near-term EPS growth.

Watch the units: the market's PEG usually uses growth as a percentage number, not as a decimal.

For example, if the P/E is 30x and expected growth is 30%, the usual convention is 30 / 30 = 1. Not 30 / 0.30 = 100. That alone tells you the PEG is a practical rule for comparison, not a clean valuation formula.

By simplifying the discounted cash flow formula and translating it into the P/E, we lost the ability to distinguish between short-term growth and terminal growth. The PEG tries to "fix it" by introducing this new short-term growth variable.

And it is worth knowing its factory defects:

• Value is not linear in growth. Doubling growth does not double fair value, so dividing the P/E by growth has no theoretical basis. The famous "PEG = 1 is fair" is a rule of thumb, not a formula.
• It ignores r completely. Mind the nuance: risk is implicit in the price (and therefore in the traded P/E), but the PEG gives you no way to use it. Two companies with the same P/E and the same growth have the same PEG even if one is far riskier. The risky one deserves a lower P/E and, therefore, a lower justified PEG. By treating the same PEG as equally attractive, you erase that difference in risk.
• It punishes low growth too much. A company growing at 0% has an infinite PEG, but it is not worth zero. A stable, safe business can be a great investment with a horrible PEG.

Being a "reference" multiple does not make it useless. That is not the lesson. You have to understand what the multiple does and why it exists to know whether it makes sense and whether it is truly useful.

Maybe the lesson in one sentence is:

Some multiples estimate value more directly. Others help you compare and screen companies.

The PEG belongs more to the second group.

A common mistake: market cap or enterprise value?

And, for those who have been paying attention, a reasonable doubt may be left over from the sales section:

If sales belong to the whole operating business, why did we end up with P/S and not EV/Sales?

To answer it properly we have to talk about the most common mistake when using multiples: forgetting what exactly we are valuing.

When we discount cash flows, the result is not "the value of the company" in the abstract. It is the value that corresponds to the cash flow you are plugging into the formula.

It can be:

• Equity value / market cap: what the shareholders' slice is worth.
• Enterprise value: what the whole operating business is worth.

It all depends on whether the cash flows belong only to the shareholders, or also belong to more people.

Back to the lemonade stand. If it has no debt and no excess cash, there is no difference: the business and the shares are worth the same.

But if the stand is worth $100,000 as an operating business, has $40,000 of debt, and holds $10,000 of excess cash, things change:

$100,000 operating business − $40,000 debt + $10,000 excess cash = $70,000 for the shareholders

The operating business is worth $100,000, but the shareholders' slice is worth $70,000. At the end of the day, the bondholders also deserve their share.

The relationship is this:

Equity Value = Enterprise Value − Debt + Excess cash

Equity Value = $100,000 − $40,000 + $10,000 = $70,000

Said the other way around, if you start from market cap:

Enterprise Value = Market Cap + Debt − Excess cash

Since multiples are shortcuts of the discounted cash flow, each multiple has to use the correct numerator:

• Dividends, EPS, net income, FCFE → Market cap / Equity Value
• EBITDA, EBIT, unlevered FCF → Enterprise Value
• Sales → usually EV/Sales to compare operating businesses; P/S only if you make an equity-side reading using the net margin.

The discount rate changes too. Shareholder flows (FCFE, dividends, EPS, or net income) go with the cost of equity. Whole-business flows (unlevered FCF, EBIT, or EBITDA) go with the WACC or with the required return for the entire enterprise.

That is why we use "Market Cap / Earnings" and not "EV / Earnings". If you use accounting earnings, the interest on the debt has already been subtracted: that profit belongs only to the shareholders. EV, on the other hand, represents the value for everyone who puts up capital, shareholders and creditors together. Dividing one by the other is comparing the flow of one part against the value of the whole. It does not add up.

Remember:

• If you use shareholder cash, you compare against market cap.
• If you use business cash before debt, you compare against enterprise value.

With this we can now answer the sales question.

The short answer: our derivation of the P/S is not wrong. But it is an equity-side derivation.

We took sales all the way to EPS1 using the net margin: earnings after interest, taxes, and capital structure.

Therefore, the consistent numerator is P or market cap.

Not EV.

That version is correct from a valuation standpoint. But it is too complicated for comparing businesses.

If you are comparing businesses using a sales multiple, it is almost certainly because the company has no earnings or they are not representative. The P/S mixes operating sales with debt, cash, interest, and leverage. It forces you to make assumptions about the net margin to interpret it correctly.

And that is too much work if all you want is a quick comparison.

To compare companies using a sales multiple, the more useful version is often EV/Sales.

That way you compare apples to apples:

• the value of the whole operating business: enterprise value
• against a metric of the whole operating business: sales

And, for a quick comparison, you avoid mixing capital structure and net margin from the very first step.

In short: P/S can be consistent if you are using equity-side metrics, but EV/Sales is usually cleaner for comparing operating businesses.

Conclusion

The summary is this:

To close and sum up, the full map looks like this:

Multiple	Implied flow	Consistent numerator	Hidden variables	Main danger
P/D	dividend that reaches the shareholder	Market cap	r, g, sustainability of the payout	ignoring the cash that is retained and never paid out
P/FCFE	normalized FCFE / collectible cash	Market cap	r, g, reinvestment, debt	using the wrong or non-normalized FCF
P/E	EPS → cash via payout	Market cap	payout, r, g, accounting quality	earnings ≠ normalized cash
P/B	Book value → EPS via ROE	Market cap	ROE, payout, r, g, leverage	non-economic book value / inflated ROE
P/S	sales → margin → cash	Market cap; EV in EV/Sales	margin, cash conversion, debt/cash, g	sales without margin / mismatched multiple
PEG	no direct flow; P/E + EPS growth	n/a (derived from P/E)	EPS growth, duration, reinvestment, ROIC	expensive or unprofitable growth

Both in the first post and in this one we have been almost always on the shareholder's side. We were working with the market cap.

We have talked about dividends, FCFE, EPS, P/E, P/B, PEG, and P/S. All of those multiples try to answer a similar question: how much is the slice of the business that belongs to the shareholders worth.

And, although we have already given a few hints, sometimes we want to look at something else: not just the shareholders' slice, but the whole operating business.

That is where other multiples come in: EV/Sales, EV/EBITDA, EV/EBIT, and all the cousins that use enterprise value in the numerator.

The logic does not change. We are still doing compressed DCFs.

What changes is the flow we are valuing, the numerator we use, and the rate that makes sense to apply.

And now, when someone says the market has "compressed the multiple" of a company, you know how to translate it: the market is assuming less cash, less growth, or more risk. Nothing strange is going on. The assumptions just changed.

In the next post we switch sides: we move from market cap to enterprise value.