How Do You Know When a Stock is Expensive?

Get smarter on investing, business, and personal finance in 5 minutes.

This week's newsletter is an adaptation from my recent YouTube video on How Do You Know When a Stock is Expensive.

How do you know if a stock is expensive?

What if it's a stock like Palantir that trades at 71x sales?

A lot of times investors think that there's no way we can possibly figure out how to value a business like that because it's a growth business and earnings are growing so quickly, and so you can't really use traditional valuation methods in order to understand a company like Palantir.

Don't let people lie to you about that.

That is not true, and that is actually one of the most common debates I get in with all sorts of so-called “growth investors.”

I'm putting growth investors in quotes here because basically all investing you've ever done in your life required some amount of growth in order for the investment to work.

We typically call growth investors those that require a lot of growth.

But that doesn't mean that we should loosen the standards of valuation and have no idea as to how to value these businesses just because they are growing so quickly.

Instead, what I'm going to walk you through in this newsletter is exactly how you can value a business like Palantir, a very high growth business.

It is a very simple way to use multiples to see what is currently priced into the stock today.

The stock market is a discounting machine: every company has a stream of cash flows, and what you are paying for as an investor is this stream of cash flows discounted back to today.

The market is pretty efficient at roughly figuring out how much it should discount these future cash flows back to today.

So as an investor, you basically need to take the opinion that a stock is going to grow more or do something different than what the market is going to expect it to do in order for you to get an outsized return.

So let us get into it using Palantir as a sort of case study.

The Simple Multiple Method.

Now, this is going to be kind of the rule of thumb way of doing this with multiples.

It's a very easy way to approximately figure out what is priced into the stock.

And right now, I can already see some sort of objections to this idea that, “Well, the future is very unknowable, and how could you really know how many years Palantir is going to grow at a very high growth rate for?”

And what I would tell these investors is that you don't.

And so, you have the decision of not investing in that business.

But if you do invest in that business, whether you make these assumptions explicit or not does not matter.

Either way, they are already implied in the price you're paying for a stock.

Right now, Palantir trades at $172 a share, which gets them a diluted market cap of around $440 billion.

That is against current trailing 12-month sales of $6.2 billion.

And that is a sales multiple of about 70x.

Or instead, if you prefer to look at it on earnings, that's about 145x trailing earnings.

Whenever you encounter a very high multiple, you sort of intuitively know that that means the company has to grow a lot into the future in order to rationalize that multiple.

So, what does that mean earnings need to do in the future?

Not to just make money as an investor today, but in order to support the current valuation.

They're making assumptions on future cash flows, and whether they're doing this through a DCF or a multiple, they're discounting those cash flows back to today to get a stock price.

And a multiple is a shorthand for a DCF.

A lot of times people bash the DCF as being too technical.

And I'll admit, I don't that often do a DCF compared to how often I use multiples.

But you don't have a right to use a multiple unless you understand what you're really doing with it.

Why are you paying 145x earnings for a company like Palantir?

You're doing that because you believe earnings are going to grow a lot more into the future, and that is going to support the current valuation.

A DCF handles the question of how much growth you need very well.

But let's get into the math to figure out what's priced in for Palantir.

We're going to take Palantir as an example, but you could do this for any business.

Why All Stocks Eventually Trade at 10x Earnings.

So first, all you have to do is you just start with the current market cap of the company.

Now, with that market cap, you're getting this very large multiple, earnings multiple, 145x.

Most stocks, over time they go ex-growth and they're no longer growing, they trade at 10x.

In fact, it's kind of a pretty well-known quip that eventually all stocks trade at 10x.

Why 10x?

Well, when a stock is fully ex-growth, which means that it's no longer growing at all, it should basically return the discount rate.

So, a 10x multiple, if you invert that, that's basically going to be 1/10, that's a 10% earnings yield.

And if you're not growing at all, that's your return.

The only return you make as an investor in this scenario is what the current earnings of the business are returned back to you.

And even if the company is growing 1%, 2%, 3% at that point, that's basically what inflation is, so it still is basically a 10% return in real terms.

So, all stocks eventually will trade at 10x earnings when they are done growing.

For some companies, this future may be 30 years away, 50 years away.

For other companies, maybe investors are worried that it's very imminent.

We covered Adobe stock on this channel, and I think one of the reasons why it traded at such a low multiple was because investors worried that a lot of the growth opportunities would dry up, and growth would continue to decelerate, which would eventually leave you with what people call a melting ice cube.

Now, that's not my opinion on Adobe, I'm just trying to help you understand why a stock will trade at 10-12x earnings.

If you understand that this is the destiny for all businesses, eventually, when they stop growing, then we can start to build a better foundation for understanding what multiple to pick for a stock.

Now, the problem is very few stocks ever actually go fully ex-growth.

They usually are growing a little bit, maybe a mid to high single digit growth rate, and they can maintain that for a very long period of time.

That presents a little bit of a problem for the math we are trying to do here, because if I'm going 50 years out in order to put a 10x multiple on this business, then it really doesn't work that well.

It's too far into the future; you're not really going to be comfortable with those assumptions.

So instead, we take a little bit of a middle ground, where right now, we know that the historical average multiple of the S&P 500 is about 17x.

If you invert that, (1/17), you get a 5.8% earnings yield.

If you bought $100 worth of a stock, $5.80 of that is going to be in earnings.

Most companies do a mix of retaining some of those earnings and paying out the remainder.

The amount they pay out, we call a dividend.

If you're looking at an average sort of S&P 500 dividend, it could be about 2%.

Of that 5.8% earnings yield, 2% is being paid out as a dividend, a payout ratio of roughly 35%.

The remainder of that 5.8% earnings yield gets retained by the business, and they use that to reinvest back into the business to continue to grow.

So you add the mid to high single-digit growth rate to the dividend yield, growth plus dividend yield, and that's a rough approximation of what your return is going to be over the long term.

Now you're starting to get a little bit of an understanding: 10x is a multiple for a company without any growth, and then we move up to this market historical average 17x multiple for a company that's growing mid to high single digit growth.

Picking a Multiple for Palantir.

So now what if we want to assume a little bit more growth than that?

This is going to really be key to this entire discussion of understanding what is currently priced into the stock of Palantir.

There's going to be a range, and I'm not here to tell you that I know the right answer to this, but what I'm going to pick in this math is 25x.

That's already a premium multiple to what the S&P 500 trades at, and if you are trading at 25x, investor expectations at that point are high single digit to low double digit growth rates.

So what we're going to do when we're doing this Palantir math is basically try to figure out how much earnings growth is currently being priced into the Palantir stock to get to the point that it's now in this lower growth stage, where it's only warranted to get a 25x multiple.

They're going to keep growing, 80%, 70%, whatever percent it is, for a period of time, and then eventually growth is going to slow.

This is going to happen inevitably, especially as companies get bigger and bigger: in dollar terms, growing 10% on $80 billion, that's $8 billion, but right now with them at $6 billion, 10% is only $600 million.

Instead, we want to pick a multiple where we're not having to go out that far into the future.

As an investor, you can have more of a sense of growth over the next 3 to 5 years than over the next 50.

If you ever pull up a Wall Street model, they're never going out that far; they're usually going out 3 to 5 years.

So now at a 25x multiple, it's still a premium multiple for a stock.

It just assumes it's past this era of hypergrowth.

The Math Behind What's Priced In.

Step One: Divide by the Multiple.

As we mentioned, Palantir has a market cap of $440bn.

The first step of this math is you just divide by the multiple we picked, 25x: $440 billion divided by 25x gets you earnings of $17.6 billion ($440bn/25 = $17.6bn).

I am basically making the claim that if you buy the stock today, you need earnings of at least $17.6 billion in order to support the entire valuation of Palantir.

If you were going to argue with me and say Palantir deserves a much higher multiple because it's going to grow so much more for so much longer, that is fine, but save that for later on.

Right now, what we're trying to figure out is what a more conservative investor would be comfortable assuming, and a 25x multiple is more conservative.

Step Two: Gross Up for Taxes.

The next steps are pretty straightforward.

From $17.6 billion, we're going to get it to a pre-tax number by dividing by one minus the tax rate.

Assume the tax rate's 20%, so you divide by 0.8 (1-0.20 = 0.8).

That's going to get us to $22 billion in pre-tax earnings ($17.6/0.8 = $22bn).

Step Three: Back Into Revenue Using a Margin.

The next step is I want to get up to revenue, so we're going to have to assume a margin.

Right now, Palantir has a margin of around ~40% percent.

But they've had really good operating leverage, so I'm going to give them the benefit of the doubt and say they're going to get to a 60% operating margin.

With that 60% operating margin, you take that pre-tax earnings of $22 billion and divide by it, and now you get revenue of $37 billion (22bn/0.6 = $37bn).

I am now going to make the claim that if you are an investor in Palantir today, $37 billion of revenue is already priced into the stock.

Last 12 months, they did about $6.2 billion in revenue.

Revenues have to 6x to $37 billion in order for you to just support the current price you're paying for Palantir.

Because remember, that gets us to just a 25x multiple.

Now, as an investor, if you want to actually make a lot of money in Palantir from here, you're going to have to assume revenues in the future are much higher than $37 billion.

Adding a Timeframe.

Now, we could go a step further and break this down into more assumptions by adding a timeframe.

Let’s assume it will take Palantir 5 years' time for them to reach $37 billion in sales.

So if it takes them 5 years to go from $6 to $37 billion in revenue, that's going to be a growth CAGR of low 40% range, at which point we're now saying they're not going to be growing as quickly, so they get that 25x multiple.

So now we're already going 5 years out into the future.

Where you make money as an investor in Palantir is what happens beyond these 5 years.

They're saying at the end of 5 years, it's still going to be growing very, very quickly, and so 25x is the wrong multiple because we need a higher multiple that assumes even higher growth after this 5-year period.

Instead, maybe it's still growing 30%, and so it deserves an even higher multiple at that point.

Here's how you do that math: right now, the stock market has a multiple of about 21x.

Let's say you're 5 years out into the future, and you're saying, I think that even once Palantir hits $37 billion in revenue, they're going to still grow 30% a year for at least 3 more years.

So, what multiple should I put on it?

Well, one way is you can take what that market multiple I just said was, and you could grow it out by that 30% growth for 3 years, and you'd get 46x.

If we invert it, you're taking 46x, dividing by 30% growth for 3 years, and that will get you to 21x.

So, the return in this scenario would be 46x divided by 25x, which is 84% upside.

And so another way to kind of frame this is saying that the market is only pricing in the next 5 years of growth, whereas this investor is looking at the next 8 years of growth and saying that this period of high growth is going to last even longer than what the market is currently expecting.

A Multiple is a Shorthand for a DCF.

Now, with this math of how we're trying to figure out what is priced in, you have a lot of leeway as an investor in terms of what was the right multiple to pick.

Maybe some people would say, I don't want to do 25x, I want to do 20x.

Maybe other people would say, let's go with 35x and figure out how much growth needs to happen, then assume a 35x multiple on that exit year, and look at the next period of growth and do the similar sort of math.

There's lots of different ways that you could run this math, but it's a little confusing if you're trying to do this with multiples.

And I think this is one of the reasons why growth investors very often say you can't do this sort of math with a multiple.

It just doesn't work, so one, that's not true, I just showed you how to do it, but it is confusing.

But what if there's a better way to do this to be more explicit about all of the periods of growth, instead of just arbitrarily picking one growth rate?

What if Palantir can go 50%?

This is now getting at what a reverse DCF is supposed to do.

A discounted cash flow model allows you to model out future cash flows and discount them back to today by your discount rate.

A lot of times, people have trouble saying, "What is the correct discount rate?"

And they're right to question that, because the discount rate you pick is essentially the valuation of the business: you're really arguing that this is the right return an investor should get.

So if you pick, say, a 10% discount rate, and the stock price you're getting is below the current stock price, what you're basically saying is the stock should move up in price enough so that it's priced to only get a 10% return into the future.

That's basically the logic behind the DCF.

A lot of people pick 10% in no small part because Warren Buffett picked 10%; he says he looked for a 10% pre-tax return, and it's kind of in line with what the stock market has historically returned.

This is a fine way to do it, but it's not how I do it.

I prefer the reverse DCF because with it, you don't actually assume a discount rate; instead, the discount rate is going to be the output of the model.

What you assume is growth rates and margins: what do I think the company's going to grow revenue on, and what do I think margins are going to be in the future?

I find it's easier for most investors to have an opinion on that than on what the right discount rate is.

If you do it that way, then your output is the discount rate, and you can decide if that return is high enough for you.

But I'll hand it to you, multiples are very convenient, and DCFs are kind of unwieldy.

I myself don't always do a DCF, I do multiples very often.

The reason why the two work is because at its core, using multiples is a discounted cash flow model.

All the assumptions people think make a DCF too complicated still exist when you are doing a multiple.

If you are paying a multiple of earnings, like Palantir at 145x earnings, you have a lot of assumptions already embedded in there.

If you go to a DCF to see what those assumptions are, all you are doing is becoming aware of what the assumptions are.

You were already assuming them in the multiple you were paying, you just weren't aware of it.

Maybe there are some Palantir investors looking at that going, “Wait, there's at least $37 billion in revenue that I'm paying for today, and I need it to do even better than that under this math?”

You need a lot of earnings in the future for a stock to support an almost half a trillion dollar valuation.

But I think too often people think it's so complicated and hard to do that it's not worth doing, and instead they take a shortcut, kind of unaware they're basically doing the same thing as the DCF.

The Margin Assumption.

Now, the one other assumption I kind of glossed over was the margin assumption.

That should be tied to the multiple you're picking: if you're picking a relatively low, more mature multiple, like a 25x multiple, then that should assume the company is not investing as much for growth and is already nearing a mature margin profile, so your margin should be a little bit higher.

If you disagreed with the way I was doing my math and wanted to pick 35x or 40x multiple, maybe you want to lower that margin a bit, because in order for a company to support a 40x multiple, it still has to be growing really quickly.

And if it's growing really quickly, in all likelihood, they're still investing a good amount through the P&L, so that's going to depress what operating margins are going to be.

And there's a lot more I could say about all of this, and I will in the future.

But the key takeaway I want you to get from this is that a multiple is a shorthand for a DCF and has embedded assumptions with in that.

But whether you make these assumptions explicit or not does not matter.

Either way, they are already implied in the price you're paying for a stock.

For more on How Do You Know When a Stock is Expensive, check out the video below.

YouTube FAQ

Nothing in this newsletter is investment advice nor should be construed as such. Contributors to the newsletter may own securities discussed. Furthermore, accounts contributors advise on may also have positions in companies discussed. Please see our full disclaimers here.