Someone who reads my blog sent me this email:
Hi Geoff – Great article on analysis paralysis. However, I’m curious why you calculate FCF yield the way you do…
For you, free cash flow yield = 10-year average free cash flow margin / Price-to-sales.
For just about every other source, FCF yield = free cash flow / price.
I assume you use the margin figure to smooth out growth effects over the 10-year period.
But I’m not sure why you use price to sales as your denominator rather than market price?
A lot of people have asked me this question.
You can use 10-year average free cash flow divided by market cap to get a similar average earning number.
As for using free cash flow divided by price — my problem with that is that I never use a one-year number.
The 10-year average is good. In cases like Omnicom (OMC, Financial), it will yield a similar number. Maybe 10% instead of 13% or something like that. But since it’s a trailing average and Omnicom is usually growing over time — being a little lower is always expected.
My reason for choosing the 10-year average free cash flow margin divided by price-to-sales is to focus investors on the business.
I don't think it's safe to assume that a business will simply earn what it is earning this year or even what it earned on average over the last decade.
You should always look past the earnings to what is delivering those earnings. You should always be thinking in terms similar to a DuPont analysis.
If I'm looking at a bank's earning power, I need to be thinking about their cost of deposits. If I'm looking at an insurer, I need to be thinking about their combined ratio. If I'm looking at Dreamworks (DWA, Financial), I need to be thinking about their production costs and their distributor's marketing costs. Then compare that to the potential audience size for their kind of movies.
You want to look at how the company makes money. It only makes money if it can gather deposits cheaply enough (banks). Or draw a big enough audience to justify a $300 million movie (Dreamworks). Or maintain its monopoly pricing (Microsoft).
I want to get investors thinking like businessmen.
You make money by getting a return on something. In intangible asset businesses, it's often a return on sales. The profit margin. This is usually how you value Kraft (KFT, Financial) and Omnicom and FICO (FICO, Financial) and so forth.
But not always.
I would not pay attention to Dreamworks’s price-to-sales ratio. Sales are not predictable at Dreamworks.
Dreamworks is a future project and past library business. That is where the value is.
You need to value those two things. Not sales. Revenues will vary a lot at Dreamworks. Dreamworks is different from Kraft or Omnicom or VCA Antech (WOOF, Financial).
For Dreamworks, you look at the movies they’ve already made and the movies they are going to make. The movies they are going to make — 2.5 a year from now on — can be analyzed like any capital expenditures.
The old movies they own but no longer show on the balance sheet under “Film and Inventory Costs” are more complicated. They have value. You have some nice characters from those movies. You can monetize them through television shows and toys and things like that.
So there may never be a perfectly precise formula for valuing Dreamworks. But that doesn’t mean you can’t draw a line in the sand.
For example, we can say Dreamworks is worth more than $15 a share. The only way Dreamworks is worth less than $15 a share is if future productions destroy value.
They could. That’s a real risk. It’s something you’d have to analyze.
But if we look at the balance sheet — and we keep in mind that computer animated productions from Dreamworks and Pixar have historically been worth at least what they were carried at as inventory — we see the value of DWA is already $15 a share or higher.
When we say Dreamworks is worth at least $15 a share — we are really saying their assets (intangible though they may be) will support $1.50 a share in earnings.
We’ve talked about this idea of asset-earnings equivalence before.
In a tangible asset business, you’re often looking at the return on tangible assets. This is how you would normally value railroads, utilities, airlines (if you dare), and even something like Carnival (CCL).
Return on equity is what ultimately matters.
But it’s important to start by using the return on tangible assets and then look at how much leverage you can use. If you just look at the return on equity — you’ll often be measuring leverage more than anything else.
Sometimes it's none of those things.
I used Dreamworks as an example. I would actually value Dreamworks on a project basis. It has a library and it makes new movies.
I would never look at price-to-sales at Dreamworks. Nor does book value matter — because everything older than about three years has been written off the balance sheet. So, you'd be leaving Shrek off the balance sheet if you did something like that.
What you want to do with every business — with every stock — is think like an owner. Think like a businessman. Earnings don't appear out of nowhere. Returns come from economic (not accounting) goodwill, from sales, from tangible assets, etc.
Earnings come from real exploitable things the company controls.
When you buy a share of Kraft — are you buying earnings or sales?
It sounds like a silly question.
I like to think you are buying sales. You are buying products that have a place in the minds of customers. You are buying customers habits.
Those habits may change. But those habits are your starting point for analyzing Kraft.
Why do you think Nintendo is so cheap right now?
Because people are convinced habits will change. The iPad and smart phones will destroy handheld gaming. They may be right.
So is it right to think of a share of Nintendo as being backed by earnings, free cash flow, cash, or sales?
Personally, I like to think of something like a share of Nintendo as being backed by cash plus a business. And when I look at the business, I look more at sales than anything else.
You can look at market share. Or units sold. None of that is wrong.
The key is to focus on things the business controls and can exploit.
It helps if those things are easy to analyze.
One of Warren Buffett’s best bits of advice is: “Don’t spend time on companies that don’t lend themselves to valuation”.
All of this may sound confusing.
Why look at so many different measures?
How is this better than just using average free cash flow?
It's better because it draws you into thinking about the business.
At long-term margin "x" the company is worth this. At long-term margin 1.5x its worth something else. And at 0.5x it’s worth something else.
Let's say it's a video game company and "x" is 10%. Is 15% possible? Maybe. But other than Nintendo only Activision — and this is quite recent — has free cash flow margins like 15%. Historically, you would have to say that having 15% free cash flow margins in video games is tough to do.
What about 5%?
5% is doable.
If you have good properties and you do intelligent things with them you should turn $1 of sales into 5 cents of free cash flow in the video game business.
So now we have a way of looking at video game companies. If your hurdle rate is 10% and you don’t want to pay up for growth — you start by looking for a price-to-sales ratio of 1. That’s because 10% free cash flow margins seem possible in the video game business.
What about 1.5 times sales?
That’s a different story.
A price-to-sales ratio of 1.5 looks too high. I wouldn’t want to pay 1.5 times sales for a video game company.
It might work out. You could achieve high single digit returns in a video game stock even if your starting point is paying 1.5 times sales. But there’s no margin of safety.
I want a good chance of earning 10% on my money from day one.
So I want to pay no more than the stock’s per share sales. That would be for a company I was confident in.
For a company I wasn’t confident in, I’d need a bigger margin. It would have to trade under its sales per share.
Looking at the stock this way allows you to step back from that one company and look at the whole industry. At the company’s peers. At everybody’s long-term history.
You see a wider range of outcomes.
That’s what a margin of safety is all about
It’s about succeeding even when some things go wrong.
If you buy an above average video game company at a price-to-sales ratio that would work out well for you even if the company only had industry average margins from now on – you have a margin of safety.
You’re paying an average price for an above average company.
That extra quality is your margin of safety.
You can’t look at things that way if you just use free cash flow. To compare companies you need to use a return on something. Return on assets. Return on sales.
Using the free cash flow margin and the price-to-sales ratio gives you a deeper look into what you’re buying.
Using average earnings hides assumptions. It hides the fact you are actually betting both on a certain level of sales and a certain margin.
For the stock market as a whole, I never want to just buy earnings. I want to buy sales or tangible equity. Because the entire universe of public companies as a whole is going to mean revert. You can't assume abnormally high or low margins, abnormally high or low ROE for everyone.
You never want to buy an entire stock market based on the P/E ratio. This is supported by past experience. Which shows one year trailing P/Es have a poor ability to predict future returns. This is not true of long-term average earnings.
We’re not looking for earnings. We’re looking for earning power.
We don't necessarily need to know what a stock earned last year or last decade. We need to know what it will earn over the next decade.
How do we do that?
For your average business, we just use tangible book. We expect — on average — that businesses will earn returns based on their assets.
Now, what about your not average business?
What about your special business?
Your extraordinary business?
You wouldn't expect to start a new business competing with FICO using the same amount of tangible assets and earn a similar return to FICO.
So, we need another way to value unique businesses. Businesses that have accumulated valuable things — like brands — that are hard or impossible to replace.
Where the free cash flow margin doesn't vary much, I've proposed price-to-sales as the appropriate measure. Instead of assuming a certain return on book value we simply assume a certain return on sales.
The key warning is this: the company must be able to maintain its margins.
It must be able to avoid price competition.
And that’s tough.
Some industries do away with price competition. Movies don’t compete on price. They compete on attendance.
And some companies (FICO, Dun & Bradstreet, Moody's, etc.) are in a position where a competitor offering a substitute for a lower price is really no substitute.
These companies have pricing power.
If you are insulated from price competition, then price-to-sales matters. It can be used with your long-term average free cash flow margin. It tells you if the future is like the past — and it will be unless price competition emerges — here is what the company's earning power will be.
The key is to understand what you buy.
You have to understand the assumptions you are making. Assuming a certain level of earnings is a big assumption.
Start with the assets in the business or the current level of sales. Make your assumptions from there.
Don't assume you’re entitled to a certain stream of earnings.
A tech company never is. A tech company's margins come from being in a certain place on the technological frontier.
If they are no longer on the edge of new, they can't charge the same prices. As their products age, those sales will no longer deliver profits.
They need to maintain their position relative to the newest and best stuff out there or they will not maintain their earnings.
By focusing on sales and margins you start to ask yourself why a company's margins are what they are.
And if you can't convince yourself future margins will be sufficient to justify the price-to-sales ratio you are paying, then you can’t buy the stock.
Think of it this way.
Reverse my approach.
Instead of starting with margins, start with your hurdle rate and the company's price-to-sales ratio.
Let's look at Microsoft.
Say your hurdle rate is 10%. And let's say Microsoft's price-to-sales ratio is 3. That would mean Microsoft needs a free cash flow margin of 30% going forward to guarantee you a 10% return.
This statement — as always — is more conservative than accurate since the company might grow. Microsoft also has surplus cash. We are ignoring both those facts.
Okay, so can Microsoft deliver 30% free cash flow margins in the future?
That’s the question.
It's done something like that in the past. So, buying Microsoft stock at today’s price is justified based on the past.
But isn't it better to ask yourself not just what the price to free cash flow ratio is based on the past — but actually look at the margin needed to justify buying the stock at today’s price.
If you’re confident Microsoft can achieve 30% margins in the future, you should buy the stock.
If Microsoft delivers 30% free cash flow margins, the stock should provide you with double-digit returns.
But what if Microsoft can only achieve 10% or 15% margins?
That would be bad. You might not make anything in the stock if you paid 3 times sales for something that will only have 10% to 15% margins in the future.
And what about 5% margins?
That would ruin you.
My hope is to get you thinking along those lines.
Think about whether margins will be 5%, 15%, or 30%. And what that will mean for you if you buy the stock at today’s price.
I think this moves us toward Graham’s idea of earning power.
It’s a certain amount of assets, or sales, or other exploitable things the business controls. And then it’s the return on that thing. In the future. For the long-term.
That’s earning power.
Follow Geoff at Gannon On Investing
Hi Geoff – Great article on analysis paralysis. However, I’m curious why you calculate FCF yield the way you do…
For you, free cash flow yield = 10-year average free cash flow margin / Price-to-sales.
For just about every other source, FCF yield = free cash flow / price.
I assume you use the margin figure to smooth out growth effects over the 10-year period.
But I’m not sure why you use price to sales as your denominator rather than market price?
A lot of people have asked me this question.
You can use 10-year average free cash flow divided by market cap to get a similar average earning number.
As for using free cash flow divided by price — my problem with that is that I never use a one-year number.
The 10-year average is good. In cases like Omnicom (OMC, Financial), it will yield a similar number. Maybe 10% instead of 13% or something like that. But since it’s a trailing average and Omnicom is usually growing over time — being a little lower is always expected.
My reason for choosing the 10-year average free cash flow margin divided by price-to-sales is to focus investors on the business.
I don't think it's safe to assume that a business will simply earn what it is earning this year or even what it earned on average over the last decade.
You should always look past the earnings to what is delivering those earnings. You should always be thinking in terms similar to a DuPont analysis.
If I'm looking at a bank's earning power, I need to be thinking about their cost of deposits. If I'm looking at an insurer, I need to be thinking about their combined ratio. If I'm looking at Dreamworks (DWA, Financial), I need to be thinking about their production costs and their distributor's marketing costs. Then compare that to the potential audience size for their kind of movies.
You want to look at how the company makes money. It only makes money if it can gather deposits cheaply enough (banks). Or draw a big enough audience to justify a $300 million movie (Dreamworks). Or maintain its monopoly pricing (Microsoft).
I want to get investors thinking like businessmen.
You make money by getting a return on something. In intangible asset businesses, it's often a return on sales. The profit margin. This is usually how you value Kraft (KFT, Financial) and Omnicom and FICO (FICO, Financial) and so forth.
But not always.
I would not pay attention to Dreamworks’s price-to-sales ratio. Sales are not predictable at Dreamworks.
Dreamworks is a future project and past library business. That is where the value is.
You need to value those two things. Not sales. Revenues will vary a lot at Dreamworks. Dreamworks is different from Kraft or Omnicom or VCA Antech (WOOF, Financial).
For Dreamworks, you look at the movies they’ve already made and the movies they are going to make. The movies they are going to make — 2.5 a year from now on — can be analyzed like any capital expenditures.
The old movies they own but no longer show on the balance sheet under “Film and Inventory Costs” are more complicated. They have value. You have some nice characters from those movies. You can monetize them through television shows and toys and things like that.
So there may never be a perfectly precise formula for valuing Dreamworks. But that doesn’t mean you can’t draw a line in the sand.
For example, we can say Dreamworks is worth more than $15 a share. The only way Dreamworks is worth less than $15 a share is if future productions destroy value.
They could. That’s a real risk. It’s something you’d have to analyze.
But if we look at the balance sheet — and we keep in mind that computer animated productions from Dreamworks and Pixar have historically been worth at least what they were carried at as inventory — we see the value of DWA is already $15 a share or higher.
When we say Dreamworks is worth at least $15 a share — we are really saying their assets (intangible though they may be) will support $1.50 a share in earnings.
We’ve talked about this idea of asset-earnings equivalence before.
In a tangible asset business, you’re often looking at the return on tangible assets. This is how you would normally value railroads, utilities, airlines (if you dare), and even something like Carnival (CCL).
Return on equity is what ultimately matters.
But it’s important to start by using the return on tangible assets and then look at how much leverage you can use. If you just look at the return on equity — you’ll often be measuring leverage more than anything else.
Sometimes it's none of those things.
I used Dreamworks as an example. I would actually value Dreamworks on a project basis. It has a library and it makes new movies.
I would never look at price-to-sales at Dreamworks. Nor does book value matter — because everything older than about three years has been written off the balance sheet. So, you'd be leaving Shrek off the balance sheet if you did something like that.
What you want to do with every business — with every stock — is think like an owner. Think like a businessman. Earnings don't appear out of nowhere. Returns come from economic (not accounting) goodwill, from sales, from tangible assets, etc.
Earnings come from real exploitable things the company controls.
When you buy a share of Kraft — are you buying earnings or sales?
It sounds like a silly question.
I like to think you are buying sales. You are buying products that have a place in the minds of customers. You are buying customers habits.
Those habits may change. But those habits are your starting point for analyzing Kraft.
Why do you think Nintendo is so cheap right now?
Because people are convinced habits will change. The iPad and smart phones will destroy handheld gaming. They may be right.
So is it right to think of a share of Nintendo as being backed by earnings, free cash flow, cash, or sales?
Personally, I like to think of something like a share of Nintendo as being backed by cash plus a business. And when I look at the business, I look more at sales than anything else.
You can look at market share. Or units sold. None of that is wrong.
The key is to focus on things the business controls and can exploit.
It helps if those things are easy to analyze.
One of Warren Buffett’s best bits of advice is: “Don’t spend time on companies that don’t lend themselves to valuation”.
All of this may sound confusing.
Why look at so many different measures?
How is this better than just using average free cash flow?
It's better because it draws you into thinking about the business.
At long-term margin "x" the company is worth this. At long-term margin 1.5x its worth something else. And at 0.5x it’s worth something else.
Let's say it's a video game company and "x" is 10%. Is 15% possible? Maybe. But other than Nintendo only Activision — and this is quite recent — has free cash flow margins like 15%. Historically, you would have to say that having 15% free cash flow margins in video games is tough to do.
What about 5%?
5% is doable.
If you have good properties and you do intelligent things with them you should turn $1 of sales into 5 cents of free cash flow in the video game business.
So now we have a way of looking at video game companies. If your hurdle rate is 10% and you don’t want to pay up for growth — you start by looking for a price-to-sales ratio of 1. That’s because 10% free cash flow margins seem possible in the video game business.
What about 1.5 times sales?
That’s a different story.
A price-to-sales ratio of 1.5 looks too high. I wouldn’t want to pay 1.5 times sales for a video game company.
It might work out. You could achieve high single digit returns in a video game stock even if your starting point is paying 1.5 times sales. But there’s no margin of safety.
I want a good chance of earning 10% on my money from day one.
So I want to pay no more than the stock’s per share sales. That would be for a company I was confident in.
For a company I wasn’t confident in, I’d need a bigger margin. It would have to trade under its sales per share.
Looking at the stock this way allows you to step back from that one company and look at the whole industry. At the company’s peers. At everybody’s long-term history.
You see a wider range of outcomes.
That’s what a margin of safety is all about
It’s about succeeding even when some things go wrong.
If you buy an above average video game company at a price-to-sales ratio that would work out well for you even if the company only had industry average margins from now on – you have a margin of safety.
You’re paying an average price for an above average company.
That extra quality is your margin of safety.
You can’t look at things that way if you just use free cash flow. To compare companies you need to use a return on something. Return on assets. Return on sales.
Using the free cash flow margin and the price-to-sales ratio gives you a deeper look into what you’re buying.
Using average earnings hides assumptions. It hides the fact you are actually betting both on a certain level of sales and a certain margin.
For the stock market as a whole, I never want to just buy earnings. I want to buy sales or tangible equity. Because the entire universe of public companies as a whole is going to mean revert. You can't assume abnormally high or low margins, abnormally high or low ROE for everyone.
You never want to buy an entire stock market based on the P/E ratio. This is supported by past experience. Which shows one year trailing P/Es have a poor ability to predict future returns. This is not true of long-term average earnings.
We’re not looking for earnings. We’re looking for earning power.
We don't necessarily need to know what a stock earned last year or last decade. We need to know what it will earn over the next decade.
How do we do that?
For your average business, we just use tangible book. We expect — on average — that businesses will earn returns based on their assets.
Now, what about your not average business?
What about your special business?
Your extraordinary business?
You wouldn't expect to start a new business competing with FICO using the same amount of tangible assets and earn a similar return to FICO.
So, we need another way to value unique businesses. Businesses that have accumulated valuable things — like brands — that are hard or impossible to replace.
Where the free cash flow margin doesn't vary much, I've proposed price-to-sales as the appropriate measure. Instead of assuming a certain return on book value we simply assume a certain return on sales.
The key warning is this: the company must be able to maintain its margins.
It must be able to avoid price competition.
And that’s tough.
Some industries do away with price competition. Movies don’t compete on price. They compete on attendance.
And some companies (FICO, Dun & Bradstreet, Moody's, etc.) are in a position where a competitor offering a substitute for a lower price is really no substitute.
These companies have pricing power.
If you are insulated from price competition, then price-to-sales matters. It can be used with your long-term average free cash flow margin. It tells you if the future is like the past — and it will be unless price competition emerges — here is what the company's earning power will be.
The key is to understand what you buy.
You have to understand the assumptions you are making. Assuming a certain level of earnings is a big assumption.
Start with the assets in the business or the current level of sales. Make your assumptions from there.
Don't assume you’re entitled to a certain stream of earnings.
A tech company never is. A tech company's margins come from being in a certain place on the technological frontier.
If they are no longer on the edge of new, they can't charge the same prices. As their products age, those sales will no longer deliver profits.
They need to maintain their position relative to the newest and best stuff out there or they will not maintain their earnings.
By focusing on sales and margins you start to ask yourself why a company's margins are what they are.
And if you can't convince yourself future margins will be sufficient to justify the price-to-sales ratio you are paying, then you can’t buy the stock.
Think of it this way.
Reverse my approach.
Instead of starting with margins, start with your hurdle rate and the company's price-to-sales ratio.
Let's look at Microsoft.
Say your hurdle rate is 10%. And let's say Microsoft's price-to-sales ratio is 3. That would mean Microsoft needs a free cash flow margin of 30% going forward to guarantee you a 10% return.
This statement — as always — is more conservative than accurate since the company might grow. Microsoft also has surplus cash. We are ignoring both those facts.
Okay, so can Microsoft deliver 30% free cash flow margins in the future?
That’s the question.
It's done something like that in the past. So, buying Microsoft stock at today’s price is justified based on the past.
But isn't it better to ask yourself not just what the price to free cash flow ratio is based on the past — but actually look at the margin needed to justify buying the stock at today’s price.
If you’re confident Microsoft can achieve 30% margins in the future, you should buy the stock.
If Microsoft delivers 30% free cash flow margins, the stock should provide you with double-digit returns.
But what if Microsoft can only achieve 10% or 15% margins?
That would be bad. You might not make anything in the stock if you paid 3 times sales for something that will only have 10% to 15% margins in the future.
And what about 5% margins?
That would ruin you.
My hope is to get you thinking along those lines.
Think about whether margins will be 5%, 15%, or 30%. And what that will mean for you if you buy the stock at today’s price.
I think this moves us toward Graham’s idea of earning power.
It’s a certain amount of assets, or sales, or other exploitable things the business controls. And then it’s the return on that thing. In the future. For the long-term.
That’s earning power.
Follow Geoff at Gannon On Investing
