Someone who reads my blog sent me this question:
“You seem to really focus on looking at a (company’s) historical figures when estimating free cash flow. If you are buying… stocks on a P/FCF basis, how do you think about how a given (company) will generate future cash flows, since that is what you're buying, and how these future cash flows relate to the P/FCF trailing multiple you use to buy the stock?
…I believe a…FCF yield is relatively worthless if the future free cash flows are not maintainable or growing. It would be like buying a bond at an assumed 10% (yield-to-maturity) that misses a few coupon payments. The multiple doesn't live up to reality.”
I agree that past free cash flow is worthless. But so is estimated future free cash flow unless it is more accurate than the past. If I knew what the future free cash flows would be, that would be great. But I can only guess what the future free cash flows will be.
Based on studying past stock market price to 10-year and 15-year average earnings, I've seen the market is no better at predicting future earnings growth beyond 1 or 2 years than the past record alone is. If this is true for the whole market, it’s probably true for many of the parts of that whole.
People frequently do worse predicting the future earnings of individual stocks than they would if they just focused on the average of the last 10 or 15 years of earnings.
Now, obviously, I am focusing on stocks that have 10 year or 15 year records. So, I’m looking for companies that have been in existence for at that long. I’m looking for companies that have been public that long. And I’m looking for companies that have been in the same basic business, earning positive operating income and free cash flow for most of the last 10 or 15 years.
Very often, we’re talking about companies that have had positive operating income and free cash flow in 13 of the last 15 years or even 15 of the last 15 years.
Most stocks haven’t. So, I’m throwing out most stocks for being too unreliable. Too new. Too hard to analyze.
True, the future trend won’t always look good at the moment I buy. In fact, it seems I’m always buying things where next year is expected to be worse than this year.
So – maybe – we’re talking about speculative stocks in the sense that the future doesn’t look good. But we’re not talking about speculative businesses in the sense that they are new, unprofitable, or entering markets that are unsettled.
Most of the estimates I see analysts and bloggers make are not good. They are only likely to be accurate in the extremely short-run (1-3 years out). And often that's not what's going to determine the value of the stock. Correctly predicting free cash flow will decline at an advertising agency this year doesn't help you much valuing a stock that may have a much higher - and more normal - free cash flow in 2014, 2015, and 2016.
I haven't seen people who are good at correctly predicting 2014-2016 free cash flow without relying heavily on the 2001-2010 record.
That doesn't mean I just take past free cash flow and project them into the future.
For one example of how you can look at the future using the past, see my article about DuPont analysis for value investors. Using a DuPont analysis, you can break down the profit margin, asset turns, and leverage taking 10-year averages. If you feel one of those averages clearly needs to be changed, go ahead and change it. But you should know why the profit margin the next 10 years will be lower than the profit margin was these past ten years.
And I think it’s extraordinarily hard to predict what the next decade’s average profit margin will be unless it happens to bear some resemblance to the last decade’s average profit margin.
That doesn’t mean profit margins will tend to stay steady over 20 years. It just means that if you don’t see some stability, some pattern, some logic to what profit margins were from 2001-2010, I think it’s impossibly hard to predict what profit margins will look like from 2011-2020.
If you think margins in some business will get worse over the next 10 years, that’s fine. You can substitute that new, lower profit margin into the next 10 years if you want to. But, to think that’s a better estimate is crazy.
I can’t think of any situation where I would feel more comfortable in my own estimate of the average profit margin over the next 10 years than I would in just taking the actual recorded average profit margin of the least 10 years.
I can think of situations where I would feel uncomfortable with either the past record or my own estimate as a guide to the future. But, I definitely can’t think of any situations where I’d be uncomfortable with the past and yet somehow comfortable with my future estimate.
Now, what you can do is invert the process. Don’t estimate the intrinsic value using some future profit margin. Instead ask: How bad can this get? What’s the worst future scenario that sill justifies the current stock price?
But that’s about all you should do.
I don’t think you should actually project anything. I don’t think you should really have an actual number in mind for 2014 earnings or cash flows.
That’s crazy.
I don't think most estimates will be much better than just taking current sales and using the 10-year average profit margin, asset turns, and cash conversion combined with the current quarter's leverage ratio and using that to predict the cash return on equity and therefore the expected 10-year return.
In other words, despite what we know about current trends and what we think we know about the future, about the only thing I think investors need to know about the future is whether or not it’s likely to bear some resemblance to the past.
If the future is going to be totally unlike the past, I just cross off that stock entirely.
Otherwise, I ask how much worse can the future be before I start losing money on this investment.
Growth companies are different.
But I don't value growth companies like Apple (AAPL, Financial). I look at companies that are doing the same thing year after year. Maybe they’re doing it better maybe they’re doing it worse. But the “it” they’re doing is the same “it” from one year to the next.
I look at a lot of companies with just one business segment. If you were valuing a stock with lots of business segments, maybe you would need to break each segment down separately. But event hen I would certainly rather do something like a DuPont analysis based on past 10-year ratios rather than take an analyst or blogger’s estimates of future free cash flows.
I basically look at the last 10 years and then try to think if the next 10 years will be better, worse, or the same. For a homebuilder, it's too tough to use the last 10 years, so I throw homebuilders out. Same with newspapers. Stocks like that just get thrown out.
For Barnes & Noble (BKS, Financial), I asked how big was the cushion. What was my margin of safety? I figured that if you cut historical free cash flow margins in half and did it on 50% lower sales, you could still support the stock at the price I was buying.
Therefore, the questions were: could they avoid bankruptcy, could they close stores fast enough, and would the market for printed books bought in stores shrink faster than Barnes & Noble's market share gains due to lost competitors like the soon to be bankrupt Borders (BGP, Financial), and general retailers like Wal-Mart (WMT, Financial) that will cut shelf space for books quicker than bookstores when sales decline.
I just asked those questions. I didn't do any projecting.
I said how far can the past free cash flow fall to cover the current price? And what would that take in terms of lost volume, margins, etc. I didn't do a projection and then calculate a value. Instead, I took the current market value and worked backwards to see if people's predictions for the industry were dire enough to harm that low stock price.
When Barnes & Noble committed heavily to the Color Nook, the calculus changed. I sold the stock.
Because now the past record had little bearing on free cash flow. Barnes & Noble’s shareholders weren't going to keep that free cash flow. Instead, that free cash flow was going to be spent on the Nook. And I had no record to use to evaluate investment in the Nook. I only had past data on investment in bookstores.
Is every $1 investment in the Nook worth 50 cents, 20 cents, 10 cents, or two dollars in future market value?
Who knows?
Definitely not me.
So once Barnes & Noble deviated from the past like that, I sold the stock. Some people would have tried to estimate the free cash flow eventually produced by selling e-books or something. I didn't.
That's what I mean by relying on the past record.
I hear what you're saying about buying a bond. But, remember, people like Ben Graham didn't buy high yield bonds for the yield itself. They bought for price appreciation. So, if you make sure you’ve got a fat margin of safety, you can still get price appreciation if the coupon shrinks.
If you buy something at a 30% free cash flow yield and it declines to – and stabilizes at – a 15% free cash flow yield on your original cost, chances are you made at least a 50% return over the period you owned that stock. Because even a stable, no-growth business will eventually trade at a price that gives it a 10% free cash flow yield. If you sell then, you made 50% on a company that’s now only half the business it used to be.
Now, I know what you’re thinking. The key words are “and stabilizes”. How can I know a declining business will ever stabilize?
I can’t.
But I can’t know a growth business will stabilize either. Like I’ve said before, the fact that Apple is growing today is no guarantee that it won’t be making less money on less sales in 10 years.
Apple was growing once before. And it was losing money in 2001. So, the fact that it’s making money and growing today isn’t some impenetrable armor.
The question isn’t just how fast will it Apple grow? It’s also whether or not Apple will even be able to change fast enough to stay in place for a full 10 years.
So how do you defend against the future?
You insist on a big, fat margin of safety.
The problem with predicting the future isn’t that you can’t do it. You can often predict the future quite clearly. I can predict the population of the United States next year. I’m not going to be off by more than a fraction of one percent. But that’s not good enough to get someone to pay me for it.
I can predict the future of a water company. But no one is willing to sell me a water company at a 10% or 15% unleveraged free cash flow yield. So, my prediction doesn’t do me any good. All my prediction gets me is the knowledge of exactly how much someone is asking me to overpay for a leveraged utility.
Predicting doesn’t count. Handicapping counts.
Is the price of Facebook high enough to equalize the odds of an investment in Facebook succeeding versus the odds of an investment in Blyth (BTH, Financial) succeeding?
I have no doubt Facebook is a better, faster growing business than Blyth.
Facebook has tremendous competitive advantages: it’s got mindshare, it’s got routine use, it has tremendously scalable operating margins (they could double from here),
However, the price of Facebook – if it ever goes public – will be very high. The price of Blyth is very low. Both sellers and buyers of Blyth shares think the next 10 years will be worse than the past 10 years. There’s no argument on that point. Decay is already priced into Blyth just at growth is already priced into Facebook.
That’s the handicapping part of it. How much do you have to overpay for Facebook and underpay for Blyth to equalize the odds of investment success.
It’s tough. And here’s why:
In the cases where I can predict future growth of the free cash flow yield, I'm very rarely getting a yield that's good enough today.
So I think it's much smarter to insist on a certain free cash flow yield - like 10% on the 10-year average past free cash flow - before buying any stock. If you limit yourself to choosing from stocks that are priced to provide a 10% initial return, then you can look for stocks that will grow their coupon to owners over time.
But buying at a 5% or 7% free cash flow yield and then depending on growth just to get you to 10% doesn't sound safe to me.
I'm happy to buy something that will grow. I'm just not happy paying for growth.
I’d rather focus on situations where there are two separate defenses against loss:
1. The initial return
2. The expected growth
However, in good times for the stock market, there are very few situations that meet both requirements. In bad times – like early 2009 – there are plenty of stocks that meet both requirements.
Today, there are very few well known stocks that meet both requirements. So, you have to:
A) Look for unknown stocks
B) Accept stocks that won’t grow but have low enough prices to make up for it
C) Accept stocks that are overpriced but can grow enough to make up for it
The best approach is “A”. Find stocks that are small, foreign, unlisted, etc. These are basically what Ben Graham called secondary stocks. The second best approach is either “B” or “C”. Graham believed in “B”. Phil Fisher believed in “C”.
Both men did very well using their own approach.
I don’t think I’m a better judge of future growth than other people. So I don’t buy overpriced stocks that look like they’ll grow fast enough to offset the initial overpricing.
If you want to learn that approach, you should start by reading Phil Fisher’s Common Stocks and Uncommon Profits.
Personally, I don't buy into situations where the investment depends entirely on the future growth.
I will buy into situations with an adequate initial return and no expected growth. But never the reverse – an inadequate initial return but adequate future growth.
In reality, most stocks offer very little real free cash flow growth. You basically get earnings growth of 1% to 2% plus inflation. So, if you find a business that can keep up with inflation but will never grow its unit volume, that's often a good choice if it's got a good enough initial return priced in.
Even Buffett's investment in See's Candies was based on price increases. They really weren't betting on an increase in the pounds of chocolate sold.
Of course, Buffett also got an adequate initial return on his investment in See’s Candies.
I'd rather own the clearly strongest competitor in a no-growth industry than one of the two or three strongest competitors in a fast growing industry. It's hard to predict growth industries. Past margins and things like that may not be a good guide to the future.
Warren Buffett and Peter Lynch have talked about this a lot. Read either of them to get a clear idea of why fast growing industries can be bad places to look for investments.
The ideal investment would be in a company that is quickly growing its market share in an industry where the overall pie is staying the same. If you can find a decently cheap stock like that, you can retire rich.
But rapid market share gains are usually tough to predict in established industries.
My favorite investment is in an industry with almost no change where the initial return you are getting is adequate and the past record is a good guide to the future.
I can't find many investments like that. That's why I never own more than 4 or 5 stocks anymore.
From past experience, I know I’m only good at handicapping certain situations, so those are the only situations I bet on.
I think Warren Buffett is the same way.
There’s no evidence that Warren Buffett actually makes future predictions. He thinks about the future. But that doesn’t mean he predicts the future.
He just looks for businesses that don’t change.
And then he looks at the past.
Follow Geoff at Gannon On Investing
“You seem to really focus on looking at a (company’s) historical figures when estimating free cash flow. If you are buying… stocks on a P/FCF basis, how do you think about how a given (company) will generate future cash flows, since that is what you're buying, and how these future cash flows relate to the P/FCF trailing multiple you use to buy the stock?
…I believe a…FCF yield is relatively worthless if the future free cash flows are not maintainable or growing. It would be like buying a bond at an assumed 10% (yield-to-maturity) that misses a few coupon payments. The multiple doesn't live up to reality.”
I agree that past free cash flow is worthless. But so is estimated future free cash flow unless it is more accurate than the past. If I knew what the future free cash flows would be, that would be great. But I can only guess what the future free cash flows will be.
Based on studying past stock market price to 10-year and 15-year average earnings, I've seen the market is no better at predicting future earnings growth beyond 1 or 2 years than the past record alone is. If this is true for the whole market, it’s probably true for many of the parts of that whole.
People frequently do worse predicting the future earnings of individual stocks than they would if they just focused on the average of the last 10 or 15 years of earnings.
Now, obviously, I am focusing on stocks that have 10 year or 15 year records. So, I’m looking for companies that have been in existence for at that long. I’m looking for companies that have been public that long. And I’m looking for companies that have been in the same basic business, earning positive operating income and free cash flow for most of the last 10 or 15 years.
Very often, we’re talking about companies that have had positive operating income and free cash flow in 13 of the last 15 years or even 15 of the last 15 years.
Most stocks haven’t. So, I’m throwing out most stocks for being too unreliable. Too new. Too hard to analyze.
True, the future trend won’t always look good at the moment I buy. In fact, it seems I’m always buying things where next year is expected to be worse than this year.
So – maybe – we’re talking about speculative stocks in the sense that the future doesn’t look good. But we’re not talking about speculative businesses in the sense that they are new, unprofitable, or entering markets that are unsettled.
Most of the estimates I see analysts and bloggers make are not good. They are only likely to be accurate in the extremely short-run (1-3 years out). And often that's not what's going to determine the value of the stock. Correctly predicting free cash flow will decline at an advertising agency this year doesn't help you much valuing a stock that may have a much higher - and more normal - free cash flow in 2014, 2015, and 2016.
I haven't seen people who are good at correctly predicting 2014-2016 free cash flow without relying heavily on the 2001-2010 record.
That doesn't mean I just take past free cash flow and project them into the future.
For one example of how you can look at the future using the past, see my article about DuPont analysis for value investors. Using a DuPont analysis, you can break down the profit margin, asset turns, and leverage taking 10-year averages. If you feel one of those averages clearly needs to be changed, go ahead and change it. But you should know why the profit margin the next 10 years will be lower than the profit margin was these past ten years.
And I think it’s extraordinarily hard to predict what the next decade’s average profit margin will be unless it happens to bear some resemblance to the last decade’s average profit margin.
That doesn’t mean profit margins will tend to stay steady over 20 years. It just means that if you don’t see some stability, some pattern, some logic to what profit margins were from 2001-2010, I think it’s impossibly hard to predict what profit margins will look like from 2011-2020.
If you think margins in some business will get worse over the next 10 years, that’s fine. You can substitute that new, lower profit margin into the next 10 years if you want to. But, to think that’s a better estimate is crazy.
I can’t think of any situation where I would feel more comfortable in my own estimate of the average profit margin over the next 10 years than I would in just taking the actual recorded average profit margin of the least 10 years.
I can think of situations where I would feel uncomfortable with either the past record or my own estimate as a guide to the future. But, I definitely can’t think of any situations where I’d be uncomfortable with the past and yet somehow comfortable with my future estimate.
Now, what you can do is invert the process. Don’t estimate the intrinsic value using some future profit margin. Instead ask: How bad can this get? What’s the worst future scenario that sill justifies the current stock price?
But that’s about all you should do.
I don’t think you should actually project anything. I don’t think you should really have an actual number in mind for 2014 earnings or cash flows.
That’s crazy.
I don't think most estimates will be much better than just taking current sales and using the 10-year average profit margin, asset turns, and cash conversion combined with the current quarter's leverage ratio and using that to predict the cash return on equity and therefore the expected 10-year return.
In other words, despite what we know about current trends and what we think we know about the future, about the only thing I think investors need to know about the future is whether or not it’s likely to bear some resemblance to the past.
If the future is going to be totally unlike the past, I just cross off that stock entirely.
Otherwise, I ask how much worse can the future be before I start losing money on this investment.
Growth companies are different.
But I don't value growth companies like Apple (AAPL, Financial). I look at companies that are doing the same thing year after year. Maybe they’re doing it better maybe they’re doing it worse. But the “it” they’re doing is the same “it” from one year to the next.
I look at a lot of companies with just one business segment. If you were valuing a stock with lots of business segments, maybe you would need to break each segment down separately. But event hen I would certainly rather do something like a DuPont analysis based on past 10-year ratios rather than take an analyst or blogger’s estimates of future free cash flows.
I basically look at the last 10 years and then try to think if the next 10 years will be better, worse, or the same. For a homebuilder, it's too tough to use the last 10 years, so I throw homebuilders out. Same with newspapers. Stocks like that just get thrown out.
For Barnes & Noble (BKS, Financial), I asked how big was the cushion. What was my margin of safety? I figured that if you cut historical free cash flow margins in half and did it on 50% lower sales, you could still support the stock at the price I was buying.
Therefore, the questions were: could they avoid bankruptcy, could they close stores fast enough, and would the market for printed books bought in stores shrink faster than Barnes & Noble's market share gains due to lost competitors like the soon to be bankrupt Borders (BGP, Financial), and general retailers like Wal-Mart (WMT, Financial) that will cut shelf space for books quicker than bookstores when sales decline.
I just asked those questions. I didn't do any projecting.
I said how far can the past free cash flow fall to cover the current price? And what would that take in terms of lost volume, margins, etc. I didn't do a projection and then calculate a value. Instead, I took the current market value and worked backwards to see if people's predictions for the industry were dire enough to harm that low stock price.
When Barnes & Noble committed heavily to the Color Nook, the calculus changed. I sold the stock.
Because now the past record had little bearing on free cash flow. Barnes & Noble’s shareholders weren't going to keep that free cash flow. Instead, that free cash flow was going to be spent on the Nook. And I had no record to use to evaluate investment in the Nook. I only had past data on investment in bookstores.
Is every $1 investment in the Nook worth 50 cents, 20 cents, 10 cents, or two dollars in future market value?
Who knows?
Definitely not me.
So once Barnes & Noble deviated from the past like that, I sold the stock. Some people would have tried to estimate the free cash flow eventually produced by selling e-books or something. I didn't.
That's what I mean by relying on the past record.
I hear what you're saying about buying a bond. But, remember, people like Ben Graham didn't buy high yield bonds for the yield itself. They bought for price appreciation. So, if you make sure you’ve got a fat margin of safety, you can still get price appreciation if the coupon shrinks.
If you buy something at a 30% free cash flow yield and it declines to – and stabilizes at – a 15% free cash flow yield on your original cost, chances are you made at least a 50% return over the period you owned that stock. Because even a stable, no-growth business will eventually trade at a price that gives it a 10% free cash flow yield. If you sell then, you made 50% on a company that’s now only half the business it used to be.
Now, I know what you’re thinking. The key words are “and stabilizes”. How can I know a declining business will ever stabilize?
I can’t.
But I can’t know a growth business will stabilize either. Like I’ve said before, the fact that Apple is growing today is no guarantee that it won’t be making less money on less sales in 10 years.
Apple was growing once before. And it was losing money in 2001. So, the fact that it’s making money and growing today isn’t some impenetrable armor.
The question isn’t just how fast will it Apple grow? It’s also whether or not Apple will even be able to change fast enough to stay in place for a full 10 years.
So how do you defend against the future?
You insist on a big, fat margin of safety.
The problem with predicting the future isn’t that you can’t do it. You can often predict the future quite clearly. I can predict the population of the United States next year. I’m not going to be off by more than a fraction of one percent. But that’s not good enough to get someone to pay me for it.
I can predict the future of a water company. But no one is willing to sell me a water company at a 10% or 15% unleveraged free cash flow yield. So, my prediction doesn’t do me any good. All my prediction gets me is the knowledge of exactly how much someone is asking me to overpay for a leveraged utility.
Predicting doesn’t count. Handicapping counts.
Is the price of Facebook high enough to equalize the odds of an investment in Facebook succeeding versus the odds of an investment in Blyth (BTH, Financial) succeeding?
I have no doubt Facebook is a better, faster growing business than Blyth.
Facebook has tremendous competitive advantages: it’s got mindshare, it’s got routine use, it has tremendously scalable operating margins (they could double from here),
However, the price of Facebook – if it ever goes public – will be very high. The price of Blyth is very low. Both sellers and buyers of Blyth shares think the next 10 years will be worse than the past 10 years. There’s no argument on that point. Decay is already priced into Blyth just at growth is already priced into Facebook.
That’s the handicapping part of it. How much do you have to overpay for Facebook and underpay for Blyth to equalize the odds of investment success.
It’s tough. And here’s why:
In the cases where I can predict future growth of the free cash flow yield, I'm very rarely getting a yield that's good enough today.
So I think it's much smarter to insist on a certain free cash flow yield - like 10% on the 10-year average past free cash flow - before buying any stock. If you limit yourself to choosing from stocks that are priced to provide a 10% initial return, then you can look for stocks that will grow their coupon to owners over time.
But buying at a 5% or 7% free cash flow yield and then depending on growth just to get you to 10% doesn't sound safe to me.
I'm happy to buy something that will grow. I'm just not happy paying for growth.
I’d rather focus on situations where there are two separate defenses against loss:
1. The initial return
2. The expected growth
However, in good times for the stock market, there are very few situations that meet both requirements. In bad times – like early 2009 – there are plenty of stocks that meet both requirements.
Today, there are very few well known stocks that meet both requirements. So, you have to:
A) Look for unknown stocks
B) Accept stocks that won’t grow but have low enough prices to make up for it
C) Accept stocks that are overpriced but can grow enough to make up for it
The best approach is “A”. Find stocks that are small, foreign, unlisted, etc. These are basically what Ben Graham called secondary stocks. The second best approach is either “B” or “C”. Graham believed in “B”. Phil Fisher believed in “C”.
Both men did very well using their own approach.
I don’t think I’m a better judge of future growth than other people. So I don’t buy overpriced stocks that look like they’ll grow fast enough to offset the initial overpricing.
If you want to learn that approach, you should start by reading Phil Fisher’s Common Stocks and Uncommon Profits.
Personally, I don't buy into situations where the investment depends entirely on the future growth.
I will buy into situations with an adequate initial return and no expected growth. But never the reverse – an inadequate initial return but adequate future growth.
In reality, most stocks offer very little real free cash flow growth. You basically get earnings growth of 1% to 2% plus inflation. So, if you find a business that can keep up with inflation but will never grow its unit volume, that's often a good choice if it's got a good enough initial return priced in.
Even Buffett's investment in See's Candies was based on price increases. They really weren't betting on an increase in the pounds of chocolate sold.
Of course, Buffett also got an adequate initial return on his investment in See’s Candies.
I'd rather own the clearly strongest competitor in a no-growth industry than one of the two or three strongest competitors in a fast growing industry. It's hard to predict growth industries. Past margins and things like that may not be a good guide to the future.
Warren Buffett and Peter Lynch have talked about this a lot. Read either of them to get a clear idea of why fast growing industries can be bad places to look for investments.
The ideal investment would be in a company that is quickly growing its market share in an industry where the overall pie is staying the same. If you can find a decently cheap stock like that, you can retire rich.
But rapid market share gains are usually tough to predict in established industries.
My favorite investment is in an industry with almost no change where the initial return you are getting is adequate and the past record is a good guide to the future.
I can't find many investments like that. That's why I never own more than 4 or 5 stocks anymore.
From past experience, I know I’m only good at handicapping certain situations, so those are the only situations I bet on.
I think Warren Buffett is the same way.
There’s no evidence that Warren Buffett actually makes future predictions. He thinks about the future. But that doesn’t mean he predicts the future.
He just looks for businesses that don’t change.
And then he looks at the past.
Follow Geoff at Gannon On Investing
