Someone emailed me this question:
“When you combine a 3 or 5 year time frame with FCF yield analysis, does it actually make sense to include a growth factor that only reflects reality for that time period? Or does doing so inflate the multiple the company deserves to trade at, given that the sustainable growth profile is likely lower than the short term profile?”
Growth is a factor in the multiple the company eventually trades at. A not especially good business, that doesn't grow in real terms, etc. would have an end-point P/E assumption in more like the 10-15 P/E range.
A good business, that does grow a little in real terms would have more like an end-point P/E assumption in the 15-25 P/E range.
You would also incorporate growth into the analysis during the period you are analysis. So, you'd project forward EPS growth for the years during which you are analyzing (like 3, 5, etc.).
Yes, it does matter.
For example, NACCO has inflation adjusted price agreements. So, if inflation is 3% a year, you'd need to add an additional 3% a year to the figures you are looking at both in terms of the dividends, buybacks, etc. they might do and the final EPS level (at the end of 5 years or whatever).
So, imagine NC's FCF was $5 a share.
Over a 5 year time period, you'd assume $5 * 1.03^5 = $5.80/FCF in 2024. It does matter. Because, say you were using even just a 10x P/E at the end of that period. Well, an additional 80 cents of earnings adds and additional $8 to the end stock price you expect to sell at after 5 years.
To put it another way, even the difference between a company that does increase EPS with inflation and that doesn't increase EPS with inflation is something like an annual return difference of 3% (or 2%, or whatever inflation might be). This makes a big difference in whether you outperform or not over time.
So, it does matter. And over longer time periods, it matters more. By the time you're looking at like 15-year holding periods, it makes a huge difference whether you are assuming a company will manage to increase prices at least as fast or inflation - or won't.
The sustainable growth rate is really what you use to determine the expected P/E you sell at in the final year of your analysis.
It's worth mentioning something though. It's often as important - sometimes more important - what the return on capital of the business is than the growth rate.
One way to think of ROC is that it's the price of growth. So, a company that increases its growth rate at higher levels of return on capital matters a lot more than a company that grows with a low return on capital.
To put it simply: Growth - no matter how fast - really doesn't add value if the ROE is low. And a high ROE - no matter how high - really doesn't add value if the company has literally no growth.
It's only the combination of a high ROE (that is, low retained earnings each year) with growth that matters.
Take something like Kraft (which Buffett says he overpaid for). The ROE there is very high. If it grew in line with overall GDP - it might be worth a P/E of 25x at the end of your analysis period. So, you could say it's worth 25x P/E when you sell it if you expect it to always grow at least as fast as nominal GDP (which might be 4%, 5%, 6%, etc.). But, if you really think it'll grow at 0% - then, that makes a huge difference in the ending P/E ratio. If the growth rate was LITERALLY going to be 0% NOMINAL forever - the P/E ratio would need to be like 10x.
The reason for this is obviously that the FCF of the company plus the growth rate has to be in line with returns on the S&P 500. If, historically, stocks returned 10% a year - then, a business that will literally never grow (so, the equivalent of just a bond that pays out all earnings each year) would need to be very cheap to keep pace with the S&P 500. You'd need to buy it at like a 10x P/E just to match the market.
On the other hand, something that grows at least as fast as nominal GDP - let's say 6% indefinitely - is probably worth no less than 25x FCF. For some very capital light businesses, FCF basically is EPS. For example, ad agencies. So, if you believed an ad agency would grow at the rate of nominal GDP forever - then, you should assume that when you sell it the P/E you sell at shouldn't be lower than 25x.
I know that sounds exceedingly high. But, if you check the historical record from like the 1970s through the first decade of the 2000s - you can see it's not wrong. You wouldn't be worse off paying 25x EPS for an ad agency than you would be in the S&P 500. Now, however, the industry hasn't - these last 10 years or so - grown nearly as fast as the economy.
I just use ad agencies as a good example from the past. More recent examples would be software companies like Microsoft's Office and Windows business and then businesses like Facebook and Google. If you believed they would keep their competitive position - then, your analysis could include an end point where you sell in 3, 5, 10, or even 15 years at 25x EPS. This isn't wrong if you're right about the company's competitive position. In theory, a search engine / video site etc. that's dominant won't grow at less than nominal GDP in 2035.
There are plenty of companies that can't grow like that. For example, a supermarket - even a totally dominant one locally - would normally grow at the rate of just inflation plus population growth (over time, people spend less and less of their monthly income at supermarkets and more on entertainment, services, etc. as they get richer). Groceries are a basic necessity that don't keep pace with even the overall economy.
So, assume inflation is like 2-3% and population growth is 0-1% in an area. Then, your growth rate for EXISTING supermarkets can only be like 2% to 4% a year in EPS growth. As a result, the P/E of the stock must eventually be low (once it runs out of new stores to open) and the company must eventually pay out a lot of earnings in dividends, buy back stock, etc.
For chains of stores, restaurants, etc. it's often good to use the "saturation" point.
So, a company might have 100 stores today and say it will increase store count by 5% a year indefinitely. But, it also says it will "reach saturation" around 200 stores in the country. You can figure out by working backwards from how long it would take for 100 to grow to 200 at a rate of 5% annual growth that the best way to analyze this is to use a 15-year growth phase that ends with a not especially high P/E at the end.
This makes sense. In theory, chains of stores, restaurants, etc. could justify very high P/E ratios for now. But, at the end of their growth phase they really shouldn't have P/E ratios higher than like a "normal" 15x P/E or so. A chain with a 50x P/E might not be overvalued if it has another 20 years of fast growth ahead of it in store count. But, once it reaches saturation and stops opening many new stores - then, that P/E needs to contract to a more typical 15x P/E.
This is not true for very asset light companies that can raise prices. Things like Buffett's See's Candies actually can always justify a P/E of 25x.
To become a Focused Compounding free member, go to FocusedCompounding.com and type your email address in the box.
Free members get a weekly memo from me and access to stock write-ups by others.