Someone emailed me this question:
"What do you do when a stock looks cheap on an absolute basis (i.e., expected forward return), but only fairly priced on a relative basis (normalized revenue and margins plus EV/Sales, EV/EBIT vs. their own historical multiples)? For example, I own four Japanese stocks that appear to offer mid-teen forward returns but are trading closer to a [50th percentile] level relative to their own historical multiples ... am I just asking for trouble?"
No. You're not necessarily asking for trouble. It depends. There are some checks you can run to see whether you are making a mistake. One, what is the long-term total return in the stock versus the market? A stock can be priced too cheaply for many, many years. For a stock to have been truly fairly valued at many times in the past, it would need to have subsequently recorded a total return similar to the stock market. That's how a stock market works. A stock market is a handicapping system. If, over the last 30 years, the stock you are looking at returned 11% a year while the market returned 8% a year, then we know the stock was priced too cheaply at the start of that 30-year period.
You could argue that it is priced too expensively now. But that's really the only other possible explanation. If a stock outperforms the market over some period of years, it either had to start that period priced too cheaply or it had to end that period priced too expensively. Of course, the historical outperformance can be the result of both these factors. It makes sense to look at which multiples a stock traded in the past.
But investors often cite that figure without subsequently investigating the historical returns in the stock versus the historical returns in the market. Yes, if the stock has - over decades and decades - tended to deliver a compound total return roughly equal to the overall stock market, and it is now priced at about the midpoint of its past historical range of price multiples, that means it's fairly valued. However, there are stocks that some people will point out now trade at the same level they have for 20 years and yet those stocks have returned 14% a year while the market has returned 7% a year.
I've seen investors who are arguing against buying a stock make exactly that argument. For example, I remember looking at a stock that was priced a bit below book value and people said it is always priced a bit below book value. That was true. And it was a fine argument. However, the stock had long returned 10% to 12% a year over many decades. The market doesn't do 10% to 12% a year over many decades. The simplest explanation here is that the stock shouldn't have been trading for less than a price-book (P/B) of 1. In fact, it probably should have traded for a bit of a premium over book value.
The first check to perform is always to look at both the average multiple at which the stock has traded in the past and to look at the total return the stock has provided when it traded at that multiple. I talked about this a little in the newsletter issue I wrote on America's Car-Mart ( CRMT ). America's Car-Mart was not cheap when I picked the stock. It was priced pretty much in line with the multiples at which it had traded in many of the 15 years prior to when I picked the stock. However, if you looked at the return you would have gotten by buying America's Car-Mart instead of buying the Standard & Poor's 500 - you'd see that in a great many of those years you would have done several percentage points better each year by picking Car-Mart over the S&P 500. In other words, Car-Mart had been priced too cheaply in the past.
Of course, the reverse can be true too. A stock can have a good historical annual return because it is now priced too expensively. A business can also change over time in such a way that although it trades at a similar EV/Sales ratio to what it did in the past, it is now expensive. Lately, I've been looking at food companies. We are talking about stocks like Mondelez ( MDLZ ), General Mills ( GIS ), Kellogg ( K ), Campbell Soup ( CPB ), Hershey (HSY), McCormick (MKC), etc. Some of these stocks trade at the same enterprise value to sales ratios that they did in the past.
However, the 10-year growth records at many of these companies is poor. If you try to calculate what the prices of these stocks would need to be to give them annual returns that would be as good or better than the stock market - many of them don't seem cheap. This is because while they were once capable of paying out dividends and stock buybacks of 4% or 5% a year while growing 4% or 5% a year - they are now only capable of paying out dividends and buying back stock at 4% or 5% a year while growing by maybe 1% a year at best. This changes the math dramatically.
If these companies will now be no-growth businesses instead of growing at close to the rate of nominal gross domestic product (GDP) - they should be valued a lot less cheaply in the future than they were in the past. In fact, it's possible that a stock that once deserved an enterprise value to sales ratio of 3 should now trade at an enterprise value to sales ratio of 2. A truly no-growth stock has to trade at a pretty low price-earnings (P/E) multiple. If food companies really couldn't grow at all in the future, they would need to trade at much lower multiples than they did in the past. In the past, they deserved above market P/E ratios. However, in the future, they may deserve below market P/E ratios. That's how important growth is to the value of a stock. You could easily come up with a multiple contraction of 25% to 50% just because a business that had grown at the same pace as the overall economy will now stop growing at all.
Let's talk about the opposite situation. Right now, airlines are in the reverse situation of food companies. Warren Buffett ( Trades , Portfolio ) has bought stock in each of the four biggest airlines in the U.S. He has bought roughly equal amounts of each stock. Obviously, he's excited by the changes he is seeing in the competitive landscape for airlines. Airlines are different than railroads. However, Buffett's purchases follow the same pattern he used when investing in the railroad industry. A lot of airlines - I'd exclude Southwest (LUV) from this category - tended to trade at fairly low multiples of enterprise value to sales, EBITDA, etc. This was especially true whenever the industry had a good period. There's an excellent article about the chicken industry over at Bloomberg right now. It talks about the amount of data sharing - and therefore possible implicit collusion - between the major chicken producers. This is an important story to follow. The chicken industry has a short cycle. But it is brutally and often irrationally competitive.
Good times will always turn to bad times. That has nothing to do with the product economics of chicken. If there were one day just one or two or five big chicken producers and they all shared data with each other on their very, very recent - like what they did last week and last month at each plant - plans, the chicken industry would suddenly become a very good industry. The chicken industry is a bad industry as long as it is perfectly competitive. As long as producers keep secrets from one another and don't trust each other to do what is in the best interest of chicken producers as a group - this isn't an industry you should focus on as an investor.
However, if you could be certain that chicken producers were able to carry out implicit cooperation that would benefit their shared competitive position - then it would cease to be a perfectly competitive industry and start being a very attractive industry. As an example, look at Sanderson Farms (SAFM) stock. Right now,Sanderson Farms might not be cheap if this is going to be a perfectly competitive industry in the future. However, if you somehow knew that the chicken industry was going to feature some sort of implicit cooperation between all the major chicken producers - if you knew they knew they could trust each other not to cheat in a way that would benefit their own firm in the short term but hurt the group in the long term - then you should definitely buy Sanderson Farms stock right now.
If Sanderson Farms is still in a perfectly competitive industry, it might be already fairly priced on a cyclical basis. However, if Sanderson Farms is now in an industry where there will be cooperation instead of rivalry among chicken producers - then the stock is a home run of an investment. The same is pretty much true of the big airlines. I'd say the situation isn't as clear cut because stocks like Sanderson Farms are a lot cheaper than stocks like Southwest in the sense that Southwest stock seems to have a belief that "this time is different" baked into it while Sanderson Farms doesn't. In other words, investors seem more convinced that the airline industry in the U.S. has become less competitive, more collusive and more long-term rational than the chicken industry.
There's evidence for some sort of cooperation - I'm using the term "cooperation" to mean friendly, rational moves that are not explicitly prearranged - in both industries. The airline industry has a very long history of trying to implicitly make changes in such a way that everyone will follow suit. Airlines know they can make more money if they reduce the benefits they provide passengers and increase the fares they charge provided all of the major carriers follow suit.
There have been many times in the past when you could follow the announcements made by each company and see there was an attempt by one company to send up a trial balloon and see if the other airlines would all follow in lock-step. In many cases, this attempt failed. One airline would announce a certain fare hike. A couple of airlines would follow suit. And then a fourth airline would make a big point to investors that it saw no need for a fare increase and planned to ignore the action taken by its competitors. When that kind of thing happens, it should reinforce your belief that the industry is competitive instead of cooperative. In any market, firms are rivals who will both compete and cooperate.
Even in industries that are considered competitive, there is usually some cooperation. There are laws against certain kinds of cooperation. And some industries - like those with very long-lived assets that can be easily moved - are difficult to coordinate. Other industries - like those with short-lived assets that can't be easily moved - are very easy to coordinate. For example, a commodity that will expire worthless in a few days and which has a low value-to-weight ratio can easily support a low-competition, high-cooperation industry structure.
The bad thing about railroads is they last pretty much forever. The great thing about railroads is that you can't move them. That's always been the problem in both the airline industry and the cruise ship industry. You can move the assets. Airlines have the added awful quality of having an asset - airplanes - that are easy to finance. It's much harder to get someone to build you a world-class cruise ship and borrow the money you need to own it than it is to lease a passenger jet.
The historical valuations put on cruise lines, airlines and chicken producers do all matter. However, they are only reliable as long as a few facts are pretty much the same now as they were in the past. One, the stock needs to have performed roughly in line with the stock market for a long time. If the stock has outperformed the market, that's a sign it was priced too cheaply in the past. If the stock has underperformed the market, that's a sign it was priced too expensively in the past. Two, the future growth rate of the business needs to be similar to what it was in the past.
If we go back three decades, Southwest and Walmart (WMT) could be counted on to grow really, really quickly over the next 10 years. They can't be counted on to do that from now until 2027. They need to be valued at a lower multiple today than they may have traded at three decades ago. And then, lastly and perhaps most importantly, the industry structure has to be the same now as it was in the past. Long ago, railroads might have been worth no more than book value. Today, they may be worth two or three times book value. The same could be true of airlines and chicken producers as I just explained. A firm in a perfectly cooperative industry (they don't really exist, but imagine one) could easily be worth three times what a firm in a perfectly competitive industry had traded at in the past.
As an example, if you really could be sure that chicken producers are now and will continue to cooperate in the best long-term interest of the group without losing lawsuits or facing antitrust action related to this cooperation - then I'd say you should increase the enterprise value-to-sales ratio you'd pay for a chicken producer by two or three times. I have no doubt that under cooperative conditions, the value of a chicken producer's stock is at least double what it used to be.
This has to be what Buffett sees in airlines. I think he's probably right about the big four airlines. However, I would caution that Buffett has to find a way to invest billions of dollars a year. You don't. Even if Buffett preferred the competitive structure of other industries, he wouldn't buy those stocks. For example, there's no way for Buffett to put enough money to work in the chicken industry. The top airlines have a much higher combined market cap than the top chicken producers. You can consider industries Buffett can't. It's true that railroads and airlines are more rational and more profitable businesses than they used to be. It's also true that these are huge industries. And a big part of what attracted Buffett to these stocks is simply the amount of money he could put to work in those industries.
So, yes, past multiples matter. But make sure you always check to see that a stock has performed in line with the market over its long-term past - otherwise, the historical multiple will deceive you. Also, check to see if growth prospects are better or worse now than they were in the past. Finally, spend a ton of time thinking about the degree of competition and cooperation in the industry. Has it changed?
As an industry becomes less competitive and more cooperative, the stocks in the industry deserve higher multiples. Conversely, as an industry becomes more competitive and less cooperative, the stocks in the industry deserve lower multiples. Never forget that. I've talked to a lot of investors about the historical multiples at which a stock has traded. They're often very knowledgeable about the quantitative facts. However, they almost never mention the common sense qualitative situation. Companies and industries change over time. You need to account for that.
Ask Geoff a Question
Start a free seven-day trial of Premium Membership to GuruFocus.Premium Members
This article first appeared on GuruFocus .