The price-to-sales ratio (PSR; sometimes P/S) is a good example of groundbreaking capital markets technology I pioneered that was powerful in its day yet isn’t so much now. I had uncovered a way no one had yet used to tell if a stock might be over- or undervalued—it became the subject of my 1984 book called Super Stocks. While Ben Graham made passing mention of the relationship between price and sales being potentially interesting, the first published work anywhere on the relationship was mine. I’m very proud of that—just like I am of my third-grade school report on Guatemala. But otherwise, neither is noteworthy today. Just memories.
But 30 years ago, if you could simply screen for low-PSR stocks (which wasn’t easy), you could beat the market more often than not. After my book and subsequent exposure, the PSR became widely used and even, off and on, part of the required curriculum for the CFA exam. Most analytical websites include the PSR today. But as capital markets technology and a forecasting tool, the PSR has become largely priced into the market. Even a great discovery becomes obsolete with popularity and time. If it becomes popular, it loses its power. It’s always time to be working on the next discovery.
For the uninitiated (in the event the name didn’t give it away), the PSR tells you a stock’s price relative to its per-share sales. It’s just like a P/E but uses annual revenue or sales where the P/E uses earnings. A stock selling for $25 with $25 in sales per share has a PSR of 1. Pretty straightforward. This may not sound revolutionary now—like trying to imagine a time before someone said, “Hey, what would happen if we divided a stock’s price by its earnings?” But when I first started writing about it, no one had done it.
For the purposes of this book, there is no need for an in-depth rehash of PSRs or my first book, Super Stocks. But there are a couple of points I should make. First, about the book. Reading old investment books is useful and a great way to learn the canvas of what developed in terms of capital markets technology—when, how and by whom.
Finding Earnings When There Were None
Second, anyone who plans on making money in stocks knows you want to buy a stock before it becomes “in favor.” The requisite flip side is you must buy the stock when it seems out of favor. The trick is knowing which stock will be in favor soon while currently seeming doggy. How can you know that? That was how I got interested in PSRs initially. My early PSR evolution translates directly into how to think about developing capital markets technology.
Investors have long used P/Es to look for cheap stocks. One hundred years ago—plus! But some emerging companies may not have any earnings to report. Even established companies may be profitless during cyclical downturns and times of individual corporate crises, and those can be interesting times to consider a stock—when it’s out of favor. You can’t get a P/E in those scenarios because you can’t divide something by zero. Sometimes a company has a P/E of 1,000 because its earnings have almost completely disappeared. Sometimes a company has a P/E of 5 because of temporarily high profit margins that can’t be sustained.
But even if a firm doesn’t have earnings, it still has sales (or at least it should, or it’s in big trouble). This is where Question Two came into play for me decades ago—what could I fathom others couldn’t? It made sense any stock with a low price relative to its previous 12-month sales would rise—if its future earnings might become large enough to make the current PSR translate into a low future P/E ratio (or in other words, a high future earnings yield). What people don’t like today, they would tomorrow. How tough is that? People would eventually discover this undervalued stock was posting super revenue and fat future profit margins and was cheap—and they’d come around and drive the price up. Hence, if you were rational, you’d want to buy a stock when its price relative to the company’s sales was low but didn’t appear cheap based on P/E. Low relative to the market but, more important, low relative to its category and low relative to its future earnings.
When I wrote Super Stocks, I defined low PSR stocks as generally ones where the company’s total market value was less than 75% of the company’s total annual revenue. I defined high PSR stocks as ones where the market value of the company was more than three times its annual revenue.
There was the theory, but that’s all it was. At the time, there were no sources citing PSRs like there are now. There were no databases. Bloomberg.com and Morningstar.com didn’t exist to neatly calculate a PSR for every stock for me like they (and many other sources) do now. I extrapolated data from publicly available information and built my own data to arrive at my ratio. Back then, there was actually money to be made by compiling data because data was still scarce and expensive. Today, data is essentially free. If you’re younger, you may have difficulty realizing just how hard data was to get. In 1981, I paid Goldman Sachs $20,000 for a simple one-time screen of the New York Stock Exchange (NYSE) based on current PSRs. It was that expensive for something anyone can get now for free, near instantly. Historical data had to be built by hand, which was effectively prohibitive unless you knew exactly what you wanted to do.
When I back-tested historical PSRs against subsequent stock market returns, my theory held up. I built data against several stock universes ranging from a 1970s Tech universe compiled by the former investment banking firm, Hambrecht & Quist, to general 1930s stocks based on Moody’s data retrieved by hand. Stocks with lower PSRs did far and away better than those with higher ratios. Not every single time with every single stock, but enough to provide me with a reliable indicator for forecasting and a good basis for a market bet. In other words, stocks with low PSRs were superior stocks, which ultimately led to my book title.
Before I wrote Super Stocks, I used this nifty new technology with a fair amount of success. Effectively, my usage of PSRs propelled my career in many ways. You may again wonder why I would advertise something giving me a competitive edge in a mass-produced book. You may think I should have kept what I knew about the PSR a secret so I could have maintained my advantage. Not so! Wrong way to think.
Any advantage you have is likely temporary. Behind you, there is someone else looking for what you just found. I knew I had discovered something neat, but I hadn’t done anything magical or prohibitively mathematically complicated or both. I hadn’t done anything anyone else couldn’t do. All I did was look at available data (hard-to-get data, but available data nonetheless) in a new way and do a little back-testing and a little linkage to fundamental theory. Other people were bound to see what I saw sooner or later. Probably sooner! So I kept putting the idea out there to see if anyone else was biting. For a while, no one was.
After I published the book, people bit, but it took a long time. For about 10 more years, I had the PSR pretty much to myself. For example, in 1997 James O’Shaughnessy wrote a best-selling book titled What Works on Wall Street in which he analyzed all the common ratios to see which might lead to higher subsequent returns.3 He labeled the PSR “the King of the Value Factors,” and he claimed his analysis (at the time) showed PSRs generated higher subsequent returns than any other single ratio. Jim’s handwritten inscription in the copy he gave me reads, “Boy did you ever come up with a good ratio! Think how horrible us poor money managers would feel if instead of market cap, S&P had based their index on low PSR stocks.” O’Shaughnessey’s book propelled PSRs further into prominence, and soon, the PSR lost most of its power. It got priced.
Was I upset my innovation became widely used and therefore thoroughly priced? Not at all! I was thrilled. First, if it really was good, it was inevitable—at least this way it happened on my terms and I was ready for it. There was no way the dot-com era would arrive with all its free data without someone stumbling on the PSR and popularizing it. I got a good run from that technology and was able to see when it was time to move on. By then, I was off on other new things I’d never dreamed of when I first did the PSR. I wasn’t so overconfident (a cognitive error) in the 1970s and early 1980s to assume I had out-thought all investors for all time. If the PSR was clever in the early 1980s, in a world of CP/M-based PCs with 5 ¼-inch floppy discs and no hard drives, the guys who created the Commodore 64 were pretty clever, too. The subsequent wave of 1980s and 1990s electronics would wipe out the world of expensive data. It would make the PSR visible to everyone. Markets evolve, and so must we.
If the PSR no longer works much, why waste ink telling you about it? First, because some people are still convinced this antiquated piece of technology still works as it did. With any ratio like this, once popularized, sometimes it works and sometimes it doesn’t—just enough to keep people interested—same with the P/E, dividend yield, price to book, you name it. You can even take the most nonsensical ratios and find times when they appear to generate excess returns. Just looking at the companies with the highest cash per share will sometimes give you market-beating stocks. Looking at companies with the lowest cash per share will sometimes beat the market, too. But neither has anything to do with being able to beat the market in the long term.
Over the past 20 years or so, low-PSR stocks have been a bit less volatile than both low-P/E stocks and the market and have no long-term, risk-adjusted, excess return—now. They’re priced. However, when value stocks have outperformed growth, low-PSR stocks have generally beaten both the market and value and been more volatile than the market overall. When growth stocks beat value stocks, low-PSR stocks have lagged the market and value.
What can you do with this knowledge right now—the way you’ve learned to think in Chapters 1 through 3? How can you use what I’ve told you to make rational, workable bets here and now—at least until making those bets becomes well known and popular? Let’s do a Question Two. What can you fathom no one else fathoms about this? We saw in Chapter 2 how to know when growth stocks are likelier to best value stocks, and vice versa. I just told you when value stocks lead the market, PSRs have still had extra oomph relative to value but not when growth stocks lead historically. So I just told you when you’re likelier to get some temporary excess return by using PSRs as a screening device. The PSR isn’t a uniform tool for all seasons—it’s a tool to use temporarily when incorporated with other successful capital markets technology. Your goal is to figure out when value stocks will lead the market and then incorporate low PSR stocks into your stock selection—but not when growth stocks lead the market. When growth leads the market, you want your stocks to be higher PSR stocks. And like any good capital markets technology, be sure to keep testing this in the future to ensure the relationship hasn’t broken down.
The Stock Market Is Not a Ball-Peen Hammer—But It Can Hit You Hard
Briefly consider this another way: Some category of stock outperforms the market for five years. A preponderance of investors jumps on its bandwagon, doing whatever it was that was so successful in those years. But those things stop working for the next five years or so, leading investors to think they will never work again. Because investors think they won’t work anymore, it’s very possible for them to start working again. They’re no longer discounted into pricing but simply ignored because of cognitive error. This is what happens to low PSR, low P/E, dividend yield and others. They come through long periods where they haven’t worked, so they’re ignored for another long period. Then, when value comes back into favor, they can and do work temporarily. Traditional craftsmen hate these kinds of very real market phenomena because they want their tools to work the same way all the time.
This further illustrates the importance of continued testing and ongoing innovation. I repeatedly tested my PSR technology before deploying it so I wouldn’t be guilelessly relying on it when it became priced and useless. I gave up on it long ago as a primary tool and have since developed many other capital markets technologies because developing the next new thing is what the game is all about. But it is still a secondary tool that can be used at times.
Sure, you can use the Three Questions a few times and stumble on a thing or two others don’t know. But if you don’t make the Three Questions a part of your way of thinking, always, you will eventually lose any advantage you might have stumbled on.
Let’s take another sidestep to Warren Buffett. A quality many miss about Mr. Buffett is his ability to morph. If you read his materials from the 1960s, he said very different things than in the 1970s and early 1980s. Early on, he was buying dirt-cheap stocks by simple statistical standards and typically smaller stocks—which would today be referred to as small-cap value (although that term didn’t exist until the late 1980s). Later, he bought what he called “franchises.” Then he entered a period of buying great managements of big companies and being a long-term holder—otherwise thought of as big-cap growth today—that many ascribed to the influence of my father coupled with Charlie Munger.
When Mr. Buffett was buying Coke and Gillette, you couldn’t quite reconcile those activities with the kinds of things he owned two decades earlier. Then, amazingly, 12 years ago, at just the right time, he was buying smaller things dirt cheap again just as value came back into play as the twenty-first century began. While Buffett never lost the core of what he was doing or what he was looking for, he tactically morphed steadily over the decades. Trying to freeze his tactics from any decade and replicate them in the next few would never have led you to his actual actions. There’s nothing wrong with that. It’s as it should be. That he doesn’t develop capital markets technology is just his way because—I think—he is mainly intuitive and, in that regard, very rare. But whether developing capital markets technology or being instinctual like Mr. Buffett, morphing, adapting and changing are fundamental to success. Stagnancy is failure long term. Since I don’t know how to be instinctual, I rely on the Three Questions and building capital markets technology.