Super Stocks: A New Assessment of Ken Fisher’s Pioneering Book

This article is the fourth in a series about screens designed by famous investors. The first, on Benjamin Graham, can be found here; the second, on William O’Neil, can be found here; the third, on Joel Greenblatt, can be found here; and for an overview of the subject, see my article “Can Screening for Stocks Still Generate Alpha?”

Kenneth L. Fisher finished writing his very first book, Super Stocks, in September 1983, when he was 32 years old. Back then, it made sense to write,

First, get a personal computer. . . . It would help to have two disk drives. Then get a modem and communications software. . . . Through a normal push-button telephone line and modem, [personal] computers can access broad data-base services such as “The Source” by Reader’s Digest. . . .

Kenneth Fisher, Super Stocks

Despite some dated stuff like this, it’s still a good read, full of good advice. And surprisingly, Fisher’s system still works, despite enormous changes in the market over the last 38 years. That’s more than I can say for most stock market writers.

What Is a “Super Stock”?

Fisher defines it right up front.

Fisher was a true pioneer here. He wasn’t kidding when he opened the book with the words “THIS BOOK OFFERS CONCEPTS NEVER BEFORE PRESENTED.” Those concepts were two value ratios: the price-to-sales and the price-to-research ratios. At the time he was writing, no other investors used these, and nor did academics. James O’Shaughnessy, author of What Works on Wall Street, named Fisher as the first person to define and use the price-to-sales ratio, and the price-to-research ratio is still on very few people’s radars. He was also one of the first people to identify small-cap value as a universe worth looking at. It’s one of the many reasons that Fisher became one of the richest Americans, a billionaire whose firm, Fisher Investments, now manages $160 billion.

Super Stocks is a guide for discretionary stockpickers. Unlike Benjamin Graham or Joel Greenblatt, Fisher did not lay out a very specific step-by-step strategy for picking stocks. There are only two hard-and-fast rules in the book: never buy a stock with a price-to-sales ratio greater than 3, and if you’re looking for Super Stocks, avoid any stocks with a price-to-research ratio greater than 15.

Fisher doesn’t give any rules for liquidity in his book. After all, in order to buy a stock back then, you had to call your broker, so stocks weren’t really that liquid to begin with, compared to today. So I’ll use some off-the-cuff rules here. If you stick with stocks with a market cap of $30 million, a price greater than $1, and an average daily dollar volume greater than $50,000, you still have between 150 and 250 stocks to choose from if you follow Fisher’s rules. How you choose them is the subject of most of the book.

Most of the criteria that Fisher identifies are unmeasurable. Five of the key criteria are (and I use his terminology here):

• growth orientation,
• marketing excellence,
• an unfair competitive advantage,
• creative personnel relations, and
• the best in financial controls.

Of these, only the last is somewhat quantifiable. If the company has negative income or negative cash flow, there should be more than sufficient working capital to cover it, and the debt-to-assets ratio should be relatively low.

Fisher also places a good deal of emphasis on margins, but a super stock doesn’t have to have strong margins when one buys it. Instead, it has to have strong evidence that its margins will be good. Fisher believes that a company’s market share, along with the market share of its largest competitor and the growth of the industry, has a lot to do with its margin potential.

There are a few other quantifiable odds and ends in Fisher’s approach—he favors companies in industries without giants; he favors companies with few analysts. But these factors are not as important as his two value ratios, the five criteria above, and the possibility of strong future margins.

Creating and Backtesting a Screen for Super Stocks

Let’s first stick with Fisher’s two-rule system—low price-to-sales and low price-to-research (and research is best defined as R&D expenditures)—and use the minimum liquidity rules I gave above. Fisher is a long-term buy-and-hold investor, so let’s try a rolling backtest with a two-year holding period. If we assume slippage of 0.5% per round-trip transaction, the average two-year return per stock that passes this test since 1999 is 36.44%, compared to the S&P’s average two-year return of 15.55%. (I used Portfolio123 to perform this test, relying both on Compustat and FactSet data.) If we look at just the past ten years’ worth of two-year returns (which means starting twelve years ago, in January 2009), the average return goes up to 44.58%, compared to 30.09% for the S&P.

Fisher suggests that we should really focus on stocks with a price-to-sales ratio of less than 1.5 and a price-to-research ratio less than 10. Stocks with price-to-sales between 1.5 and 3 aren’t forbidden, but are not preferred; the same goes for stocks with a price-to-research ratio between 10 and 15. These two changes narrow our universe considerably: now we’re buying only about 100 stocks at a time. And now our results are even higher: the average stock since 1999 gets a two-year return of 45.65% and the average stock since 2009 gets a two-year return of 52.03%.

If we narrow our universe even more and just buy the ten stocks with the lowest price-to-sales and price-to-research ratios, we get even better returns. Because Fisher places more emphasis on the price-to-sales ratio than the price-to-research ratio, I’ve ranked these stocks according to the product of the square of the price-to-sales ratio and the price-to-research ratio, and chosen the lowest.

The average of these top-ten stocks goes up to 47.26% since 1999 and 79.43% since 2009. Moreover, the former per-stock results reflected the fact that you bought a lot more stocks during a down market than during an up market; this 10-stock backtest simulates an even portfolio, and therefore it’s harder to get good per-stock numbers from it.

If we try to add some of Fisher’s other criteria using a screen, we’re necessarily extrapolating wildly from what he wrote. Using a ranking system loosely based on them may be a bit better, but when I tried this, it didn’t improve backtested results, and it’s once again a real extrapolation. The simple value system that Fisher came up with in 1983 still works as a general system for buying stocks.

Sell Rules

But holding stocks for two years is quite arbitrary. Fisher actually lays down some hard and fast sell rules. Only sell a stock when its sales ratio is between 3 and 6—or if you judge, based on the above criteria, that it’s no longer a super stock.

We can simulate the first part of this strategy using Portfolio123’s simulation engine. We’ll use the same liquidity rules as we used above, and the narrower limits—price to sales less than 1.5 and price to research less than 10. We’ll buy every stock that passes those limits every week and hold until either the company gets bought, goes out of business, or rises to a price-to-sales of 3 or higher. If we run this since January 1999, we get a record of buying close to 3,000 stocks—quite a huge sample. Their average return is 70.52% with an average holding period of 3.5 years. If you held the S&P 500 for 3.5 years over this period your average return would be 30.77%. The portfolio itself would vary quite hugely in size, sometimes holding over 600 stocks and sometimes fewer than 300; the annualized return of its equity holdings would be 18.33%.

We can also simulate the strategy over only the last ten years, buying stocks from January 2011 until now. The average return is now 62.73% over an average holding period of a little over three years. The annualized return of its equity holdings is 15.38%.

A Look at Past Screen Results

Three years ago today, the top ten stocks ranked by the product of the square of the price-to-sales ratio and the price-to-research ratio included three Chinese companies about which reliable information would have been difficult to come by. Besides these, the top ten stocks were Eastman Kodak (KODK), Renesola (SOL), Centrus Energy (LEU), Unisys (UIS), Radisys (RSYS), Ophthotec (which later changed its name to Iveric Bio (ISEE)), Voxx International (VOXX), Perion Network (PERI), Toshiba (TOSYY), and Aviat Networks (AVNW). Some of these stocks turned out to indeed have Super Stock returns: 500% or 600%; some of them just sat there; and one of them—Kodak—tanked, losing 20%. Would it have been possible three years ago, using Ken Fisher’s methodology, to separate the wheat from the chaff?

Perhaps not. Out of these ten stocks, the “Super Stocks,” in terms of their return, turned out to be Renesola, Centrus Energy, and Perion; back in January 2018, the potential of the first two of these companies didn’t look a whole lot better than Kodak’s or Toshiba’s. Of these ten stocks, only Ophthotec and Aviat had improving margins at this point, with Perion looking hopeful as well. Here are my very rough (and rosy) guesses about how these stocks would have met Fisher’s criteria in January 2018:

If you had blindly invested in all ten stocks and held for the next three years, you would have earned 194%, compared to only 47% with the S&P 500. If, on the other hand, you had stuck only with the stocks with higher “Super Company” potential—Ophthotec, Perion, and Aviat—you would have earned 190%, which isn’t much different.

This, of course, doesn’t prove anything. It’s a sample of ten stocks out of about 3,000 that passed the screen over the last 22 years. Perhaps if I had the energy, time, and research tools to go through all 3,000 I would find that Fisher’s criteria actually made an appreciable difference. Or perhaps I’d find that they didn’t.

What Companies Pass the Screen Today?

The ten highest ranked stocks according to this system (square the price-to-sales ratio, multiply by the price-to-research ratio, and look for the lowest numbers) are—if we exclude companies that are under M&A conditions or have filed for bankruptcy—Tenneco (TEN), Weatherford (WFTLF), Power Solutions Intn’l (PSIX), L.S. Starrett (SCX), Goodyear (GT), Nissan (NSANY), Ion Geophysical (IO), Diebold Nixdorf (DBD), Ford (F), and Cooper-Standard (CPS). Interestingly, half of these companies are in the automobile industry. Are any of them “super stocks”? If we look at the first part of Fisher’s definition—stocks that increase 3 to 10 times in value in three to five years—I would guess that some of them will be. But if we look at the second part, which concerns “Super Companies,” I doubt it.

Conclusions

Last year I wrote about the value of the ratio of R&D to market cap, so it was a great pleasure for me to discover now that Ken Fisher had advocated using the same ratio back when I was an undergrad. Of the many books on investing I’ve read, Super Stocks is one of the more prescient. Fisher figured out how to beat the market by using innovative takes on value ratios, and his strategy has been working well ever since. I find that quite impressive. Joel Greenblatt’s 2005 Little Book that Beats the Market, like Super Stocks, offered a rather simple two-factor investing model; but that formula stopped working about five years after he published it. The formula Ken Fisher offers in Super Stocks has been working for over 35 years.