A Valuation Rule of Thumb

What kind of earnings multiple should an investor be willing to pay for a decent company? While this may be different for every investor, Whitney Tilson shares his rule of thumb and explains how he hopes to refrain from overpaying for even the best companies. "With few exceptions," wrote Peter Lynch, "an extremely high P/E ratio is a handicap to a stock, in the same way that extra weight in the saddle is a handicap to a racehorse."

Format for Printing

Format for printing

Request Reprints


By Whitney Tilson
July 31, 2001

I'm a value investor. This simply means I attempt to buy stocks at steep discounts to their intrinsic value, which Warren Buffett defines as "the discounted value of the cash that can be taken out of a business during its remaining life." To do this, all I have to do is plug estimates of companies' future free cash flows into an Excel spreadsheet, apply a reasonable discount rate to determine intrinsic value, and buy stocks that trade at a goodly discount to this figure. What could be easier?

Not so fast. As Buffett notes, "intrinsic value is an estimate rather than a precise figure, and it is additionally an estimate that must be changed if interest rates move or forecasts of future cash flows are revised. Two people looking at the same set of facts, moreover... will almost inevitably come up with at least slightly different intrinsic value figures."

Given this uncertainty, it's critical to buy stocks only when their valuation is so low that there's a huge "margin of safety," to use Ben Graham's famous saying. Consider this exchange from the 1996 Berkshire Hathaway (NYSE: BRK.A) annual meeting:

"Warren talks about these discounted cash flows," said Vice Chairman Charlie Munger. "I've never seen him do one."

"It's true," replied Buffett. "If [a company's value] doesn't just scream out at you, it's too close."

When is a stock screaming?
Unfortunately, stocks don't scream, "I'm really, really cheap. Buy me!" Instead, investors have to figure this out for themselves. While every investment opportunity is unique, I find it helpful to have rules of thumb regarding what I might be willing to pay for different types of businesses. One I use: Pay no more than 10 times earnings for a decent business, and no more than 20 times earnings for even the greatest business.

But this statement is extremely broad, and carries with it numerous caveats. Before applying it, one must think about these questions:

  • How does one define "decent" or "great" business? I've shared thoughts on this topic in articles written in Feb. 2000 and this March that discuss sustainable competitive advantage.
  • Should one use trailing earnings or future earnings estimates to generate a price-to-earnings (P/E) ratio? I generally use trailing earnings rather than analysts' estimates -- unless I have good reason to believe future earnings will be materially different.
  • Does the company have any debt or excess cash?  If so, adjust for this, since one should compare earnings to enterprise value, not market capitalization. (Enterprise value is market cap + debt - excess cash.)
  • Is free cash flow significantly different from net income? If so, depending on the circumstances, you may want to use the former as the "E" in the P/E. (I've written a series of columns on free cash flow.)
  • Is this a cyclical stock, which tends to move with economic cycles? If so, ignore my rule of thumb. As Peter Lynch noted in Beating the Street, "buying a cyclical... when the P/E ratio has hit a low point is a proven method for losing half your money in a short period of time. Conversely, a high P/E ratio, which with most stocks is regarded as a bad thing, may be good news for a cyclical. Often, it means that a company is passing through the worst of the doldrums, and soon its business will improve."
  • Are there any unusual gains or losses that one should adjust earnings for? Be careful following a company's guidance in this area -- an increasing number of companies seem to report "one-time charges" every year. What's one-time about that?
  • Does one need to adjust for stock options?

In addition to all of these caveats, there are no studies endorsing the numbers I've picked. This is my personal rule of thumb, based on my own experience, which I hesitate to share because I can't defend it in an objective, rigorous way. 

I am doing so anyway, however, because I find that many people don't have the experience to know when a stock might be cheap; and perhaps more importantly, how risky it is to pay more than 50 times earnings (in some cases more than 100 times, even today!) for stocks such as those I've expressed concerns about in previous columns, among them Krispy Kreme (NYSE: KKD), Siebel Systems (Nasdaq: SEBL), Starbucks (Nasdaq: SBUX), and Manugistics (Nasdaq: MANU).

Starbucks vs. Outback Steakhouse
Let's use a specific example to see my rule of thumb at work. When I compared Starbucks and Outback Steakhouse (NYSE: OSI) in February, I warned investors away from the former due to its valuation, which was 61.5x trailing EPS at the time. Instead, I suggested considering Outback, which was trading at 12.6x trailing earnings.

Outback is what I'd call a prototypical "decent" business. It has steady growth, consistently high returns on equity, a clean balance sheet, and substantial free cash flows it uses to repurchase shares. "It's quite likely," I said in the column "that this is a good business (in a lousy industry, sure) that has encountered some temporary difficulties, causing its stock to become attractively priced. That's the sort of situation I like." (Incidentally, I never ended up buying the stock because I wanted to buy it even cheaper -- at no more than 10x trailing earnings -- but the stock didn't fall to that level. In fact, it's risen 10% since then.)

There's little doubt that Starbucks, with its rapid growth and powerful brand name, is a better business than Outback. But it's far from perfect -- and even if it were, I would never consider buying it with a P/E ratio more than triple my maximum of 20x. Starbucks, I concluded, was "priced for perfection," and while it was possible "the company could grow into the stock's nose-bleed valuation," the odds were "not very good, in my opinion." Since then, Starbucks is down 26%.

Other examples
Though the evidence is mostly anecdotal and short-term in nature, a look back at stocks I've discussed in columns over the past year and a half is perhaps also instructional. (More historical information can be found in a recent column and discussion board post.) The lesson, to me, is clear: If you buy solid companies with stocks trading at, say, seven to 12x earnings, good things are likely to happen. Conversely, if you buy the most popular stocks -- which are invariably trading at 50 to 100x earnings -- prepare to lose your money.

To quote Lynch again, this time from One Up on Wall Street: "If you remember nothing else about P/E ratios, remember to avoid stocks with excessively high ones. You'll save yourself a lot of grief and a lot of money if you do. With few exceptions, an extremely high P/E ratio is a handicap to a stock, in the same way that extra weight in the saddle is a handicap to a racehorse."

My goal in this column is not to toot my horn about some successful short-term stock picks, nor to argue that one should be locked into rigid P/E ratios when thinking about valuation and evaluating investments. Instead, I hope I've given some general sense of what reasonable valuation multiples look like, and -- perhaps more importantly -- how critical it is to avoid overpaying for the stocks of even the best companies.

-- Whitney Tilson

Guest columnist Whitney Tilson is Managing Partner of Tilson Capital Partners, LLC, a New York City-based money management firm. He did not own shares of the companies mentioned in this article at press time. Mr. Tilson appreciates your feedback at To read his previous columns for The Motley Fool and other writings, visit