Taking the Measure of a Stock

Discounted cash flow tells what other methods don't

In July, 2000, as Ariba (ARBA ) soared above $130, Aswath Damodaran, a New York University finance professor, decided the Internet software company was worth a little more than half that--a valuation that has since proved far too generous. What did Damodaran know that the stock market didn't? A better question might be: What method did he use to value the stock that gave him an answer so different from the market's?

While sell-side stock analysts were using price-to-sales ratios to justify high-flying prices--often dividing the price by a sales number bloated on wildly optimistic projections--Damodaran and others like him slogged through something called the discounted-cash-flow model. The method takes a lot more work than simply dividing a stock's price by its per-share sales or earnings, then deciding if the result is reasonable. But for your effort, you get a picture of how such variables as future growth and interest rates will affect the value of the stock. Also, instead of trying to come up with a company's value, some use the model as a reality check, plugging in the current stock price to calculate how fast a company would have to grow to justify it.

TIME-TESTED. Discounted cash-flow models aren't an untested device frantically developed in the aftermath of last year's market bubble. Money managers and academics have been using them for decades, coming up with numerous variations: the dividend-discount model (best suited for companies paying dividends), the free-cash-flow-to-the-firm model, and so on. Investment bankers also use these models to price companies involved in mergers or acquisitions.

The models, with their various permutations, are all an attempt to do the same two things: First, they look at factors such as growth rates and profit margins to project how much money a company can generate in the future. Then, they discount that cash to come up with a present value. How much of a discount depends on the interest rate available from a risk-free investment, the company's cost of capital, and the riskiness of the stock.

What these models really give you is an appreciation for what drives stock values. Changes in the long-term growth rate seem to have the greatest impact on growth companies, with next year's earnings projection and, of course, changes in interest rates, also making a big difference. Consider the numbers generated by a discount model offered by Bloomberg Financial Markets (table). McDonald's (MCD ), which recently traded in the market at $27.50, is valued at almost $24, making it a little overpriced. The model assumes a long-term growth rate of 11.5% and projected earnings per share next year of $1.70. Cut the growth rate by 30%, and Mickey D's valuation plunges to $17; lift the growth rate 30%, and the stock value climbs to $34. In a like manner, raising interest rates by one percentage point shoves the valuation down to $18; a one percentage point drop boosts the value to about $35.

Manually working through all these numbers might bring back bad memories of math class. Although Bloomberg's helpful model is available only to those paying $1,600 or more a month for its financial information service, fortunately there are also a few free Web sites with valuation models. They let investors simply plug in the numbers or adjust ones already there. The programs then complete the calculations.

Damodaran's Web site (www.stern.nyu.edu/˜adamodar/) might seem a bit too much like a textbook for some investors. But it's a feast for valuation buffs. He offers nine different discounted-cash-flow models set up in Excel spreadsheets. There's also informational material explaining the nuances of the models and telling users how to estimate the various inputs. And if discounted-cash-flow models aren't enough for you, there are sections on other approaches to valuation.

For most investors, ValuePro.net will be easier to use. There's only one model, and it's all on one page. The site, also free, was set up by Gary Gray, a visiting professor at Pennsylvania State University. He, along with two other Penn State faculty members, Patrick J. Cusatis and J. Randall Woolridge, wrote Streetsmart Guide to Valuing a Stock, a useful book for investors who want to work through their own valuation model. (The book costs $29.95 and is published by The McGraw-Hill Companies, which also publishes BusinessWeek.)

The ValuePro model is relatively simple (table). Type in a ticker symbol, and the model will come up with appropriate figures drawn from the company's financial statements, analysts' earnings forecasts, and the stock and bond markets. Of course, you may want to change some of the inputs, particularly the growth rate, which is based on Wall Street analysts' often too-optimistic forecasts. For JDS Uniphase (JDSU ), a maker of optical networking equipment, analysts predict 33% earnings growth, despite the meltdown in the telecom sector, the main customer for its products. Cut the JDS growth rate in half, and the model slices the stock's value by two-thirds.

GARBAGE IN? Even with a relatively simple model such as ValuePro's, the investor's ability to make good projections is absolutely critical. "Theoretically, the discounted-cash-flow model is the perfect measure," says Bala Iyer, director of quantitative research at Banc One Investment Advisors. "The problem with it is that it relies a lot on forward-looking forecasts, and because of that it's susceptible to error."

That's not the only problem. The model isn't suited for short-term investing--since it focuses on long-term value. And just because the model says a stock is worth $50 doesn't mean it will trade for that anytime soon. Using a good cash-flow model may help you avoid buying into a bubble, but it might make you miss those big runups that can be profitable (if you sell at the peak). Also, focusing too much on just numbers might cause you to overlook an unusual opportunity like, say, Microsoft (MSFT ) a dozen years ago. The company never looked cheap, but its ability to dominate the PC software market made it a financial powerhouse and investor's dream.

With all the caveats, Damodaran still argues that discounted-cash-flow models make the best valuation tools. He says analysts who rely on price-earnings ratios also make assumptions about growth when they decide what p-e is justifiable for a stock. They just don't bother doing it explicitly. Without weighing all the elements that are in the discounted-cash-flow model, says Damodaran, valuation becomes a beauty contest--with stocks compared with each other rather than judged on intrinsic value. "If the companies you are comparing your company to are all overpriced," says Damodaran, "what you end up with is a stock that drops by 60% or 65%." That's something easier to imagine now than it was two years ago. Besides, he says, focusing only on earnings puts investors at the mercy of companies adept at jiggering the bottom line. Cash flows are more difficult to manipulate.

It's the sell-side analysts--the talking heads most likely to pop up on cable TV--who are enamored of relative valuation methods such as price-earnings and price-sales ratios, Damodaran says. The same firms' money managers, serving their well-heeled private clients, may opt for more conservative valuation methods. Christopher Wolfe, an equity strategist at JPMorgan Private Bank, which manages $310 billion for rich investors, says its discounted-cash-flow model alerted its portfolio managers last year to the tech-stock bubble. "Unfortunately, our sell side was very aggressive," he said.

SHIFTING SANDS. Even its strongest advocates say discounted cash flow is no sure bet. And it's a moving target as well. If the company blows its latest quarterly earnings, if its major customer files for bankruptcy, or if interest rates take a dramatic turn, you have to rerun the numbers. Says Damodaran: "Any time your expectations change, your stock's value is going to change."

Remember Damodaran's valuation of Ariba at $72? The stock recently traded at $7.73, although the professor now estimates it's worth about twice that. Keep in mind that an investor using a cash-flow model would not have bought Ariba at all when it was trading in the triple digits. A tool that helps you skirt disasters like that is extremely valuable. If you don't believe it, just ask an investor who purchased Ariba at $130.

By Carol Marie Cropper

    Before it's here, it's on the Bloomberg Terminal.