"NOTHING HAPPENS UNTIL

somebody sells something." That bit of inspiration is a staple at seminars for everyone from auto dealers to insurance agents to software salesmen. You can even buy it, framed or as a lapel pin, and it appears on more than 2,000 Web pages. Lots of marketing gurus claim they coined the phrase. But the best candidate is the late Arthur "Red" Motley, who began his career selling manure in Minnesota and made a fortune at "Parade" magazine, the Sunday newspaper supplement he created.

I'm telling you about Motley's aphorism, hokey as it may sound, because it's also true. With the exception of a few ambitious biotech outfits, every business depends on sales. There are times when investors focus too closely on the revenue side. And, of course, the pendulum swings the other way: Remember when Internet companies soared based on clicks, pops and other dubious measures? Today it's earnings and balance sheets that are under the microscope, and revenue often gets less attention than it deserves.

So I'm going to revisit one of my favorite value yardsticks, the price/sales ratio. I'll explain how this tool can help you analyze companies more effectively, and I'll tell you how one mutual fund has used price/sales screens to outpace the market by 2 to 1 over the past decade. I'll also report on new research about companies that beat earnings estimates. Which surprises are more significant, those that come because of lower expenses or because of higher sales? Given what I've said so far, the answer should be easy. But the market doesn't always catch on and that disconnect provides a nifty opportunity that I exploited to compile this month's table of stocks.

Like price/earnings and price/cash flow, price/sales is a way of measuring value. These yardsticks tell you how much profit, cash flow or revenue you get per dollar you invest. Cheaper is usually better, and cash flow from operations is my favorite benchmark. But price/sales has some special advantages.

First, it's a snap to calculate. Just divide a company's market value (share price times shares outstanding) by its annual revenue. Second, it's stable and doesn't bob up and down like earnings or cash flow. And it works even when a company is losing money and other indicators go off the charts. Some reference points: The average price/sales ratio for the S&P 500 is 2.3, a number that ranges from 16.4 at eBay to 0.04 at Delta Air Lines. That means you've got to pay $16.40 for every revenue dollar at the Internet auction outfit, but you can get a revenue dollar for just 4 cents if you buy Delta.

Price/sales analysis can be particularly useful when comparing companies in the same industry. Take supermarkets. P/E ratios are all over the aisle. But the big three (Albertsons, Kroger, Safeway) sell for nearly identical P/S ratios of 0.24. Whole Foods, meanwhile, has a P/S of 1.46 (nearly twice as high as Wal-Mart's). But you can buy small-cap competitor Wild Oats for a P/S of just 0.20. Bargains, anyone?

Sales are also less subject to manipulation than earnings. Spotting chicanery can be easy. Just monitor accounts receivable, which is the balance sheet entry for unpaid bills. If receivables are growing faster than sales, it means either that customers are in trouble or that they're being pressured to load up on inventory.

Nearly a decade ago, James O'Shaughnessy published "What Works on Wall Street," a bestseller chockfull of stock-performance statistics. One of his insights was that portfolios with low P/S ratios and strong momentum (in terms of share price gains) were consistent winners. O'Shaughnessy created a family of mutual funds to capitalize on his ideas, which led to several successful columns for me.

Neil Hennessy another money manager who likes quantitative strategies purchased O'Shaughnessy's funds in 2000. Hennessy Cornerstone Growth is O'Shaughnessy's original low price/sales portfolio, and managing the fund is a snap: Screen for stocks with P/S ratios under 1.5 and rising annual earnings, buy the 50 with the best recent price gains, and hold them for 12 months.

The approach may sound simplistic, but the results are good enough to embarrass both active managers and indexers. Since 1996, Cornerstone Growth has racked up an average annual return of better than 17%. That's roughly twice the comparable figure for the S&P 500 and the small-stock Russell 2000 and powerful evidence that P/S is a value benchmark worth watching. Despite top-tier performance, however, Hennessy's fund has only $1 billion in assets. One drawback is a 1.25% expense ratio, pricey for a portfolio that runs on autopilot.

Among academics, meanwhile, revenue has become a moderately hot topic. Recent research has centered on earnings surprises, which result from higher-than-anticipated sales or lower-than-anticipated expenses. As you might expect, revenue growth can boost share prices more than cost-cutting. Earnings increases caused by higher sales also tend to be more persistent. But investors don't discriminate as well as they should.

A series of working papers, including two by Narasimhan Jegadeesh at Emory University and Joshua Livnat at New York University, highlights a profit opportunity. More than 30 years ago, economists discovered that estimate-beating stocks are good buys. They tend to outperform the market by something like two percentage points in the three or four months following the earnings announcement. Given the track record of earnings-beaters, why don't investors bid up their prices as soon as the news of their earnings is released? The best explanation is stick-in-the-mud conservatism.

Jegadeesh and Livnat, however, discovered that the proceeds from a surprise strategy can be nearly twice as fat more like four percentage points over four months if you zero in on companies that beat both earnings and revenue estimates. Not only is the market too skeptical about earnings surprises, but it also fails to understand the difference between expense-driven and revenue-driven gains. Most databases track consensus estimates for sales as well as earnings. So these numbers aren't difficult to find.

To create this month's table, I tried to piggyback on the professors' work. I'm writing in February, so to avoid stale data I looked only at companies that have reported results for the December quarter. I also wanted stocks whose consensus estimates for both earnings and revenue are available. That's a universe of nearly 600 firms.

In line with academic convention, I ranked these companies two ways by earnings surprise and by revenue surprise. (The most common measure of surprise is the difference between quarterly results and the consensus estimate; if the consensus is $1 and earnings are $1.10, the surprise is 10%.) I looked for companies in the top 20% on both lists. Then I applied my normal size minimum, lopping off stocks with a market value below $500 million.

My eight finalists are remarkably diverse whether you're talking size, industry or price. Everyone knows Apple, Bear Stearns, Deere and Exxon, though they have little else in common. Lower-profile outfits like Commercial Metals and Valero are in commodity markets, which makes them unusually cheap in price/sales terms. And the products they sell, steel and re?ned gasoline, respectively, have been on a tear. But there are tech and growth stocks, too: M-Systems makes flash memory for PCs and cell phones, and Resources Connection provides temporary accountants and other specialists to help with things like Sarbanes-Oxley chores.

The common thread, of course, is that each of these firms posted earnings gains that beat expectations. They also racked up a significant revenue surprise. That puts them in a very select group. To add a corollary to Red Motley's insight that nothing happens until somebody sells something: And if the market doesn't catch on immediately, that spells opportunity for those who do.