BRIGHT ECONOMISTS WIN

Nobel Prizes. Successful money managers get rich-and write bestsellers about how to invest. Even financial writers pontificate on TV talk shows. But there are no famous accountants.

Maybe there should be. I'm not talking about corporate number crunchers the miscreants at Arthur Andersen, HealthSouth or WorldCom. I mean smart researchers who know how to squeeze value from balance sheets and income statements. When growth stocks were hot, technology gurus were stars. Now that dividends are back in style, why not put an occasional accountant in the spotlight?

My favorite rising star is Richard Sloan, a 39-year-old professor at the University of Michigan. In his most influential work, Sloan develops a straightforward stock-picking strategy with the strongest results I've ever seen. Hypothetical gains average close to 20% a year. Sloan is also intensely practical. His students manage money in real time, and his research buy and sell recommendations included is easily accessible on the Web.

Sloan's landmark paper, published in 1996, documents what's called the accrual anomaly. It takes a bit of explaining, but you won't feel cheated when you finish this column. The key is realizing that every company's earnings have both a cash and a noncash component. Suppose earnings at my widget factory were $800,000. But cash flow, the increase in my bank balance, was $1 million. That $200,000 difference comes from lots of messy adjustments depreciation, changes in inventories and receivables, bad-debt allowances. Collectively, they're called accruals.

Critics, including many economists, see this distinction as pointless niggling. But Sloan discovered a powerful link between these noncash earnings and stock prices. For years value investors have been tracking cash flow along with earnings. Sloan, who audited speculative mining companies in his native Australia before getting an accounting Ph.D., looked at this relationship in a more creative way. He ranked companies based on the size of noncash earnings relative to total assets, a statistic I call an accruals ratio. Then he measured how their shares performed in the year after these results were announced.

On average, companies with the highest accruals ratios (lots of noncash earnings) are losers, big time. Companies with strongly negative accruals ratios (cash flow that exceeds earnings) are consistent winners. In Sloan's latest work, based on back-testing for the years 1961 through 1999, a strategy that exploits this gap produces portfolio gains of more than 17% a year.

Sloan draws two conclusions here: The first is nothing new, at least to investors who are into fundamentals. When you see lots of noncash items on a company's earnings statement, stay away. On the other hand, start getting interested whenever cash flow exceeds reported earnings.

The other message from Sloan's research is more surprising. Even though accruals provide valuable information about stock prices, few people pay attention. That's why Sloan's strategy racks up those outsize returns his analysis identifies winners and losers before others catch on. Sloan thinks this is because investors focus too closely on earnings and ignore valuable information in other accounting data. But it might just be that they're lazy.

Sloan lectures regularly to professional investors. His results have been circulating in the academic community for nearly a decade, and others have tried unsuccessfully to poke holes in his work. Another study, by Brian Bushee at Wharton, tested several stock-picking strategies under real-world conditions. Most were duds; Sloan's wasn't. Yes, potential gains are diminishing, but very slowly. Many market anomalies disappear almost as soon as they're discovered.

www.alphaseeker.com. You can also visit a student Web site (www.earningstorpedo.com that highlights companies with particularly weak earnings numbers. Another site (www.valuedog.com features bargains based on the work of Joseph Piotroski at the University of Chicago. Piotroski, about whom I've written several columns, studied at Michigan with Sloan.

If you're more ambitious, it's not too difficult to calculate an accruals ratio Sloan's benchmark on your own. Sloan derives this statistic from balance sheet entries, but the annual statement of cash flows is a neat shortcut. Here's an example using trailing 12-month data for Russell, one of the companies in my table.

Start with net income. In Russell's case it's $35.1 million. Then subtract cash flow from continuing operations ($149 million, mostly gains from depreciation) and cash flow from investing (-$36.6 million, mostly capital spending). The result closely matches Sloan's definition of total accruals. In Russell's case, it's -$77.3 million, a good sign. When accruals are negative, a company isn't bulking up its reported earnings with lots of noncash charges.

The next step is to put this number in context. Sloan looks at accruals relative to a company's size. Russell has total assets of $991.7 million. Divide assets into accruals, express the result as a percentagex, and you have Russell's accruals ratio, -7.8%. These days, accruals ratios for nonfinancial companies split down the middle half positive, half negative (doing this stuff for banks and insurers is tricky, so I'm ignoring them). The extremes are where Sloan's strategy kicks in. Anything above 5% is a sell; below -5% is a buy.

This isn't a magic bullet, of course. You can't invest in one or two stocks based on accruals information alone and expect to get rich. Sloan's results depend on the law of averages large portfolios, buying winners and selling losers. Out of curiosity, however, I did some research. Anyone who ran Sloan's numbers for Enron or WorldCom in 2000, well before the stocks collapsed, would have been very nervous. Both companies had accruals ratios above 10%, deep into alarm territory.

www.simplystocks. com, a new fundamental-analysis package. It's a creature of the global economy: More than 300 accountants and other employees in India gather electronic data from SEC filings. Subscribers screen and generate reports on the Web. Basic service is $40 a month, with a five-day free trial.

To create the accompanying table, I used trailing 12-month cash flow results numbers that are tough to get elsewhere. As a first cut, I combed the 10,000-company Simplystocks' universe for accruals ratios of -5% or lower. Fewer than one company in 10 passed the test. Then I focused on midcap stocks with market values between $500 million and $1.5 billion. Finally, I looked for the lowest price/earnings and price/book ratios, and I wanted companies with less debt than equity.

I'm happy with my eight finalists. Though there are two apparel makers, the group is comfortably diverse. As you can tell by their negative accruals ratios, all of these companies have more cash flow than earnings. On average, they're cheaper than the S&P 500. They've been growing faster. And all but one offers a dividend. (Benchmark Electronics is excused, based on its 38% revenue growth.)

Even though these companies may not seem exciting, boring can be beautiful. Consider this lesson from Richard Sloan. He likes stock screens and often gives students lists similar to my tables. They pick companies that look interesting and make class presentations based on their research. Nearly always, says Sloan, the stocks that everyone ignores are the best performers.