Call up a corporate earnings statement and prepare to read a lot about the top and bottom lines. Revenue and net income routinely dominate the discussion of a company's quarterly results as corporate press releases and subsequent media write-ups focus on whether sales and profit have met the expectations of Wall Street analysts.

But research by Jan Barton, an associate professor of accounting at Emory University's Goizueta Business School, suggests that subtotals near the center of the income statement, such as operating income, have a much stronger association with contemporaneous stock returns than do the top-line and bottom-line numbers. This applies not just to U.S. companies, but also to firms around the world, says Barton, who has analyzed financial statements from nearly 20,000 firms based in 46 countries.

So, did the research reveal a single magic metric? Did one key item in the middle of the income statement reliably provide investors with the best signal to company performance regardless of industry quirks, country of origin or the vagaries of international accounting rules?

Alas, Barton says, it's not that simple. In a paper entitled, Which Performance Measure Attributes Do Investors around the World Value the Most-and Why? slated for publication in The Accounting Review in May, he and co-authors Grace Pownall, professor of accounting and associate dean of the doctoral program at Goizueta, and Bowe Hansen of the University of New Hampshire, argue that no one single performance measure can serve as a Rosetta Stone for investors as they shop for stocks around the world. Still, Barton says, the research suggests that helpful metrics tend to be those that quickly and directly reflect information about a firm's future cash flow. And the researchers' observations about the complexity of assessing value across the globe provide important caveats for investors, as well as for groups that set accounting standards.

At least in the case of stock valuation, there's really no way to say, 'This is the dominant performance metric,' notes Barton. So, the next step is to say, what are the attributes that make one particular performance measure important for valuation?

Performance measures that reflect more closely and timely changes in current and expected future operating cash flows are more useful to equity investors, write Barton and his co-authors. On the other hand, performance measures that are more persistent, predictable, smooth, and conservative are less useful for equity valuation.

Back in 2001, the Financial Accounting Standards Board (FASB) and the International Accounting Standards Board initiated a project to explore whether they should change the reporting requirements for financial statements in ways that might de-emphasize the importance of the bottom line. As FASB noted in a background document at the time, many companies had been emphasizing pro-forma earnings numbers, and the board was concerned about the potential for self-serving selection of those indicators that put performance in the best light. With similar concerns in mind, Barton and his colleagues set out to provide guidance to standard setters not just in the U.S., but also across the globe, on whether other metrics might be more helpful to investors.

The researchers started by amassing a huge set of data on the performance of 26,479 firms from around the world between 1996 and 2005. They whittled down the data to exclude companies that lacked complete information for a variety of financial metrics, and focused the analysis on 19,784 firms in 46 countries. Researchers calculated how well company stocks had performed during the time period. What they wanted to see was which of the following eight financial metrics showed the strongest correlation with stock performance: sales; earnings before interest, taxes, depreciation and amortization (EBITDA); operating income before taxes; income before taxes; income before extraordinary items and discontinued operations; net income; total comprehensive income; and operating cash flows.

Barton and his co-authors found that the value relevance of each of the eight performance measures varied substantially across countries. Income before taxes was the most value relevant measure in 25 of the 46 countries, including Australia, Canada, the United States and France. However, each of the other performance measures, with the exception of sales, was the most value relevant in at least one country. In Germany and the United Kingdom, for example, EBITDA was the most value relevant performance measure. But in Japan, operating income before taxes was most value relevant. As the researchers sliced and diced the analysis further, it became increasingly difficult to find a clear winner in the hunt for the most value relevant metric.

If we had stopped there, it wouldn't have really solved the problem, Barton says. So, the next step was to change the question to: What are the attributes that make one particular performance measure important for valuation?

The academic literature describes numerous attributes of financial reports that render the documents useful to investors-things like conservatism, timeliness and predictability. The researchers decided to take their data set and see if it could help establish which of seven commonly cited attributes are, indeed, the most useful. Their first observation: Many of the seven attributes were strongly correlated with one another. For example, companies that scored highly in terms of the persistence of their earnings also scored highly in terms of predictability and smoothness.

So, the researchers grouped the attributes and identified two factors: one they called a sustainability factor, and the other a cash flow articulation factor. Then they looked for associations between each factor and value relevance. They found a negative link between value and earnings reports with a high sustainability factor. And they found a positive link between value and reports that captured information about cash flows.

Overall, our results suggest that performance measures become more value relevant when they incorporate economic value added in a timely, unbiased fashion and when their accruals can be articulated with the firm's underlying cash flow, the researchers wrote. In contrast, performance measures that are smoother, more predictable and persistent, and less conservative are less useful for valuation; such attributes lower the value relevance of a performance measure. These inferences apply similarly to performance measures reported by firms from the 46 countries represented in our sample.

The conclusion amounts to something of a negative judgment on companies like General Electric, which Barton says has a history of reporting smooth earnings from quarter to quarter. Some companies contend that by smoothing results they are screening out noise that otherwise would lead to quarterly fluctuations that aren't reflective of the underlying business.

Taken to the extreme, the impulse to smooth earnings can cross some dangerous lines. In August 2009, the Securities and Exchange Commission (SEC) filed civil fraud charges alleging that GE misled investors by reporting materially false and misleading results in its financial statements. Regulators said GE used improper accounting methods to increase its reported earnings, and the company agreed to pay a $50 million penalty to settle the charges without admitting wrongdoing. A much more extreme case is the demise of telecommunications giant WorldCom. In an effort to boost earnings, WorldCom overstated earnings before income taxes and minority interests by approximately $3.8 billion in 2001 and 2002, according to the SEC. By capitalizing and deferring costs-rather than expensing and immediately recognizing them-the company violated accounting rules and gave the false impression of profitability, the government alleged. WorldCom executives wound up in prison as a result of the scandal, and the company sought bankruptcy court protection in 2002.

Investors don't always value smooth performance or measures that capture smoothness, because too often smoothness comes from managing earnings, Barton says. What I value more is information that's timely-information that, good news or bad news, tells me now what cash flows are going to be in the future.

Unfortunately, the conclusion doesn't point to an easy answer for groups such as FASB, Barton says. How can an accounting standards group mandate attributes from a quarterly earnings report? There's no easy answer for investors, either.

As an investor, the worst thing you can do is say: I will always focus on net income, or operating income, or operating cash flows, Barton says. It's really a matter of trying to understand how the numbers are put together-according to the local accounting rules, and the incentives that managers have to put those numbers together-combining them with other information like the firm's business strategy, and then making judgments.