Accounting items can predict stock gains-research

* Noncash earnings seen as key to stock performance

* Researchers tweak older stock-picking strategy

By Dena Aubin

NEW YORK, Feb 9 (Reuters) - Stock investors fixated on earnings could improve their returns by paying more attention to some often-overlooked accounting items, new research has found.

The research, led by a University of Michigan accounting professor, centers on accruals, the noncash part of earnings, such as changes in accounts receivable. Though cash has not changed hands, companies book such transactions as part of earnings under accrual accounting rules.

In simplest terms, the researchers' trading strategy involves buying stocks of companies with the least accruals and selling short companies with the most.

Use of accruals for stock picking is not new, but researchers claim to have found a way to refine the strategy by looking at accruals as a percent of earnings, instead of the traditional way, which relates accruals to a company's size.

"It's about the composition of earnings and what percent were due to accruals," said Professor Russell Lundholm, co-author of the report.

Accruals are essentially a way of comparing a company's cash and net income. When accruals are high, net income far exceeds cash because the company's income is being bolstered by accounting items.

STOCKS STILL

Lundholm points to examples where a high level of accruals preceded poor stock performance.

Accruals at Monsanto Co (MON.N), the world's biggest seed producer, were 58 percent of earnings for the previous four quarters when the company reported results on Jan. 6, 2010, according to Lundholm. That was in the top 10 percent of all U.S. companies.

Since then, its shares have dropped 13 percent, while the S&P 500 index is up about 16 percent.

Using the strategy to pick or short stocks over the past two decades would have earned 11.7 percent better returns than just buying shares of similarly sized companies, the research found. That result is about 5 percentage points better than the older accrual method, Lundholm found.

A focus on the quality of earnings is certainly not a novel idea. Many investors already scrutinize a company's financials to be sure net income is not phantom and will eventually be realized in cash.

Even so, too few investors are making use of accruals in their trading decisions, Lundholm said.

"I can pretty much assure you not enough people are doing it already. If they were I wouldn't be able to systematically pick overvalued stocks," he said.

Some strategists questioned whether using a single factor to pick stocks is a good idea.

"I'm sure it has some value, but none of this has value by itself without fundamental analysis," said Robert Williams, head of research at Sage Advisory Service in Austin, Texas.

Steve Wood, chief market strategist at Seattle-based Russell Investments, also said investors need more information than any single measure can provide.

"If there were one single metric which would explain future stock performance, it would be recognized and priced out of the market," he said.

20 YEARS OF DATA

Results of the research, published in the current issue of the American Accounting Association's Accounting Review, are based on data on all publicly traded U.S. companies, excluding financial firms and utilities, between 1989 and 2008.

Co-authors were Matthew Van Winkle of Voyant Advisors and Nader Hafzalla, a former student of Lundholm's, now deceased.

Richard Sloan, a former University of Michigan researcher, proposed picking stocks by looking at accruals relative to a company's assets about 15 years ago -- the higher the accrual ratio, the worse a company's stock was likely to perform, he found.

Many researchers were skeptical when Sloan's research was published because it meant that many stocks were being mispriced with respect to accruals -- something that should not happen in efficient markets.

"It could be that the market is now completely efficient with respect to accruals -- that enough quantitative investors and well-informed investors who went to good business schools have all learned this," said Lundholm. "I just don't think so." (Editing by Steve Orlofsky)