Accounting anomaly gets traded to death, proving its existence

Call it closure. A new paper co-authored by Mark Soliman, an associate professor of accounting at the University of Washington Foster School of Business, charts the demise of a hotly debated accounting anomaly which, employed as an investment strategy, reaped huge returns for attentive hedge funds until the market got wise.

It’s also the end of a long chapter in Soliman’s life that led him from academia to high finance and back again.

“I’ve come full circle,” he says.

Once upon a time

The story begins with the discovery of “Sloan’s accruals anomaly,” outlined in a landmark 1996 paper in The Accounting Review. Richard Sloan, then a professor of accounting at the University of Michigan, discovered something curious on the balance sheet. Specifically, something curious about accruals, the non-cash portion of earnings that include estimates of depreciation, inventories, receivables and the like.

When a firm reports a large percentage of earnings as accruals, Sloan found that it tends to see lower earnings in future years.

Soliman studied under Sloan as a doctoral student at Michigan in the early 2000s and collaborated on several papers expanding upon the accruals anomaly. Eventually, they created a trading strategy based on the anomaly in a paper that later earned the American Accounting Association Notable Contributions to the Accounting Literature Award.

“If you know that a company is going to have lower earnings next year, well, you just got a crystal ball,” Soliman says. “That’s the future. And anyone who knows the future has a money machine in their back pocket.”

The doubters

At first, though, few paid any attention to the accruals anomaly.

“It seemed to surprise auditors, analysts, investors—the entire market,” Soliman says.

So even after published research unveiled the predictive power of high accruals on the earnings report, the financial markets did not respond appropriately, which is to say, efficiently. Higher accruals today mean lower earnings next year which should result in lower share prices today. But even when cued, the market was unmoved.

This initial lack of market response sparked widespread skepticism in the academic world. Accounting scholars published a torrent of critical papers discrediting the accruals anomaly as a statistical error or earnings management rather than an aberration of substance. If it were real, they argued, why wouldn’t the market trade away any investing advantage the instant it was discovered?

Soliman says that the answer lies, at least initially, in ignorance, skepticism and lack of ability to exploit.

The believers

And then, as the new century dawned, hedge funds and high-frequency trading emerged as important market movers. Hedge funds invested more broadly—in equities, bonds, commodities, derivatives—than mutual funds in an effort to achieve positive returns whether the market was up or down.

The most prescient of hedge fund managers put faith in the accruals anomaly. They traded short (betting on a loss) on companies reporting high accruals. The earliest to act were rewarded handsomely. “I can’t tell you how many hedge fund guys I met who told me how many millions of dollars they made because of the accruals anomaly,” Soliman says.

Not surprisingly, the market caught on quickly. Increasingly powerful computing accelerated the pace of trading. And, in short order, widespread adoption and lightning-fast transactions led to a rapid demise of the accruals anomaly as a productive trading strategy.

Field test

At some point after the discovery of the accruals anomaly, all of Sloan’s original collaborators left their faculty positions to join hedge funds, along with legions of doctoral students who could mine the source papers for their new employers. Eventually, even Soliman joined Citadel Funds as vice president of accounting-based research in 2006.

By that time, however, returns from the anomaly were drying up rapidly.

“Higher accruals are still predictive of future earnings,” Soliman says. “It’s just that everybody gets it now. The window of the anomaly has shortened to the point where it has become almost instantaneous. The numbers come out, a million hedge funds hit it, and the stock price drops accordingly in a matter of seconds. That’s an efficient market.”

Soliman didn’t get rich off the accruals anomaly. But he did find a silver lining to its decline.

Proof of existence

“The demise that we document in this paper puts to rest any doubt that this was some statistical flaw in the data,” says Soliman. “The anomaly was real, and people made a lot of money on it.”

After witnessing this first-hand in the world of high finance, Soliman joined the Foster School in 2007. His latest—and probably last—paper on the accruals anomaly is really a post mortem which, ironically, vindicates its existence.

He says the entire experience also demonstrates the impact that academics can have in the world. “A few of us left universities for hedge funds and traded the heck out of this anomaly until we made it disappear,” he said. “Then we all went back to academic jobs again, a little bit wiser to the ways of the world.”

Closure. The end.

“Going, going, gone? The demise of the accruals anomaly,” by Mark Soliman, John Hand and Jeremiah Green, is published in the May 2011 issue of Management Science.