Damage caused by fraud at public companies soared in the past decade, with 347 separate cases leading to at least $120 billion in cumulative misstatements and misappropriations.

It’s a big jump from 294 cases in the decade from 1987 to 1997, and a tripling of the average loss, from $4.1 million per episode in the prior decade to $12.05 million in the more recent decade.

Those are the findings of a recent comprehensive fraud study, Fraudulent Financial Reporting: 1998-2007, sponsored by the Committee of Sponsoring Organizations. The study is meant to provide a comprehensive analysis of fraudulent financial reporting incidents investigated by the Securities and Exchange Commission from 1998 to 2007, updating a similar study by COSO 10 years earlier.

The study found not only that the magnitude of fraud increased, but so did the size of the company typically caught up in a fraud scheme. In the 1999 study, the typical fraud target company had median assets and revenues under $16 million while in the more recent decade, the median jumped to just under $100 million.

The most common fraud method was to toy with revenue recognition or overstate existing assets or capitalization of expenses, the study found. A little more than one-fourth of companies changed auditors sometime after the last clean financial statements and the last fraudulent ones, compared with only 12 percent of companies without fraud claims that changed audit firms.

In 89 percent of fraud cases in the most recent decade, the SEC named the CEO and/or CFO for some role in the fraud scheme, up from 83 percent of cases in the earlier decade. In the last 10 years, only about 20 percent of CEOs and/or CFOs were indicted, and more than 60 percent of those indicted were convicted.

Dana Hermanson, a co-author of the study and an accounting professor at Kennesaw State University, said the study authors hope their work “will encourage additional resources to better understand organizational behaviors, leadership dynamics, and other important aspects of the financial reporting process that may have an impact on fraud prevention, deterrence, and detection.”

Mark Beasley, another co-author and a professor at North Carolina State University, said additional research would be particularly helpful to sort out what impact Sarbanes-Oxley has had on fraud. The most recent decade takes in years before and after the Sarbanes-Oxley Act came into play, but the study doesn’t establish a correlation.

Mark Beasley, a professor at North Carolina State University and a member of the COSO board, said more research is needed to better understand processes surrounding the board of directors and audit committee. “We need to determine if there are certain board-related processes that strengthen the board’s oversight of risks affecting financial reporting,” he said.