Jan. 26, 2006

SOX fails to stop stock-option manipulation by executives

ANN ARBOR, Mich.—Sarbanes-Oxley has been less effective in putting a halt to stock-option manipulation than regulators originally hoped, say researchers at the University of Michigan's Ross School of Business.

A new study by finance professors M.P. Narayanan and H. Nejat Seyhun suggests that although the prompt-disclosure requirement imposed by the Sarbanes-Oxley Act of 2002 (SOX) has partly curtailed managerial influencing of grant day stock prices by executives seeking to increase the value of their option grants, such activities still occur.

"Backdating of grant dates and camouflaged timing appear to be practiced even after SOX, especially by smaller firms," Narayanan said. "In order to further restrict this behavior, the Securities and Exchange Commission needs to enforce the SOX two-day reporting rule and, if possible, limit the use of unscheduled option grants to legitimate purposes only."

Under Sarbanes-Oxley, executive option grants must be reported to the SEC within two business days of the grant date—a more stringent reporting requirement imposed in the wake of widespread corporate scandals. Since the exercise price is usually set at the stock price on the grant date, managers want to keep that price as low as possible, in hopes of reaping higher gains later when they exercise their options, the researchers say.

Narayanan and Seyhun analyzed data on all option grants by officers and directors of publicly traded firms between Jan. 1, 2000, and Aug. 31, 2004. Their findings reveal that 24 percent of the 569,000 option grants reported to the SEC by insiders after SOX regulations went into effect Aug. 29, 2002, were reported late, with 10 percent of the grants being reported more than one month later.

Small firms, which tend to have less official scrutiny, weaker corporate governance and lower institutional ownership, exhibited significantly higher average reporting lags (the time between the presumed grant date and the date of filing the grant with the SEC) post-SOX, averaging 19 business days compared to 12 days for the overall sample, the researchers say.

Top executives also appeared more in violation of SOX reporting requirements on average than other executives and larger grants provided more incentive to influence grant date stock prices, they say.

The key evidence consistent with backdating is the positive relationship between reporting lags and post-grant date stock returns, Narayanan and Seyhun say. If there is no backdating, these two variables should be unrelated. Since backdating creates an automatic reporting lag and is worthwhile only if stock prices have been rising, it will result in a positive relation between reporting lags and post-grant date stock returns.

The study shows that the longer the delay in reporting option grants, the greater the financial gains. Reporting lags of two-to-22 business days brought a 30-day post-grant date market-adjusted stock return of 3.5 percent, compared to a 1.5 percent return for grants reported promptly (within two business days). When grants were reported with a delay of more than 22 business days, the 30-day return rose to 8.7 percent and the 90-day return hit 25 percent in cases where the grant size was greater than 100,000 shares.

In addition to late reporting, the researchers say that unscheduled awards increase the potential for backdating. In their post-SOX sample, 51 percent of the grant awards were scheduled on a regular basis. However, that number dropped to 45 percent when the reporting lag ranged from two-to-22 business days and to 41 percent with lags greater than 22 days.

Related Links:

M.P. Narayanan

Nejat Seyhun

Contact: Bernie DeGroat
Phone: (734) 936-1015 or 647-1847