Researching a Scandal

"We found that 19 percent of grants that were unscheduled had return patterns that suggested that they had been manipulated. It's very widespread."

Randall Heron
Associate Professor of Finance

As of August 2002, companies were required by the Securities and Exchange Commission to report stock-option grants within two days. Heron and Lie figured that if companies could no longer look back over months of stock charts trying to find the most advantageous valleys, then the puzzling pattern should disappear.

Indeed, that's just what their research showed: "When companies filed immediately, there was absolutely no return pattern," Heron says. "This was compelling evidence that pretty much everything that was going on was backdating."

But they also found that one in five companies filed their reports late, and once those late reports were filed those companies were still showing the unusual return pattern, with their stock-option grants sitting at the bottom of valleys on the stock price chart. In fact, the later they filed, the stronger the pattern.

Late filers had an average stock decline of more than 3 percent in the month prior to the option grant dates, followed by a rise of more than 7 percent in the month afterward, according to the research by Heron and Lie. The research suggested that option backdating had, indeed, been occurring before Sarbanes-Oxley, and that for many companies it was continuing even afterward, disguised by late SEC filings.

Spreading the Word

The research was unbelievably compelling. So unbelievable, in fact, that the editors of one top research journal rejected the study when Heron and Lie submitted it - they simply couldn't believe the results. "We were kind of disheartened by the fact that it got rejected," Heron recalls. "But we sent it to another journal, and the referee there said the evidence was compelling, but asked why is it that nobody knows about it?"

So Heron and Lie did some more digging and uncovered anecdotal evidence of backdating. They also contacted investigators at the SEC and found that the government was starting to look into the possibility of widespread backdating, but was just at the start of the process.

Then came the Mercury Interactive case. In the summer of 2005, the board of the Silicon Valley software company announced that it had formed a committee to look into possible backdating. Its internal investigation found dozens of instances where the reported date of stock-option grants appeared to be different from the date they actually were granted. The CEO and two other top executives resigned, the company's stock price plunged, and Wall Street finally began to take notice of stock-option backdating, beginning with a major article in The Wall Street Journal.

"We were in the process of revising our paper to send back to the journal when that broke," Heron says. "We decided that we needed to let the reporters at The Wall Street Journal know that this is much bigger than just Mercury Interactive."

So they contacted the newspaper to share their research. One of the reporters to benefit was special projects editor Mark Maremont. "The academic research was extremely important in getting us interested in pursuing the matter further," Maremont says. "It broke a lot of new ground. What they did was show that backdating was likely to be going on in a lot of companies."

Of course, once the news had sunk in that the practice appeared to be widespread, the next logical question was, who else is doing it? "We didn't want to accuse any individual companies," Heron says. "So over the course of the next few months, my co-author trained some of the reporters what to look for. That's when you started seeing more things in the media, including the 'Perfect Payday' series in The Wall Street Journal in March 2006."

Maremont and two colleagues, Charles Forelle and James Bandler, received the highest accolade in their profession for their work, assisted by Heron and Lie. In April The Wall Street Journal received a Pulitzer Prize for Public Service.

The media requests kept coming in. "People were wanting to know how many companies are doing this? So we started focusing on how many companies are doing this and what percentage of option grants," Heron says.

Some companies' option grants follow a previously specified, calendar-based schedule, while others are unscheduled. Heron and Lie decided to look further into unscheduled grants, because those are the ones whose dates have the potential to be manipulated. "We found that 19 percent of grants that were unscheduled had return patterns that suggested that they had been manipulated. It's very widespread," Heron says.