Robert Jenson and Dr. Ted Christensen, School of Accountancy
Introduction
In recent years, companies have increasingly released financial performance numbers in addition to those required under generally accepted accounting principles (“GAAP”). These adjusted, or pro-forma earnings, have been a way for management to report earnings without extraordinary, one-time, or other items. Previous research suggests that management may use these pro-forma numbers to meet benchmarks such as analysts’ expectations when GAAP measurements fall short of these expectations.1 In addition, there are companies which – due to misstatements resulting from accounting errors or even fraud – are required to restate their earnings. These companies may have reported inflated earnings to reach benchmarks, only to restate earnings later. It is hard to tell the difference between “the good guys” and “the bad guys” that report pro forma earnings. The purpose of my research is to see if companies that also have restatements help to isolate managers with opportunistic reasons for reporting pro forma earnings. This is why I look at the “overlap” firms with both restatements and pro-forma numbers. So, I investigate whether companies with two strikes against them (restatements and pro-forma earnings) are worse than companies with only one strike against them.
Data
To perform research, I needed comprehensive data on earnings releases with pro-forma earnings. This data was compiled by searching PR Newswire and Business Wire from the years 1998-2006. Data regarding GAAP earnings per share (“EPS”), pro-forma EPS, and exclusions backed-out of GAAP earnings to get pro-forma earnings were used in this study. Additionally, restatement data for the same period was found observations for companies with restatements only. After merging the data, I had 13,882 press release observations for companies with only restatements, resulting in 2,155 unique companies. There were 7,297 press release observations for companies with only pro-forma earnings, resulting in 1,846 unique companies. The “overlap” – companies that issued pro-forma data and restatement data – had 4,352 observations resulting in 954 unique companies.
Hypothesis
I hoped to identify the “bad guys” – companies who manipulated earnings in order to meet benchmarks to inflate their stock price. I hypothesized that the overlap segment would report more aggressively manipulate earnings compared to the pro-forma companies and the restatement companies. To do this, I compared the overlap to each other segment individually.
Finding
In comparing the press releases of companies issuing pro-forma earnings to the overlap, the evidence for exclusions was minor, though there were some differences. In particular, it appears that the Overlap excluded restructuring, merger, below-the-line, and other items more than the pro-forma segment. However, further studies are needed to determine the statistical significance of these items, as differences are minimal. I expected to find a greater difference, though the studies show that the pro-forma earnings for each segment are similar to each other.
Also comparing the pro-forma and overlap segments, I investigated whether the pro-forma number was a profit, while the GAAP measure or earnings before extraordinary items were a loss (“EPS Special”). The results are consistent with my hypothesis, and suggest that the overlap group did use proforma earnings numbers to turn a loss into a profit.
In addition, I looked at whether the pro-forma number beat analyst expectations while their GAAP number or their Special EPS did not. I found that the overlap segment does this slightly more than the pro-forma segment.
I also compared the restatement segment to the overlap segment, to determine which segment was more aggressive in their earnings reporting. I have found that the overlap segment had to restate their earnings due to Revenue and other Core items (such as COGS, etc.) compared to companies that only issued a restatement.
Conclusion and Further Research
My results show that companies which issue pro-forma earnings but have also made restatements were slightly more aggressive in their financial reporting. I am also looking at compiling more data for these companies, such as whether a fraud was the reason for restatement. This is topic that has received increased attention during the past decade, and further research on this matter can help determine aggressive reporting decisions. The research can help researchers, analysts, and/or understand who these companies are.