Pro forma disclosure and investor sophistication: External validation of experimental evidence using archival data

Insights on how ordinary, less-sophisticated investors interpret and process management-issued pro forma earnings numbers are useful to regulators because of concerns that pro forma disclosures are misleading to ordinary investors. Two recent experimental studies [Frederickson, J. R., & Miller, J. S. (2004). The effects of pro forma earnings disclosures on analysts' and nonprofessional investors' equity valuation judgments. The Accounting Review, 79(3), 667-686; Elliott, W. B. (2006). Are investors influenced by pro forma emphasis and reconciliations in earnings announcements? The Accounting Review, 81(1), 113-133] find that the existence of a pro forma number in the earnings press release as well as the relative placement of the pro forma and GAAP earnings figures within the press release affect the judgments of less-sophisticated investors but not those of more-sophisticated investors. Experimental and archival methodologies complement one another and results that persist in both settings are likely to be robust to both internal and external validity concerns. Therefore, we complement experimental evidence using trade-size-based proxies constructed from intraday transactions data to distinguish the trading activities of less-sophisticated investors from more-sophisticated investors. Our results suggest that less-sophisticated investors rely significantly more on quarterly earnings press releases that include a pro forma number than on those that do not, while more-sophisticated investors exhibit the opposite behavior. This result is consistent with Frederickson and Miller's experimental evidence. Further, consistent with Elliott's results, we find that less-sophisticated investors rely more on the pro forma figure when it is placed before the GAAP earnings number in the press release, while more-sophisticated investors' trading behavior is unaffected by the relative placement of the two earnings metrics. We conclude that the existence of a pro forma number as well as its strategic placement in the press release generally affect the judgments of less-sophisticated (but not more-sophisticated) investors and these inferences are robust because they persist in both experimental and archival settings. (c) 2006 Elsevier Ltd. All rights reserved.