Research by the HKUST’s Charles Hsu shows that disclosures of non-GAAP earnings numbers are associated with a greater risk of stock price crashes. Working with international collaborators, Hsu found that non-GAAP disclosures enable managers to divert investors’ attention from bad news included in GAAP earnings. This mechanism, which is substitutive to earnings management, leads to greater firm overvaluation, raising the risk of an eventual crash.
Despite the existence of generally accepted accounting principles (GAAP), managers can voluntarily exclude some GAAP-mandated earnings components to emphasize the importance of other items. These announcements receive close attention from investors. Hsu and colleagues investigated the effect of non-GAAP disclosures on market crashes, noting that “ex ante, it is unclear whether non-GAAP reporting would increase or decrease crash risk.”
Some firms report non-GAAP earnings more often than others do. Hence, the researchers estimated the association over 14 years between the frequency of firms’ non-GAAP disclosures and four measures of the risk of stock price crashes. This enabled them to test two hypotheses: one regarding the effect of non-GAAP announcements on crash risk and the other regarding the relation between this effect and the practice of earnings management. As the authors note, earnings management is “the traditional form of bad news withholding in the crash risk literature,” and it can be a complement or substitute to non-GAAP disclosures.
Non-GAAP disclosures can increase crash risk by exaggerating firms’ value and prospects. By excluding GAAP items that signal bad news and emphasizing those that communicate good performance, non-GAAP disclosures may lead to inflated stock prices, which will eventually crash when the firm’s true performance does not meet investors’ inflated expectations. That prediction was confirmed by the study, which found that “the likelihood of a crash significantly increases when a firm discloses non-GAAP earnings at some point during the prior year.”
However, it could be argued that non-GAAP announcements reduce the crash risk if they do not exaggerate firm value. As the authors note, “recent research indicates that managers often disclose non-GAAP earnings for informative reasons.” For example, managers could exclude non-recurring items, which have little implication for firm value, to focus on providing investors with the metrics most relevant to the firm’s ongoing operations. Moreover, non-GAAP disclosures may increase or decrease a firm’s crash risk, depending on the sign of excluded items. Consistent with this, the study found that non-GAAP reporting actually reduced the crash risk when non-GAAP earnings were lower than GAAP earnings.
Regarding the second hypothesis, the authors found that non-GAAP disclosure could substitute for earnings management as a perception management strategy. Specifically, the positive association between income-increasing non-GAAP announcements and crash risk was mostly found in firms with low levels of opacity (abnormal accruals). That is, some managers who do not use earnings management to paint a “rosy picture” of their firms instead do so via non-GAAP disclosure, and thus mislead the market.
The study provides much-needed evidence of the relation between accounting practices and stock price crashes. Importantly, the authors note that despite the risky effects of (income-increasing) non-GAAP reporting, “our results do not imply that this increased risk fully negates the average benefit of non-GAAP reporting.” It can still be a beneficial practice when performed in good faith.