An Analysis of Small Firms’ Non - GAAP Disclosures’
Informativeness After SEC’s 2010 Updated CDI
Name: Charelle Felix Student number: 10868763
Thesis supervisor: dhr. Prof. dr. David Veenman Date: 26 June 2017
Word count: 13,365
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student, Charelle Felix, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
In this thesis, I research how the SEC’s update in 2010 in its Compliance and Disclosure Interpretations (CDI) on non-GAAP disclosures affected the informativeness of small firms’ non-GAAP measures. This update affected big firms, where previous studies show that there has been a reappearance of non-GAAP disclosures frequencies and magnitudes. Since some studies also show that non-GAAP reporting can be both informative and misleading, it is important to research non-GAAP measures informativeness. Besides that, also researching non-GAAP disclosures informativeness of small firms is relevant, since they are different compared to large companies in regards to information provision. I provide new descriptive statistics of non-GAAP reporting for small firms, and find that both non-GAAP and non-non-GAAP earnings have increased after the new CDI announcement, while exclusions have decreased. This finding suggests that small firms make less use of exclusions when GAAP earnings increase. The regression results show that recurring exclusions have no significant relation with future performance, even after the new CDI announcement, suggesting that recurring exclusions quality have not been affected after the new CDI announcement. Thus, I conclude that the CDI announcement had no significant effect on small firms’ non-GAAP informativeness.
Contents 1! Introduction+...+5! 2! Literature+review+...+9! 2.1! Corporate+disclosures+...+9! 2.2! GAAP+&+Non+–+GAAP+disclosures+...+10! 2.3! Securities+and+Exchange+Commission+...+16! 2.3.1! SEC!2003!Mandate!...!16! 2.3.2! SEC!2010!Update!...!19! 2.4! Small+cap+firms+...+21! 3! Hypothesis+development+...+23! 4! Methodology+...+25! 4.1! Sample+selection+...+25! 4.2! Research+design+&+Variable+measurement+...+25! 5! Results+...+27! 5.1! Descriptive+statistics+...+27! 5.2! The+effect+of+CDI+announcement+on+small+firms’+nonTGAAP+informativeness+...+32! 6! Conclusion+...+35! References+...+37! Appendices+...+40!
1 Introduction
In the last years there have been several concerns regarding non-Generally Accepted Accounting Principles (non-GAAP) disclosures. Non-GAAP financial figures are measured and created by managers but not according to Generally Accepted Accounting Principles (GAAP). The concerns have been in the direction of the Sarbanes Oxley (SOX) towards the U.S. Securities and Exchange Commission (SEC) to establish guidelines about non-GAAP disclosures, which resulted in the SEC implementing new non-GAAP disclosure rules in March 2003 (SEC, 2003). The rule mandated that firms have to follow additional requirements when they disclose non-GAAP earnings. The rules were intended to reduce the opportunistic use of non-non-GAAP earnings disclosures and they included that: the most directly comparable GAAP earnings has to be disclosed, a reconciliation between the GAAP and the non-GAAP number has to be disclosed, and an 8-K Form has to be given with an explanation of why the firm’s management believes that the non-GAAP disclosures are useful to investors.
Since the introduction of this regulation, some studies have found that there has been a decline in GAAP earnings disclosures and in the magnitude between GAAP and non-GAAP earnings differences (Heflin & Hsu, 2008; Campbell & Pitman, 2009). Although more recently there has been an update related to the use of non-GAAP financial disclosures, where the SEC’s division of corporate finance updated its Compliance and Disclosure Interpretations (CDIs) (SEC, 2010). In this update, the SEC suggests that companies can make adjustments for recurring items even if they do not meet the requirements of non-recurring, infrequent or unusual, which gives a little flexibility to what was prohibited in 2003 in Item 10 of Regulation S-K. Regulation S-K is a regulation under the US Securities Act of 1933, a law that governs the securities industry, which represents the integrated disclosure requirements and applies to disclosure documents filed with the SEC (SEC, 2016). The amendment of item 10 of Regulation S-K requires more detailed disclosure of SEC filings’ non-GAAP financial measures.
Since the update, a few studies’ findings show that there has been a reappearance of non-GAAP financial disclosures showing significant increases of the frequencies and magnitude (Webber et al., 2013; Black et al., 2015). Webber et al. (2013) suggest that this increase is due to the change in tone at the SEC, where it relaxed its position on non-GAAP disclosures. However, only a few studies have analyzed the non-GAAP disclosures since the 2010 update of the CDIs on Item 10 of Regulation S-K, where it is explained that companies can make adjustments for recurring items even if they do not meet the requirements of non-recurring, infrequent or unusual. Studies have mainly focused on big firms and on non-GAAP disclosures that have been
affected since the 2003 update with a time frame mostly until 2010. Researching non-GAAP disclosures is therefore important, since the use of non-GAAP information has also increased. Webber et al. (2013) indicate that half of the companies in their sample repeatedly disclosed specific types of non-GAAP adjustments after the 2010 update, which makes it important to research exclusions of non-GAAP reporting. Besides that, it is also relevant since it is believed that non-GAAP reporting can be both informative or misleading, thereby raising the question whether these measures are potentially misleading to investors and if it reintroduces the problem what SEC in the first place wanted to resolve.
Besides that, to enhance generalizability, Webber et al. (2013) suggest that future research should examine non-GAAP disclosures by smaller SEC registrants. Small cap companies are also important since they represent approximately 25% of all business entities and even though they only represent 5% of the total US equity market capitalization, they represent a big market segment of the US economy (SEC, 2006).
Research on small companies is relevant, since they are different compared to large companies in regards to information provision. Since small companies have a characteristic of the opaqueness of their available information, they try to reduce this information asymmetry with other forms of disclosure, such as non-GAAP disclosures. So it can be expected that small firms make more use of non-GAAP disclosures than large firms, thereby making it important to also research small firms non-GAAP disclosures. Campbell and Lopez (2010) have researched small cap firms’ non-GAAP disclosures, but its time frame was from before the SEC’s update in 2010, so not representing the call for generalizability from Webber et al (2013). Therefore, in this paper I research small SEC capital firms’ non-GAAP disclosures since the 2010 update of the SEC by extending the research of Webber et al. (2013), and answering the question whether the 2010 SEC update in its Compliance and Disclosure Interpretation affected the usefulness of the non–GAAP disclosures by small firms.
I follow prior research (Kolev et al., 2008) by using Institutional Brokers’ Estimate System (I/B/E/S) actual earnings to proxy for non-GAAP earnings disclosed by managers, where I obtain data from the Compustat Fundamentals Annual file and the I/B/E/S Detail Unadjusted Actual file. I collect data for the period between 2008 and 2011, so I could compare the periods before and after the new CDI. To test the hypothesis I use a model comparable as the one used from prior research (Kolev et al., 2008; Black et al., 2017) by examining how non-GAAP exclusions map into future operating performance.
I present new descriptive statistics about non-GAAP reporting for small firms, where the results show that the mean GAAP earnings per share are considerably higher after the new CDI
announcement date compared to before the new CDI announcement, while Non-GAAP earnings per share show a slightly increase. The means of the exclusions have dropped significantly, which comes from a substantial drop from recurring exclusions per share, while special items experienced a little decrease. Since in the CDI announcement, flexibility is given on adjusting recurring exclusions by indicating that in reporting non-GAAP, a charge or gain can still be adjusted even if this charge or gain cannot be indicated as a non-recurring, infrequent or unusual, the decrease of the recurring expenses after the CDI announcement was not expected. However, a reason for the decline of the recurring exclusions is that also special items have decreased while GAAP earnings have increased, suggesting small firms make less use of recurring exclusions when GAAP earnings increase.
The regression results of the test on the effect of CDI announcement on small firms’ non-GAAP informativeness show that non-GAAP earnings have a positive relation with future performance, which is in line with previous studies about big firms, where it is found that non-GAAP earnings are representative of future performances when non-non-GAAP earnings relation are positive to future performance variables. New CDI shows to be negatively related with future performance. However, the relation between recurring item exclusions and future performance is negative, but non significant, while after the new CDI, there is no relation between recurring exclusions and future performance, but this is also non significant. So these finding cannot confirm the notion that recurring exclusions are adjusted for opportunistic reasons, but they confirm that recurring exclusions are used for informative reasons. Neither can it confirm the notion that recurring exclusions are of higher quality since the recurring exclusions have no predictive power for the future earnings.
After the given results, I do not reject the hypothesis that there is no association between the announcement of the CDI in 2010 and informativeness of non-GAAP measures of small cap firms. I do not find statistically significant evidence of the recurring exclusions’ quality after the new CDI announcement being affected, since they have no predictive power for the future performance. So, I conclude that the CDI announcement did not significantly affect small firms non-GAAP informativeness.
My research contributes to the non-GAAP disclosures literature by showing how exclusions by small firms are used for informative reasons. I provide new descriptive statistics about non-GAAP reporting for small firms, since other studies mainly provide non-GAAP reporting statistics for big firms (Webber et al., 2013; Black et al., 2017) and non-GAAP statistics for periods only before the new CDI announcement (Campbell & Lopez, 2010). Campbell and Lopez (2010) researched non-GAAP behavior of small firms in the time period before the new
CDI announcement and they present descriptive statistics of the determinants of non-GAAP information importance, while I research the quality of non-GAAP information after the new CDI announcement. Webber et al. (2013) and Black et al. (2017) research non-GAAP measures after the new CDI announcement and they both find that after the new CDI announcement, non-GAAP measures have increased, and that the quality of the non-GAAP measures were not compromised by this. However, all of their samples consisted of large SEC registrants, so to enhance generalizability, my findings compliments their findings by showing that small firms’ non-GAAP measures have also increased and that the adjusted recurring exclusions of the GAAP measures were neither compromised by the new CDI announcement.
The results are in contrast of concerns of the SEC regarding the opportunistic use of non-GAAP measures and recurring exclusions, since the findings show that the non-non-GAAP measures and recurring exclusions are used for informative motives, because of the predictive power of the non-GAAP measure and the insignificant relation between the recurring exclusion and the future performance. The findings also do not confirm the notion that investors are misled by the adjusted recurring exclusions, since the findings show that the recurring exclusion that are adjusted are not significantly related to future performance, so thereby do not predict the future performance of the company, which is what the investor is interested in.
2 Literature review
2.1 Corporate disclosuresDisclosures of company information and financial information are important sources for investors to be informed about the company’s performance and its governance, which makes disclosure important for an efficient capital market (Healy & Palepu, 2001). Disclosure can be done through regulated financial reports or through voluntary communication. Regulated financial reports include yearly or quarterly financial statements, including its footnotes and the management discussion and analysis, or any other filings that are regulated, while voluntary communication can include forecasts made by management, presentations of analysts, conference calls, releases to the press, information on websites, or any other reports from the corporation (Healy & Palepu, 2001).
Corporate disclosures have two distinct roles in the capital market: a valuation role and a stewardship role (Beyer et al., 2010). The valuation role permits the evaluation of the return potential of investment opportunities by investors ex-ante, while the stewardship role allows monitoring by the investors of the capital after it has been used (Beyer et al., 2010). So the demand for corporate disclosures results from the need of reducing information asymmetries and agency problems between outside investors and managers (Healy & Palepu, 2001; Beyer et al., 2010; Choi & Young, 2015). Since managers have more information about firms’ current or expected future profitability of investments than outside investors, this creates information asymmetry where assessing the profitability of company investments opportunities will be difficult for outside investors. Agency problems arise due to the separation of ownership and decision-making responsibilities, where the decision-making responsibilities are delegated to the managers, who may have a different incentive than the investor (Healy & Palepu, 2001).
Since investors need useful information for their decision-making, financial reporting and disclosures need to provide high quality financial reporting for economic decision-making. In the conceptual framework of the Financial Accounting and Standard Board (FASB) it is noted that the objective of financial reporting is to provide information that is useful in making investment and credit decisions and assessing cash flow prospects, and provide information about the entity’s resources, claims to those resources, and changes in resources and claims (FASB, 2006). According to FASB for information to be decision-useful, it needs to contain the following qualitative characteristics: relevance, faithful representation, comparability and understandability.
2.2 GAAP & Non – GAAP disclosures
Generally accepted accounting principles (GAAP) disclosures are defined by Healy and Palepu (2001) in their review of empirical disclosure literature as regulated reporting choices presented to managers to disclose the firm’s financial performance through a commonly accepted language that the managers could use to report to users of the financial statement. Healy and Palepu (2001) also claim that regulators improve the credibility of these disclosures and that it can reduce the information gap between informed and uninformed after concerns for financially unsophisticated investors their welfare.
On the other hand, managers may choose to voluntarily disclose more information than mandated, via non-GAAP disclosures (Healy & Palepu, 2001; Bansal et al., 2012). According to the SEC (2003), non–GAAP financial disclosures are not required to be presented according to the commonly accepted language such as GAAP, rules of Commission or any other systems of regulation applicable to a SEC registrant. More specifically, a non-GAAP financial measure is a numerical measure that is adjusted to its most directly comparable measure that is determined according to GAAP, and provides additional information about the company’s historical or future financial position, performance, cash flows, or liquidity (PWC, 2014)
In appendix A an example is presented of a non-GAAP disclosure of MKS Instruments, a small cap company from the S&P 600 list. In the example, the 2002 non-GAAP financial measures or “pro forma figures” of the company is presented to show how companies presented their non-GAAP financial measures before the 2003 mandate of SEC. From the example, one of the non-GAAP measures that are adjusted is the “pro forma net income (loss) before amortization of goodwill and acquired intangible assets, and other acquisitions and disposition related charges, net income of taxes” with a loss amount of $22,404. The most directly comparable GAAP measure to this in the example is the “net income (loss)” with the loss amount of $39,537 before the adjustments. So with this example it can be noted that the non-GAAP amount looks more favorable than the non-GAAP amount with a difference of $17,133 from adjustments made.
There are several types of non-GAAP disclosures including non-GAAP income from operations, non-GAAP earnings adjusted with depreciations, taxes, amortizations, interest and special items, non-GAAP earnings per share, non-GAAP operating expenses, non-GAAP net earnings (PWC, 2014). Besides that, different aspects of the operations of a company can be disaggregated or effects of large, unusual or unique transactions such as acquisitions or dispositions can be removed. The effects of these transactions can be removed if the company believes that the information provided from it, can benefit the financial statement’s user’s
assessment of the company’s historical performance, long-term prospects, or business decisions effects. Non-GAAP earnings measures are mostly communicated through analysts’ presentations, conference calls, press releases, websites, and other corporate reports (Healy & Palepu, 2001).
If accounting regulation seems to be imperfect, several studies argue and show that managers tend to voluntary disclose non-GAAP information, since these measures better reflect actual economic performances, and also better portray performance than would with GAAP measures (Bradshaw & Sloan, 2002; Bhattacharya et al., 2003; Lougee & Marquadt, 2004; Bowen et al., 2005; Choi & Young, 2015; Leung & Veenman, 2016; Black et al., 2017).
Bradshaw and Sloan (2002) research the rise of non-GAAP measures and its implications for its interpretations. The study of Bradshaw and Sloan (2002) find that non-GAAP adjustments have increased in both frequency as magnitude, and that there is a preference for reporting non-GAAP figures that are higher than non-GAAP figures. They also show through market responses that there has been a shift away from GAAP earnings as a primary determinant of stock prices to non-GAAP earnings. Lastly they find through press release analysis that the rising attention to non-GAAP information is driven by managers’ reporting strategies.
Bhattacharya et al. (2003) examine actual pro forma press releases to analyze the reaction of market participants on pro forma earnings informativeness and persistence compared to GAAP operating income. The results of their study show that announcers of non-GAAP information report frequent GAAP losses and that they are mostly concentrated in service or high-tech industries. They also find that non-GAAP earnings are more informative and permanent than GAAP operating earnings by analyzing short-window abnormal returns and revisions in analysts’ one-quarter earnings forecasts. Thus their evidence suggests that market participants believe that non-GAAP measures better reflect actual economic performances than GAAP measures.
In the study of Lougee and Marquadt (2004) press releases are also analyzed to examine the characteristics of firms that include pro forma earnings information in their press releases, and they find that firms with low informativeness in their GAAP earnings are more likely to disclose non-GAAP earnings than other firms. They also examine whether the usefulness to investors of non-GAAP earnings varies systematically and find that investors consider non-GAAP to be more useful when GAAP earnings are less informative or when strategic considerations are missing. However they cannot determine precisely whether investor reaction to pro forma earnings ta press release date is consistent with market efficiency or mispricing, since the predictive ability tests of pro forma earnings for future profitability and returns are mixed.
In the research of Bowen et al. (2005), the determinants of the emphasis placed on non-GAAP and GAAP earnings are examined with quarterly earnings press releases, and they find that firms place more emphasis on value-relevant measures and on measures that show a more favorable firm performance. Besides that they also find that managers’ disclosure emphasis decisions are affected by the extent of media coverage a firm has. They also researched whether there has been a shift away from pro forma earnings emphasis to GAAP earnings, where they find a highly significant shift away from pro forma emphasis to a more GAAP emphasis compared from 2002 to 2001. Lastly they also examined the stock market reactions to earnings news and its influence on the emphasis placed in press release measures, where they find that there is a stronger market reaction to the surprise in an earnings measure when greater emphasis is placed on that earnings measure.
Choi and Young (2015) also research managers’ reporting incentives of non-GAAP reporting by analyzing the association between non-GAAP earnings disclosure and transitory items in GAAP earnings. Their evidence show both a statistically as an economically significant relation between disclosure trend and transitory items in GAAP earnings. They thereby conclude in their research that non-GAAP disclosures are for informative reasons when GAAP earnings beat market expectations, and that non-GAAP earnings are for strategic incentives when GAAP earnings underperform market expectations.
Leung and Veenman (2016) research GAAP losses and whether investors understand its nature and implications when non-GAAP measures are disclosed. They research this by hand collecting GAAP loss firms’ data that report non-GAAP profit. Leung and Veenman (2016) find that GAAP earnings of the loss firms are uninformative of future cash flows, while GAAP are highly informative. Their findings thus suggest that the items excluded from the non-GAAP earnings remove the predictive nature of non-GAAP earnings. Besides that, their results also show that compared to GAAP-only loss firms, loss-converters have cash flows that are significantly stronger, return to profitability is more likely, and delisting is less likely. They conclude that loss firms use non-GAAP disclosures to inform investors about future cash flows. Black et al. (2017) examine how earnings consistency and comparability are influence by the discretion afforded in GAAP reporting and find that the frequency has increased in non-GAAP reporting and that more definitions have emerged for non-non-GAAP reporting. When examining earnings consistency over time, they find that firms, who change their non-GAAP calculations over time, do this to report a more informative measure. They examine the effect of non-GAAP earnings on overall earnings comparability and find that compared to GAAP earnings, non-GAAP improve overall earnings comparability. Besides that, a new comparability
measure is also used to examine earnings comparability across industry peers and find that firms, who report non-GAAP earnings differently than their industry peers, do this to better inform their stakeholders about future performance. So Black et al. (2017) conclude that managers use non-GAAP measures for informative reasons by providing evidence that non-GAAP improves earnings comparability and by showing that firms’ non-GAAP inconsistency are also for informative motives.
However, other studies argue that opportunism is the primary motive of management to voluntarily disclose non-GAAP financial figures (Doyle et al., 2003; Bhattacharya et al., 2004; Black et al., 2009; Miller, 2009; Barth et al., 2012; Brown et al., 2012). Disclosing non-GAAP measures is understood to be opportunistic, since it benefits management and current shareholders at the expense of other investors who are not shareholders (Miller, 2009). In the research of Miller (2009), earnings forecasts and non-GAAP earnings measures are examined from the academic accounting research. In their research they find that disclosures are done with opportunistic motives, since the evidence show that some managers tend to disclose information so they can influence their firm’s share price with the intent of increasing the share price. So Miller’s (2009) findings support the notion of management voluntarily disclosing non-GAAP financial figures for the motive of opportunism, since these disclosures are only consistent with the objective of maximizing shareholder value, but not consistent with the objective of broader management responsibility theories such as maximizing stakeholders’ and the public’s benefits.
Doyle et al. (2003) research if pro forma earnings are informative or not by examining stock returns and future cash flow. They find that the excluded expenses of non-GAAP earnings predict lower future cash flows and are also negatively related to future stock returns. So the findings of Doyle et al. (2003) demonstrate that excluding certain expenses is done for opportunistic reasons, since these excluded items are important for investors to understand the future performance of the company. They conclude that the opportunistic use of non-GAAP earnings misinforms investors.
Black et al. (2009) research how different types of adjustments affect the choice between GAAP earnings and non-GAAP earnings from continuing operations by using hand collected non-GAAP data. They also research how managers meet strategic earnings targets with the use of different types of adjustments. They find that besides adjusting one-time items such as restructuring charges, managers also adjust recurring expenses like depreciation, stock based compensation, and research and development to meet strategic earning targets, suggesting that these adjustments of recurring exclusions are indicative of aggressive pro forma reporting. Lastly, Black et al. (2009) find that firms who only occasionally adjust their earnings are more likely than
firms who adjust more regularly, to use non-GAAP reporting by excluding recurring items for strategic motives.
Bhattacharya et al. (2004) research the trends of pro forma reporting by analyzing a sample of actual pro forma press releases, and find some characteristics of firms who tend to report on non-GAAP, such as firms being relatively young and concentrated in tech sectors and the business services industry. Other characteristics that they find are that non-GAAP reporting firms are less profitable, and have higher debt levels, and book-to-market ratio, and are more liquid than their peer industry firms. The results from Bhattacharya et al. (2004) also provide evidence for the notion of strategic motives by demonstrating that firms, who normally exclude certain expenses in their pro forma figures, do not usually exclude those same items in their following pro forma statement. Thereby supporting the criticism of managers’ non-GAAP decisions being motivated by their desire to meet or beat analysts’ expectations or avoiding drops of earnings.
Barth et al. (2012) research the adjustments of stock-based compensation expense in non-GAAP reporting by managers and analysts since recognizing this expense was required by the Statement of Financial Accounting Standards No. 123R (SFAS 123R). They find that managers exclude expense to increase and smooth earnings, and also to meet benchmarks, while these exclusions do not result in earnings measures that predict future earnings. On the other hand, they find that analysts’ adjustments of expenses from earnings forecasts do result in earnings future predictability. So the findings of Barth et al. (2012) conclude that opportunistic motives are the main reason for exclusion of expenses for non-GAAP figures, while the predictive ability is the main reason for street earnings’ exclusions.
Brown et al. (2012) research the effect of investors’ reaction on manager’s disclosure of non-GAAP earnings measures in earnings press releases. They find that managers tend to disclose more adjusted earning measures according to the level of investor reactions, and that this tendency is even more when the non-GAAP measure exceeds the GAAP measure. They also find that managers exclude higher levels of both recurring as non-recurring expenses from the non-GAAP earning measure, when investor reactions increase. Besides that, their results also show that managers tend to highlight the non-GAAP measures by placing it notably in the earnings press release, when investor reactions increases. Thus Brown et al. (2012) conclude that the association between investor reaction and manager’s non-GAAP disclosure decision partly reflects opportunistic motives, and that managers’ own motives also play a role in their decisions of non-GAAP disclosures.
The effects of voluntary disclosures are also researched by several studies, where it is discussed that there are three types of effects for voluntarily disclosing information: improved stock liquidity, reduced cost of capital, and increased information intermediation (Healy & Palepu, 2001; Balakrishan et al., 2014). It is argued that there is improved stock liquidity because of increase of confidence by investors in stock transactions at a fair price due to reduction of information asymmetries between the informed and uninformed caused by voluntary disclosures. The effect of reduced cost of capital comes into place when there is an abundance of disclosure or voluntary disclosure and thereby low information risks for investors. Increased information intermediation occurs when required disclosures does not reveal all information, voluntary disclosures reveals information leading to lower cost of information acquisition for analysts.
On the other hand, studies show that voluntary disclosed information can also affect the judgment of less sophisticated investors (Frederickson & Miller, 2004; Elliot, 2006; Allee et al., 2007; Bhattacharya et al., 2007). Frederickson and Miller (2004) experiment on the effects of pro forma earnings disclosure on analysts’ and nonprofessional investors’ judgments. They let both analysts as nonprofessional investors assess stock prices after reviewing background financial information and a firm’s current earnings announcement, which is manipulated to report only GAAP measures in one situation and both GAAP as non-GAAP in the other situation. Subsequently they find that nonprofessional investors assess a higher stock price if they receive an earnings announcement containing both GAAP as non-GAAP figures, than if they receive an earnings announcement containing only GAAP disclosures. Other findings were that analysts’ judgments on stock prices were not affected by any of the disclosures.
In the study of Elliot (2006), pro forma emphasis and the presence of a quantitative reconciliation are examined to find out their influence on nonprofessional investors’ and analysts’ dependence on non-GAAP information, with an experiment. The results of Elliot (2006) show that the emphasis placed on pro forma measures influences the decisions and judgments of less sophisticated investors, however that this effect is reduced when there is a quantitative reconciliation. Elliot’s (2006) findings however show that the effect is different for analysts’ judgment and decisions, where the presence of the quantitative reconciliation leads to analysts viewing non-GAAP earnings measure as more reliable and thereby increasing their dependence to pro forma disclosure when judging a firm’s earnings performance.
Allee et al (2007) use trade-size-based proxies that are constructed from intraday transactions for their study to examine the difference between nonprofessional and professional investors’ reliance on non-GAAP earnings disclosure. Their results are equal to that of Frederickson and Miller (2004), where they find that nonprofessional investors rely more on earnings press releases
if the press releases include a pro forma measure than on those that do not. For professional investors, the finds are opposite, where they rely more on quarterly earnings press releases if it does not include a GAAP figure than if it does. Allee et al. (2007) also find that if the non-GAAP measure is placed before the non-GAAP earnings measure in the press release, nonprofessional investors rely more on the non-GAAP earnings measure, while professional investors are not affected by the relative placement of these measures. Thus Allee et al. (2007) conclude in their research that both the existence of non-GAAP figures as their strategic placement in press releases can affect less sophisticated investors’ judgment.
Bhattacharya et al. (2007) also did research on this subject by examining who actually uses non-GAAP information to trade. They researched this by analyzing intraday transactions around earnings announcements that contains non-GAAP earnings disclosure. And accordingly, the findings from the research of Bhattacharya et al. (2007) also support the previous studies’ findings by showing that less sophisticated investors trade on non-GAAP measures, while the more professional investors do not.
2.3 Securities and Exchange Commission
In all countries there are different regulations that govern corporate reporting and disclosures. In the US there is the Securities and Exchange Commission (SEC) that governs corporate reporting and disclosures for companies entering the capital markets.
2.3.1 SEC 2003 Mandate
In 2003 the SEC adopted new disclosure requirements for non-GAAP financial information after its concern of the increasing use of non-GAAP disclosures and the potential for the misleading of investors (SEC, 2001; SEC, 2003; Clarke et al., 2003). The new disclosure requirements are intended to protect the investors by ensuring that they receive the right information to assess the use of non-GAAP information by companies (SEC, 2003). The requirements include Regulation G, Item 10 amendments of regulation S-K and amendments to form 8-K.
Under Regulation G, companies who publicly disclose non-GAAP financial information need to follow several mandates including disclosure and reconciliation. Public disclosures include SEC filings, earnings releases, investor presentations, and any other publicly disclosed documents related to Regulation FD. Under the disclosure mandate, non-GAAP financial measures may not be disclosed in a way that may mislead investors. The reconciliation mandate requires all companies to disclose the most directly comparable quantitative GAAP measures,
and to present a reconciliation of the GAAP and non-GAAP measures that is clearly comprehensible. Appendix B shows an example from MKS Instruments of 2016 on how they follow the disclosure mandate and the reconciliation mandate with several reconciliations of GAAP figures to non-GAAP figures.
Regulation S-K is a regulation under the US Securities Act of 1933, a law that governs the securities industry, which represents the integrated disclosure requirements and applies to disclosure documents filed with the SEC (SEC, 2016). The amendment of item 10 of Regulation S-K requires more detailed disclosure of SEC filings’ non-GAAP financial measures. Equal to Regulation G, a quantitative reconciliation of the GAAP and non-GAAP financial measures is also required by item 10 of Regulation S-K. Besides that, Regulation S-K also demands prominence and explanation. With prominence it is meant that the comparable GAAP measure should be presented with equal or greater importance than the non-GAAP measure. Explanation means that managers should explain why they believe that the non-GAAP information is useful for the investors and how the management uses the non-GAAP information.
Furthermore, Item 10 prohibits certain measures including non-GAAP liquidity measures that exclude charges or liabilities that require or will require cash settlement, except earnings before interest and taxes (EBIT) and earnings before interest, taxes, depreciation, and amortization (EBITDA)(SEC, 2003). It also prohibits the adjustment of non-GAAP performance measures that eliminate or smooth items that are identified as non-recurring, infrequent or unusual, if that adjusted charge or gain occurred within the previous two years or if the nature of the charge or gain makes it reasonably likely that the charge or gain will recur within two years. Finally, Item 10 also prohibits the use of titles or descriptions for non-GAAP figures that are the same as or similar to titles or descriptions used for GAAP figures.
Certain material corporate events should be reported on a current basis by listed companies, which can be done by filing a current report on Form 8-K whenever the company needs to inform shareholder about major events (SEC, 2012). Form 8-K amendment includes a new disclosure Item 12, which mandates that companies provide on Form 8-K, the information disclosed in their issued earnings releases and other disclosed non public information about the completed annual or quarterly periods, within five business days of its release.
There are several studies that analyzed the effects of the amendments and in analyzing how effective the disclosure regulation was in resolving the problem of increased use of non-GAAP financial information and the potential to mislead investors (Heflin & Hsu, 2008; Campbell & Pitman, 2009; Jennings & Marques, 2011). In the study of Heflin and Hsu, the consequences of the non-GAAP disclosure rules of the SEC are analyzed. Heflin and Hsu (2008)
results suggest that the disclosure regulation has been effective in reducing the opportunistic use of non-GAAP disclosures and in reducing the use of non-GAAP measures. They also conclude that the meeting-or-beating forecasts trend has been intersected by the regulations and that managers and analysts increased their emphasis on GAAP measures.
In the research of Campbell and Pitman (2009), actual corporate press releases are used for their sample, where they analyze non-GAAP financial measures, which agree to the criteria of the SEC’s definition of non-GAAP financial measures. Contrary to the previous studies, Campbell and Pitman’s (2009) findings only show a small decline in companies disclosing non-GAAP information. Besides that, their results also suggest that there was a transformation of adjustment categories used in non-GAAP earnings disclosures and also an increasing use of firm-specific measures after Regulation G.
Jennings and Marques (2011) analyze governance and the SEC disclosure regulation to show the effect on non-GAAP disclosures. Their results suggest that the regulation was effective in reducing the potential of misleading investors, but that corporate governance also plays a role. More specifically, their results show that investors are not misled by non-GAAP measures made by firms that have weaker corporate governance after the SEC regulation, while before the regulation they were misled.
Black et al. (2015) examine two measures of aggressive non-GAAP reporting to analyze how the regulation has influenced potentially misleading disclosures knowing that the SEC has changed its emphasis. Their evidence show that managers are more careful when reporting adjusted earnings measures after the period of the regulation, even though they do find some firms aggressively adjusting recurring items. So Black et al. (2015) conclude that even though the regulation has reduced aggressive non-GAAP reporting, non-GAAP adjustments that could be misleading were still used by a number of companies after the regulation.
Campbell and Lopez (2010) also investigated non-GAAP measures after Regulation G, but was to my knowledge the only study investigating non-GAAP measures for small cap companies. In their study they reveal several causes for the importance of non-GAAP financial information for small cap companies. The results in their paper suggest that a higher level of importance is placed on non-GAAP measures when GAAP earnings show to be of lower value-relevance and when institutional investors percentage are higher. Besides that, as their listing tenure increases, small cap companies place less importance on non-GAAP measures.
On the other side, several studies also show how the quality of non-GAAP has improved since the regulation (Marques, 2006; Kolev et al., 2008; Zhang & Zheng, 2011; Black et al., 2012). Marques (2006) show that even though the probability of the disclosure of non-GAAP financial
measures has declined, there still seems to be a positive market reaction from investors towards the disclosure of non-GAAP earnings.
In the study of Kolev et al. (2008), the effects of the increasing attention of non-GAAP reporting on non-GAAP earnings exclusions are examined. Kolev et al. (2008) findings show that after the SEC intervention, exclusions seem to be of higher quality. They also find that firms, who had exclusions of lower quality before the SEC intervention period, stopped releasing non-GAAP figures after the intervention. Even though these results suggest that the SEC intervention improved the overall quality of non-GAAP earnings measures, they however find evidence that the quality of special items have dropped after the SEC intervention.
Zhang and Zheng (2011) examine the impact of the requirement of Regulation G of reconciling non-GAAP earnings with GAAP earnings on mispricing of non-GAAP earnings. They find that firms with low quality of reconciling before Regulation G had mispricing of non-GAAP earnings, while they find no mispricing evidence for firms with high reconciliation quality and no evidence of mispricing after Regulation G. However they find a cross-Regulation R reduction of pro forma earnings mispricing where reconciliation quality of firms improves, while for firms who have high reconciliation quality both before and after Regulation G, there is no mispricing. So their findings suggest that mispricing is reduced when there is better reconciliation of non-GAAP earnings to GAAP earnings.
Black et al. (2012) examine the effect of investors’ perception on non-GAAP earning measures after Regulation G. They find that investors show more interest to non-GAAP information after the regulation, suggesting that investors perceive non-GAAP measures as more credible since they are more regulated. The results also show that investors disregard aggressive non-GAAP earnings disclosure both before as after the regulation, but that after the regulation this is even more for non-GAAP disclosures that are potentially misleading. Their findings also suggest that the average quality of non-GAAP reporting has increased because of the regulation of non-GAAP, thereby filtering out the ones who are potentially misleading. So the findings of Black et al. (2012) correspond with previous findings that the quality of non-GAAP has improved after the SOX regulatory environment and that it also affected the investor’s perception towards non-GAAP measures.
2.3.2 SEC 2010 Update
The SEC’s Division of Corporation Finance updated its Compliance and Disclosure Interpretations (CDI’s) related to non-GAAP financial measure use in 2010 (SEC, 2010). This turned the direction of the SEC, thereby signaling a substantially increasing flexibility about non-GAAP financial measures inclusion in SEC filings regarding Item 10 of Regulation S-K (Webber
et al., 2013). As mentioned above, Item 10 of Regulation S-K prohibited certain measures that adjusted non-GAAP measures by eliminating or smoothing items identified as non-recurring, infrequent or unusual. However, in the SEC’s 2010 update in its CDIs related to non-GAAP financial measures, there is another statement after the question, whether the prohibition is based on the description of the charge or gain or on the nature. Answering the question, the SEC (2010) explains that the prohibition is based on the description of the adjusted charge or the adjusted gain. They continue explaining that it is only appropriate to indicate that a charge or gain is non-recurring, infrequent or unusual if it meets that specified criteria. But if the registrant cannot indicate the charge or gain as being non-recurring, infrequent or unusual, does not mean that the charge or gain cannot be adjusted. So registrants could make adjustments if they believe this is necessary as long that it follows Regulation G and other requirements of Item 10(e) of Regulation S-K (SEC, 2010).
Consequently with this statement SEC suggests that companies can make adjustments for recurring items even if they do not meet the requirements of non recurring, infrequent or unusual. This flexibility granted by the SEC is in respond to “two-tiered reporting system” coming from the difference between Regulation G and Item 10 of Regulation S-K, where companies could disclose certain non-GAAP financial measures in their press release, while that is not possible in their SEC filings (Webber et al., 2013).
After this update there are only a few studies that have analyzed the effect it would have (Webber et al., 2013; Black et al., 2017). Webber et al (2013) did research after the update and their results show that there was an increase of companies using non-GAAP financial measures. However, they do note that the increase can be due to several factors such as the economic downturn in 2008 resulting in companies using non-GAAP information to either “mask poor performance” or to provide better information to their investors. Besides that, they also indicate that half of the companies in their sample repeatedly disclosed specific types of non-GAAP adjustments. Therefore this also raises the question whether these measures are potentially misleading to investors and thereby re-introduce the problem what SEC in the first place wanted to resolve.
Black et al. (2017) also researched what the effect would be on the quality of non-GAAP earnings by examining the effect on earnings consistency and comparability. Equal to Webber et al. (2013), they also find that there is an increase in non-GAAP reporting frequency. They also find that compared to GAAP earnings, non-GAAP earnings improved comparability of earnings. Besides that, they used a new measure of non-GAAP consistency and thereby concluded that companies changed their non-GAAP calculation for informative reasons.
Thus from the research from Webber et al. (2013) and Black et al. (2017) it can be concluded that after the new CDI announcement, non-GAAP measures have increased, and that the quality of the non-GAAP measures were not compromised by this. However, all of their samples consisted of large SEC registrants, so the question remains if the 2010 CDI update also affected the smaller SEC registrants in the same way as their results.
2.4 Small cap firms
There are different definitions for small cap firms. The Standard & Poor U.S. Indices Methodology identifies small cap firms as public companies with unadjusted market capitalization of US$ 400 million to US$ 1.8 billion, while the Advisory Committee on Smaller Public Companies (ACSPC) of the SEC’s cutoff of small cap market capitalization is from US$ 128.2 million to US$ 787.1 million (SEC, 2006; S&P, 2017).
Although according to ACSPC the percentage of total U.S. equity market capitalization of the small cap companies is only 5%, the percentage of small cap companies in all U.S. public companies is still substantial amounting to 25.9% (SEC, 2006). Additionally, the commissioner of SEC also stresses the importance of small cap companies in the Forum on Small Business Capital Formation by stating that “Small business are the engine that drives the U.S. economy” and that “There is no debate that the success of small businesses is essential to the sustained growth of our greater economy” (SEC, 2014).
Even though small cap firms have some similarities with big corporate firms, some studies show that they are different in several ways (Pettit & Singer, 1985; Ang, 1991; Van Auken & Holman, 1995; Kumar & Francisco, 2005). One important difference is the difference in information asymmetries, where small firms tend to have more information asymmetry than big firms. Since small firms have a lack of long credit history, are relatively young in the capital market, and they have less or no previous financial statements compared to big corporate firms (Kumar & Francisco, 2005). Resulting in small cap firms having more constraints accessing the capital market compared to the large corporations (Van Auken & Holman, 1995). To reduce the information asymmetries it could then be expected that small cap companies’ disclosure behavior would thus be different than the disclosure behavior of large firms.
The study of Campbell and Lopez (2010) researched what influenced the use of non-GAAP information in small firms by using a sample of companies in the S&P Small Cap 600 index. They found that small firms place a higher level of emphasis on using non-GAAP financial information when GAAP earnings are less significant. The results of their study also suggest that small cap companies place less emphasis on non-GAAP earnings when the percentage of
institutional investors is higher and when their listing tenure increases. Looking at the findings of the study of Campbell & Lopez, it can thus be concluded that small cap firms place a level of emphasis on non-GAAP information to reduce the information asymmetries between the firm and investors.
3 Hypothesis development
Since small cap firms have more information asymmetry than large firms, it can be expected that they make more use of non-GAAP disclosures to reduce the information asymmetry, so they can be more accessible for the capital market. In other words, the use of non-GAAP disclosures by small firms is expected to inform their investors and prospective investors about profitability of their potential recent and future investment opportunities.
According to the FASB (2006), for the information to be useful for the investors’ decision making, it needs to contain several qualitative characteristics: relevance, faithful representation, comparability and understandability. Consistency is a means to reach the goal of comparability qualitative characteristic of financial information. Consistency refers to both consistency in accounting policies and procedures from period to period within an entity as in a single period from firm to firm.
The update of the CDI in 2010 can affect informativeness of non-GAAP measures, since comparability can be affected. With the update of the CDI, firms have the possibility to adjust recurring items even if they do not meet the requirements of non-recurring, infrequent or unusual. This could lead to inconsistency in adjustments of recurring expenses, and thereby affect the comparability qualitative characteristics of the financial figures.
Inconsistently adjusting recurring items can be done to inform investors about profitability of their potential recent and future investment opportunities. Since small cap firms try to reduce their information asymmetry by informing their investors with non-GAAP measures, they would vary their adjustments of recurring items from time to time to inform their investors if they believe that those recurring items are important to be included in the earnings or not for the investors’ decision-making regarding future performances. On the other hand, adjusting recurring items from time to time can also be done for managers’ opportunistic reasons.
Thus, since the announcement of the CDI in 2010 can lead to informativeness as well as opportunism, I state the following hypothesis in null form:
H0: There is no association between the announcement of the CDI in 2010 and informativeness of non-GAAP measures of small cap firms.
The hypothesis is presented in the Libby box in figure 1. The two boxes on the upper row represent the conceptual independent and dependent variables, while the two middle boxes represent the operational variables. The control variables are presented in the lowest box on the right.
FIGURE 1 Libby box H0
CDI announcement Small cap firms’ non-GAAP
informativeness
Earnings announcement date after CDI announcement = 1
Otherwise = 0
There is no relation between small firms’ recurring expenses and future operating
performance
Total assets, market-to-book ratio, leverage, and loss
4 Methodology
4.1 Sample selectionI follow prior research (Kolev et al., 2008) by using I/B/E/S actual earnings to proxy for non-GAAP earnings disclosed by managers. I obtain data from the Compustat Fundamentals Annual file and the I/B/E/S Detail Unadjusted Actual file. The period tested ranges from 2008 through 2011, which allows for comparing the periods before and after the new CDI, which is in 2010. I identify small firms as the firms who are listed in the S&P Small 600 index and have a company ID number in the Compustat Fundamentals Annual file between the years 2008 and 2011, where I obtain a total of 565 companies’ ID numbers.
After merging the data from Compustat and I/B/E/S, I obtain a sample of 2,025 observations for domestic US-firms with key Compustat variables such as total assets, total ordinary equity, common shares used to calculate earnings per diluted share, common shares outstanding, earnings per diluted share excluding extraordinary items, total liabilities, special items, and street earnings per share data from I/B/E/S. Since the nature of financial firms differs from those of nonfinancial firms, I follow prior studies (Campbell & Lopez, 2010; Leung & Veenman, 2016) by excluding them from the sample. The full sample remains with 1,364 observations with non-missing values for each of the variables needed to test the hypothesis.
4.2 Research design & Variable measurement
Non-GAAP informativeness dependent variable is operationalized with a model comparable as the one used from prior research (Kolev et al., 2008; Black et al., 2017) by examining how non-GAAP exclusions map into future operating performance, and the conceptual independent variable of CDI announcement is operationalized with the variable NewCDI:
Future performancet+1 = a0 + a1Non-GAAP Earningst + a2New CDIt + a3Recurring Exclusionst + a4New CDI * Recurring Exclusions + a5Log (Total Assets)t + a6Market-To-Book Ratiot + a7Leveraget + a8Losst + Et+1
Future performance is defined as operating earnings per share from a firm in the following year, which is measured with the Compustat data item “Earnings per share (diluted) – from operations”. Non-GAAP Earnings are defined as the I/B/E/S actual street earnings per share. New CDI is defined as 1 if earnings announcement is after 11 January 2010, and 0 if otherwise.
Recurring Exclusions are defined as the total per share excluded recurring amount in calculating non-GAAP EPS. The total non-GAAP exclusions per share are calculated with total
non-GAAP earnings per share actual street value from I/B/E/S less earnings per share (Diluted) excluding extraordinary items. The total per share excluded recurring amount is calculated by subtracting the calculated total non-GAAP exclusions per share with special items per share. Special items per share are measured by dividing the Compustat total “special items” with Compustat’s “Common shares used to calculate earnings per share – diluted”. Exclusions related to recurring expenses (gains) are coded as a negative (positive) value. New CDI * Recurring Exclusions relates to the interactions between New CDI and Recurring Exclusions.
Similar to the previous study from Black et al. (2017), the following control variables that are associated with future performance are included: log of total assets, market-to-book ratio, leverage, and an indicator variable if the firm reports lost. Log of total assets is the log of the firm’s total assets as measured by Compustat. The market-to-book ratio is measured by dividing the Compustat data item “Price Close” with the calculated book value per share. The book value per share is measured by subtracting the Compustat data items “Total assets” with “Total liabilities” and by dividing the difference with data item “Common shares outstanding”. The leverage is measured by dividing the Compustat data item “Liabilities- total” with the data item “Common/Ordinary equity”. Loss is an indicator variable if the firm reports a loss, where it gets an indicator of 1 if the GAAP earnings is less than 0, and an indicator of 0 if otherwise.
5 Results
5.1 Descriptive statistics
The dataset of the sample provides new descriptive statistics about non-GAAP reporting for small firms, since other studies mainly provided non-GAAP reporting statistics for big firms (Black et al., 2017) and non-GAAP statistics for periods only before the new CDI announcement (Kolev et al., 2008; Campbell & Lopez, 2010). Campbell and Lopez (2010) is the only study to my knowledge that researched non-GAAP behavior of small firms, however the sample was in the time period before the new CDI announcement. Besides that, the variables for the descriptive statistics of their sample are not similar to this research, since they researched the determinants of non-GAAP information importance, and not the quality of non-GAAP information. On the other hand, Black et al. (2017) do have similar variables and a similar time period for their descriptive statistics, nevertheless the sample differs since they researched big cap firms. Kolev et al. (2008) also performed a similar study as Black et al. (2017) and my research, however their research was in the time period before the new CDI announcement and consisted of a sample of only big firms.
The descriptive statistics of this research for each variable of the whole sample is presented in table 1. The mean (median) GAAP earnings per share of the sample is 0.687 (0.790) and non-GAAP earning per share is 1.182 (1.005), which is much lower compared to the means (medians) of Black et al. (2017). Black et al. (2017) present a mean (median) GAAP earning per share of 3.06 (2.62) and a non-GAAP earning per share of 3.50 (3.04). This difference can be explained due to the difference of the sizes of the companies, since the sample of Black et al. (2017) includes only big firms, while this sample covers small cap firms. The mean (median) of total exclusions per share is 0.474 (0.090), special item per share is -0.420 (-0.004), and recurring exclusions is 0.885 (0.183). The total exclusions per share and special item per share averages of Black et al. (2017) only differ slightly from this sample, which is -0.42 (0) and -0.39 (-0.10).
Table 2 presents the descriptive statistics of the differences of the variables before and after the new CDI announcement. The mean GAAP earnings per share are significantly higher after the new CDI announcement date compared to before the new CDI announcement, with a mean of 1.139 after 11 January 2010 compared to 0.404 before 11 January 2010. The median GAAP earnings per share have also improved after the new CDI, rising from 0.660 to 0.960. Likewise, mean Non-GAAP earnings per share have also improved after the new CDI, increasing from 1.152 to 1.230, while the median non-GAAP earnings per share have slightly increased from 1.000 to 1.050.
TABLE 1 Descriptive Statistics
N= 1,364
Variable Mean Std. Dev. P25 P50 P75
GAAP Earnings p/s 0.687 2.272 0.130 0.790 1.670 Non-GAAP Earnings p/s 1.182 1.644 0.390 1.005 1.880 Total Exclusions p/s 0.474 1.901 -0.120 0.090 0.730 Special Items p/s -0.420 1.434 -0.201 -0.004 0.000 Recurring Exclusions p/s 0.885 3.076 -0.119 0.183 0.972 Future Performance p/s 1.122 1.601 0.305 0.955 1.845 New CDI 0.385 0.487 0.000 0.000 1.000
Log Total Assets 6.539 1.126 5.766 6.396 7.278
Market-To-Book 2.084 1.968 1.062 1.517 2.330
Leverage 2.152 3.795 0.376 0.821 1.820
Loss 0.215 0.411 0.000 0.000 0.000
Notes: The continuous variables are winsorized at the 1st and 99th percentiles. Variables definitions:
GAAP Earnings p/s = Earnings per share (diluted) – Excluding extraordinary items. Non-GAAP Earnings p/s = I/B/E/S reported actual street earnings per share.
Total exclusions p/s = Non-GAAP exclusions – total share calculated: non-GAAP earnings less GAAP earnings.
Special Items p/s = Special items per share.
Recurring Exclusions p/s = Recurring exclusions per share calculated: total exclusions less special items. Future performance = Operating earnings per hare in the following year.
New CDI = Indicator variable that equals 1 if the observation falls before 11 January 2010, and 0 if otherwise.
Log Total Assets = Log of total assets.
Market-To-Book = Market –to-book ratio of common equity.
Leverage = Leverage calculated with total liabilities divided by total assets.
Loss = An indicator variable that equals 1 if earnings before extraordinary items per share is less than 0, and 0 if otherwise.
Total exclusions means have changed significantly from 0.726 in exclusion gains before 11 January 2010 to 0.072 in exclusion gains after 11 January 2010, whereas the medians of total exclusions do not differ a lot between these two periods from 0.180 to 0.010 after 11 January 2010. It can be seen that the big drop of the mean of the total exclusions comes from also a substantial drop from recurring exclusions per share, while special items per share also experienced a drop in expenses. The means of special items per share have decreased after the new CDI announcement from -0.577 to -0.169, while the medians show almost no difference. Recurring exclusions’ means likewise fall considerably after 11 January 2010 from 1.298 in recurring exclusions gains to 0.226 in recurring exclusions gains whilst the medians show a
slighter difference from 0.256 before 11 January 2010 and 0.081 after. Future performance means have minor differences across the periods, whereas the medians differ from 0.960 to 0.930 after the new CDI.
The drop of the recurring exclusions after the CDI announcement date of 11 January 2010 is not as expected, as it was expected that recurring exclusions would increase, since the new CDI would make the possibility of easily adjusting a charge or gain as a recurring exclusion if a charge or gain cannot be indicated as a non-recurring, infrequent or unusual. However, the results show that both recurring exclusions as special items have decreased since the new CDI announcement date, while GAAP earnings have increased after the new CDI, suggesting that small firms make more use of exclusions when GAAP earnings declines. This could also explain why after the CDI announcement, there was the unexpected decline of recurring exclusions.
TABLE 2
Descriptive Statistics Before and After New CDI
N=839 N=525
Before 11 Jan. 2010 After 11 Jan. 2010
Variable Mean Median Mean Median
GAAP Earnings 0.404 0.660 1.139 0.960 Non-GAAP Earnings 1.152 1.000 1.230 1.050 Total Exclusions 0.726 0.180 0.072 0.010 Special Items -0.577 -0.002 -0.169 -0.005 Recurring Exclusions 1.298 0.256 0.226 0.081 Future Performance 1.121 0.960 1.124 0.930
Log Total Assets 6.528 6.389 6.558 6.408
Market-To-Book 1.958 1.413 2.285 1.626
Leverage 2.343 0.889 1.846 0.715
Loss 0.262 0.000 0.139 0.000
Notes: see Appendix C for variables definition. The continuous variables are winsorized at the 1st and 99th percentiles.
Table 3 presents a correlation matrix between the variables in the sample. From the matrix it can be perceived that total exclusions are negatively correlated with GAAP earnings per share (-0.655), while total exclusions are positively correlated with non-GAAP earnings (0.217). This is in line with previous research for big firms, where previous research shows that when GAAP earnings are low, it is more likely that firms will exclude items to reach to a non-GAAP measure.
When looking into the different types of exclusions, it can be noted that there is a difference of correlation, where special items are positively correlated to both GAAP earnings (0.659) and non-GAAP earnings (0.085), while recurring exclusions are negatively correlated with GAAP earnings (-0.717) and positively correlated with non-GAAP earnings (0.090). This also confirms the previous finding mentioned above for the comparison before and after the new CDI announcement, that recurring exclusions decrease when the GAAP earnings increase.
I also find that future performance is less positively correlated with GAAP earnings (0.536) than with non-GAAP earnings (0.670). This also agrees to the findings of Kolev et al. (2008), where they show a more positive correlation between future performance and non-GAAP earnings than with non-GAAP earnings. This also suggests that for both small firms as big firms, the non-GAAP earnings are more value relevant than the GAAP earnings. On the other side, future performance is negatively correlated with total exclusions (-0.048), which means that future performance decreases when total exclusions increases, suggesting that non-GAAP exclusions have a value relevant effect for future earnings. This is in contrast to previous studies’ results, where they find that non-GAAP figures that are excluded are less value relevant than those that are included in the non-GAAP figures (Bradshaw and Sloan, 2002). Looking into the different exclusions, future performance is positively correlated with special items (0.063), while it is negatively correlated with recurring exclusions (-0.059).
New CDI is positively correlated with GAAP earnings (0.157), while it is also positively correlated with non-GAAP earnings (0.023), however the correlation between new CDI and non-GAAP earnings is non significant (p= 0.391). Lastly, comparing the different types of exclusions’ correlations with the new CDI, I find that new CDI is positively correlated with special items (0.138), while it is negatively correlated with recurring exclusions (-0.170). The correlation between the new CDI and recurring exclusions is inconsistent with what was expected that would happen from the new CDI. It was expected that there would be an increase of recurring exclusions because of the flexibility of the new CDI on recurring exclusions in contrary with the prohibition of recurring expenses in calculating non-GAAP figures before the new CDI. However, looking at the positive correlation between the new CDI and GAAP earnings, the unexpected result of the recurring expenses can be explained. Since GAAP earnings increase after the new CDI, this results in a decrease of recurring exclusions after the new CDI, since recurring exclusions also have e negative correlation with GAAP earnings.
