Are the U.S. Securities and Exchange Commission new Compliance and Disclosure Interpretations (May 17, 2017) on reporting non-GAAP disclosures effective (less bias and misinterpretation)?

(1)

Thesis

Are the U.S. Securities and Exchange Commission new Compliance and Disclosure Interpretations (May 17, 2016) on reporting non-GAAP disclosures effective (less bias and misinterpretation)?

Name: Delilah Salook Student number: 11063696

Thesis supervisor: David Veenman Date: 25 June 2017

Word count: 20995, 0

MSc Accountancy & Control, specialization Accountancy and Control Faculty of Economics and Business, University of Amsterdam

(2)

2

Statement of Originality

This document is written by student Delilah Salook 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.

(3)

3 Abstract

Because of concerns that non-GAAP financial measures were subjected to bias or misinterpretation, regulation is introduced. Reporting of non-GAAP earnings changed with the introduction of non-GAAP regulation. I examine whether the U.S. Securities and Exchange Commission new and revised Compliance and Disclosure Interpretations on reporting non-GAAP disclosures are effective (less bias and misinterpretation). I hand-collected data on non-GAAP measures from the 8-K earnings press releases and find, other than the presentation of the (adjusted) earnings per share, no change if you compare the period before and after the new Compliance and Disclosure Interpretations. There was no decline in the frequency of companies that reported non-GAAP and there was no decline in the magnitude of exclusions under non-GAAP after the new Compliance and Disclosure Interpretations. Regarding the presentation, 41% of the companies shifted from not presenting the (adjusted) earnings per share in the correct way to presenting it in the correct way. Because there is no change present, it can be argued that non-GAAP disclosures are not used to mislead or manage the perception of the investors or that the managers’ who mislead or manage the perception of the investor also stayed the same. Overall, I conclude that the new Compliance and Disclosure Interpretations didn’t have much effect.

(4)

4

For the functioning of an efficient capital market, corporate disclosure is critical. Disclosures consist of regulated financial reports prepared under General Accepted Accounting Principles (GAAP), as well as voluntary disclosures. Managers do not voluntarily disclose all their private information, so there is room for regulation with regards to disclosure in capital markets. Beyer et al. (2010) gives two main reasons for the regulation of disclosure, misalignment of entrepreneurs’ or management’s and investors’ incentives and the public goods aspect of disclosures which results in free-rider problems.

Managers have the incentive to disclose information to separate themselves from other managers with less favorable information. Bhattacharya et al. (2003) stated that managers provide non-GAAP disclosures to provide a clearer picture of the core earnings of the company that they believe will continue in the future. According to Kim and Verrecchia (1994), voluntary disclosures can have potential benefits, such as reducing information asymmetry among informed and uninformed investors which results in higher stock liquidity and lower cost of equity capital.

Non-GAAP financial measures are voluntary disclosures which adjust the reported GAAP numbers. Baumker et al. (2014) report that GAAP earnings are GAAP earnings that exclude cash or non-recurring items and generally exceed GAAP earnings, which in turn causes regulatory scrutiny. In the past there has been a lot of controversy about non-GAAP measures. Prior research done by Entwistle et al. (2006) shows that non-GAAP disclosures can be used either to (1) communicate transparency or to (2) deceive or mislead. Kolev et al (2008) provide evidence that non-GAAP disclosures tend to be more value-relevant than GAAP earnings and also fulfill a valuation role. Bhattacharya et al. (2003) find that market participants believe that non-GAAP earnings are more representative of core earnings than GAAP operating income. However, Kolev et al. (2008) also find evidence that some managers use non-GAAP earnings opportunistically.

Because of the concerns that non-GAAP financial measures were subjected to bias or misinterpretation, section 401(b) of the Sarbanes-Oxley Act instructed the U.S. Securities and Exchange Commission to introduce regulation of non-GAAP financial information. Regulation G was then introduced. A lot of studies, among which Entwistle et al. (2006), examined whether firms’ reporting of non-GAAP earnings changed with the introduction of non-GAAP regulation. They find that non-GAAP disclosures are used in a less biased, prominent and potentially misleading manner after regulation. In contrast, Baumker et al. (2014) provide some evidence that there continues to be an opportunistic component of non-GAAP reporting following Regulation G.

(7)

7

In extension to these studies, in this paper I will examine whether the new and revised Compliance and Disclosure Interpretations are effective (less bias and misinterpretation). The new Compliance and Disclosure Interpretations are the latest regulation issued by the U.S. Securities and Exchange Commission on May 17, 2016 with the intention to clarify guidance on the use of non-GAAP financial measures.

I initially obtained 674 quarterly 8-K earnings press releases from companies from the Standard & Poor’s 500. My final sample that I hand-collected consist of 314 firms so in total 628 observations. I compared the first quarter of 2016 with the third quarter of 2016. I find that there was no decline in the frequency of companies that reported non-GAAP after the new Compliance and Disclosure Interpretations. I also did not find a significant increase, so the frequency of companies that reported non-GAAP after the new Compliance and Disclosure Interpretations stayed exactly the same. However, the control variable size of the firm has a low statistically significant effect on non-GAAP earnings. Firm size is positively related to non-GAAP earnings, thus bigger firms are more likely to report non-GAAP earnings and smaller firms are less likely to report non-GAAP earnings. The control variable age of the firm has a highly statistically significant effect on other non-GAAP measures. The age of the firm is negatively related to other non-GAAP measures. Older firms report less other non-GAAP measures and younger firms report more other non-GAAP measures.

I also find that there was no decline in the magnitude of exclusions under non-GAAP after the new Compliance and Disclosure Interpretations. I also did not find a significant increase, so the magnitude of exclusions under non-GAAP after the new Compliance and Disclosure Interpretations stayed the same. Other result shows that the control variable age of the firm has a significant effect on how many other non-GAAP measures there were presented in the earnings press release reports. The age of the firm is negatively related to how many other non-GAAP measures there were presented. The magnitude of other non-GAAP measures is smaller for older firms and bigger for younger firms. The control variables size of the firm and age of the firm have a low significant effect on the difference between non-GAAP and GAAP earnings per share. Firm size is positively related to the difference between non-GAAP and GAAP earnings per share. Bigger firms are more likely to have a greater difference and smaller firms are more likely to have a smaller difference between non-GAAP and GAAP earnings per share. The age of the firm is negatively related to the difference between non-GAAP and GAAP earnings per share. When firms get older, the difference between non-GAAP and GAAP earnings per share gets smaller. Younger firms have bigger differences between non-GAAP and GAAP earnings per share. The control variables loss and earnings volatility have a highly significant effect on the

(8)

8

difference between non-GAAP and GAAP earnings per share. Loss and earnings volatility are both positively related to the difference between non-GAAP and GAAP earnings per share. Firms with a higher loss and earnings volatility also have a bigger difference between non-GAAP and GAAP earnings per share. Firms with a lower loss and earnings volatility have a smaller difference between non-GAAP and GAAP earnings per share.

I also researched if there was a significant change in the non-GAAP measures that stands out and in the presentation in between and in the last part of the financial statements. I find no statistically significant difference for all, so I conclude that there was no change in non-GAAP measures that stands out and in the presentation in between and in the last part of the financial statements.

The only significant change that was found during my whole research is the place of the presentation of the earnings per share and the adjusted earnings per share. The correct way to present it, is first GAAP earnings and then non-GAAP earnings. A big shift took place from not presenting it in the correct way to presenting it in the correct way, about 41% of the companies shifted to the correct way of presenting (non-)GAAP earnings.

So all taken together, other than the presentation of the (adjusted) earnings per share, there is no change if you compare the period before and after the new Compliance and Disclosure Interpretations. These findings can be argued, I will explain two views. The first one is that non-GAAP disclosures are not used to mislead or manage the perception of the investors. I expected a decline in non-GAAP disclosures after the new Compliance and Disclosure Interpretations if non-GAAP disclosures are used to mislead or manage the perception of the investors. I expected this outcome because of the new Compliance and Disclosure Interpretations, managers would be more careful and less motivated to mislead or manage the perception of the investors. These findings suggest that before the new Compliance and Disclosure Interpretations, non-GAAP disclosures are also not used to mislead or

manage the perception of the investor because there was no change. However, the second view is that

these findings can also suggest that the managers’ who mislead or manage the perception of the investor also stayed the same because the new Compliance and Disclosure Interpretations didn’t have much effect. Thus, within this view, the new Compliance and Disclosure Interpretations wasn’t effective (less bias and misinterpretation).

In my study, managers’ decision to disclose non-GAAP measures is not affected at all. Like Entwistle et al. (2006) reported, if managers would use non-GAAP disclosures to inform investors, there would be no decline present in reporting non-GAAP measures, and this shows in my research. Neither a decline nor an increase in the frequency and magnitude of non-GAAP measures are present or they

(9)

9

can be both present but are offsetting each other. I did not expected a reduction for a long time in reporting non-GAAP measures after the new Compliance and Disclosure Interpretations because Black et al. (2017) reported that over the last decade the propensity for managers to disclose non-GAAP earnings has increased significantly. I expected an initial reduction, just like after regulation G but this effect wasn’t present at all. My results supports the research done by MarketWatch (2017) which also stated that the new Compliance and Disclosure Interpretations didn’t have much effect.

The implication of this study is that regulation doesn’t always have an effect. In my study the new Compliance and Disclosure Interpretations didn’t have much effect. There was, other than the presentation of the (adjusted) earnings per share, no change. The question is whether regulation leads to the desired outcomes, if there is no visible observed effect present questioned are raised about the effectiveness of the regulation. Regulators, practitioners and standard setters have to raise themselves the question if this is the outcome they desire. The implication for future research is to focus more on the age, size, loss and earnings volatility of firms because these variables does have an effect on non-GAAP measures. Another implication for future research is to go more in depth why regulation doesn’t have much effect or if they actually do have an effect but nothing changes. As in my research, the answer to this question can change the results.

The contribution of this research will be to extend existing knowledge and this extension will be about a new regulation in a new timeframe. It is relevant for the U.S. Securities and Exchange Commission but also for the companies itself to know, in addition to prior literature, if regulation really changes reporting and what this change will be, but also to what extent. Another contribution is the limited research done based on recent hand-collected data about the new Compliance and Disclosure Interpretations. Most prior research discusses other regulations than the new Compliance and Disclosure Interpretations and this regulation in combination with datasets based on recent hand-collection is limited.

2. Background literature

2.1. Financial reporting and disclosure

Financial reporting and corporate disclosure are means for management that are important to communicate firm performance and governance to outside investors. Healy and Palepu (2001) provide a framework for analyzing managers’ reporting and disclosure decisions. For the functioning of an efficient

(10)

10

capital market, corporate disclosure is critical. Disclosures are provided by firms through regulated financial reports which includes the financial statements, management discussion, footnotes and analysis and other regulatory filings. In addition to regulatory disclosure, some firms engage in voluntary communication, such as press releases, analysts’ presentations and conference calls, management forecasts, internet sites and other corporate reports. Disclosures about firms by information intermediaries is another form, such as financial analysts, the financial press and industry experts.

In a capital market economy, information and incentive problems hinder the efficient allocation of resources. The optimal allocation of savings to investment opportunities is an important challenge for any economy. New entrepreneurs but also existing companies would like to attract household savings to fund their business ideas. So a logical consequence is that savers and entrepreneurs would like to do business with each other, matching savings to business investment opportunities.

However, matching savings to business investment opportunities is complicated because of two reasons, these are "information problem" and "agency problem". "Information problem" means entrepreneurs generally have better information than savers about the value of business investment opportunities and incentives to overstate their value. Because of this, when savers make investments in business ventures they face an "information problem". According to Jensen and Meckling (1976) "Agency problem" means that when savers have invested in business ventures, entrepreneurs have an incentive to expropriate their savings. Savers usually do not intend to play an active role in its management, thus the responsibility is delegated to the entrepreneur. After the investment of the saver, the entrepreneur that is self-interested has an incentive to make decisions that expropriate the funds of the savers’. Examples of expropriation are excessive compensation, acquiring perquisites or making investments or operating decisions that are harmful to the interest of outside investors. To mitigate these two problems, disclosure between managers and investors are very important but also the institutions that are created to facilitate credible disclosures.

Managers do not voluntarily disclose all their private information, so there is room for regulation with regards to disclosure in capital markets. According to Beyer et al. (2010) two main reasons are given for the regulation of disclosure. The first reason is misalignment of entrepreneurs’ or management’s and investors’ incentives, which can make it difficult for managers to transfer information that is credible. Regulation, like for example the disclosure requirements from the U.S. Securities and Exchange Commission enforcement division are mechanism that allow firms to commit to certain levels of disclosure but also to improve the credibility of disclosed information. Disclosure requirements, auditors and accounting standards are also mechanisms to improve the credibility of disclosed information.

(11)

11

The second reason is the public goods aspect of disclosures which results in free-rider problems, which occurs when people enjoy a good service without paying anything. In a free market, if enough people can enjoy a good without paying for the cost there is a danger that the good will be under-provided or not provided at all. In this case, even though additional information would improve social welfare, it creates situations in which managers’ incentives to voluntary disclose information are not sufficient. This is also the case why regulation of certain disclosures can be good. The regulation of disclosure depends on the kind of information that a firm voluntarily discloses or that can be produced by other participants of the market.

General Accepted Accounting Principles (GAAP) is a regulation which consist of accounting standards, conventions and rules that publicly traded and regulated companies are mandated to follow when they report their financial information. Companies without external investors are not obligated to follow GAAP. Companies use these to measure their financial results. These results include net income but also how companies record assets and liabilities. In the United States, the U.S. Securities and Exchange Commission has the authority to establish GAAP, however they have historically allowed the private sector to establish the guidance (U.S. Securities and Exchange Commission (SEC), 2016). 2.2. Voluntary disclosures

Beyer et al. (2010) argue that if managers do not disclose any information, investors take the conclusion that the information that is not disclosed would have caused them to revise their beliefs about firm value downwards because they conclude that the information is bad. Because of this reason, managers have the incentive to disclose information to separate themselves from other managers with less favorable information.

The authors discuss six different conditions under which firms voluntarily disclose all their private information. These are that (1) disclosures are costless; (2) investors know that firms have private information; (3) all investors interpret the firms’ disclosure in the same way and firms know how investors will interpret that disclosure; (4) managers want to maximize their firms’ share prices; (5) firms can credibly disclose their private information and (6) firms cannot commit ex-ante to a specific disclosure policy.

(12)

12

If a manager wants to maximize firm value and disclosure is costly, the firm will only disclose information if it is sufficiently favorable. It is favorable when it reveals that asset values are expected to be high and/or low risk (Lanen and Verrecchia, 1987; Jorgensen and Kirschenheiter, 2003).

(2) Probabilistic information endowment

If the investor does not know that the manager has private information, firms with unfavorable information do not disclose because from the investors’ point of view they are the same as firms without information (Pae, 2002).

(3) Uncertain investor response

All investors interpret and react to managers’ (absence of) disclosures. The investors’ interpretation and reaction is known to managers. Investors are allowed to have private information about the demand for the firm’s products. Investors who have private information that indicate that the demand for the firm’s product is high will interpret high inventory levels as positive news to meet the future demand and vice versa. Because of this uncertainty about the interpretation of the investor about disclosure causes managers to disclose information that is sufficiently high or low (Dutta and Trueman, 2002).

(4) Uncertain disclosure incentives

Managers want to maximize the firm’s stock price and investors know that this is the objective of the manager but this condition may not hold when managers have incentives to maximize stock prices but also sometimes have incentives to minimize stock prices. When disclosure is absent, investors price the firm at a weighted average of good and bad news and thus can prevent full disclosure (Einhorn, 2007).

(5) Non-verifiable disclosure

Firms can only make truthful disclosures. However, some of the information is being shared through informal communication channels, through these channels firms do not have to tell the truth. Investors can learn from voluntary disclosures that may not be truthful and what disclosure strategies managers follow but this depends on whether misrepresentation is costless, also known as cheap-talk models. If there are no costs, managers will disclosure information that will lead the investor to value their firm to the closest objectives of the manager. When misrepresentation is costly it is also known as costly state falsification models. Within this model, disclosures do not have to be truthful, false reporting is costly to managers (Fischer and Stocken, 2001; Beyer, 2009).

(6) Ex-ante commitment to disclosure strategies

Firms cannot commit in a credible way to a disclosure policy prior to receiving private information. For example, risk-aversion occurs when people are subject to risks of for example a loss. Because of this, people demand insurance which allows them to share the risk. Each person has private information of

(13)

13

the probability of incurring a loss. If they cannot commit to a disclosure policy ex-ante, the person with low probability will lower their insurance premium. Because of this, insurers will interpret non-disclosure as an indication that the person will incur a loss with high probability. Thus, the insurance premium will decrease for people with low-risk and increase for people with high-risk and this reduces risk-sharing opportunities. Another example is the decision of managers to commit ex-ante to a disclosure strategy versus when they cannot. When the selling price is positive, managers who are able to commit ex-ante choose not to disclose any information but managers who cannot commit ex-ante disclose their private information. Ex-ante disclosure reduces the informed firm’s profit because it allows the uninformed firm to increase its output when market demand is high. This shows that disclosure can reduce health because it has the ability to destroy risk-sharing opportunities and disclosure must have some real effect that lead to benefits that outweigh the costs of disclosure (Dye, 2001; Verrecchia, 2001).

Voluntary disclosures can have potential benefits, as Kim and Verrecchia (1994) stated that voluntary disclosure can help to reduce information asymmetry among informed and uninformed investors which result in higher stock liquidity and lower cost of equity capital. It will result in higher stock liquidity because the investors of firms that disclose at a high level are confident that any stock transaction occur at a "fair price" and thus this will increase the liquidity of the stock. It will result in lower cost of equity capital because managers provide voluntary disclosure to reduce the cost of capital. Barry and Brown (1984) argue that when disclosures are not perfect, investors bear the risk of the future payoffs from their investment. When the risk cannot be diversified, investors will demand a return for bearing the information risk. Thus, firms with high disclosure and low information risk are likely to have a lower cost of equity capital than firms with low disclosure and information risk that is high. Another benefit of voluntary disclosure is increased information intermediation. If disclosures of management’s private information are not fully revealed through the required disclosures, the costs of information acquisition for analysts and their supplyare lower through voluntary disclosure.

2.3. Non-GAAP

2.3.1. Use of non-GAAP measures

Firms have the choice to supplement their financial reports by reporting non-GAAP financial measures, which are voluntary disclosures. These measures adjust the reported GAAP numbers, which

(14)

14

provide additional insight into the business. By providing additional insight, this can help users to better understand the company’s underlying operational performance, financial position or liquidity.

Non-GAAP measures are often used by companies, but across industries and between companies their comparability varies. This can lead to a lack of transparency with regard to the decision why companies use non-GAAP measures, inconsistency and subjectivity in how the measures are calculated. Without proper explanation and context, non-GAAP financial measures can be subject to bias or misinterpretation (PriceWaterhouseCoopers, 2014).

2.3.2. Difference non-GAAP measures and GAAP measures

Baumker et al. (2014) and Bhattacharya et al. (2003) report that non-GAAP earnings are GAAP earnings that exclude non-cash or non-recurring items. A non-cash item is an expense that is reported on the income statement of the current accounting period but there is no related cash payment during this period, examples are depreciation and amortization. A non-recurring item is unlikely to occur again

in the normal course of business, like for example litigation fees and repair costs for damage caused by

natural disasters. By using non-GAAP measures companies can choose to exclude some GAAP expenses from the GAAP numbers, which are then published in press releases. Companies report this non-GAAP number in the same press release with their standard GAAP number. Non-GAAP earnings generally exceed GAAP earnings, which in turn causes regulatory scrutiny.

2.3.3. Adjustments non-GAAP disclosures

Bhattacharya et al. (2003) categorized the adjustments that are made with regards to non-GAAP disclosures in the press releases by companies within the next categories: (1) depreciation and amortization costs; (2) stock based compensation costs; (3) merger and acquisition costs; (4) acquired in-process research and development costs written-off; (5) gains or losses on asset dispositions; (6) "below the line" items; (7) adjustments to the number of shares outstanding used in the denominator of the earnings per share calculation and (8) other specific adjustments. Zhang and Zheng (2011) included three additional categories which are (9) adjustments related to amortization of intangibles and goodwill, (10) restructuring charges and (11) write-offs due to asset impairment.

Bhattacharya et al. (2003) investigates whether market participants perceive non-GAAP earnings to be more informative and more persistent than GAAP operating income. The adjustments that occur

(15)

15

most often with regards to non-GAAP disclosures are routine expenses, such as depreciation and amortization. These expenses should be included in operating income under GAAP, under non-GAAP this result in higher income figures. Non-GAAP operating income result more often in a profit than GAAP operating income. Because firms use different adjustments to arrive at the non-GAAP numbers, these numbers may not be comparable across firms. Black et al. (2017) explore how the discretion afforded in non-GAAP reporting influences earnings consistency and comparability. They find that the most common non-recurring exclusions in dollar value relate to restructuring, tax resolution and acquisition. These exclusions are reported across various sectors and represent the largest non-recurring exclusions. With regards to the recurring items which are items that are likely to occur again in the normal course of a business, the most commonly excluded items relate to investments, amortizations of intangibles and stock-based compensation. These exclusions are clustered in a few sectors and represent the largest recurring exclusions.

Isidro and Marques (2013) report the international evidence on the impact of board quality and compensation on the voluntary disclosure of non-GAAP earnings numbers. They observe that firms adjust for recurring items such as R&D expenses, stock-based compensation, depreciation expenses and tax-related values, more often when the compensation contracts of the directors are linked to market performance. This finding suggests that firms are aware that capital markets react to non-GAAP earnings disclosures and they use them strategically to increase their market value. As a result, this increases performance based compensation. Board quality, however, has no effect on the decision to adjust items that are viewed as recurring. A possible explanations is that board members do not have detailed information on how often excluded items can recur. Their overall finding is that the nature of the adjustments varies substantially, the common adjustments are those related with goodwill, gains and losses on sale of assets and special items. Special items are identified as an unusual or infrequent expense or income that a company does not expect to recur in future years. A special item is reported in the financial statement but is separated from other categories of expense or income. By doing this, investors can more accurately compare the numbers of the company across accounting periods.

2.3.4. Companies that report more non-GAAP measures

Bhattacharya et al. (2004) provide empirical evidence on trends in non-GAAP reporting. They find that firms that tend to use more non-GAAP measures to supplement their reports are service and high-tech industries, these are relatively young firms. These firms are significantly less profitable, more

(16)

16

liquid and have higher debt levels, Price-to-Earnings ratios and book-to market ratios than other firms in their own industries. Isidro and Marques (2013) find that manufacturing, materials and electronics and transportation and communication industries tend to report more non-GAAP earnings disclosures. 2.3.5. Relation non-GAAP measures and agency theory

Prior research by Entwistle et al. (2006) studies the misleading disclosure of non-GAAP earnings. Their study shows that non-GAAP disclosures can be used either to (1) communicate transparency or to

(2) deceive or mislead. By using a standard agency model, they expect that companies possessing bad

news would be more likely to use disclosures designed to mislead. Prior research also uses the two competing disclosure motives for reporting non-GAAP earnings: to provide value relevant information or to manage investors’ perceptions.

With regards to value relevance, Kolev et al. (2008) examined the effects of intensified scrutiny over non-GAAP reporting on the quality of non-GAAP earnings exclusions. They provide evidence that non-GAAP disclosures tend to be more value-relevant than GAAP earnings and also fulfill a valuation role. This suggest that income statement items that are less value relevant are excluded from non-GAAP measures. According to the findings of Bhattacharya et al. (2003) non-GAAP earnings are more informative to investors and a more permanent measure of firm profitability to analysts than GAAP operating earnings. Their overall findings suggest that market participants believe that non-GAAP earnings are more representative of core earnings than GAAP operating income.

On the other hand, there is also evidence by Kolev et al. (2008) that some managers use non-GAAP earning opportunistically. By doing this, managers’ misuse the non-non-GAAP numbers to present their company healthier to the outside world than the actual true value. In this way, managers present the numbers more positively than it actually is. Thus, by using non-GAAP earnings opportunistically, managers mislead the investors. For example, Bowen et al. (2005) examine the determinants of emphasis placed on non-GAAP earnings within earnings press releases and they find that firms emphasize the metric that represent firm performance that is more favorable. Another example is from the study of Marques (2006), who examined the effect of two regulatory interventions related to disclosure of non-GAAP measures. She discusses two possible explanations of the disclosure of non-non-GAAP. One possible explanation is the view that some managers mislead investors by excluding the effects of some negative events from income that are likely to recur in the future and thus the non-GAAP earnings that are excluded are not really non-recurring. Other evidence from the study of Bhattacharya et al. (2004) is that

(17)

17

some managers also appear to use non-GAAP earnings measures to meet earnings benchmarks, thus to meet or beat the expectations of analysts’.

2.3.6. Viewpoint manager regarding non-GAAP measures

From the managers’ point of view, Bhattacharya et al. (2003) stated that they provide non-GAAP disclosures with the reason to provide a clearer picture of the core earnings of the company that they believe will continue in the future. Proponents argue that removing non-cash and non-recurring items, such as one-time charges or unusual items from earnings and thus from GAAP earnings, will reduce noise in the earnings measure. Another view is from Entwistle et al. (2006), who stated that from the manager point of view non-GAAP earnings enables investors to make better decisions because these disclosures contain information that is useful and thus improves GAAP measures. According to Black et al. (2017) managers also argue that non-GAAP earnings are customized earnings that better portray core performance than GAAP-based earnings and thus are beneficial to investors. They find evidence that firms’ quality of non-GAAP reporting increases as non-GAAP calculations become less comparable to sector peers. Thus, firms that deviate from the norm of the sector have non-GAAP metrics that better inform investors about their core performance. Bowen et al. (2005) find that managers use emphasis in the earnings press release as a disclosure tool and this emphasis influences current and potential investors. Managers are intentionally emphasizing non-GAAP metrics in earnings press releases, so this is not random. However, the perceived net benefit of emphasizing on these metrics declined in 2002, after attention of the U.S. Securities and Exchange Commission and other negative publicity.

2.3.7. Viewpoint investor regarding non-GAAP measures

From the investor’s point of view, Bradshaw et al. (2016) examined how measurement error in earnings expectations affects prior evidence regarding investors’ preferences for GAAP versus non-GAAP earnings. They provide strong evidence that investors find non-non-GAAP earnings to be more informative than GAAP earnings, so investors respond more to non-GAAP than to GAAP earnings. This evidence is inconsistent with prior studies that expressed concerns about measurement error and the validity of results. These results in prior literature indicated that investors prefer non-GAAP earnings but this result might not be valid due to these concern. While there was a lack of separate GAAP and non-GAAP forecasts in prior research, they used analysts’ non-GAAP forecasts as the expectation for

(18)

18

both GAAP and non-GAAP earnings. This has hindered researchers’ ability to resolve these concerns. Within the research of Bradshaw et al. (2016) this concern is removed because GAAP earnings forecasts are now provided.

2.3.8. Need for regulation around non-GAAP disclosures

Before regulation of non-GAAP information was introduced by the U.S. Securities and Exchange Commission, questions were raised about the quality of financial reporting because of weak corporate governance, aggressive accounting practices and earnings restatements. To restore public confidence in the stock markets and to improve the accuracy and reliability of corporate disclosures the U.S. Congress acted out the Sarbanes-Oxley Act. Because of the concerns that non-GAAP financial measures were subjected to bias or misinterpretation, section 401(b) of the Sarbanes-Oxley Act instructed the U.S. Securities and Exchange Commission to introduce regulation of non-GAAP financial information. This was done with the effort to ensure that such reporting is not done in a misleading manner and is useful (Entwistle et al., 2006). In March 2003 the U.S. Securities and Exchange Commission implemented the new non-GAAP earnings disclosures rules which was named Regulation G.

2.3.9. Regulation G

The new rules of regulation G were that if a company discloses non-GAAP earnings it must also (1) disclose the most directly comparable GAAP earnings number, (2) disclose a reconciliation of the non-GAAP number to the GAAP number and, (3) supply, within five days an explanation of why management believes the non-GAAP number is useful to investors (Heflin and Hsu, 2008). Because GAAP information must be presented with the same prominence as non-GAAP information, the U.S. Securities and Exchange Commission believes emphasis is an important disclosure tool (Bowen et al., 2005).

Entwistle et al. (2006) examined whether firms’ reporting of non-GAAP earnings changed with the introduction of non-GAAP regulation by examining Standard & Poor’s 500 (S&P 500) companies between 2001 and 2003. S&P 500 measures the value of stocks of the 500 largest corporations by market capitalization listed on the New York Stock Exchange. They find that the frequency of reporting non-GAAP disclosures declined following regulation. They also find that non-non-GAAP disclosures are used in a less biased, prominent and potentially misleading manner after regulation.

(19)

19

However, Heflin and Hsu (2008), who examined the impact of the U.S. Securities and Exchange Commission’s regulation of non-GAAP disclosures, and Jennings and Marques (2011), who examined the joint effects of corporate governance and regulation by the U.S. Securities and Exchange Commission on the disclosure of manager-adjusted non-GAAP earnings numbers in the United States, find slightly other results. They find that the proportion of companies releasing non-GAAP financial measures has decreased only moderately regarding the frequency of special- and other-item exclusions and the probability disclosed earnings meet or beat forecast. The same literature also finds a decline in the magnitude of the exclusions and the association between returns and forecasts errors. Regulations have reduced the opportunistic use of non-GAAP earnings disclosures, investors are not misled anymore which is consistent with the intentions of the U.S. Securities and Exchange Commission.

Kolev et al. (2008) also find consistency with the U.S. Securities and Exchange Commission’s objective to improve the quality of non-GAAP earnings figures. They find that exclusions are of higher quality after the intervention of the U.S. Securities and Exchange Commission into non-GAAP reporting. They also find that firms that have stopped releasing non-GAAP earnings numbers following U.S. Securities and Exchange Commission intervention had lower quality exclusions in the period before the intervention. Another positive effect, according to Marques (2006), after regulation G is that investors had a positive market reaction to the disclosure of non-GAAP earnings. Firms claim that they disclose non-GAAP numbers to give investors a clearer picture of their results by removing unusual or transitory items, which are temporarily items. If investors agree to these adjustments, he expect and the results support this, that investors will consider non-GAAP financial measures reliable and they will react positive to the market.

In contrast to previous literature, Baumker et al. (2014) investigate the disclosure of non-GAAP earnings following regulation G, by focusing on transitory gains. They provide some evidence that there continues to be an opportunistic component of non-GAAP reporting following Regulation G. Gains are less likely to be taken out of earnings when there are no concurrent transitory losses. This provides some evidence that managers are more likely to limit disclosures when it reduces non-GAAP earnings relative to when it increases non-GAAP earnings.

After an initial decline in non-GAAP reporting following Regulation G, numerous studies like Bradshaw et al. (2016), Black et al. (2017) and Leung and Veenman (2016) report that the frequency of non-GAAP disclosures recovered but also increased dramatically in recent years and is currently at a record high level. An explanation is because of the positive results of non-GAAP literature which

(20)

20

examines managers’ motives for disclosing non-GAAP information and investors’ preference for GAAP versus non-GAAP earnings.

2.3.10. Need for latest regulation: new Compliance and Disclosure Interpretations

After regulation G, the U.S. Securities and Exchange Commission took a "hands-off approach" toward the use of non-GAAP financial measures by public companies listed in the United States. Until the U.S. Securities and Exchange Commission implemented a task force in July 2013. This was done with the intention to scrutinize non-GAAP disclosures for possible enforcement actions and the U.S. Securities and Exchange Commission issued the new Compliance and Disclosure Interpretation on May 17, 2016. So probably there was a need to intervene again and thus, knowing if this intervention leads to the desired outcomes. The new Compliance and Disclosure Interpretations are issued with the intention to clarify guidance on the use of non-GAAP financial measures. Because the release date is quite recent, up until this point in time, the effect on reporting non-GAAP disclosures has not been researched much. This research will hopefully provide some useful insights if this new and revised Compliance and Disclosure Interpretations are effective.

2.3.11. New Compliance and Disclosure Interpretations

The new Compliance and Disclosure Interpretations are a continued effort of the U.S. Securities and Exchange Commission to enhance the quality of non-GAAP disclosures, following its increasing concern about firms’ use of non-GAAP financial measures (Leung and Veenman, 2016). So the new Compliance and Disclosure Interpretations are considered effective when the quality of non-GAAP disclosures is improved. It is interesting but most importantly relevant to know if the new Compliance and Disclosure Interpretations also will have an initial decline in reporting non-GAAP and to what extent or if there still will be an opportunistic component.

In prior literature, articles were published about the interventions of the U.S. Securities and Exchange Commission and if this had an effect on reporting non-GAAP disclosures, like for example reporting less disclosures.But in this literature authors mainly talked about the interventions of the U.S. Securities and Exchange Commission that was adopted before the new Compliance and Disclosure Interpretations, which was regulation G and item 10(e) of regulation S-K as part of its directive under the Sarbanes-Oxley Act of 2002 to regulate the use of non-GAAP financial measures. So regulation G

(21)

21

has been researched extensively. For example, what the changes were before and after this regulation with regards to reporting non-GAAP disclosures (Heflin and Hsu, 2008; Zhang and Zheng, 2011; Kolev et al., 2008).

The contribution of this research will be to extend existing knowledge and this extension will be about a new regulation in a new timeframe. It is relevant for the U.S. Securities and Exchange Commission but also for the companies itself to know, in addition to prior literature, if regulation really changes reporting and what this change will be, but also to what extent.

To provide more insight in what the new Compliance and Disclosure Interpretations hold there will be an explanation what companies should do and what companies should avoid regarding the new landscape of the Compliance and Disclosure Interpretations in the appendix as stated by the U.S. Securities and Exchange Commission (2016), which is described in exhibit A. To name an example, the new Compliance and Disclosure Interpretation states that non-GAAP metrics that are inconsistently calculated across quarters can be considered misleading (U.S. Securities and Exchange Commission (SEC), 2016).

2.3.12. Examples earnings press release reports

In the appendix, two different examples of earnings press release reports are shown. The first one is exhibit B, this is an earnings press release report from before the U.S. Securities and Exchange Commission new Compliance and Disclosure Interpretations (May 17, 2016). This earnings press release is from the third quarter of 2013 from Facebook. This earnings press release shows that the focus is on the non-GAAP numbers because these numbers are reported in the first ten sentences of the press release.

The second earnings press release report is again from Facebook, exhibit C. This is an earnings press release report that is in line with how an earnings press release report regarding non-GAAP should look. This is an earnings press release report of the third quarter of 2016, so this is after the U.S. Securities and Exchange Commission new Compliance and Disclosure Interpretations (May 17, 2016). The focus is not on non-GAAP reporting, because the non-GAAP numbers are reported after ten sentences in the earnings press release report. In these earnings press release reports from Facebook it is clear that there is a shift in emphasis on non-GAAP measures from the first ten sentences in 2013 to the end of the report in 2016.

(22)

22

3. Hypotheses

When firms disclose non-GAAP earnings, they must consider different aspects of non-GAAP reporting. The first one is whether firms report a non-GAAP measure, so the frequency of non-GAAP disclosures. If firms decide to report a non-GAAP measure, this research will provide insights about whether this has been affected by non-GAAP regulation. The second one is whether to exclude a (non-)GAAP item from the GAAP disclosures, which then result in the non-GAAP disclosures. Within this research, the magnitude of exclusions means the difference between GAAP earnings per share and non-GAAP earnings per share but also the number of non-non-GAAP measures. So this research will also provide insight if the magnitude of the exclusions have been affected by non-GAAP regulation.

3.1. Frequency of non-GAAP disclosures

Regulation doesn’t prohibit the use of non-GAAP disclosures, it also does not state that the reduction of non-GAAP disclosures would be in the best interest of shareholders. Non-GAAP disclosures can be useful and the U.S. Securities and Exchange Commission supports the use of these numbers as long if it is not potentially misleading. The focus is therefore to ensure that investors are not misled by non-GAAP disclosures. When non-GAAP disclosures are used to inform investors, a decline in reporting non-GAAP numbers would not be expected because this should not affect their decision to disclose non-GAAP numbers. On the other hand, when it is used to mislead or manage the perception of the investors, then the frequency of reporting non-GAAP would decline (Entwistle et al., 2006). So after regulation, and in this case the new Compliance and Disclosure Interpretations, I expect that less firms will disclose non-GAAP earnings. Thus, the next alternative hypothesis is:

H1: The frequency of companies that reported non-GAAP declined after the new Compliance and Disclosure Interpretations.

The research of Entwistle et al. (2006) was done between 2001 and 2003, this was the first intervention of the U.S. Securities and Exchange Commission to regulate the use of alternate earnings measures which was the Sarbanes-Oxley Act. They find that the proportion of firms reporting non-GAAP earnings declined from 77 to 52 percent. By 2003, the use of these alternate earnings is used in a less biased manner, the proportion of firms that used non-GAAP to increase reported income has

(23)

23

become smaller but also the magnitudes are reduced from 2001 till 2003. Firms also present non-GAAP earnings in press releases in a potentially much less misleading and prominent manner. In other past literature Kolev et al. (2008) finds that firms that stopped releasing non-GAAP earnings numbers after the U.S. Securities and Exchange Commission intervened, had lower quality exclusions in the period before the intervention.

The second intervention of the U.S. Securities and Exchange Commission was regulation G. Heflin and Hsu (2008) find that regulation G reduces the frequency of non-GAAP earnings disclosures, the association between returns and forecast errors and the magnitude of amounts excluded from GAAP earnings.

In contrast to previous literature, other literature discusses that it is possible that the interventions of the U.S. Securities and Exchange Commission gave non-GAAP disclosures credibility and this in return, would lead to an increase in the frequency of disclosure of these measures. Another alternative is that neither a decline nor an increase are present or they can be both present but are offsetting each other. If this is the case, then there is no change in the frequency of disclosing non-GAAP financial measures (Marques, 2006).

However, according to Black et al. (2017) the frequency of non-GAAP reporting has increased with 35% in recent years, over the last decade the propensity for managers to disclose non-GAAP earnings has thus increased significantly. Because of this increasing trend in the past years, the expectation that the proportion of companies that report non-GAAP will decline after the new Compliance and Disclosure Interpretations will probably also be an initial reduction, just like after regulation G.

With regards to the new Compliance and Disclosure Interpretations a study was done by MarketWatch (2017). This research shows that the new Compliance and Disclosure Interpretations didn’t have much effect. They stated that reporting non-GAAP earnings numbers in the press releases stayed the same or in some cases, even increased. Shortly after the new Compliance and Disclosure Interpretations was known (the first and second quarter), companies remove at least one metric from their earnings report but after a while (in the third quarter) they added them again. So the effects of the new Compliance and Disclosure Interpretations according to MarketWatch were short lived.

3.2. Exclusions of non-GAAP disclosures

The magnitude of the difference between GAAP earnings per share and non-GAAP earnings per share is likely to be greater for firms that use non-GAAP disclosures to manage the perception of the

(24)

24

investor in comparison to firms that use non-GAAP disclosures to provide supplemental information. Firms may also be more likely to emphasize non-GAAP earnings if the differences between GAAP and non-GAAP earnings are larger (Bhattacharya et al., 2004; Bowen et al. 2005; Entwistle et al., 2006). The expectation is that after the new Compliance and Disclosure Interpretations the opportunity to manage the perception of the investor though reporting non-GAAP earnings that are higher than GAAP earnings would decline. The same goes for the number of non-GAAP financial measures. Thus, the next alternative hypothesis is:

H2: The magnitude of exclusions under non-GAAP declined after the new Compliance and Disclosure Interpretations.

Jennings and Marques (2011) conducted a comparable study with regards to regulation G. If the concerns of the U.S. Securities and Exchange Commission were justified, they expected that exclusions were partially persistent for firms with weaker corporate governance. They repeated these comparisons for 2003, the first year in which regulation G was effective. The persistence of excluded items should decline in 2003 relative to prior regulation for firms with weaker corporate governance. They find no evidence that firms with strong corporate governance that reported non-GAAP adjustments mislead investors before or after the U.S. Securities and Exchange Commission intervened. On the other hand, investors of firms with weaker corporate governance were misled by non-GAAP adjustments to GAAP earnings before the U.S. Securities and Exchange Commission intervened. Non-GAAP adjustments were both persistent and positively associated with future returns, similar results were find for adjustments that are ex ante recurring and that just meet or beat analysts’ forecast. After the U.S. Securities and Exchange Commission intervened they find no evidence that investors continue to be misled by non-GAAP adjustments. This included adjustments that are ex ante recurring or that just meet or beat analysts’ forecast.

Kolev et al. (2008) also find that with regards to non-GAAP reporting exclusions are of higher quality, thus more transitory following the U.S. Securities and Exchange Commission intervention. However, they find evidence that the quality of special items has decreased after the intervention. This suggest that managers shifted recurring expenses to special items, which improves non-GAAP earnings. Within the category differences with regards to exclusions, Black et al. (2017) examinedthe extent to which a firm’s non-GAAP calculation differs from peer firms. They constructed a new measure of

(25)

25

earnings comparability based on firm-specific earnings exclusions. They find that firms deviate from the sector norm to provide earnings metrics that inform stakeholders better about future core operations.

4. Data and research design

4.1. Initial data collection

I initially obtained a sample of 337 companies. These companies are manually taken from the Standard & Poor’s 500. More specifically, I obtained the press releases from these companies. These are called 8-K filings, or 8-K earnings press releases. I got these 8-K filings from the EDGAR Company filing search on the website of the U.S. Securities and Exchange Commission. EDGAR collects electronic documents to help investors get information. In the process of collecting my data I removed all the financial companies from the Standard & Poor’s 500, so my sample are all non-financial companies.

Because I wanted to research the quarterly earnings report before and after the intervention of

the U.S. Securities and Exchange Commission with regards to the new Compliance and Disclosure Interpretations, I have to look at the period before May 17, 2016 and after this date. So this will be a comparison of the first quarter of 2016 with the third quarter of 2016. So I will only use two quarterly earnings reports of 2016 of the same company. I intentionally left a quarter in between, because this is when the intervention happened but also to give the companies some time to adjust their quarterly earnings reports. In my opinion this will be the best representation of the effect of the new Compliance and Disclosure Interpretations on reporting non-GAAP disclosures.

Only press releases from companies which quarter ended in March and September are taken into the sample so the release date of the press releases were mostly in April or May if the quarter ended in March. The release date of the press releases were mostly in October or November if the quarter ended in September. The links to the 8-K earnings press releases before and after the intervention were put in the excel file that I created.

My excel file also contained the tabs Central Index Key, Committee on Uniform Securities Identification Procedures number, Compustat ID, name, industry, date of the data, fiscal year, fiscal quarter, data quarter, release date and period. The Central Index Key is used on the U.S. Securities and Exchange Commission computer systems to identify corporations and individual people who have filed

disclosure with the U.S. Securities and Exchange Commission. With the Central Index Key people can

(26)

26

Procedures number identify a company or issuer and the type of financial instrument.Compustat ID is

a unique six-digit number key assigned to each company in the Capital IQ Compustat database. The tab period received a 0 or a 1 for each company. 0 stand for the first quarter of 2016, so the period before the intervention of May 17, 2016. Period 1 stands for the third quarter of 2016, this is the period after the intervention of May 17, 2016.

4.2. Non-GAAP data collection

With regards to collecting non-GAAP data I set up different columns in my Excel file with the information I would like to collect from the 8-K earnings press releases about specific non-GAAP data. The first and main question and thus first column of my Excel file is, if there are non-GAAP earnings in the 8-K earnings press releases. Non-GAAP earnings means adjusted earnings per share. I only searched for non-GAAP and GAAP earnings for the specific quarter, in the 8-K earnings press releases this is the number beneath the "Three Months Ended". The second column is, if there are other non-GAAP measures present in the 8-K earnings press releases. A couple of examples are adjusted net income, adjusted Earnings before Interest, Taxes, Depreciation and Amortization (EBITDA) and free cash flow. I wrote these measures in the next tab. But more important is the number of other measures that were presented. I also wrote down in two different columns the exact number of the non-GAAP earnings per share and the GAAP earnings per share. I wanted to know if the non-GAAP earnings per share, (adjusted) funds for operations per share and distributable cash flow before certain items per share are standing out in the 8-K earnings press releases. If the adjusted earnings per share were shown in the headline, bold, summed up, under the heading highlights or in the first table without any text before it, I considered the per share amounts standing out. I also wanted to know if the order in presenting non-GAAP earnings per share, (adjusted) funds for operations per share and distributable cash flow before certain items per share has changed after the intervention, so if the non-GAAP per share amount or if the GAAP earnings per share was presented first. The last two columns of my excel file consist of the presentation of non-GAAP measures in the financial statements, particularly if the non-GAAP measure is presented in the beginning or middle or in the last part of the financial statements. To answer most of these questions I used a 1 and a 0. 1 means yes and 0 means no. To answer the columns how many other non-GAAP measure are presented, non-GAAP earnings number per share and GAAP earnings number per share I used numbers. To answer the column which other non-GAAP measures are presented I used words.

(27)

27

The next step in collecting my data was searching for non-GAAP measures in the 8-K earnings press releases. To minimize the burden of manual data collection, I used a text search of non-GAAP keywords. The keywords to search through the 8-K earnings press releases are the next words: "non", "non-GAAP", "non GAAP", "non-U.S. GAAP", "non U.S. GAAP", "adjusted", "pro-forma", "pro forma", "EPS", "earnings per share", "adjusted EPS", "adjusted earnings", "adjusted earnings per share", "diluted" and "adjusted diluted". A lot of the columns in the excel file focus on non-GAAP earnings disclosures but I also wanted to focus on other non-GAAP measures, for example "adjusted EBITDA". The keyword "adjusted" helped me a lot to find the other non-GAAP measures.

I had some setbacks while collecting my data. One of them was that the links to the 8-K filings sometimes didn’t work. Every now and then I ended up with an empty website or the website wouldn’t load at all. I went to the EDGAR Company filing search on the website of the U.S. Securities and Exchange Commission and typed in the Central Index Key of the companies from which I couldn’t collect data. Through this way I found the 8-K filings with text and the financial statements. I found these either on another place on the website or in the sheets of the company.

Not all companies report non-GAAP earnings per share or another non-GAAP measures. From the 337 companies, 22 companies didn’t report any non-GAAP measures before and after the new Compliance and Disclosure Interpretations. These companies didn’t report non-GAAP earnings nor any other non-GAAP measures. These companies have no added value for my dataset because there was no change before and after the new Compliance and Disclosure Interpretations, neither in time and non-GAAP measures. I only researched companies that report non-non-GAAP measures because with regards to my hypotheses, I am only interested in the change and magnitude of non-GAAP reporting and this only regards companies that report non-GAAP measures. Thus, my dataset only consist of companies that report non-GAAP earnings and/or non-GAAP measures.

I removed the companies that didn’t report any non-GAAP measures from my sample, so eventually 315 companies reported non-GAAP earnings per share and/or another non-GAAP measure. In Compustat, there was missing data from one company, so I did not include this company in my dataset. Because each company has two observations, in the end my dataset consist of 628 observations in total. 4.3. Libby boxes

(28)

28

To give a better understanding of my research I will add some Libby boxes. The first Libby box is about my first hypothesis which is: the frequency of companies that reported non-GAAP declined after the new Compliance and Disclosure Interpretations.

-FIGURE 1- New Compliance and

Disclosure Interpretations Theory

Frequency companies that report non-GAAP measures

Opera tional iz ing Indepe ndent V ariabl e V ariabl e

One quarter before and after the new Compliance and Disclosure Interpretations Testing association Opera tional iz ing Depe ndent V ariab le - Non-GAAP earnings - Non-GAAP other measures C ontr ols

- Book-to-market assets ratio - Earnings volatility

- Log firm age

- Log total assets (firm size) - Loss

(29)

29 4.3.2. Libby box second hypothesis (H2)

The second Libby box is about my second hypothesis which is: the magnitude of exclusions under non-GAAP declined after the new Compliance and Disclosure Interpretations.

-FIGURE 2- New Compliance and

Disclosure Interpretations

Magnitude of exclusions under non-GAAP Opera tional iz ing Indepe ndent V ariabl e

One quarter before and after the new Compliance and

Disclosure Interpretations Testing association

Opera tional iz ing Depe ndent V ariab le

- How many other non-GAAP measures - Difference non-GAAP

and GAAP earnings per share

-

C

ontr

ols

- Book-to-market assets ratio - Earnings volatility

- Log firm age

- Log total assets (firm size) - Loss

- Sales growth Theory

(30)

30

5. Results

5.1. Means

5.1.1. Means dummy variables

-TABLE 1-

Variable (0/1) _{Obs Mean} _{Dev. Min Max}Std.

Non-GAAP earnings 628.000 0.834 0.372 0.000 1.000

Non-GAAP other 628.000 0.978 0.148 0.000 1.000

Stand out 628.000 0.572 0.495 0.000 1.000

First EPS then EPS adjusted (first EPS then

adjusted) 545.000 0.791 0.407 0.000 1.000

Non-GAAP measures in between financial statement

(FS in between) 623.000 0.446 0.497 0.000 1.000

Non-GAAP measures in last part financial statement

(FS last part) 623.000 0.692 0.462 0.000 1.000

Table 1 presents the number of observations, mean, standard deviation, minimum and maximum for each of the dummy variable in the table.

From the sample of 314 companies, so all 628 observations taken together, before and after the new Compliance and Disclosure Interpretations 83.40% companies report non-GAAP adjusted earnings per share. 97.80% companies report other non-GAAP measures. From table 1, I will only use the variables non-GAAP earnings and other non-GAAP measures to test my first hypothesis which is the frequency of companies that reported non-GAAP declined after the new Compliance and Disclosure Interpretations. The other variables in table 1 are not used to test the hypotheses. However, the other variables will be tested and the results will be shown.

Approximately half of the companies, 57.20% have in their 8-K earnings press releases adjusted earnings per share that were shown in the headline, bold, summed up, under the heading highlights or in the first table without any text before it. From the companies that reported adjusted earnings per share 79.10% first report diluted earnings per share and then adjusted earnings per share, this is consistent with the new Compliance and Disclosure Interpretations. 44.60% of the companies that reported non-GAAP measures, reported these measures in between the financial statements and 69.20% reported these measures in the last part of the financial statements. This is the correct way to present the non-GAAP measures because these measures are presented separate at the end of the financial statements. This is clearer and will lead to fewer misinterpretations.

(31)

31 5.1.2. Means other variables

-TABLE 2-

Variable _Obs _Mean _{Std. Dev.} _Min _Max

How many non-GAAP other measures 628.000 3.912 2.886 0.000 16.000

Non-GAAP earnings per share 523.000 1.096 0.943 -0.660 5.190

GAAP earnings per share 627.000 0.855 1.018 -2.000 5.100

Table 2 presents the number of observations, mean, standard deviation, minimum and maximum for each of the variable in the table.

The average of how many other non-GAAP measures were presented in the 8-K earnings press releases is 3.91. So this would be on average around 4.00 other non-GAAP measures. The average of non-GAAP earnings per share and GAAP earnings per share is respectively 1.09 and 0.85. The average of non-GAAP earnings per share is higher than GAAP earnings per share. This supports the study of Baumker et al. (2014) that reported that non-GAAP earnings generally exceed GAAP earnings. I will use the variables how many other non-GAAP measures and the difference between non-GAAP and GAAP earnings per share to test my second hypothesis which is the magnitude of exclusions under non-GAAP declined after the new Compliance and Disclosure Interpretations.

5.1.3. Means dummy variables by period

-TABLE 3- Period Non-GAAP earnings Non-GAAP other Stand out

First EPS then EPS adjusted FS in between FS last part 0 0.834 0.978 0.580 0.588 0.452 0.683 1 0.834 0.978 0.564 0.996 0.440 0.701 Difference 0.000 0.000 -0.016 0.408 -0.012 0.018

Table 3 presents the means of non-GAAP earnings (yes/no), other non-GAAP measures (yes/no), if the non-GAAP measures stand out (yes/no), if first earnings per share are shown and then the adjusted earnings per share (yes/no), if non-GAAP earnings are presented in between the financial statements or if the non-GAAP earnings are presented in the last part in the financial statements (yes/no). These means are divided between two periods, before the new Compliance and Disclosure Interpretations (period 0) and after the new Compliance and Disclosure Interpretations (period 1).

The average of non-GAAP earnings, 83.40% stays the same before and after the new Compliance and Disclosure Interpretations. So the companies that report non-GAAP earnings per share stays the same before and after the new Compliance and Disclosure Interpretations. This also applies to the other

Are the U.S. Securities and Exchange Commission new Compliance and Disclosure Interpretations (May 17, 2017) on reporting non-GAAP disclosures effective (less bias and misinterpretation)?

Thesis

Contents