The Basel Accords and their Impact on the Informativeness of Non-GAAP Measures Reported by the European Banking Industry

(1)

University of Amsterdam

The Basel Accords and their Impact on the

Informativeness of Non-GAAP Measures

Reported by the European Banking

Industry

Name: Ahmed Boukachar Student number: 10753443

Thesis supervisor: drs. J.F. Jullens Date: 20 August 2018

Word count: 16,518

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

(2)

2

Statement of Originality

This document is written by the student, Ahmed Boukachar, 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

The question whether the Basel Accords have an impact on the way European banks report non-GAAP measures is answered on the basis of an empirical research. Only a few researches have focused on Europe, despite the increase in both the complexity of accounting and the use of non-GAAP reporting. Furthermore, the majority of these researches exclude banks and other financial institutions from their sample to focus on a more homogenous set of firms. This research provides mixed but strong evidence supporting the notion that the Basel Accords incentivize banks to use non-GAAP reporting as a way to manage earnings. Namely, banks are making more adjustments to compute the non-GAAP earnings, while the excluded components are partially recurring items, when they have a relatively high CET1 or Leverage Ratio. Therefore, a potential negative effect of the Basel Accords on the way non-GAAP measures are reported by banks is observed. However, this is not the case, when differentiating between banks having a relatively low or high Liquidity Coverage Ratio. Therefore, this research can be informative for the BCBS, CESR and the ESMA. As a result, investors can make decisions based on information that does not accurately reflect the financial reality of the financial institution. To prevent the latter, regulators can decide to introduce the regulation of non-GAAP reporting. Furthermore, the Basel Accords have such a potential effect on the profitability of banks, especially in terms of Return on Equity, that they are incentivized to manage their earnings. The Basel Committee on Banking Supervision should take this into account, when adjusting the minimum required levels of their standards in the future. This research is not only comparing the informativeness of these non-GAAP measures reported by banks and non-financial firms, but also contributes to the literature by providing insights on the relation between the Basel Accords and the use of non-GAAP reporting.

(3)

3

Contents

1 Introduction 4

2 Literature Review 6

2.1 Bank Regulations 6

2.1.1 The First Basel Accord 6

2.1.2 The Three Pillars of Basel II 7

2.1.3 Basel III: Bank Supervision after the Financial Crisis 9

2.2 The Effects of the Basel Accords 10

2.3 The Role of Non-GAAP Measures 11

2.3.1 An Alternative Way of Reporting 11

2.3.2 A Tool for Managers 13

2.3.3 The Informativeness for Investors 16

2.4 Non-GAAP Regulations 18

2.4.1 The Intervention by Regulators 18

2.4.2 The Impact of the Regulations 19

3 Hypotheses Development 22

4 Research Design 25

4.1 Regression Models and Procedures 25

4.2 Sample Selection 27

5 Empirical Results and Discussion 29

5.1 Descriptive Statistics 29

5.2 Relative and Incremental Informativeness 30

5.3 The Impact of Having a Relatively High Standard Ratio 32

5.4 Robustness Tests 34

6 Conclusion 36

(4)

4

1 Introduction

The Royal Bank of Scotland (RBS) reported an Adjusted Annual Operating Profit of £4,818 million for the year ended on 31 December 2017. However, RBS (2018) excluded restructuring, litigation and conduct costs to report this measure. This way of reporting is also known as non-GAAP reporting, since it is not in compliance with the Generally Accepted Accounting Principles (GAAP). Under GAAP, RBS (2018) reported an Annual Operating Profit of only £2,239 million, which is almost 54% lower than its non-GAAP counterpart. This observation is in line with the findings of Bradshaw and Sloan (2002), who argue that non-GAAP reporting usually leads to a higher performance measure. Consequently, the question that may arise is what motives firms or their managers have to use non-GAAP reporting. Andersson and Hellman argue that non-GAAP reporting is a reaction on the increasing complexity of accounting (2007), which can be found in Europe too after the adoption of the International Financial Reporting Standards (IFRS) in 2005. Therefore, managers argue that they report non-GAAP measures to provide an understandable view of the financial position of the firm. RBS (2018), for example, states that these measures exclude certain items, which the management believe are not representative of the underlying performance of the firm. The result is a situation where a substantial part of the information is based on non-GAAP reporting (Marques, 2010).

On the other side, Isidro and Marques (2013) found that managers use non-GAAP reporting to give a too opportunistic view of the financial position of the firm. This opportunistic use of these exclusions increases, if managers are less able to apply other forms of earnings managers to meet or beat analysts’ expectations (Doyle, Jennings and Soliman, 2013). Therefore, it can be concluded that the use of non-GAAP measures is a substitute for other forms of earnings management. In addition, Doyle et al. (2013) argue that managers use non-GAAP reporting to meet or beat analysts’ forecast. However, Black, Christensen, Kiosse and Steffen (2017) argue that the quality of these non-GAAP measures has increased after the intervention by the Securities and Exchange Commission (SEC). In the same vein, Brown and Sivakumar (2003), Entwistle, Feltham and Mbagwu (2010) and Venter, Emanual and Cahan (2014) provide evidence showing that non-GAAP measures are both relatively and incrementally informative for investors. Furthermore, there are only a few researches focusing on the use of non-GAAP reporting by European firms, despite the lack of non-GAAP regulation in Europe (Guillamon-Saorin, Isidro & Marques, 2017). However, Isidro and Marques (2013) argue that European firms adopt more consistent non-GAAP disclosures than

(5)

5

their US counterparts. In addition, they rarely exclude recurring earnings components (Choi, Lin, Walker and Young, 2007).

Furthermore, among others, the majority of these researches exclude financial institutions such as banks from their sample. One of the reasons is that these firms are subject to specific regulations (Marques, 2006). In other words, there is a lack of literature about the informativeness of non-GAAP measures reported by financial institutions. However, there is a reason to expect that banks use non-GAAP reporting in a different way. Namely, the Basel Accords as published by the Basel Committee on Banking Supervision (BCBS) should have a negative impact on their profitability (Ozili, 2015). More specifically, McKinsey (2012) expects that the Return on Equity (ROE) in the four biggest retail-banking markets in Europe will decrease from 10.0 to 6.0 percent. European banks could use non-GAAP reporting as a form of earnings management to compensate for their decreased profitability as suggested by Isidro and Marques (2013) and Doyle et al. (2013), since there is no form of non-GAAP regulation in Europe (Guillamon-Saorin et al., 2017). However, as already indicated, the relation between the Basel Accords and non-GAAP reporting is not covered yet by the prior literature. With the aim to fulfill this recognized gap, the main research question is whether the Basel Accords affect the way non-GAAP measures are reported by banks. In summary, the objective of this research is not only to compare the informativeness of these non-GAAP measures reported by banks with the findings of prior literature, but also to gain insights on the relation between the Basel Accords and non-GAAP reporting.

This research provides mixed but strong evidence supporting the notion that the Basel Accords incentive banks to use non-GAAP reporting as a way to manage earnings. Namely, banks are making only more adjustments to compute the non-GAAP earnings, while the excluded components are partially recurring items, when they have a relatively high CET1 or Leverage Ratio. Therefore, a potential negative effect of the Basel Accords on the way non-GAAP measures are reported by banks is observed. However, this is not the case, when differentiating between banks having a relatively low or high Liquidity Coverage Ratio. These findings could be informative for European regulators, when deciding whether they should introduce non-GAAP regulations or not.

The next section will review the prior literature regarding the Basel Accords and non-GAAP reporting more in-depth. Based on this review, the hypotheses and expectations will be formulated and developed in the third section. Thereafter, the research design will be explained. The results will be discussed in the fifth section of this research. Finally, a conclusion will be reached in the sixth and final section.

(6)

6

2 Literature Review

This section will start with a short but detailed overview of the three Basel Accords and how they could negatively affect a bank’s profitability. Based on prior literature, it is reasonable to expect that banks will manage their earnings to undo these negative effects in their financial statements. Therefore, non-GAAP reporting and how it can be used as a tool to manage earnings will be explained. This section will also discuss whether these non-GAAP measures are informative for investors and how non-GAAP regulations affect this alternative way of reporting.

2.1 Bank Regulations 2.1.1 The First Basel Accord

As mentioned earlier, this research investigates the influence of the Basel Accords, which are the result of the activities of the Basel Committee on Banking Supervision (BCBS) on the reporting of non-GAAP measures. According to Jablecki (2009), it became clear that the international operations of banks lacked supervision after the fall of the Long Island’s Franklin National Bank and the Herstatt Bank in 1974. The Committee was established by the Group of Ten (Belgium, Canada, France, Germany, Italy, Japan, the Netherlands, Sweden, the United Kingdom and the United States of America), Luxembourg and Switzerland in response to this banking crisis with the purpose to provide its members with a forum for discussing cooperative approaches to the supervision of multinational banks (Zharing, 2005). Moreover, Barr and Miller (2006) argue that the BCBS evolved into a forum for harmonizing national supervision and capital standards for banks, which has led to more sophisticated guidelines for capital adequacy. Barr and Miller (2006) state that the first Basel Accord is one of the most successful international regulatory initiatives ever attempted. The Accord was adopted by the members of the Committee, leaving discretion for national authorities to set higher requirements, if desired (BCBS, 1988). However, Jablecki (2009) notices that the Basel I is adopted for the internationally operating banks as well.

Basel I is primarily directed towards assessing capital and credit risk, while capital adequacy also depends on the quality of a bank’s assets (BCBS, 1988). To keep the framework simple, the assets are categorized into only five groups according to their credit risk. Namely, 0% for, for example, cash and claims on central governments, 10% for claims on domestic public-sector entities, 20% for claims on multilateral development banks and banks incorporated in the OECD, 50% for loans fully secured by mortgage on residential property and

(7)

7

100% for claims on the private sector and on banks incorporated outside the OECD. Furthermore, the BCBS (1988) required that a bank holds a minimum percentage of their risk-weighted assets in capital. Hereby, the emphasis is placed on equity capital and disclosed reserves. However, the Committee argues that there are also other important components of a bank’s capital base. Capital is therefore defined in two tiers. Namely, Tier 1 which consist of equity capital and disclosed reserves and Tier 2 which consist of the other elements of capital (BCBS, 1988). Under the first Accord, the ratio of total capital, Tier 1 and Tier 2, to the risk-weighted assets should be at least 8.0%, while the ratio of Tier 1 Capital to the risk-risk-weighted assets should be at least 4.0% by the end of 1992 (Table 1, Panel A).

2.1.2 The Three Pillars of Basel II

According to the BCBS (1999), Basel I only established minimum levels of capital for internationally operating banks. In addition, the Committee also wanted to address other risks, such a market and operational risk, instead of only mainly focusing on credit risk. The BCBS argues that the first Basel Accord has helped to strengthen the stability of the banking industry. However, the financial world has changed significantly to the point where the calculation of the capital ratios, using Basel I, may not be a good indicator of the bank’s financial stability (BCBS, 1999). The proposals to reform the first Basel Accord led to the second Basel Accord, Basel II, which is based on three fundamental Pillars as described in Table 1, Panel C. Namely, the first Pillar addresses the minimum capital requirements, while the second and third Pillars are committed to the supervisory review process and market discipline (BCBS, 2004).

According to the first Pillar, the total capital ratio, Tier 1 and Tier 2, should be at least 8.0%, while Tier 2 capital is limited to 100% of the Tier 1 Capital (Table 1, Panel A) (BCBS, 2004). In addition, banks have the choice between three approaches to compute their capital requirements, while requirements for market and operational risk are also introduced. The first approach, the Standardized Approach, measures risk in a standardized manner with the use of external assessments. For example, claims on central banks with a Triple A rating are assigned a risk weight of 0%, while a weight of 150% is assigned when rating below B- is considered. Claims with a rating varying from A+ to B- are assigned different risk weights. The two alternative approaches provided by the Committee (2004) are known as the Internal Rating Based (IRB) Approaches and allow banks to use their own rating systems. The difference between the Foundation IRB Approach and the Advanced IRB Approach is that the

(8)

8

Table 1

Summary of the Basel Accords

Panel A: Minimum required levels

Basel I Basel II Basel III

Common Equity Tier 1 Ratio 2.0% 7.0%*

Tier 1 Ratio 4.0% 4.5% 8.5%*

Total Capital Ratio 8.0% 8.0% 10.5%*

Leverage Ratio 3.0%

Liquidity Coverage Ratio 100%

Net Stable Funding Ratio 100%

Panel B: Calculation methods

Standard Ratio Equation

Common Equity Tier 1 Ratio** Common Equity Tier 1 Capital / Risk-Weighted Assets Leverage Ratio Tier 1 Capital / Total Exposure

Liquidity Coverage Ratio High-Quality Liquid Assets / Total Net Cash Outflows over the next 30 Days Net Stable Funding Ratio Available Amount of Stable Funding / Required Amount

Panel C: The Basel Pillars

Pillar Addressed Topic

First Pillar Minimum Capital Requirements

Second Pillar Supervisory Review Process

Third Pillar Market Discipline

* Including the required conservation buffer of 2.5%.

** The Tier 1 Ratio and Total Capital Ratio are computed in the same way. However, the numerator of the Tier 1 Ratio consists of the Common Equity Capital and disclosed reserves, while the numerator of the Total Capital Ratio consists of both the Tier 1 and Tier 2 Capital.

assumptions under the first approach are provided by supervisors, while banks are allowed to determine the assumptions themselves under the Advanced IRB Approach. The two alternative approaches provided by the Committee (2004) are known as the Internal Rating Based (IRB) Approaches and allow banks to use their own rating systems. The difference between the Foundation IRB Approach and the Advanced IRB Approach is that the assumptions under the first approach are provided by supervisors, while banks are allowed to determine the assumptions themselves under the Advanced IRB Approach.

The supervisory review process, as addressed in the second Pillar, is intended to ensure that banks have adequate capital and to encourage them to develop better risk management techniques (BCBS, 2004). One of the principles of the review process is that supervisors should evaluate banks’ internal capital adequacy assessments on a regular basis. However, the

(9)

9

supervisors should not act as the bank management instead of reviewing the quality of the bank’s risk controls. On the other hand, the Committee (2004) argues that supervisors should intervene to prevent capital from falling below the minimum levels or when they are concerned that a bank is not meeting the requirements. To complement the first two Pillars, the BCBS (2004) introduced the third Pillar for disclosure requirements. Hereby, the aim is that more relevant information is available for market participants to evaluate, since banks have discretion in assessing their financial stability due to the reliance on internal methodologies.

2.1.3 Basel III: Bank Supervision after the Financial Crisis

The Basel Committee on Bank Supervision (2010) published the first version of Basel III in response to the market failures during the financial crisis of 2008. Furthermore, the aim is to improve the ability to absorb shocks arising from financial and economic crises. Building on the same three Pillars as introduced in Basel II, the Committee is raising the resilience of the banking industry by improving both the quality and quantity of the regulatory capital base of the framework. First of all, the Common Equity Tier 1 (CET1) ratio must be at least equal to 4.5%, while also a conservation buffer of 2.5% is required (BCBS, 2017). Therefore, banks are required to have a minimum CET1 Capital ratio of 7.0%. In addition, the minimum level of the Tier 1 Capital ratio increased to 6.0%, while Total Capital, Tier 1 and Tier 2, must be at least 8.0% of the risk-weighted assets (Table 1, Panel A). Hereby, the Tier 1 Capital consists of the 4.5% CET1 Capital plus an extra 1.5% of Additional Tier 1 (AT1) Capital (BCBS, 2010). One of the reasons the financial crisis became so severe was that the banks had built up on- and off-balance sheet leverage according to the Committee (2010), while there were not enough liquidity buffers. Due to the credit losses, banks were forced to reduce their leverage in a manner that reinforced the downward pressure on asset prices. Therefore, the BCBS (2010) decided to introduce a minimum Leverage Ratio (LR) requirement of 3.0% with the aim to constrain the leverage in the banking industry. The LR is computed by dividing the Tier 1 Capital by the total exposure of the bank, which follows the accounting measure of exposure (Table 1, Panel B). In addition, the Committee (2010) introduced the Liquidity Coverage Ratio (LCR) and the Net Stable Funding Ratio (NSFR). First, according to the BCBS (2013), the objective of the LCR is to promote the short-term resilience of a bank’s liquidity risk profile and requires to hold sufficient high-quality liquid assets to cover the total net cash outflows over the next 30 days (Table 1, Panel B). However, the LCR is introduced on the 1 January 2015 with a minimum of 60%. This minimum will rise to reach a level of 100% on 1 January 2019. In addition, the NSFR’s objective is to mitigate the risk of future funding stress (BCBS,

(10)

10

2014). As reported in Table 1, Panel B, this ratio can be computed by dividing the available amount of stable funding by the required amount and should be at least equal to 100% from 1 January 2018. The Basel Committee on Bank Supervision (2017) finalized Basel III at the end of last year. The effect of the Basel Accords will be discussed in the next paragraph.

2.2 The Effects of the Basel Accords

According to Ozili (2015), the literature predicts that the Basel Accords should have a negative impact on the performance of banks. McKinsey (2012), for example, expects that the Return on Equity (ROE) in the four biggest retail-banking markets in Europe will decrease from 10 to 6 percent. In addition, Santos (2001) concludes that higher capital requirements negatively affect bank development and credit expansion due to increased fixed and operation costs. It is also suggested that higher capital requirements may lead undercapitalized banks to engage in risk-taking activities (Calem and Bob, 1999). Consequently, this could lead to negative effects on the bank’s profitability, especially in terms of ROE. The same can be concluded concerning the liquidity rules of the Basel Accords according to McKinsey (2010), since they will lead to a shortfall in short- and long-term funding of respectively 1.3 and 2.3 trillion euros.

A few studies are also focusing on the banks’ stock prices to investigate the impact of the Basel Accord. Eysell and Arshadi (1990), for example, find that the stock prices of publicly traded banks decrease at the time of the announcement of the Basel Committee’s intention to impose a pre-determined minimum level of risk-adjusted capital. Furthermore, those banks with capital levels which are deficient relative to the mandated levels suffer the largest relative losses. The negative effect on the stock prices suggests that investors assume that the Basel Accords will have a negative impact on the profitability of banks. For example, Chiuri, Ferri and Majnoni (2002) find that the enforcement of bank capital requirements led to a reduction in bank loan supply, which is a major source of a bank’s interest income, in emerging economies. Therefore, they argue that the phasing in of higher capital requirements should be managed carefully in order to avoid a decrease in the credit supply. However, Allen, Chan, Milne and Thomas (2012) argue that the impact should be less than expected by the banking industry. Meanwhile, they state that there are more difficulties in ensuring a coordinated adoption of the Basel Accords, while they agree that the higher capital requirements could potentially limit the availability of credit and reduce economic activity.

On the other hand, Pasiouras, Tanna and Zopoundis (2009) argue that the second and third Pillars, respectively higher supervisory power and market discipline, have a positive impact on both the cost and profit efficiency. In the same vein, Ayadi, Naceur, Casu and Quinn

(11)

11

(2016) conclude that the Basel Accords do not have an effect on performance, when a bank already complies to all the regulations and standards, while higher standards could have a potential effect. However, Pasiouras et al. (2009) also found that the first Pillar of the Basel Accords, capital requirements, have a positive effect on cost efficiency, but a negative impact on the profit efficiency of banks. According to Pasiouras et al. (2009), the positive effect on cost efficiency could be explained by the decrease in the probability of financial distress due to the higher capital requirements, reducing the need for costly risk management activities. A change in the portfolio towards less risky assets leads to lower returns, which can be a possible explanation for the negative impact on profit efficiency. In line with these findings, Bitar, Pukthuanthong and Walker (2018) argue that higher capital ratios may have a negative impact on both the profitability and efficiency of the banks’ operations. Therefore, it can be argued that higher requirements hinder banks to efficiently benefit from economies of scale and scope (Classens and Klingebiel, 2001).

These effects cause a trade-off between meeting the Basel Standards and meeting shareholders’ expectations, since meeting the Basel Accords could affect the profitably in such way that shareholders’ expectations are not met anymore and vice versa. According to Liu and Ryan (2006), banks have an incentive to manage earnings through provisions for loan losses to meet both analysts’ forecasts and capital requirements. Nevertheless, because of the stricter disclosure requirements, the Central Bank of Spain (2006) argues that this form of earnings management becomes less effective after the European adoption of IFRS. However, Doyle, Jennings and Soliman (2013) state that non-GAAP reporting acts as a substitute for other forms of earnings management. This topic will be discussed more in depth in the following paragraphs.

2.3 The Role of Non-GAAP Measures 2.3.1 An Alternative Way of Reporting

Non-GAAP earnings can be seen as a collective name for the different alternative performance measures that are used. Isidro and Marques (2008) investigated the most frequently reported non-GAAP measures by 321 European firms (Table 2). They conclude that the majority of these measures are derived from the official GAAP measure. Albring, Cabán-Garcia and Reck (2010) argue that an alternative earnings number can be useful if it contains information about the ongoing operations of a firm. Therefore, an argument for the use of non-GAAP earnings is that it offers the opportunity to provide more information than would be possible under GAAP. Namely, the GAAP measures are corrected for non-recurring components, such as impairment

(12)

12

Table 2

Summary of the most frequently reported non-GAAP measures

Description Frequency in %

Non-GAAP Earnings per Share (EPS) 16.26

Earnings before Interest and Tax (EBIT) 16.26

Non-GAAP net income 13.24

Earnings before Interest, Tax, Depreciation and Amortization (EBITDA) 12.47

Non-GAAP EBITDA 9.39

Free Cash Flow 9.04

Non-GAAP EBIT 8.97

Non-GAAP income from operations 6.80

Other cash measures 4.27

Non-GAAP income from continuing operations 1.68

Non-GAAP EPS from continuing operations 1.05

Note: This table is obtained from Isidro and Marques (2008).

charges, in order to compute the non-GAAP measures. This is in line with the findings of Entwistle, Feltham and Mbagwu (2006). In the same vein, Entwistle, Feltham and Mbaqwu (2010) indicate that the non-GAAP measures are computed by disregarding non-recurring or non-cash costs. These exclusions are irrelevant according to the management, and thus are able to distort the underlying financial position of the firm. It can also be concluded that the non-GAAP measures help investors to predict the future performance based on these actual recurring earnings (Entwistle, Feltham & Mbagwu, 2010).

In addition, there are also unofficial subtotals, such as EBIT and EBITDA, which can be seen as non-GAAP measures. However, Allee, Bhattacharya, Black and Christensen (2007) do not define these measures as non-GAAP measures in their research, since these subtotals are, in general, reported in the income statement, which is prepared in accordance with GAAP. Moreover, reporting these subtotals is not compulsory according to the International Accounting Standards Board (IASB), but this could be the case under IFRS if it provides relevant information and if it faithfully represents the financial position of the firm. On the other hand, Grant and Parker (2002) argue that there is no standard for subtotals such as EBIT and EBITA. Firms can decide for themselves how these subtotals are computed, because it is not a mandatory part of their financial reports. These subtotals are, like the other non-GAAP measures, not comparable between different reporting firms. Therefore, it can be argued that these subtotals can be very misleading for investors (Grant & Parker, 2002).

(13)

13

The above findings of Isidro and Marques are in line with the definitions of non-GAAP measures as formulated by the SEC and the Committee of European Securities Regulators (CESR). First of all, the SEC’s (2002) Regulation G defines a non-GAAP measure as an adjusted numerical measure, which reflects the historical or future financial position of a firm. A non-GAAP measure excludes or includes costs that respectively are or are not included in the official GAAP measure. The SEC mentions operating income that has been adjusted for certain costs, which are defined as non-recurring, as an example of a non-GAAP measure. A non-GAAP measure is, therefore, a figure that deviates from the official GAAP measure. The CESR (2005) also defines GAAP measures in the same way as the SEC. Namely, a non-GAAP measure is derived from the official non-GAAP number, whereby adjustments are made for earnings that are non-recurring (CESR, 2005). Non-GAAP earnings can also be based on alternative sources according to the CESR and include, for example, additional performance indicators, such as production and activity levels or forward-looking indicators.

There are also non-GAAP measures that are computed and published by analysts. These non-GAAP measures are also known as the street earnings or I/B/E/S earnings (Curtis, McVay & Whipple, 2014). The relationship between these different types of non-GAAP measures will be discussed in more detail in the next paragraphs. As indicated earlier, managers can defend the use of non-GAAP measures due to the complexity of accounting. On the other hand, various researchers provide evidence that these measures are opportunistic with the aim to meet or beat analysts' expectations (Doyle, Jennings & Soliman, 2013). The following paragraphs will discuss the reasons behind the reporting of non-GAAP measures and the informativeness for investors. Subsequently, the measures taken by the SEC, the CESR and the European Securities and Markets Authority (ESMA) to prevent the reporting of opportunistic or misleading non-GAAP measures will be addressed.

2.3.2 A Tool for Managers

The discussion about the true motives behind non-GAAP reporting started after it became clear that the non-GAAP measures are, in general, higher than the GAAP measures (Bradshaw & Sloan, 2002). To begin with, Graham, Harvey and Rajgopal (2005) assume that managers use their discretion within GAAP to meet or beat the earnings expectations. In addition, Bartov, Givoly and Hayn (2002) have argued that investors reward firms that meet or beat analysts' expectations. Kasnik and McNichols (2002) also providing evidence that meeting expectations result in a significantly more positive result on the stock exchange. However, it is shown that this reward is driven by excessive positive reactions from investors to good news (Zarowin,

(14)

14

1989). If the management of the GAAP measures is not sufficient to meet or beat the expectations, non-GAAP measures are reported (Black & Christensen, 2009). However, recurring costs are also excluded in order to report a higher non-GAAP measure. Doyle et al. (2013) also argue that firms report non-GAAP measures to outline an opportunistic view, so that analysts' expectations are met or even surpassed. They conclude that the probability that a firm meets or beats these expectations rises by 14%, when income-increasing exclusions are used to compute the non-GAAP measures.

In addition, the opportunistic use of these exclusions increases, if managers are less able to apply other forms of earnings managers to meet or beat analysts’ expectations (Doyle et al., 2013). Therefore, it can be concluded that the use of non-GAAP measures is a substitute for other forms of earnings management. In the same vein, Black, Christensen, Joo and Schmardebeck (2017) conclude that managers do not report non-GAAP measures, when the GAAP results already meet the expectations. When the use of earnings management is already successful, the probability that a firm will introduce non-GAAP measures remains small. On the other hand, this probability increases significantly, when the expectations are not met. This suggests that the reporting of non-GAAP measures is not only a tool to meet analysts' expectations, but that it is also a relatively cheap way compared to earnings management techniques within GAAP according to Black et al. (2017).

Isidro and Marques (2015) prove that, in countries where efficient legislation and enforcement is in place, investors are strongly protected, financial markets are well developed and the information is well communicated and disseminated, there is a higher probability of reported non-GAAP measures meeting the expectations, if the GAAP measures do not meet them. They interpret these results as an indication that managers are experiencing more pressure to achieve the short-term objectives in institutionally and economically developed countries. In addition, the emphasis on market performance is higher in these countries according to Brown and Higgins (2001). Burgstahler, Hail and Leuz (2006) argue that, despite the higher pressure to meet the expectations, managers have more difficulty in manipulating the GAAP earnings in these developed countries, because there is more efficient reporting regulation, the risk of litigation is high and there is intensive public supervision on the reporting of financial information. As a result, managers are looking for unregulated and difficult to verify alternatives, such as current or forward-looking non-GAAP measures. This conclusion is consistent with the findings of Brown and Higgins (2005). Namely, these suggest that managers make more use of forecast guidance to meet the expectations of analysts in countries where investors are strongly protected, because the management of reported earnings is made more

(15)

15

difficult. It is also more likely that managers disregard recurring costs, such as R&D, depreciation and stock-based compensation, in order to arrive at non-GAAP earnings in institutionally and economically developing countries (Isidro & Marques, 2015). This is in line with the findings of Black and Christensen (2009), who concluded that managers disregard recurring costs in order to computed the non-GAAP measures, so that the analysts' expectations are met.

Black, Black and Christensen (2014) approached the reporting of non-GAAP measures by managers from the Agency Theory. According to Datar, Kulp and Lambart (2000), this theory suggests that compensation contracts can influence the motives of managers. For example, Holthausen, Larcker and Sloan (1995) argue that compensation in the form of bonuses can encourage managers to focus on short-term profitability and the manipulation of earnings. On the other hand, term performance plans help managers to concentrate more on long-term growth and less on earnings management. Black et al. (2014) claim and prove that compensation contracts can encourage managers to report potentially misleading non-GAAP measure. However, managers are less likely to report these misleading measures, if the compensation contracts include a long-term performance plan.

Isidro and Marques (2013) have conducted a similar research earlier. However, in contrast to Black et al. (2014), they focused their research on Europe instead of the United States. Furthermore, instead of focusing on the compensation contracts of the mangers, they examined the relation between the reporting of non-GAAP measures and the contracts of the board of directors. Furthermore, they focus on the relationship between non-GAAP reporting and compensation contracts of the management instead of the managers. They conclude that non-GAAP reporting is encouraged, if the compensation contracts are linked to the market performance. In this case, there is also a greater tendency to mention the non-GAAP measures in the titles of the press releases. In addition, more adjustments are made for recurring costs, while managers try to avoid to report a reconciliation between the GAAP and non-GAAP measures (Isidro & Marques, 2013). These studies suggest that managers report non-GAAP measures to provide an opportunistic view of the financial position of the firm. Bradshaw and Sloan (2002) have shown that the non-GAAP measures are generally higher than their GAAP counterparts. It can be argued that non-GAAP reporting is not only a substitute for earnings management, but that it is also a relatively cheap alternative. The informativeness of these non-GAAP measures for investors will be discussed further in the next paragraph.

(16)

16

2.3.3 The Informativeness for Investors

Ball and Brown (1968) were one of the first who showed that stock prices are strongly correlated with the performance of firms. This performance was traditionally measured on the basis of the GAAP figures, including net income and earnings per share. However, Collins, Maydew and Weiss (1997) prove that the value relevance of the GAAP measures has decreased for investors. According to Barth (2000), value relevance means that a performance measure is associated with a certain value, in this case the share price. Thus, he states that a performance measure is informative, if it significantly influences the share price. According to Francis and Schipper (1999), there is indeed a correlation between the official GAAP measures from the annual reports and the share prices. However, they also observe that this informativeness decreased in the period from 1952 up to and including 1994.

Albring et al. (2010) argue that an alternative measure can be useful if it contains information about the ongoing operations of the firm. In addition, investors use this information to predict future earnings. The accuracy of the prediction depends on the persistence of the earnings. Collins and Kothari (1989), therefore, came to the conclusion that investors value persistency. As a result, managers are motivated to report GAAP measures, excluding non-recurring costs out of the measure. Among others, Bhattacharya et al. (2003) and Venter, Emanuel and Cahan (2014) have proven that these non-GAAP measures are more persistent compared to their GAAP counterparts. Therefore, Bradshaw and Sloan (2002) observed a shift from the focus on GAAP to the non-GAAP measures. For this reason they did not only examine the informativeness of the GAAP measures, but also those of the non-GAAP alternative. One of their conclusions is that the informativeness of the GAAP measures has decreased over the period from 1986 to 1997. In addition, however, they show that the informativeness of the non-GAAP measures has increased for investors, while Bhattacharya et al. (2003) came to the same conclusion.

In the previous paragraph, it is argued that managers use non-GAAP measures to outline an opportunistic view of the financial position of the firm. Young (2014) also states that there is a dilemma for investors and regulators. On the one hand, managers must have the ability to incorporate relevant information in the non-GAAP measures. On the other hand, it must be ensured that they do not use their discretion to report a too opportunistic performance measure. Therefore, it can be argued that there is a trade-off between informativeness and reliability. According to Andersson and Hellman (2007), the use of non-GAAP measures by managers affects the expectations of professional investors, including analysts. However, this finding contradicts the results of previous research. For example, Frederickson and Miller (2004)

(17)

17

investigated the impact of non-GAAP measures on analysts' expectations of stock prices. They did not find a significant difference between the forecasts based on the GAAP measures and those based on their non-GAAP alternative. They argue that this is due to the trend that analysts rely on well-defined valuation models instead of on accounting measures.

Elliott (2006) comes almost to the same conclusion, but he also investigates the effect of a reconciliation between the GAAP and non-GAAP measures. However, he comes to the conclusion that analysts do give higher ratings and are willing to invest more funds in firms, if a reconciliation is provided. Therefore, the presence of a reconciliation leads to an increase in the confidence of analysts in the non-GAAP measures. Although these results indicate that analysts are not easily influenced by these measures, Andersson and Hellman (2007) have come to a different conclusion. They investigated the impact of non-GAAP measures on the EPS forecasts made by analysts. Analysts predict a significantly higher EPS when they receive both the GAAP and non-GAAP measures compared to when they only have the GAAP measures at their disposal. The researchers acknowledge, if the reporting of non-GAAP measures leads to analysts no longer focusing on the GAAP measures and thus assessing a firm differently, as the study suggests, a potential risk in trusting in accounting.

If professional investors are already influenced by the opportunistic non-GAAP measures, then the question arises what the influence of these measures have on the less developed investors. Even though Frederickson and Miller (2004) conclude that professional investors are not influenced by the non-GAAP measures, they show that this is the case for the less developed investors. Therefore, they argue that these less developed investors perceive non-GAAP measures to be more favorable, leading to a higher attached valuation of the firm. They suggest that this is the result of an unintended cognitive effect instead of being based on the relevant information that is incorporated in the non-GAAP measures. Elliott (2006) suggests the same, since she believes that this is because managers emphasize the non-GAAP information. Therefore, the less developed investors assume that the provided information is reliable. However, as argued earlier, the influence on the decision-making can be mitigated by a reconciliation between the GAAP and non-GAAP measures.

Bhattacharya et al. (2003) arguing that investors believe that the non-GAAP measures are more representative of a firm's core earnings than its GAAP counterpart. However, as indicated earlier, managers are using non-GAAP measures in a strategic manner, which leads to an opportunistic view of the financial reality of the firms. Doyle, Lundholm and Soliman (2003) are concluding that investors are misled by these measures, because they do not understand that certain exclusions are associated with lower future earnings performance. In

(18)

18

summary, it can be concluded that the less developed investors are influenced by the non-GAAP measures in particular. In addition, that these measures are often reported to provide a too opportunistic view of the financial position is argued as well. However, these findings are mainly based on literature that is provided before the introduction of regulation, and more specifically before the introduction of the SEC’s Regulation G. Therefore, the intervention by regulators and its impact will be discussed in the next paragraphs.

2.4 Non-GAAP Regulations

2.4.1 The Intervention by Regulators

Bradshaw and Sloan (2002) were one of the first indicating that the use of non-GAAP measures has increased significantly. They also came to the conclusion that investors prefer these non-GAAP measures to its non-GAAP counterpart. The SEC issued Regulation G, which is effective since 28 March 2003, in order to prevent managers using non-GAAP measures to mislead investors. According to Marques (2006), the SEC seeks to ensure greater transparency and consistency, so that investors can better understand the non-GAAP measures without being misled. First of all, a general disclosure requirement has been formulated in Regulation G. This requirement means that a firm is not allowed to disclose a non-GAAP measure that, in conjunction with the accompanying information, contains a material misstatement. In addition, this requirement also applies if the misstatement is not disclosed, so that the non-GAAP measure can be presented as non-misleading. Second, the SEC also included a specific reconciliation requirement. This means that firms have to provide a reconciliation between these measures, which explains what caused the difference between the GAAP and non-GAAP measure. The reconciliation requirement also applies to forward-looking non-GAAP measures. In contrast to the United States, there is a lack of regulation regarding non-GAAP reporting in Europe (Guillamon-Saorin, Isidro & Marques, 2017). However, shortly after the introduction of IFRS, the CESR did make recommendations regarding the use of non-GAAP measures (2005). These recommendations were made on the basis of four qualitative requirements with which financial information must comply according to IFRS. The CESR, therefore, wants both GAAP and non-GAAP information to be understandable, relevant, reliable and comparable. First of all, the CESR recommends defining alternative measures. This means that the terminology used and the calculation method must be explained. In other words, this means that it must be defined which components are excluded and included in the non-GAAP measures. According to the CESR, a clear disclosure is the key to understanding the relevance of these non-GAAP measures.

(19)

19

The CESR also recommends to present the non-GAAP measures as additional information next to the audited GAAP measures, while it is important to explain the difference between these measures. The latter can be done by means of a reconciliation, so that investors are provided with sufficient information to fully understand the results and the financial position of the firm. This recommendation is similar to the reconciliation requirement of Regulation G issued by the SEC. Furthermore, comparable information must be provided to ensure that the non-GAAP measures can be compared over time. It is important that the period to which the alternative measure relates corresponds to the related GAAP period. The definition of the non-GAAP measure must also be consistent over the entire reporting period in order to prevent investors from making decisions based on incorrect assumptions. On the other hand, the reasons should be explained for investors, when the exceptional situation arises that a firm decides to redefine the alternative measure. The CESR also concludes that firms tend to place more emphasis on the non-GAAP measures compared to its GAAP counterpart. She recommends emphasizing the GAAP measures in order to ensure that investors are not misled. Furthermore, the CESR expects firms to explain the internal use of non-GAAP measures, so that investors understand the relevance of the non-GAAP information. This explanation can only be used if it is presented together with the accompanying non-GAAP measures.

The European Financial Reporting Advisory Group (EFRAG) (2009), an organization that provides the European Commission with technical advice regarding to accounting, indicated that the non-GAAP measures of large European firms are inconsistent and obscure. The ESMA (2015) has definitively converted the above recommendations into directives, which are effective since 3 July 2016, as a result of these findings. However, these guidelines are very similar to the recommendations made by the CESR in 2005. For example, the ESMA also states that a non-GAAP measure must be defined. In doing so, it should be taken into account how this alternative measure has been computed. Furthermore, according to the ESMA, a reconciliation between the GAAP and non-GAAP measures must be provided, while taking into account the corresponding accounting period. Finally, it can be concluded that the recommendations of the CESR and ESMA are very similar to the requirements of Regulation G. The impact of these regulations and guidelines will be discussed in the next paragraph.

2.4.2 The Impact of the Regulations

Marques (2006) came to the conclusion that the introduction of Regulation G led to a decrease in the reporting of non-GAAP measures. Namely, the number of firms using these non-GAAP measures decreased from 63% to 50% in 2003, immediately after the introduction of Regulation

(20)

20

G. Entwistle et al. (2006) came to the same conclusion, but they observed a decrease from 77% in 2001 to 54% in 2003. This indeed suggests that managers reported non-GAAP measures, before the introduction of Regulation G, to influence the perception of investors or to mislead them. The introduction of Regulation G made the latter more difficult, making it less advantageous for managers to continue reporting non-GAAP measures. Furthermore, these results are in line with the findings of Heflin and Hsu (2008), since they also conclude that the introduction of Regulation G led to a decrease in the reporting of non-GAAP measures. On the other hand, Kolev, Marguardt and McVay (2008) have reached a different conclusion, when the exclusions are split up into special items and other exclusions. Namely, they show that the quality of the special items has decreased after the introduction of Regulation G. This suggests that managers are redefining recurring costs as special items more often under the regulation. The decrease in the reporting of non-GAAP measures, right after the introduction of Regulation G, can be explained on the basis of the findings of Kolev et al. (2008). They prove that firms that stopped reporting non-GAAP measures reported low-quality exclusions in the period prior to the regulation. The exclusions are of high quality if they can be defined as non-recurring. In line with Heflin and Hsu (2008), this explains why managers are less willing to report non-GAAP measures after the introduction of Regulation G. On the other hand, Kolev et al. (2008) conclude that the quality of the exclusions has increased after the intervention by the SEC. However, it can be argued that this decrease in the reporting of non-GAAP measures was temporary. Namely, Brown, Christensen, Elliott and Mergenthaler (2012) found that the reporting of non-GAAP measures increased significantly since 2004. Entwistle et al. (2006) prove that managers have started using non-GAAP measures less to increase earnings. The number of adjustments has also decreased. Furthermore, firms have started reporting non-GAAP measures in a less prominent and potentially misleading way. This suggests that managers have become more reserved with reporting non-GAAP measures. In line with the findings of Black, Christensen, Kiosse and Steffen (2017), Regulation G has, therefore, had a positive effect on the quality of non-GAAP measures. For example, the case that recurring items are excluded to compute the non-GAAP measures is less likely to happen. Managers from the United States have also used non-GAAP measures less as a tool to meet or beat analysts' expectations. This suggests that, after the introduction of Regulation G, it is less likely to be the case that managers exclude recurring components when calculating the non-GAAP measures. Therefore, it can be assumed that investors are not being misled anymore by the non-GAAP measures after the intervention of the SEC it can be assumed that investors are not being misled anymore by the non-GAAP earnings.

(21)

21

The aforementioned studies focus on disregarding recurring costs and the influence of Regulation G. Baumker, Biggs, McVay and Pierce (2014), however, focus on non-recurring revenues. They show that 88.5% of the firms report the non-recurring revenues in press releases, while only 34% are disregarding these revenues when reporting the non-GAAP measures. Prior to the introduction of Regulation G, this percentage was still 62%. According to Baumker et al., this decrease suggests that Regulation G has unintentionally led to the including of non-recurring revenues in the non-GAAP measures. Regulation G has also resulted in managers not reporting non-GAAP measures, when these measures are reporting a lower profit than the official GAAP measures. Although it has been argued earlier that Regulation G has led to an increased quality of the exclusions, the research conducted by Baumker et al. shows that managers are still able to report opportunistic GAAP measures when it comes to non-recurring revenues. Jennings and Marques (2011) also investigated the relationship between non-GAAP reporting and corporate governance. They confirm that investors were misled in the period prior to the introduction of Regulation G. However, no evidence found that this is still the case after the introduction. Furthermore, Marques (2006) shows that investors respond positively to both the GAAP and non-GAAP measures. However, after the introduction of Regulation G, non-GAAP measures are also considered as relatively informative compared with its GAAP counterpart.

As indicated earlier, non-GAAP reporting in Europe, in contrast to the United States, is not regulated (Guillamon-Saorin et al., 2017). In addition to the fact that the exclusions consist mostly of costs and that Europe is characterized by a relatively weak investor protection (La Porta, Lopez-de-Silanes & Shleifer, 2006), it can be suggested that there is a potentially greater chance of investors being misled in Europe compared to the United States. In line with this reasoning, Guillamon-Saorin et al. (2017) argue that non-GAAP measures are not only reported to reflect the core earnings of the firm, but also to influence the perception of investors with regard to the profitability. In summary, various researchers were indicating that investors are misled by the non-GAAP measures (Doyle et al., 2003). For example, it has been concluded that investors consider the exclusions to be relevant, even when they are reported to provide an opportunistic presentation of the firm's financial position (Doyle et al., 2013 & Guillamon-Saorin et al., 2017). However, this is not the case anymore, since it depends on the investors’ financial knowledge according to Christensen, Drake and Thornock (2014). The hypotheses will be formulated in the next section, while these are based on the findings discussed during the literature review.

(22)

22

3 Hypotheses Development

Ball and Brown (1968) were one of the first who showed that stock prices are strongly correlated with the performance of firms. However, Collins et al. (1997) prove that the informativeness of the GAAP measures has decreased for investors. On the other hand, Bhattacharya et al. (2003) and Venter et al. (2014) have proven that the alternative non-GAAP measures are more persistent and informative compared to their GAAP counterparts. In the same vein, Bradshaw and Sloan (2002) observed a shift from the focus on GAAP to the non-GAAP measures. In addition, they argue that the informativeness of the non-GAAP measures has decreased, while the informativeness of the non-GAAP measures has increased for investors. Furthermore, in the previous section it became clear that the magnitude of the adjustments decreased over time, which results in a smaller difference between the GAAP and non-GAAP measures (Heflin & Hsu, 2008). In the same vein, managers are more reserved with reporting these non-GAAP measures according to Black et al. (2017). These findings suggest that the chance that recurring costs are excluded decreased. Hereby, it is assumed that investors will take the exclusions into account, if they still contain any recurring components based on the findings of Doyle et al. (2013). Guillamon-Saorin et al. (2017) come to the same conclusion, even when managers emphasize opportunistic non-GAAP measures.

However, among others, the majority of these researches exclude financial institutions such as banks from their sample (e.g., Marques, 2006; Bentley, Christensen, Gee & Whipple, 2018). One of the reasons is that these firms have to comply with specific regulations (Marques, 2006). Therefore, the exclusion of banks and other financial institutions leads to a sample with a more homogenous set of firms (Leung & Veenman, 2018). In other words, there is a lack of literature about the informativeness of non-GAAP measures reported by financial institutions. Therefore, in order to be able to compare the informativeness of non-GAAP measures reported by banks with prior literature regarding non-financial firms, the following null hypotheses are formulated:

H1: Non-GAAP measures reported by banks are not relatively more informative

than their GAAP counterparts for investors.

H2a: Non-GAAP measures reported by banks are not incrementally informative for

investors.

H2b: The adjustments to the GAAP measures that are made to compute the

(23)

23

As mentioned above, Ayadi et al. (2016) state that the Basel Accords will not have an impact on performance, since the majority already complies with all the regulations and standards. Therefore, a more interesting question is what the effect would be, if the Basel Committee on Bank Supervision decides to increase the minimum levels of the standard ratio´s that should be maintained by banks. In line with the findings of Ozili (2015), Ayadi et al. (2016) argue that a negative impact can be expected, when higher standards are introduced. McKinsey (2012), for example, expects that the ROE of European banks will decrease by 40 percent. In addition, Santos (2001) concludes that higher capital requirements negatively affect bank development and credit expansion due to increased fixed and operation costs.

Since this effect will result in a negative effect on the regulated GAAP measures, it could be expected that managers report higher non-GAAP measures to cover up these effects. Namely, it can be argued that non-GAAP reporting is both a substitute and a cheaper alternative for earnings management based on the findings of Doyle et al. (2013) and Black et al. (2017). In line with this reasoning, Guillamon-Saorin et al. (2017) argue that non-GAAP measures are not only reported to reflect the core earnings of the firm, but also to influence the perception of investors with regard to the profitability of the firm. As indicated earlier, non-GAAP reporting in Europe, in contrast to the United States, is not regulated (Guillamon-Saorin et al., 2017). In addition to the fact that the exclusions consist mostly of costs and that Europe is characterized by a relatively weak investor protection (La Porta, Lopez-de-Silanes & Shleifer, 2006), it can be suggested that there is a potentially greater chance of investors being misled in Europe compared to the United States.

An important topic that should be addressed, is the fact that a majority of the stock listed European banks are operating internationally and are therefore also listed on the New York Stock Exchange (NYSE). Kolev et al. (2008) argue that the quality of the adjustments that are made to compute the non-GAAP measures has increased after the intervention by the SEC. These adjustments are of high quality, if they can be defined as non-recurring. Therefore, it can be concluded that Regulation G guarantees the reliability of the reported non-GAAP measures. The question arises whether the cross-listings on the NYSE affects the way non-GAAP measures are computed and reported by European banks. Regarding to this question, Lang, Raedy and Wilson (2006) find more evidence of earnings management for cross listed firms from countries with a weaker investor protection, which suggest that the SEC regulations do

(24)

24

not supplant the effect of the local environment. Therefore, it is assumed that the cross listing of European banks does not affect the way they compute and report their non-GAAP measures. These expectations result in the following hypotheses, stated in the alternative way:

H3a: The magnitude of the adjustments made by banks that have a relatively

high standard ratio is not equal to the magnitude of these adjustments made by banks that report a relatively low standard ratio.

H3b: The adjustments made by banks contain recurring components, when they

have a relatively high standard ratio.

(25)

25

1

NGEARNS

t

+ ꜫ

t (1.2),

where

Δ

Pt represents the change of the stock price from approximately one month prior and

one month after the announcement or press release of the of the earnings of the latest quarter. In line with the interpretation of Brown and Sivakumar (2003), this means that investors have enough time to incorporate their expectations and to react on the actual measures. EARNSt and

NGEARNSt equals respectively the earnings surprise of the reported GAAP and the non-GAAP

measure, namely the earnings on a per share basis. In other words, it equals the extent to which the GAAP and non-GAAP measure (meets or) beats the expectations, while the forecasts as recorded in the I/B/E/S database are used to define the earnings surprises.

Furthermore, the market-to-book ratio is often used to control for growth (GROWTHt)

to ensure that the results are not biased. Ettredge, Kwon, Smith & Zarowin (2005) expect a positive coefficient for this variable, because firms with high market-to-book ratios tend to have high capitalized future earnings. Another variable which can serve as a control, for the size of the

firm and the information environment, is the logarithm of the number of analysts (ANALYSTSt)that

are following the firm (Entwistle et al., 2010). A positive coefficient is predicted for this variable, because larger firms have a relatively better information environment. Furthermore, Basu (1997) finds that bad news, negative earnings, are timelier than good news. To control for the timeliness of earnings, a variable (TIMELINESSt) is included, which represents the

contemporaneous return. This variable is equal to one fornegative GAAP earnings and zero for positive or no earnings. The coefficient of this variable should be negative because negative earnings are both timelier and less persistent. In addition, Hayn (1995) also finds that firms which report losses are valuated differently. Therefore, he also controls for this different

(26)

t (2.2).

t

+ ꜫ

t (3).

where DIFFEXCLt stands for the difference between EARNSt and NGEARNSt as a result of the

adjustments made to compute the non-GAAP measure. In other words, the adjustments are equal to the components that have been left out of consideration in order to come up with the non-GAAP earnings, namely the exclusions. The non-GAAP earnings are incrementally informative to the exclusions, if a1 significantly differs from zero (Venter et al., 2014). In that

(27)

27

case, it provides investors with more information by splitting up the GAAP earnings into the ‘recurring’ non-GAAP earnings and the ‘non-recurring’ exclusions. On the other hand, the exclusions contain incremental information for investors, if its coefficient, a2, significantly

differs from zero (Jennings, 2001). The incremental informativeness of the exclusions can also be tested by using the partial F-test, where the non-GAAP model (2.2) is compared with model (3). The exclusions are incremental informative, if the explanatory power of model (3) is higher than that of model (2.2).

The sample is divided into two groups to test the third hypothesis. First, a group of banks that are reporting a relatively low standard ratio is recognized. Consequentially, the other group consists of banks that are having a standard ratio that is higher than the average maintained ratio. This process is repeated three times for banks reporting their the CET1, Leverage or Liquidity Coverage Ratio. Banks having a relatively high standard ratio are making more adjustments to the GAAP measures to compute their non-GAAP counterparts, when their average exclusions are significantly higher than the average adjustments made by bank from the other group. The significance can be tested by using the T-test for comparing the means of two samples. The procedure of the second hypothesis can be used to test whether the exclusions contain recurring components.

4.2 Sample Selection

The initial sample consists of 360 firm-quarters reflecting the last six years, since the publication of the revised version of Basel III. As indicated earlier, the sample consists of only European banks that are listed on the main indices of Europe, which include the Dutch AEX 25, the Belgian BEL 20, the French CAC 40, the German DAX 30, the Spanish IBEX 35, the British FTSE 100, the Italian FTSE MIB 25, the Swedish OMXS 30 and the Norwegian OSEBX 73. However, not all the observations are usable, since the banks are selected randomly. Both the GAAP and the non-GAAP measure, on an earnings per share basis, are hand-collected. These data can be found in the press releases or the quarterly reports as published by the firms. In the case that a firm does not report non-GAAP earnings per share, the closest non-GAAP measure, as summarized by Marques (2006), is used to compute this financial measure. Moreover, banks are required to report at least one non-GAAP measure in their quarterly report to be included in the final sample. This results in the drop of 37 unusable firm-quarters. In addition, both the analysts’ forecasts of quarterly earnings (EPS) and the number of analysts following the firm are obtained from the I/B/E/S database of Thomson Reuters. Moreover, the firm should be followed by at least one analyst during the fiscal quarter to be