Navigation – Plan du site

AccueilNuméros27-1Consequences of the Financial Mar...

Consequences of the Financial Markets Authority Guidelines on the Relevance of Non-GAAP Earnings

Conséquences des recommandations de l’Autorité des Marchés Financiers sur la pertinence des résultats non-GAAP
Davrinche Grégoire et Dumas Guillaume
Cet article est une traduction de :
Conséquences des recommandations de l’Autorité des Marchés Financiers sur la pertinence des résultats non-GAAP [fr]

Résumés

Dans cet article, nous étudions les conséquences des recommandations de l’AMF sur la pertinence des résultats non-GAAP (RNG). Ces recommandations imposent aux entreprises divulguant des RNG le respect d’un ensemble de bonnes pratiques, visant à en améliorer la transparence et la comparabilité. À partir d’un échantillon de 441 RNG communiqués entre 2011 et 2017, nous constatons une diminution de la pertinence des RNG suite à l’intervention du régulateur. Dans le même temps, les investisseurs réagissent davantage aux RNG après 2015, lorsque ces indicateurs reflètent une meilleure « qualité » de communication (i.e. conformité aux recommandations de l’AMF). Enfin, nous constatons une diminution de la dispersion des prévisions d’analystes financiers après 2015, ce qui suggère une diminution de l’asymétrie d’information suite à l’intervention de l’AMF.

Haut de page

Texte intégral

1. Introduction

  • 1 The term GAAP refers to Generally Accepted Accounting Principles. Thus, GAAP earnings are financial (...)
  • 2 According to the “Autorité des Marchés Financiers” (FMA), an APM is "a financial indicator, histori (...)

1Non-GAAP earnings (hereafter NGE) are defined as “financial indicators of corporate performance that are not prepared in accordance with generally accepted accounting rules and/or principles” (Bradshaw and Sloan, 2002, p. 42)1. NGE are part of what regulators refer to as Alternative Performance Measures (APMs)2. Labeled by firms under a variety of names (e.g. adjusted net income, recurring EBIT, etc.), these indicators are voluntarily disclosed alongside financial statements (e.g. earnings releases, IPO prospectuses, etc.), and are adjusted for items considered by management to be non-recurring and/or not relevant for investors’ decision-making (e.g. restructuring charges).

2Over the last twenty years, NGE disclosure has grown steadily. In the United States, approximately 97% of S&P 500 firms disclosed metrics of this type in 2017 (Chen et al., 2021). A similar trend is observed in Europe, where almost 75% of firms disclose NGE. This is particularly true for France as recent studies show (Clinch et al., 2019; Davrinche, 2023; Guillamon-Saorin et al., 2017). On the one hand, managers attest that NGE provide investors with better information on the firm's financial situation, a finding validated by some of the academic literature (Bradshaw et al., 2018; Guillamon-Saorin et al., 2017; Leung and Veenman, 2018). On the other hand, stock market regulators and accounting standard setters have expressed concern about the risk that NGE could end up occupying too prominent a position in the eyes of investors, compared with the results derived from financial statements (SEC, 2001; IOSCO, 2016; Hoogervorst, 2015, 2018). This concern is reinforced by academic studies revealing the adoption of disclosure strategies that could potentially mislead investors (Black and Christensen, 2009; Chen et al., 2022; Doyle et al., 2013).

  • 3 The Committee of European Securities Regulators (CESR) was the body responsible for assisting the E (...)

3This situation has led regulators to coordinate the disclosure of NGE in order to improve transparency and comparability. This process began in the US when the Security and Exchange Commission (SEC) introduced Regulation G in 2003. In Europe, following an initial initiative launched in 2005 by the Committee of European Securities Regulators (CESR)3, the European Securities and Markets Authority (ESMA) issued guidelines in 2015 for firms on the disclosure of NGE. According to the European regulator, compliance with these recommendations is supposed to “improve the comparability, reliability and/or comprehensibility of the indicators communicated” (ESMA, 2015, p. 5). ESMA also requires its guidelines to be implemented by the competent national authorities (ESMA, 2015, p. 6). France is therefore one of the countries directly affected by the European framework. On December 3rd, 2015, the “Autorité des Marchés Financiers” (hereafter FMA) incorporated these guidelines into its General Regulation which came into force from July 2016 (FMA, 2015).

4The purpose of this article is to assess: 1/ whether the publication of the FMA guidelines has led to an improvement in the “quality” of NGE disclosures, and 2/ whether this intervention has improved the relevance of these indicators to investors. Of particular interest is the question of the effectiveness of the regulatory framework applicable in France. In the US, there is consensus in the literature that Regulation G has been beneficial: it has led to an improvement in the relevance of NGE, greater transparency in disclosures and a reduction in practices likely to mislead investors (Black et al., 2012; Jennings and Marques, 2011; Kolev et al., 2008; Marques, 2006).

5However, there is no a priori evidence that these results can be transposed to the European and French context. For example, the few studies on this topic conducted in Europe show that the earlier 2005 recommendations issued by the CESR had a negligible effect on NGE disclosure (Hitz, 2010; Isidro and Marques, 2015). More recently, Jana and McMeeking (2021) show that only 62% of German firms comply with ESMA guidelines. It therefore seems appropriate to question the effect of this same regulation on French firms. In particular, there are reasons to doubt the FMA's ability to monitor the correct application of its recommendations and to sanction firms that fail to comply (Leuz, 2010). Unlike the SEC, the French regulator does not have the power to enforce this regulation. From the point of view of the relevance of NGE, FMA intervention is likely to produce effects contrary to the regulator's intentions. Indeed, if the regulatory framework is perceived as insufficiently monitored, it could prevent the signals from ‘good’ firms from being considered credible (Akerlof, 1970; Spence, 1973) and, at the same time, increase investors’ distrust of NGE (Black et al., 2012).

  • 4 This measure of “quality” differs from the definition of Dechow et al., (2010, p. 345), who assess (...)

6Our sample is composed of French firms listed in the SBF 120 index between 2011 and 2017. This period includes the two years following the FMA's implementation of the ESMA guidelines (2016 and 2017 annual results). NGE were hand-collected from press releases on the annual results. We measured disclosure quality using the score devised by Jana and McMeeking (2021). This score captures the degree of compliance of each NGE with the various recommendations of the regulator4. Finally, the relevance of NGE was measured by the investors’ reaction at the time of the earnings announcements.

7Over the entire study period, the quality of NGE disclosure can be described as “average”. For instance, only 2.7% of the firms in our sample fully comply with FMA guidelines. Nevertheless, we observed a significant increase in disclosure quality after 2015. This improvement mainly concerns the definition of NGE, the explanation of how the NGE were calculated and the presentation of comparative information. Furthermore, we find a decline in the relevance of NGE after FMA guidelines were issued: investor reaction to the announcement of NGE is less pronounced after 2015, suggesting increased investor mistrust of these indicators. However, NGE are more relevant after 2015, when the disclosure quality of these indicators became stronger. In the absence of a framework perceived as sufficiently binding, the degree to which firms comply with the guidelines therefore appears to be a particularly credible signal for investors. Our results are made more robust by the elimination of cross-listed firms that are subject to US regulations from the sample. Further analysis also reveals that after 2015, NGE are not more associated with the firm's stock price. This corroborates the notion that the regulatory framework put in place by the FMA has, in itself, had relatively limited effects. Nevertheless, we note a reduction in the dispersion of financial analysts' forecasts, which suggests reduced information asymmetry following the intervention of the French regulator.

8This article contributes to the previous literature in several ways. To begin with, it is a continuation of the studies on the impact of regulation on NGE disclosure which focus almost exclusively on the impact of Regulation G in the US context (Black et al., 2012; Black et al., 2017; Heflin and Hsu, 2008; Kolev et al., 2008; Marques, 2006). To our knowledge, only Jana and McMeeking (2021) have studied this issue in Europe. However, their study is confined exclusively to the period subsequent to the application of the ESMA guidelines (2016-2017). Moreover, these authors focus on identifying the determinants that explain firms’ compliance with these guidelines. Our study, on the other hand, highlights the consequences of this intervention for investors who are the main users of NGE (Black et al., 2018). This issue has remained largely unexplored so far. More specifically, this article documents the impact of regulation in France, one of the European countries most concerned by NGE disclosure (Clinch et al., 2019). As Leuz and Wysocki, (2016, p. 530) note, the consequences of regulation need to be studied beyond the US context in order to better understand the vectors that contribute to the effectiveness of the rules/good practices put in place by the authorities. In this sense, our results suggest that the framework applicable in France is perceived by firms as not very enforced. This is reflected in the only partial application of FMA guidelines and in increased investor mistrust of NGE. Finally, our study is likely to be of interest to regulatory authorities, and to ESMA in particular, for whom compliance with the guidelines published in 2015 is of major importance (ESMA, 2017, 2018). In particular, our results highlight the importance of equipping regulatory authorities with the tools to monitor and enforce the correct application of the relevant guidelines.

9The article is organized as follows. In the first section, we outline the institutional framework applicable to NGE disclosure (2). We then develop the relevant literature (3), leading to the formulation of our research hypotheses (4). We will then present the research methodology (5) and the results of the study (6). Finally, the last section will be devoted to the presentation of additional analyses (7).

2. Institutional Framework

2.1. Overview of Regulations in the United States

10In the US, NGE have been regulated since the early 2000s. In 2001, the SEC, the US stock market regulator, issued a statement warning investors of the potential dangers associated with the proliferation of NGE (SEC, 2001). The following year, the Sarbanes-Oxley Act was adopted, imposing new rules on corporate accounting and financial transparency. The Act was also the origin of the first regulation to formally regulate NGE disclosure. In 2003, the SEC issued Regulation G (hereafter Reg. G.). This regulation specifies how NGE should be disclosed by American firms in their public communications (e.g. press releases, investor presentations, etc.). Its aim is to increase the transparency of disclosed NGE and their comparability with financial statements. Hence, Reg. G. requires firms disclosing these indicators: (i) to present directly comparable GAAP earnings and (ii) to provide a reconciliation table explaining the adjustments used to convert GAAP earnings to NGE (SEC, 2003).

11The provisions of Reg. G. have led to the amendment of item 10(e) of Regulation S-K (hereafter Reg. S-K.), which sets out the accounting and financial disclosure requirements for listed firms. As a result, NGE covered by regulated disclosures (e.g. Form 10-Q or 10-K) are subject to stricter disclosure requirements. In addition to the requirements of Reg. G., firms must present NGE in the same way as GAAP earnings, and must justify why the NGE are useful to investors. Finally, according to the US regulator, the NGE must not “mislead users by misrepresenting or omitting information that is essential to their understanding” (SEC, 2003, §A3). Reg. S-K. provides a more exhaustive list of prohibited practices for the disclosure of NGE. Table 1 sets out the scope and obligations of each of these regulations.

Table 1. Reg. G. and Reg. S-K. (item 10(e))

Table 1. Reg. G. and Reg. S-K. (item 10(e))

2.2. Existing guidelines in Europe/France

12The existing framework in the US has contributed to the emergence and spread of regulation at the international level (IOSCO, 2002).

13At the European level, a first attempt at harmonization emerged in 2005. The Committee of European Securities Regulators (CESR) published a best-practice guide to the communication of alternative performance measures (applicable to NGE). These recommendations, strongly inspired by Reg. G. and Reg. S-K., require compliance with the following principles: definition of indicators; stability of indicators over time; explanation of the reasons why disclosure of indicators is relevant; reconciliation of indicators with results from regulated information; presentation of indicators with the same emphasis as results from regulated information (CESR, 2005). Following in CESR’s footsteps, the “Autorité des Marchés Financiers” (FMA) published a press release the same year in which it reiterated that it was not opposed to the use of NGE (FMA, 2005). However, the authority pointed out that “the use of NGE should not lead to the neglect of commonly used and recognized indicators”. In 2010, the FMA reiterated the principles set out by CESR in 2005, stressing the need for firms disclosing NGE to provide accurate, precise and fair information to investors (FMA, 2010).

14On October 5th, 2015, the European Securities and Markets Authority (ESMA) issued “Guidelines on Alternative Performance Measures”. These guidelines apply to all APMs disclosed outside financial statements (e.g. earnings releases, IPO prospectuses) by firms whose securities are listed on a regulated market in Europe. ESMA's intervention marks a turning point in the oversight of NGE in Europe and France. Firstly, these guidelines are issued by a European supervisory authority of a legal nature (CESR reported to the European Commission) and broaden ESMA's scope for action. Secondly, the European regulator imposes the implementation of its guidelines by the competent national authorities. On December 3rd, 2015, the “Autorité des Marchés Financiers” (FMA) incorporated ESMA's guidelines into its General Regulation (FMA, 2015). The FMA guidelines reproduce word for word all the provisions laid down by the European authority and that apply to NGE disclosed on or after July 3rd, 2016. The regulatory framework in force in France therefore derives from a supranational regulatory authority.

15According to the FMA, firms disclosing APMs should :

  • Define the APMs used and their components as well as the basis of calculation adopted.

  • Disclose the definitions of all APMs used in a clear and readable way.

  • Reconcile the APMs to the most directly reconcilable line item, subtotal or total presented in the financial statements.

  • Explain the use of the APMs in order to allow users to understand their relevance and reliability.

  • Not display APMs with more prominence, emphasis or authority than measures directly stemming from financial statements.

  • Provide comparatives for the corresponding previous periods.

  • Define APMs in a consistent way over time. In exceptional circumstances, where issuers or persons responsible for the prospectus decide to redefine an APM, the issuer should explain the reasons why these changes result in reliable and more relevant information on the financial performance and provide restated comparative figures.

Figure 1 summarizes the regulatory setting of NGE disclosure since the early 2000s.

Figure 1 summarizes the regulatory setting of NGE disclosure since the early 2000s.

2.3. Scope of the Frameworks and Regulators’ Capacity for Action

16Basically, the guidelines published in 2015 by the ESMA (and implemented in France by the FMA) are consistent with the current regulatory framework in the US. Indeed, almost all the recommendations required by the SEC have been adopted by the European regulator. This consistency reflects the coordination efforts undertaken by the various authorities to harmonize disclosure practices worldwide. However, two points of divergence are worth highlighting.

17Firstly, the legal status of the two frameworks reflects the differences in the ability to enforce the practices. In the US., the disclosure requirements are governed by regulations in their own right (Reg. G. and item 10 of Reg. S-K.), whereas in France they take the form of simple guidelines (or recommendations). The terms and conditions prescribed by the SEC are more restrictive than those of the FMA. For example, the SEC goes so far as to prohibit the exclusion of expenses and/or income likely to recur within two consecutive years (SEC, 2003, B2). On this same point, the French regulator remains much vaguer, simply stating the fact that firms “must not wrongly characterize items as non-recurring, infrequent or unusual” (FMA, 2015, p. 6). More generally, the scope of the SEC's regulatory framework is broader than that of the FMA. For example, Reg. S-K. prescribes how NGE presented in addition to regulated information are to be disclosed. Although auditors are not responsible for auditing these indicators, this gives them the opportunity to offer an opinion on their consistency with the results derived from the audited financial statements (Black and Christensen, 2018).

  • 5 According to Audit Analytics, the percentage of comment letters sent to firms concerning non-GAAP d (...)

18In addition, the US regulatory authority has more substantial means at its disposal to oversee the application of its regulations. In 2013, an audit committee was created to monitor compliance with Reg. G. (Rapoport, 2013). The SEC can issue “comment letters” to firms that fail to comply with the conditions set out in Reg. G. or Reg. S-K. The authority can thus demand explanations of the use of certain non-GAAP adjustments and ask the entities concerned to make changes to their financial communication5. Lastly, the SEC has on several occasions sanctioned firms for failing to comply with its regulations (e.g. SafeNet in 2009). ESMA's means of intervention are much more limited. Created in 2011, ESMA is a relatively young organization. Unlike the SEC, the European regulator does not have a specific body responsible for enforcing the application of the guidelines, this competence being delegated to each of the national authorities involved in the regulation. Consequently, the supervision of the application of these regulations is complicated by the difficulties of coordination between the various European nations.

3. Related Literature

3.1. Disclosure of NGE: Current State of the Art

19Over the past twenty years, the disclosure of NGE has become widespread in corporate financial communication. These indicators are displayed in earnings releases (Bentley et al., 2018), in the narrative sections of annual reports (Bouwens, 2019; Malone et al., 2016), in conference-calls with financial analysts (Henry et al., 2020) and in IPO prospectuses (Brown et al., 2022). While this practice was still sporadic in the early 2000s (Bhattacharya et al., 2004), voluntary disclosure of NGE now concerns most firms (Bentley et al., 2018; Black et al., 2018). In the US, disclosure frequency of NGE has steadily increased in recent years, rising from 53% in 2009 to 71% in 2014 (Black et al., 2018). According to Audit Analytics, 97% of S&P 500 firms disclosed at least one NGE metric during 2017.

20The disclosure of NGE is not confined to the American context. Studies document the use of these indicators in Canada (Cormier et al., 2017, 2022), Australia (Cameron et al., 2012; Malone et al., 2016), New Zealand (Rainsbury, 2017) and South Africa (Venter et al., 2014). In Europe, this phenomenon is more recent, although the transition to IFRS in 2005 has largely contributed to the boom in disclosure of these indicators (Young, 2014). Over the period 2003-2007, Isidro and Marques (2015) note that over 60% of firms present at least one NGE in their press releases. More recently, Guillamon-Saorin et al., (2017) note a disclosure rate approaching 75% between 2009 and 2014.

21However, there is still a degree of heterogeneity, with this practice being far more common in some countries than others (Clinch et al., 2019; Isidro and Marques, 2015). In the UK, Choi et al., (2007) report that the frequency of communication of these indicators was already 40% in 1993, rising to almost 75% by the early 2000s. In Germany, Hitz (2010) found that over the period 2005-2006, just over 35% of press releases included at least one NGE. In a more recent study, Jana and McMeeking (2021) found that NGE were included in almost half the annual reports of their sample of German firms. In France, very few studies document the disclosure of NGE. However, there seems to have been a significant jump in this practice too (Aubert, 2010; Davrinche, 2023). According to Davrinche (2023), nearly 80% of SBF 120 firms disclosed NGE over the period 2011-2017.

3.2. The Motivations behind NGE Disclosure

22The proliferation of NGE in corporate financial communications has led to the emergence of a large body of literature attempting to explain the motivations behind this practice. These studies put forward two main rationales for disclosure.

23On the one hand, according to an “informative” logic, the disclosure of NGE acts as a signal to investors of the company's future prospects, thus facilitating resource allocation on the stock market (Akerlof, 1970; Healy and Palepu, 2001). The premise is that, by eliminating non-recurring items (“non-GAAP adjustments”), NGE should facilitate the prediction of future earnings thereby giving a better assessment of a company's financial value (Nichols and Wahlen, 2004). The academic literature corroborates this first hypothesis. For example, NGE are found to be more closely correlated with stock market returns than results derived from financial statements (i.e. GAAP results), suggesting a higher information content in these indicators for investors (Aubert, 2010; Bradshaw et al., 2018; Bradshaw and Sloan, 2002; Campbell et al., 2022; Guillamon-Saorin et al., 2017). NGE also more accurately reflect the company's share price and stock market profitability over the long term (Wieland et al., 2013). Finally, studies show that NGE are more closely associated with the company's future earnings and cash flows (Bhattacharya et al., 2003; Heflin et al., 2022; Leung and Veenman, 2018). The disclosure strategies adopted by managers support the idea of a desire to provide better information. For example, most of the items excluded from the calculation of NGE correspond to transitory items: these adjustments have little (or no) association with stock market returns and/or future performance (Black et al., 2021; Choi et al., 2007; Curtis et al., 2014). Lastly, managers seem more inclined to disclose NGE when the GAAP earnings are not very relevant and/or do not accurately reflect the company's economic reality (Brown et al., 2021; Choi and Young, 2015; Lougee and Marquardt, 2004; Malone et al., 2016).

24On the other hand, according to an “opportunistic” logic, NGE disclosure could be motivated by a desire of management to mislead investors about the company's economic reality. This logic draws on agency theory and posits that managers would tend to exploit to their advantage the existing information asymmetry with the company's investors/shareholders (Jensen and Meckling, 1976; Merkl-Davies and Brennan, 2017). A number of practices adopted by managers suggest a desire to paint a more favorable picture of the company. As non-GAAP adjustments are essentially composed of expenses, by excluding them, the amount of NGE is mechanically “inflated”. As a result, NGE are almost systematically higher than GAAP earnings (Aubert, 2010; Black and Christensen, 2009). Although the majority of these adjustments represent items unrelated to the company's normal business activity, they are also made up of more recurring items: this is the case of impairment charges, share-based compensation expenses and certain tax restatements (Barth et al., 2012; Bhattacharya et al., 2004; Black and Christensen, 2009; Chen et al., 2022). NGE disclosure can also be a way of tempering poor performance as calculated using regulated information when this is deemed unsatisfactory. This is particularly the case when GAAP earnings are below certain earnings benchmarks, such as analysts’ forecasts, or the previous year's earnings (Black and Christensen, 2009; Doyle et al., 2013; Lougee and Marquardt, 2004). More generally, the literature documents various disclosure strategies to influence investors. For example, executives tend to place greater emphasis on NGE (Bouwens, 2019; Bowen et al., 2005; Cameron et al., 2012; Chen et al., 2021), modify the timing of NGE disclosure (Brown et al., 2012), or modulate the tone of the discourse accompanying these indicators (Guillamon-Saorin et al., 2017; Michailesco, 2017).

3.3. Intervention by Stock Market Regulators and NGE Disclosure

3.3.1. Consequences of Reg. G. and Reg. S-K on NGE Disclosure

25Previous studies have extensively documented the effects of SEC intervention on NGE disclosure in the US. Firstly, it has led to a reduction in the latitude exercised by executives. Studies found a decline in the frequency of NGE disclosure immediately after 2003, when Reg. G. was implemented (Heflin and Hsu, 2008; Marques, 2006; Nichols et al., 2005). According to Nichols et al., (2005), almost two-thirds of firms stopped disclosing these indicators after 2003. At the same time, there has been a decline in: (i) the magnitude of non-GAAP adjustments, (ii) the use of “recurring” adjustments, and (iii) the use of adjustments that allow the firm to achieve earnings benchmarks (Black et al., 2017; Chen, 2010; Heflin and Hsu, 2008; Kolev et al., 2008). Taken together, these results suggest that the regulation has significantly reduced the adoption of disclosure strategies considered as “opportunistic” (and therefore inappropriate).

26Secondly, the intervention of the US regulator has been accompanied by greater transparency in the disclosure of NGE, implying that the firms are conforming with the obligations imposed on them by Reg. G. and Reg. S-K. Thus, there has been a notable increase in the number of firms providing a quantitative reconciliation to explain the non-GAAP adjustments used in calculating NGE (Marques, 2010; Zhang and Zheng, 2011). For example, Marques (2010) shows that this practice increased from 27% in 2001 to 74% in 2003. Similarly, the number of firms giving NGE more prominence than GAAP earnings (a practice prohibited by the SEC) decreased after the SEC's intervention (Bowen et al., 2005; Entwistle et al., 2006; Marques, 2010).

27Finally, Reg. G. seems to have increased the relevance of NGE in the eyes of investors. Several studies document a more pronounced investor reaction to NGE announcements in the period following the SEC intervention (Black et al., 2012; Marques, 2006). According to the authors, these results reflect greater confidence by investors in NGE due to: (i) greater NGE transparency following the introduction of the obligations imposed by Reg. G. and (ii) the risk of SEC sanctions for firms that fail to meet these obligations. Moreover, it would appear that investors are less likely to be misled (Chen, 2010; Black et al., 2012; Jennings and Marques, 2011). For example, Black et al., (2012) find that since 2003, investors are better able to identify and sanction opportunistic disclosure strategies (e.g. use of recurring adjustment). Overall, the SEC's intervention appears to have had the desired effects, namely: 1/ a reduction in practices likely to mislead investors, and 2/ an improvement in the relevance of reported NGE.

3.3.2. Consequences of CESR Recommendations and ESMA / FMA Guidelines

28Relatively few studies have examined the impact of regulation outside of the US context. In New Zealand, Rainsbury (2017) shows that the recommendations put in place in 2012 by the Financial Markets Authorities (FMA) have had a negligible impact on NGE disclosure. Over the long term, however, Carvajal et al., (2022) note a reduction in the use of opaque adjustments (i.e. not properly explained by management), suggesting an improvement in disclosure transparency.

29In Europe, Isidro and Marques (2015) note a marked reduction in the frequency of communication following the implementation of the CESR recommendations in 2005. Nevertheless, they find that only 36% of firms provide a full reconciliation of NGE, while 83% continue to place greater emphasis on NGE. In the German context, Hitz (2010) studied the evolution of disclosure strategies following the implementation of the CESR recommendations. He found an increase in the use of non-GAAP adjustments, a decrease in disclosure transparency and an increase in the emphasis placed on NGE after 2005. These results suggest that CESR's intervention failed to constrain disclosure strategies considered to be opportunistic.

30Finally, Jana and McMeeking (2021) studied the disclosure of Alternative Performance Measures (including NGE) in the period following the publication of ESMA guidelines (period 2016-2017). In the German context, the authors find that some of these recommendations are fairly uniformly applied: this is the case for consistency of disclosure (94%) and presentation of NGE with the same emphasis (87%). On the other hand, relatively few firms explain why NGE provide relevant information to users (32%). Finally, the obligations to reconcile NGE (56%) and to provide comparisons with previous periods (55%) remain only partially applied. Overall, the authors find that only 10.4% of firms comply fully with ESMA's recommendations.

31In short, the study by Jana and McMeeking (2021) finds that the degree of compliance by German firms with the ESMA guidelines is mixed. However, by focusing on the period after implementation of the ESMA guidelines, it fails to capture the real impact of this intervention on NGE disclosure (with a before-after comparison). What's more, their study ignores a key question: has the implementation of ESMA guidelines resulted in a change in investor perception of NGE?

4. Hypotheses

32Considering the literature outlined above, the consequences of FMA recommendations on the relevance of NGE are not self-evident. On the one hand, we can assume that this intervention has a positive effect on investors’ overall perception of these indicators. From the point of view of signal theory, the introduction of regulation imposes an additional cost on the production of information. In this sense, it is likely to have discouraged the most opportunistic firms from disclosing NGE after 2015 (Heflin and Hsu, 2008; Kolev et al., 2008; Marques, 2006). If this is the case, firms continuing to disclose should be those whose NGE are most relevant to investors (Black et al., 2017; Jana and McMeeking, 2021).

33Furthermore, from the perspective of agency theory, regulation can be a means of compensating for the inadequacies of the governance mechanisms in place within the firm. For example, Jennings and Marques (2011) show that the improvement in the quality of NGE following the introduction of Reg. G was mainly detected in those firms with the weakest governance arrangements. In both cases, NGE should be perceived as more credible by investors who should then be able to rely more on these indicators in their investment choices.

34That said, several arguments cast doubt on the likelihood of such an effect. Firstly, the legal status of the ESMA guidelines (transposed in France by the FMA) calls into question the binding nature of the regulatory framework in Europe. The same applies to the ability of national regulators to sanction firms that fail to comply with these recommendations. Yet, the effectiveness of a regulation is intimately linked to the regulator's ability to impose costly sanctions on defaulting firms (Leuz, 2010; Leuz and Wysocki, 2016). If the cost associated with opportunistic NGE disclosure (e.g. reputational costs, risk of regulator sanction) is not sufficiently high in relation to the potential benefits of such a practice (e.g. increased stock price), regulation could encourage ‘free rider’ behavior and prevent ‘good’ firms from providing a sufficiently credible signal (Akerlof, 1970; Spence, 1973). Secondly, it is possible that FMA intervention has drawn investor attention to NGE disclosed after 2015 and increased their distrust of them. Studies thus document investors’ ability to sanction those NGE disclosures they consider to be opportunistic, in particular after the SEC’s intervention in the US (Doyle et al., 2013).

35Given these contradictory arguments, we postulate that FMA intervention has an effect on the relevance of NGE but we do not predict its direction. Consequently, hypothesis H1 is formulated as follows:

H1: FMA intervention has an impact on the relevance of NGE

36One of the objectives pursued by the FMA guidelines is to “promote the usefulness and transparency of APM included in prospectuses and regulated information” (FMA, 2015, p. 1). This transparency involves firms complying with certain procedures for disclosing NGE (e.g. definition of indicators, numerical reconciliation, justification for the use of indicators, etc.), as these procedures are intended to improve the understandability of these indicators by investors. Previous studies in the literature show that the qualitative characteristics that accompany accounting and financial information facilitate the acquisition and assimilation of the information, particularly by unsophisticated investors (Hirst and Hopkins, 1998; Maines and McDaniel, 2000). According to Li (2008), the readability of annual reports is positively associated with firm performance and earnings persistence over time, suggesting a desire on the part of management to provide more relevant information to users.

37More recently, Boutant Lapeyre et al., (2022) show that the readability of extra-financial information in annual reports is associated with a reduction in information asymmetry. Furthermore, Hirst et al., (2007) find that earnings forecasts disclosed by management are perceived as more credible when broken down (or disaggregated). Elliott (2006) shows that reconciling NGE with GAAP earnings helps to temper the optimism displayed by individual investors when assessing these indicators. Similarly, it seems that NGE are more closely associated with stock returns and/or future performance when they are presented along with a numerical reconciliation (Baik et al., 2008). In light of previous literature, we postulate a positive effect of NGE disclosure quality (i.e. compliance with FMA guidelines) on the relevance of NGE. Hypothesis H2 is therefore as follows:

H2: Better NGE disclosure quality has a positive impact on the relevance of NGE

38Given the uncertain effect of FMA intervention on the relevance of NGE (cf. H1), we are justified in questioning whether NGE disclosure quality continues to play a decisive role in the period following this intervention. On the one hand, if we postulate an improvement in the relevance of NGE following FMA intervention, we assume that investors rely more on these indicators because of the credibility conferred by the newly established framework. In this case, we can assume that the disclosure quality no longer plays a decisive role, as investors are already reassured by the existence of the regulatory framework. On the other hand, if the relevance of NGE is diminished as a result of FMA intervention, we can assume that investors are more sensitive to NGE disclosure quality, insofar as this may enable them to better discriminate between ‘good’ and ‘bad’ firms. Considering these two arguments, we postulate that the effect of disclosure quality on the relevance of NGE is different after FMA intervention, although we cannot predict the direction. Hypothesis H3 is therefore formulated as follows:

H3: The impact of disclosure quality on the relevance of NGE is different after FMA intervention

5. Research Methodology

5.1. Econometric Models

5.1.1. Impact of FMA Intervention on the Relevance of NGE (H1)

39Hypothesis H1 aims to test the effect of FMA intervention on the relevance of NGE. To do this, we applied the model used by Black et al., (2012) as follows:

40Where the CAR variable corresponds to the Cumulative Abnormal Returns observed around the time of the earnings announcement for firm i and period t. It captures investors’ reactions at the time of the earnings announcement. Abnormal returns are the difference between the actual stock return of the firm at the time of the earnings announcement and the normal return that the firm should have experienced in the absence of this announcement Normal return is calculated by reference to the market model. In line with Aubert (2010), this was estimated over a window ranging from 250 days to 11 days before the event date [-250; -11]. In addition, we have chosen a 3-day event window centered around the event date [-1; 0; +1]. The event date corresponds to the publication date of the annual results press release by the firms in the sample.

41The variable SURPRISENG corresponds to the fraction of NGE not anticipated by investors at the time of the earnings announcement. This is measured by the difference between the NGE disclosed in the press release and the latest consensus on the forecast by financial analysts established by IBES prior to the earnings announcement, deflated by the share price at the close of the previous fiscal year (Black et al., 2012). This consensus is calculated by IBES as the average of the forecasts of each financial analyst following the firm. For this variable, the consensus of forecasts represents the publicly available information. Thus, the difference between NGE and this consensus captures the private information held by the executive and that they communicate to the market. The coefficient β1 represents the effect of the NGE announcement on investor reaction. If investors consider the information contained in the NGE to be useful in their decision-making, this coefficient should be positive and significant.

  • 6 For most firms in our sample, the first disclosure affected by the FMA guidelines corresponds to th (...)

42The POSTFMA (β2) variable captures the effect of the FMA intervention on investor reaction. As a reminder, the FMA guidelines were published in December 2015 and apply to NGE disclosed on or after July 3rd, 2016. The first fiscal year affected by these guidelines therefore is the one closing after this date. Thus, the POSTFMA variable has been coded 1 for earnings releases published after July 2016 (corresponding to fiscal years 2016 and 2017) and 0 for releases published before July 2016 (corresponding to fiscal years 2011 to 2015)6. The interaction term β3 captures the moderating effect of FMA intervention on the relevance of NGE. We assume that this intervention may have resulted in contradictory effects. On the one hand, it is possible that the publication of the FMA guidelines has led to greater investor skepticism about NGE. In that case, the perceived reliability of NGE, and therefore their relevance, may have declined after 2015. On the contrary, it is possible that the FMA intervention was perceived by investors as a guarantee of transparency and reliability, given the presentation obligations it established. In this case, we can expect the NGE disclosed from 2016 onwards to be considered more relevant by investors.

43Given the presence of individual effects in our sample, we performed a Hausman specification test on panel data. This led us to adopt a fixed-effects model. Finally, in order to limit problems of heteroscedasticity and autocorrelation, the error terms were estimated at the level of each firm (clustered standard-errors), in accordance with the procedure recommended by Petersen (2009).

5.1.2. Impact of disclosure quality on the relevance of NGE (H2)

44Hypothesis H2 postulates a positive association between disclosure quality and the relevance of the NGE. To test this hypothesis, we use the model proposed by Bhattacharya et al., (2003), which we have adapted to the needs of our research question. Thus, we estimated equation (2) as follows :

The variables CAR and SURPRISENG are measured in the same way as for equation (1). The variable DQS corresponds to the disclosure quality score of NGE within press releases (measurement presented in section 5.2). Finally, the coefficient of interest β3 allows us to test hypothesis H2. The interaction between SURPRISENG and DQS captures the moderating effect of disclosure quality on the relevance of NGE. In other words, if disclosure quality increases the relevance of NGE (as we postulate), the interaction term β3 should be positive and significant.

5.1.3. Joint effect of disclosure quality and FMA intervention on the relevance of NGE (H3)

45Hypothesis H3 postulates the existence of a joint effect of disclosure quality and FMA intervention on the relevance of NGE. To test this hypothesis, we estimate equation (3) as follows :

46In this model, the variables POSTFMA and DQS are introduced simultaneously. Thus, the coefficients β4 and β5 capture the respective moderating effect of: (i) FMA intervention and (ii) disclosure quality on the relevance of NGE. The interaction term β6 captures the joint moderating effect of these two variables on our dependent variable. Thus, if NGE disclosure quality has a greater (lesser) impact after the issuance of FMA guidelines, this coefficient should be positive (negative) and significant. Conversely, if the impact of disclosure quality on the relevance of NGE is independent of FMA intervention, this coefficient should be non-significant.

5.2. Measurement of the Disclosure Quality Score (DQS)

47To measure NGE disclosure quality, we refer to the FMA guidelines on the disclosure of Alternative Performance Measures (see section 2.). Following Jana and McMeeking (2021), we consider several disclosure attributes (items), compliance with which is designed, according to the regulator, to “improve the comparability, reliability and comprehensibility of APMs” (AMF, 2015, p. 1 ): (1) the presence of a definition of NGE, (2) the appropriate denomination of NGE, (3) the presence of a reconciliation with comparable GAAP earnings from financial statements, (4) the comparison of NGE over different periods, (5) the management team's justifications of the NGE disclosure, (6) the management team's justification of the relevance of the NGE in understanding the company's performance, (7) the presentation of the NGE with the same emphasis as GAAP earnings, and (8) the consistency of NGE disclosure from one period to the next.

48For each NGE, we carried out a detailed analysis of the press release to check for compliance (or non-compliance) of the NGE with each of the 8 items presented above. For example, when the NGE was explicitly defined in the press release, item no. 1 was coded as 1. On the other hand, if this same NGE was not reconciled with the comparable GAAP earnings from the financial statements, item no. 3 was coded 0. Thus, for each NGE disclosed, this coding results in a total score ranging from 0 (NGE complies with none of the items) to 8 (NGE complies with all items). This total score was then deflated by the total number of items making up the coding (i.e. 8 items) to give a measure ranging from 0 to 1. Furthermore, in order to have a score at the firm-year level, we weighted the score of each NGE included in the same press release by the total number of indicators present in that press release. Finally, the disclosure quality score for each firm-year (variable DQS) is measured as follows:

With:

j= NGE

k= FMA item

n= Number of NGE in the same press release

49In order to ensure the reliability of the disclosure quality score, we carried out double coding using a manual to help us interpret each of the items making up the score. To do this, we relied on: (i) the clarifications provided by FMA within its position and (ii) the interpretations issued by ESMA specifying the application modalities of some of the guidelines transposed by the FMA (ESMA, 2017). As a first step, each author coded the score in a separate file. In a second step, a check was carried out to ensure consistency in the database. Any disagreements over item coding were resolved by discussion between the authors. The coding manual is presented in Appendix 1. In addition, an example of press release coding (Gemalto SA, 2015 results) is presented in Appendix 2.

5.3. Sample and Data

  • 7 Considering the impact of IFRS standards on certain disclosure strategies (Barth et al., 2012; Malo (...)

50The initial sample is composed of French firms listed on the SBF 120 index between 2011 and 2017. The choice of this index gives us access to large firms which are particularly inclined to report NGE. Moreover, the SBF 120 is an index whose performance is the subject of increased attention from investors and the financial world. In this context, we can expect the relevance of the accounting figures to be conditioned by the level and quality of the financial communications of the firms making up this sample. Starting the study period with the 2011 financial year avoids potential biases linked to the consequences of the 2008 financial crisis. Furthermore, the 2017 financial year was selected as the cut-off date because of regulatory changes brought about by the implementation of IFRS 15 (standard relating to the recognition and accounting of sales) taking effect on January 1st, 20187. This period gives us two years after the introduction in 2015 of the ESMA guidelines and their transposition in France by FMA in the same year (ESMA, 2015; AMF 2015). From our initial sample, we excluded banking institutions and insurance firms due to the accounting and regulatory specificities particular to these business sectors (56 firm-years). In addition, changes in the composition of the SBF 120 index led us to exclude 57 firm-years.

Table 2. Constitution of the Sample

Table 2. Constitution of the Sample

51The NGE were hand-collected from annual earnings press releases (Chen et al., 2021; Isidro and Marques, 2015; Jana and McMeeking, 2021). Earnings releases represent a particularly relevant disclosure vehicle, insofar as they receive special attention from investors (Davis et al., 2012). Following the definition given by FMA, we consider as NGE all indicators other than those precisely defined by IFRS (IAS 1). However, unlike Jana and McMeeking, (2021), we excluded indicators linked to the company's financial position and/or cash flows from our sample. Similarly to them, we only considered historical financial performance indicators and excluded indicators for future periods from our sample. Data on stock market returns and control variables were collected from the Refinitiv Datastream database. Finally, financial analysts' earnings forecasts were collected from the IBES database. After eliminating missing data, our final sample is composed of 84 firms representing 533 firm-years. Details of the sample are given in Table 2. The variables used in this article are defined in Table 3.

Table 3. Variables Definitions

Table 3. Variables Definitions

6. Results

6.1. Sample description

52Table 4 shows the distribution of the sample by year and industry. Of the 533 press releases analyzed (total firms-years), 441 included one or more NGE, representing over 82% of the total sample. In other words, just over 8 out of 10 press releases disclosed these indicators during the study period. By way of comparison, Aubert (2010) found that only 116 press releases included this type of indicator between 1996 and 2006. This practice has therefore increased considerably in recent years and has become comparable with the level observed in the US (Black et al., 2018). In addition, we observe a slight increase in NGE disclosure over the course of the study period. Thus, 80% of SBF firms disclosed an NGE in 2011 compared with nearly 84.5% in 2017. Thus, contrary to what was observed in the US after the introduction of Reg. G. (e.g.: Heflin and Hsu, 2008; Kolev et al., 2008), the FMA's guidelines do not appear to have had the effect of reducing the frequency of NGE disclosure.

Table 4. Sample Distribution

Table 4. Sample Distribution

53NGE disclosure concerns all industries (Panel B). It can be seen that all firms in the construction, financial services, food, healthcare, telecommunications and utilities sectors disclose NGE. Disclosure rates are close to 90% for most other industries. By contrast, firms in the industrial (70%), media (76.67%) and basic resources (66.67%) industries report relatively fewer NGE. Lastly, firms in the automobile industry seem the least inclined to disclose these indicators, with a disclosure percentage of just 27% (9 year-firms out of a total sample of 33).

6.2. Descriptive Statistics

  • 8 . A one-sample mean test (t-test) reveals that the variable CAR has a mean statistically greater th (...)

54Table 5 presents some descriptive statistics. The mean value of the NG variable is € 5.434 per share (median: 4.026). This is almost twice as much as GAAP earnings (measured by net income from continuing operations), whose mean value is € 2.910 per share (median: 2.439). The ADJUST variable represents the amount of non-GAAP adjustments (calculated as the difference between NG and GAAP). These adjustments represent € 1.865 per share, or nearly 64% of GAAP earnings. In the early 2000s, these adjustments accounted for just 18% of GAAP earnings (Aubert, 2010). As in the US, the magnitude of non-GAAP adjustments used in the French context has increased significantly in recent years. Cumulative abnormal returns at the time of earnings announcements (variable CAR) averages 0.008 (median: 0.008)8. Finally, the disclosure of NGE results in an average positive surprise (variable SURPRISENG) of 0.079.

55The mean value of DQS is 0.540. As a reminder, this variable takes values from 0 to 1 (1 being perfect conformity and 0, total non-conformity). This score therefore reflects a disclosure quality that can be considered ‘average’. The median DQS is 0.500. Thus, half of our sample is in the ‘superior’ quality range, while the other half is in the ‘inferior’ quality range. A more detailed analysis of the distribution of this variable reveals that only 10% of firms have a score above (below) 80% (37.5%). Finally, the minimum and maximum DQS values (not shown) are 0.187 and 1, respectively. In other words, no firm in our sample has a zero-quality score (i.e. compliance with none of the items making up the score). On the other hand, the maximum score of 1 reflects the fact that at least some of these firms are fully compliant with these recommendations.

56Table 5 also gives details of each of the items making up the disclosure quality score. There is considerable heterogeneity in firms’ compliance levels. On the one hand, some recommendations seem to be fairly well implemented. For example, almost all NGE are accompanied by comparative information for previous periods: the variable DQS_Item4 has a mean value of 0.987. Furthermore, firms adopt consistent disclosure choices from one period to the next, as evidenced by the variable DQS_Item8 (mean: 0.939). Thus, 93.9% of NGE disclosed in N were also disclosed in N-1. Finally, most NGE (around 3 out of 4) are designated by a terminology that correctly reflects their nature and method of calculation (DQS_Item2). Other recommendations, on the other hand, seem to have been poorly, or not at all, applied by firms. For example, the variable DQS_Item1 is only 0.321, which suggests that less than a third of NGE are explicitly defined in the press releases. Furthermore, NGE are correctly reconciled with the GAAP result in less than half the cases: the mean value of DQS_Item3 is 0.405. Finally, the use of NGE is rarely justified by management (objectives pursued and relevance of NGE to understanding performance): the variables DQS_Item5 and DQS_Item6 are 0.236 and 0.170 respectively. DQS_Item7 has a mean value of 0.502. So, about half of the NGE are presented with the same emphasis as the GAAP result.

Table 5. Descriptive Statistics

Table 5. Descriptive Statistics

57Finally, Table 5 provides information on the variables used for our additional analyses (see section 7.). The variables |SURPRISENG| and σFORECASTNG have a mean of 0.086 (standard deviation: 0.356) and 0.205 (standard deviation: 0.492), respectively. The mean value of share price (PRICE) is € 60.154. SIZE and ANALYSTS (measured in logarithm) are 7.086 and 2.886, respectively. The market-to-book ratio (MTB) is 2.961 and sales growth (ΔSALES) is 4.9%. Around 11% of the firms in the sample reported an accounting loss (LOSS), while intangible assets represented an average of 20.2% of total assets (INTANG). The σROA and HORIZON variables are 0.021 and 3.964. Lastly, the mean value of earnings per share (EPS) is 33.454.

6.3. Mean Comparisons

58In order to assess whether the FMA intervention resulted in an improvement in NGE disclosure quality, we carry out mean comparisons (Student's paired-samples t-test) of the variable DQS and the various items making up the score before and after the FMA intervention. The POSTFMA= 0 sub-sample (fiscal years 2011 to 2015) comprises 302 firm-years, while the POSTFMA= 1 sub-sample (fiscal years 2016 and 2017) comprises the remaining 139 firm-years. Insofar as the DQS variable is truncated, we also perform a non-parametric test (Wilcoxon z-test) to ensure the robustness of our results. The results are presented in Table 6.

59To begin with, we observe an increase in overall NGE disclosure quality. The mean value of DQS is 0.521 (standard deviation: 0.168) for the period preceding publication of the FMA guidelines (POSTFMA= 0) versus 0.581 (standard deviation: 0.186) for the period following publication (POSTFMA= 1). This difference is significant at the 1% level for both parametric (t= -3.338) and non-parametric (z= -3.103) tests. The overall increase in disclosure quality seems to be driven by specific items. For example, we observe a particularly significant increase in the variable DQS_Item1 (presence of a definition of NGE in the press release). This variable rises from 0.280 to 0.410 (t= -2.894; z= -2.916). Similarly, the presence of a reconciliation with GAAP earnings (variable DQS_Item3) is much more frequent following the FMA intervention: this rises from 0.359 to 0.507 (t= -3.043; z= -3.040). Finally, firms appear more inclined to present comparative information (DQS_Item4) and to justify the objectives pursued by NGE disclosure (DQS_Item5). This increase is relative, however, as mean comparisons are only significant at the 10% level.

Table 6. Mean Comparisons of DQS Variables Before and After FMA Intervention

Table 6. Mean Comparisons of DQS Variables Before and After FMA Intervention

6.4. Hypothesis Testing

  • 9 The variables SURPRISENG and |SURPRISENG| (r = 0.98), NG and SURPRISENG (r = 0.69), NG and |SURPRIS (...)

60Before proceeding to hypothesis testing, Table 7 presents the correlation matrix of the variables used. Unsurprisingly, investor reaction (CAR) is positively correlated with the surprise associated with NGE disclosure (SURPRISENG). Furthermore, NGE disclosure quality (variable DQS) is positively correlated with the variables SIZE (r= 0.25) and ANALYSTS (r= 0.23). Thus, larger firms seem more inclined to provide transparent and comparable information to their users. The variable PRICE (share price) is correlated with book value of equity (BOOK) and NGE (variable NG). Analyst forecast dispersion (σFORECASTNG) is positively correlated with the occurrence of accounting losses (LOSS), earnings dispersion (σROA) and negatively correlated with the number of analysts. Finally, company size is strongly correlated with the number of financial analysts following the company (r= 0.58)9. To ensure that our statistical tests were not biased by a multi-collinearity problem, we measured the Variance Inflation Factor (or VIF) for each of our estimates.

Table 7. Correlations Matrix (Pearson)

Table 7. Correlations Matrix (Pearson)

Note : This table presents the correlation matrix of the variables. All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.

6.4.1. H1: FMA intervention and relevance of NGE

61The results of equation (1) are presented in column 1 of Table 8. To begin with, we can see that the coefficient β1 (variable SURPRISENG) is positive and significant. As a reminder, a positive and significant sign suggests that the information contained in NGE at the time of their disclosure produces a positive reaction in investors (earnings response coefficient). In other words, independently of the FMA intervention, NGE provide useful information for investor decision-making. This finding is in line with previous studies conducted in both the US and Europe (Bradshaw et al., 2018; Guillamon-Saorin et al., 2017). The POSTFMA variable is not significant. Being in the period after the introduction of FMA guidelines does not, in itself, translate into greater investor reaction. Finally, and in line with H1, we can see that the relevance of NGE is different in the period following the intervention of the French regulator. More specifically, the interaction term β3 is negative and significant. Thus, the information content of NGE disclosed after 2015 is lower: this is 0.047 (β1) - 0.014 (β3) or 0.033. Overall, these results suggest that the FMA's intervention has resulted in greater investor skepticism of NGE disclosed by firms. On average, these indicators are perceived as less reliable after 2015, which translates into a lower reaction from investors at the time of their announcement. This finding is in contrast with what has been observed in the US, where SEC intervention has led to an improvement in the information content of NGE (e.g. Marques, 2006).

Table 8. Hypothesis Testing

Table 8. Hypothesis Testing

Note : This table presents the results of equations (1), (2), and (3). All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.

6.4.2. H2: Disclosure quality and relevance of NGE

62The results of equation (2) are presented in column 2 of Table 8. The coefficient β1 is positive and significant. The coefficient β2 (variable DQS) is -0.006 and insignificant. On the other hand, the interaction term β3 appears positive and significant at the 5% level. As a reminder, this coefficient reflects the additional effect of disclosure quality on investor reaction to NGE announcements. Thus, the relevance of NGE is more pronounced the higher the communication quality of these indicators. In total, investor response is therefore 0.026 (β1) + 0.130 (β3) or 0.156. A Fisher test reveals that these coefficients are jointly significant (F= 8.72; p-value < 0.01). Overall, the results presented in Table 8 allow us to validate hypothesis H2, and suggest that the qualitative attributes accompanying NGE disclosure are perceived as a transparency signal by investors, resulting in a greater response from them (Aubert and Grudnitski, 2014; Baik et al., 2008).

6.4.3. H3: Joint effect of disclosure quality and FMA intervention

63Column 3 of Table 8 presents the results of equation (3). This tests the joint effect of FMA intervention and disclosure quality on the relevance of NGE. In other words, we seek to determine whether the effect of disclosure quality is different after publication of the French regulator's guidelines in 2015. In line with the results presented above, we find a negative (positive) effect of FMA intervention (of disclosure quality) on the relevance of NGE: the interaction term β4 is negative and significant, while the interaction term β5 is positive and significant. Finally, the coefficient of interest is the interaction term β6 (SURPRISENG*POSTFMA*DQS). This is positive and significant at the 1% level (t= 0.042; p-value < 0.01). In line with hypothesis H3, the impact of disclosure quality on the relevance of NGE is indeed different after the FMA intervention. This total impact can be calculated as the sum of the coefficients β1 (0.029), β4 (-0.013), β5 (0.064), and β6 (0.042), i.e. 0.122. A Fisher test reveals that this positive coefficient is statistically significant. Furthermore, this coefficient is larger than the cumulative effect of β1 (SURPRISENG) and β5 (SURPRISENG*DQS). In other words, the positive effect of disclosure quality on the relevance of NGE is more pronounced in the period after publication of the FMA guidelines (after 2015).

6.4.4. Analysis based on the intensity of investors’ reaction

64In Table 9, we reproduce the test of hypotheses H1, H2 and H3 as a function of the intensity of investors’ reaction. To do this, we calculate the absolute values of the RAC variable (|CAR|), decompose the values obtained into four quartiles (Q1-Q4), and recalculate equations (1), (2) and (3) for each of these quartiles.

Table 9. Hypothesis Testing (Quartile Analysis)

Table 9. Hypothesis Testing (Quartile Analysis)

Note : This table presents the results of equations (1), (2), and (3) for each quartile of the dependent variable. All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.

  • 10 This result holds for all the estimates presented in Table 9 (columns 1 to 12).

65Columns 1 to 4 of Table 9 show the results for equation (1). The coefficient of the variable SURPRISENG is significant (and positive) only for the last two quartiles of the |CAR| variable (columns 3 and 4)10. Similarly, the interaction term β3 (SURPRISENG*POSTFMA) appears to be negative and significant for the last two quartiles of the CAR variable only. Hypothesis H1 is thus validated for the largest reactions. The results for equation (2) (columns 5 to 8) are relatively similar. In line with H2, the interaction between SURPRISENG and DQS is positive and significant. However, this is only true for the last quartile of the dependent variable (column 8). Thus, disclosure quality has an effect on the relevance of NGE only when investor reaction is strongest. Finally, columns 9 to 12 present the results relating to equation (3). Like the results presented above, the interaction between SURPRISENG and POSTFMA is negative and significant for the last two quartiles only. On the other hand, we no longer find a moderating effect of disclosure quality in equation (3). Finally, the interaction between the three variables of interest (SURPRISENG*POSTFMA*DQS) is positive for the last quartile (in line with the results in Table 8) but negative for the first and third quartiles.

66Overall, the results presented in Table 9 corroborate our hypotheses, but only in cases where investors react with the greatest intensity (i.e. highest cumulative abnormal returns). For the least intense reactions, on the other hand, we find no statistically significant link with the information contained in NGE (NGE surprise), nor any moderating effect of FMA guidelines or NGE disclosure quality.

7. Robustness Checks and Additional Analyses

7.1. Elimination of Cross-listed Firms

  • 11 . An American Depository Receipt (or ADR) is “a financial security representing shares of non-U.S. (...)
  • 12 In contrast, ADRs traded on over-the-counter (OTC) markets are not subject to any special requireme (...)

67In this article, we seek to estimate the effect of a treatment (introduction of a regulation) on the relevance of NGE. This implicitly assumes that none of the firms in the sample were subjected to a similar treatment before 2015 (when the FMA guidelines were introduced). However, given that some SBF 120 firms are also listed in the US (directly, or via American Depository Receipts)11, it is possible that they are already subject to the obligations imposed by the SEC in terms of NGE disclosure (notably item 10 of Reg. S-K). In its 2003 regulation, the SEC states that “foreign private issuers will be subject to the same requirements as domestic issuers with respect to the use of non-GAAP financial measures in documents filed with the SEC (Form 20-F)” (SEC, 2003). Item 10 of Reg. S-K therefore applies to foreign firms required to file a Form 20-F (annual report) with the SEC. Specifically, these are firms whose shares and/or ADRs are traded on an organized stock exchange (e.g. NYSE, Nasdaq)12.

68To draw up the list of firms affected by the US regulations, we used the EDGAR (Electronic Data Gathering, Analysis, and Retrieval) system, available on the SEC website. For each firm in the sample, we carried out a manual search of the documents filed on EDGAR. Of the 84 firms in our sample, 57 issue “Over the Counter” ADRs and are registered with the SEC on EDGAR (Form F-6). However, only 5 firms (representing 34 firm-years) are subject to SEC financial reporting requirements (Form F-20) and, consequently, to Reg. S-K. We therefore repeated our analyses (equations 1-3), eliminating from the sample the 34 firm-years corresponding to firms already subject to Reg. S-K in the US.

69The results, presented in Table 10, are relatively consistent with those conducted on the total sample. The coefficient β1 of the SURPRISENG variable is positive and significant in columns 1 (equation 1) and 3 (equation 3). In contrast, this coefficient is insignificant in column 2. In line with H1, the interaction between SURPRISENG and POSTFMA is negative and significant (cf. column 1), reflecting a decrease in the relevance of NGE following the implementation of the FMA guidelines. Similarly, the interaction term β3 (SURPRISENG*DQS) in column 2 is positive and significant in line with H2. Finally, the interaction term β6 in column 3 (SURPRISENG*POSTFMA*DQS) is positive and significant. Like the results found in Table 8, the effect of disclosure quality on the relevance of NGE is greater after FMA intervention.

Table 10. Robustness Check (Elimination of Cross-Listed Firms)

Table 10. Robustness Check (Elimination of Cross-Listed Firms)

Note : This table presents the results of equations (1), (2), and (3) for each quartile of the dependent variable. All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.

7.2. Impact of FMA Intervention on the Value Relevance of NGE

70To make our main results more robust, we also tested the impact of the FMA intervention on another relevance proxy widely used in previous literature: the value relevance of NGE (Brown and Sikavumar, 2003; Cormier et al., 2017; Venter et al., 2014). The value relevance study involves testing the extent to which the accounting information published by the firm is reflected in its stock price. In line with Cormier et al., (2017), we mobilized Ohlson's (1995) model, which we have adapted for the needs of our study. We formulated equation (4) as follows :

71The PRICE variable represents the firm's share price (adjusted for dividend payments) at year-end. The BOOK variable represents the firm’s book value of equity, per share. In our model, we replaced earnings per share (GAAP) by NGE per share (NG variable). Finally, we introduced our binary variable POSTFMA to interact with BOOK and NG. In equation (4), the coefficient of interest is the interaction term β5 (NG*POSTFMA). This captures the moderating effect of the FMA intervention on the relevance of the NGE communicated. The results of equation (4) are presented in Table 11. Unsurprisingly, we find that the variables BOOK and NG both have positive and statistically significant coefficients. The POSTFMA variable, on the other hand, is insignificant. Furthermore, the interaction terms β4 and β5 are also insignificant. Thus, the association between NGE and stock price is neither more pronounced (nor less pronounced) after the implementation of FMA guidelines. It therefore appears that the intervention of the stock market regulator had no impact on the relevance of NGE (as measured by their association with stock price).

Table 11. Additional Analysis : Value Relevance Test

Table 11. Additional Analysis : Value Relevance Test

Note : This table presents the results of equations (5) and (6). All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %

7.3. Impact of FMA Intervention on the Quality of Analysts’ Forecasts

72Finally, we studied the impact of FMA intervention on the quality of forecasts produced by financial analysts (Gomez et al., 2022). By collecting, reprocessing and disseminating information to the financial markets, financial analysts help to reduce the cost of acquiring and processing information for investors (Healy and Palepu, 2001). However, the quality of the information provided by these analysts depends in part on the information voluntarily disclosed by firms and their managers. Previous literature thus documents a positive effect of NGE disclosure on the accuracy of analyst forecasts (Malone et al., 2016). By imposing stricter disclosure requirements on firms disclosing NGE, FMA intervention is likely to have increased the volume and quality of information available to analysts, thereby facilitating their work (Andersson and Hellman, 2007). In line with Malone et al., (2016) and Gomez et al., (2022), we measure the quality of analyst forecasts according to two variables: the accuracy (variable |SURPRISENG|) and dispersion (variable σFORECASTNG) of the consensus of analyst forecast. We regress each of these variables on the POSTFMA variable (defined above) and a set of control variables. Equations (5) and (6) are shown below.

73The |SURPRISENG| variable represents the absolute value of SURPRISENG. The more accurate analysts are, the lower the value of this variable. The σFORECASTNG variable captures the degree of divergence between analysts contributing to the forecast consensus. This is measured as the standard deviation of IBES forecasts over the year preceding the earnings announcement. These models include a set of control variables (Gomez et al., 2022). SIZE is measured by the logarithm of total assets. Growth opportunities are measured by sales growth (ΔSALES) and Market-To-Book ratio (MTB). The number of analysts (ANALYSTS) is measured by the logarithm of the average number of analysts following the company in a given year. The LOSS variable is coded 1 when net income from continuing operations is less than zero, and 0 otherwise. INTANG is measured by the amount of intangible non-current assets divided by total assets. σROA represents earnings dispersion, measured by the standard deviation of ROA over a 6-year period. Finally, HORIZON is the logarithm of the number of days between the fiscal year-end date and the results announcement date.

74The results of equations (5) and (6) are presented in columns 1 and 2 of Table 12. The coefficient of POSTFMA is insignificant for column 1 (variable |SURPRISENG|). Thus, we observe no effect of FMA intervention on the accuracy of analysts' NGE forecasts. In column 2, on the other hand, the coefficient β1 of the POSTFMA variable is negative (β: -0.066) and significant (p-value < 0.05). Thus, after the publication of the FMA guidelines in 2015, analysts produce forecasts that are relatively less divergent from one another. The greater transparency imposed on firms after 2015 may have translated into more homogeneous information being available to financial analysts, and consequently, to more convergent valuation models. Thus, this result corroborates the idea that there is less information asymmetry between firms and analysts following the FMA intervention.

Table 12. Additional Analysis : FMA Intervention and Quality of Analysts’ Forecasts

Table 12. Additional Analysis : FMA Intervention and Quality of Analysts’ Forecasts

Note : This table presents the results of equations (5) and (6). All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %

8. Conclusion

75For more than two decades, the disclosure of NGE has attracted the attention of various players in the financial sphere. These financial indicators deviate from accounting standards and principles, and are widely used by investors for decision-making purposes (Bradshaw et al., 2018; Guillamon-Saorin et al., 2017). Nevertheless, they are disclosed in ways that confer considerable latitude on management, raising fears that these indicators may be used inappropriately by firms. In 2015, the European Securities and Markets Authority (ESMA) issued guidelines aimed at increasing the transparency, comparability and, ultimately, the usefulness of NGE for investors. These guidelines were transposed into French law by the “Autorité des Marchés Financiers” (FMA) in December 2015. Based on a sample of 441 NGE disclosed by SBF 120 firms between 2011 and 2017, we study the consequences of the framework put in place by the FMA on: (i) the quality of NGE disclosed by firms (measured by the degree to which NGE comply with the guidelines) and (ii) the relevance of these indicators for investors (measured by investor reaction at the time of their announcement). This question is of major importance given the keen interest in NGE disclosure of stock market regulators, accounting standard setters, and the financial community as a whole (IOSCO, 2016; ESMA, 2018; Hoogervorst, 2018). Moreover, very little academic work has studied the consequences of regulation outside of the US context.

76Our results show that FMA intervention has led to an improvement in the quality of NGE disclosures. After 2015, firms are more inclined to define the indicators disclosed, explain how they are calculated, provide comparative information and justify the objectives pursued by disclosing such indicators. It therefore seems that the 2015 European framework (transposed in France by the FMA) has brought about more concrete effects than the previous recommendations issued by CESR in 2005. That said, many of these guidelines remain unimplemented or very little after 2015. This is particularly true of the items concerning the comprehensibility of NGE (definition, relevance, purpose). It turns out that the investors most misled by NGE are individual or unsophisticated investors, while institutional investors have the resources to process financial and extra-financial information (Elliot, 2006). Against this backdrop, it would seem necessary to make progress on quality items relating to comprehensibility, in order to improve the information received by the most vulnerable investors, and thus protect them from poor quality information. Contrary to what has been observed in the US, we witness a decline in the relevance of NGE following introduction of the FMA guidelines. The regulator's intervention does not therefore appear to have given NGE greater credibility in the eyes of investors. On the contrary, this result reflects their skepticism towards NGE after 2015.

77Finally, it seems that investors react more to NGE disclosed after 2015, when these indicators are more in line with FMA guidelines (i.e. higher disclosure quality). In other words, it seems that the degree to which firms comply with FMA guidelines is perceived as a particularly credible signal by investors. In a complementary analysis, we tested the impact of the regulatory framework on the value relevance of NGE. Our results show that these indicators are not more closely associated with stock prices after FMA intervention. This result is in line with the skepticism observed among investors at the time of the NGE announcement. Finally, after 2015, we observe a decrease in the dispersion of analysts’ forecasts, corroborating the idea of a reduction in information asymmetry between management and financial analysts.

78Our study does, however, have certain limitations that should be highlighted. Firstly, this study focuses exclusively on the short-term consequences of the FMA intervention (first two years of application). In the US, the longer-term effects of SEC intervention are more mixed. In particular, the literature documents a resurgence of opportunistic disclosure strategies (Baumker et al., 2014; Black et al., 2017; Kolev et al., 2008) and limited effects of more recent interventions by the US regulator (Bond et al., 2017). In terms of design, although our work allows us to assess investor reaction at the time of the NGE announcement (press release publication date), we are unable to assess whether this reaction reflects a rational decision or a biased valuation on the part of investors (mispricing). In this respect, previous literature shows that some investors may be misled on the date of the earnings announcement, resulting in a stock price correction in the periods following the announcement (Doyle et al., 2003 ; Landsman et al., 2007). Finally, in this article, we focus on just one category of alternative performance measures (APMs) : NGE. This choice is motivated by the fact that these are the indicators most widely used by firms to communicate their performance. However, our results are not necessarily generalizable to other types of APMs for which the decision to disclose does not always depend on the same factors (Jana and McMeeking, 2021).

79This article paves the way for future research. Firstly, in line with what we discuss above, it would be interesting to assess the longer-term consequences of the FMA intervention on NGE disclosure. In a report published in 2019, the European regulator notes that firms’ non-compliance with the guidelines published in 2015 is partly attributable to managers' poor understanding of it (ESMA, 2019). Therefore, it would be relevant to study the impact of the interpretations published by ESMA in 2017 in order to identify whether firms have better appropriated the guidelines. Secondly, the partial application of FMA guidelines and the skepticism displayed by the markets call into question the enforcement role of the French regulator and its impact on the credibility of the information provided by firms in the eyes of investors. It would therefore be interesting to study whether the coercion exercised by regulators leads to changes in the disclosure strategies adopted by firms. This question is all the more important, given that in the US the literature shows that the SEC's issuance of comment letters has contrasting effects on the quality of NGE (Donelson et al., 2020 ; Gomez et al., 2022). Finally, this study should be extended to a European sample in order to highlight the influence of institutional and cultural factors on NGE disclosure (Isidro and Marques, 2015 ; Visani et al., 2020).

Haut de page

Bibliographie

Akerlof, G. A. (1970). The Market for « Lemons » : Quality Uncertainty and the Market Mechanism. The Quarterly Journal of Economics, 84(3), 488‑500.

AMF (2010). Communications des sociétés sur leurs indicateurs financiers. Autorité des Marchés Financiers, 17 novembre 2010.

AMF (2015). Position AMF. Indicateurs alternatifs de performance. Autorité des Marchés Financiers, 3 décembre 2015.

Andersson, P., & Hellman, N. (2007). Does Pro Forma Reporting Bias Analyst Forecasts ? European Accounting Review, 16(2), 277‑298.

Aubert, F. (2010). The Relative Informativeness of GAAP and Pro Forma Earnings Announcements in France. Journal of Accounting and Taxation, 2(1), 1‑14.

Aubert, F., & Grudnitski, G. (2014). The role of reconciliation quality in limiting mispricing of non-GAAP earnings announcements by EURO STOXX firms. Advances in Accounting, 30(1), 154‑167.

Baik, B., Billings, B. K., & Morton, R. M. (2008). Reliability and Transparency of Non-GAAP Disclosures by Real Estate Investment Trusts (REITs). The Accounting Review, 83(2), 271‑301.

Barth, M. E., Gow, I. D., & Taylor, D. J. (2012). Why do pro forma and Street earnings not reflect changes in GAAP ? Evidence from SFAS 123R. Review of Accounting Studies, 17(3), 526‑562.

Baumker, M., Biggs, P., McVay, S. E., & Pierce, J. (2014). The Disclosure of Non-GAAP Earnings Following Regulation G : An Analysis of Transitory Gains. Accounting Horizons, 28(1), 77‑92.

Bentley, J. W., Christensen, T. E., Gee, K. H., & Whipple, B. C. (2018). Disentangling Managers’ and Analysts’ Non-GAAP Reporting. Journal of Accounting Research, 56(4), 1039‑1081.

Bhattacharya, N., Black, E. L., Christensen, T. E., & Larson, C. R. (2003). Assessing the relative informativeness and permanence of pro forma earnings and GAAP operating earnings. Journal of Accounting and Economics, 36(1‑3), 285‑319.

Bhattacharya, N., Black, E. L., Christensen, T. E., & Merge, R. D. (2004). Empirical Evidence on Recent Trends in Pro Forma Reporting. Accounting Horizons, 18(1), 27‑43.

Black, D. E., Black, E. L., Christensen, T. E., & Henninger, W. G. (2012). Has Regulation G Improved the Information Quality of Non-GAAP Earnings Disclosures ? Journal of Business Finance & Accounting, 18(2), 95‑145.

Black, D. E., & Christensen, T. E. (2009). US Managers’ Use of ‘Pro Forma’ Adjustments to Meet Strategic Earnings Targets. Journal of Business Finance & Accounting, 36(3‑4), 297‑326.

Black, D. E., & Christensen, T. E. (2018). Policy implications of research on non-GAAP reporting. Research in Accounting Regulation, 30(1), 1‑7.

Black, D. E., Christensen, T. E., Ciesielski, J. T., & Whipple, B. C. (2018). Non-GAAP reporting : Evidence from academia and current practice. Journal of Business Finance & Accounting, 45(3‑4), 259‑294.

Black, D. E., Christensen, T. E., Ciesielski, J. T., & Whipple, B. C. (2021). Non-GAAP Earnings : A Consistency and Comparability Crisis ? Contemporary Accounting Research, 38(3), 1712‑1747.

Black, E. L., Christensen, T. E., Kiosse, P. V., & Steffen, T. D. (2017). Has the Regulation of Non-GAAP Disclosures Influenced Managers’ Use of Aggressive Earnings Exclusions ? Journal of Accounting, Auditing & Finance, 32(2), 209‑240.

Bond, D., Czernkowski, R., Lee, Y.-S., & Loyeung, A. (2017). Market reaction to non-GAAP earnings around SEC regulation. Journal of Contemporary Accounting & Economics, 13(3), 193‑208.

Boutant Lapeyre, J., Gatignol, M., & Martinez, I. (2022). Lisibilité des informations non financières des entreprises françaises : impact de l’intégration et conséquences sur le marché financier. Comptabilité-Contrôle-Audit, 28(4), 51-99.

Bouwens, J. (2019). The Prevalence and Validity of EBITDA as a Performance Measure. Comptabilité - Contrôle - Audit, 25(1), 55‑105.

Bowen, R. M., Davis, A. K., & Matsumoto, D. A. (2005). Emphasis on Pro Forma versus GAAP Earnings in Quarterly Press Releases : Determinants, SEC Intervention, and Market Reactions. The Accounting Review, 80(4), 1011‑1038.

Bradshaw, M. T., Christensen, T. E., Gee, K. H., & Whipple, B. C. (2018). Analysts’ GAAP earnings forecasts and their implications for accounting research. Journal of Accounting and Economics, 66(1), 46‑66.

Bradshaw, M. T., & Sloan, R. G. (2002). GAAP versus The Street : An Empirical Assessment of Two Alternative Definitions of Earnings. Journal of Accounting Research, 40(1), 41‑66.

Brown, N. C., Christensen, T. E., & Elliott, W. B. (2012). The Timing of Quarterly ‘Pro Forma’ Earnings Announcements. Journal of Business Finance & Accounting, 39(3‑4), 315‑359.

Brown, N. C., Christensen, T. E., Menini, A., & Steffen, T. D. (2022). Non-GAAP Earnings Disclosure and the Valuation of IPOs. Working Paper, https://papers.ssrn.com/sol3/papers.cfm ?abstract_id =2803795.

Brown, N. C., Huffman, A. A., & Cohen, S. (2021). Accounting Reporting Complexity and Non-GAAP Earnings Disclosure. Working Paper, https://papers.ssrn.com/sol3/papers.cfm ?abstract_id =3161566.

Brown, L. D., & Sivakumar, K. (2003). Comparing the value relevance of two operating income measures. Review of Accounting Studies, 8(4), 561‑572.

Cameron, R., Percy, M., & Stevenson-Clarke, P. (2012). Do large Australian companies emphasize non-GAAP financial measures over statutory net profit (GAAP) in annual reports ? JASSA, 1, 13‑19.

Campbell, J. L., Gee, K. H., & Wiebe, Z. (2022). The Determinants and Informativeness of Non-GAAP Revenue Disclosures. The Accounting Review, 97(7), 23-48.

Carvajal, M., Lont, D. H., & Scott, T. (2022). Non-GAAP Earnings Disclosure Trends in New Zealand. Australian Accounting Review, 100(32), 19-35.

CESR (2005). Recommendation on Alternative Performance Measures. Committee of European Securities Regulators, 1er octobre 2005.

Chen, C.-Y. (2010). Do analysts and investors fully understand the persistence of the items excluded from Street earnings ? Review of Accounting Studies, 15(1), 32‑69.

Chen, H.-C., Lee, Y.-J., Lo, S.-Y., & Yu, Y. (2021a). Qualitative characteristics of non-GAAP disclosures and non-GAAP earnings quality. Journal of Accounting and Economics, 72(1), 101402.

Chen, J. V., Gee, K. H., & Neilson, J. J. (2021b). Disclosure Prominence and the Quality of Non‐GAAP Earnings. Journal of Accounting Research, 59(1), 163‑213.

Chen, N. (Xi), Chiu, P.-C., Shevlin, T., & Wang, J. (2022). Taxes in Non-GAAP Reporting : Evidence of Strategic Behavior in Selecting Tax Rates Applied to Exclusions. Management Science, Forthcoming.

Choi, Y.-S., Lin, S., Walker, M., & Young, S. (2007). Disagreement over the persistence of earnings components : Evidence on the properties of management-specific adjustments to GAAP earnings. Review of Accounting Studies, 12(4), 595‑622.

Choi, Y.-S., & Young, S. (2015). Transitory earnings components and the two faces of non-generally accepted accounting principles earnings. Accounting & Finance, 55(1), 75‑103.

Clinch, G., Tarca, A., & Wee, M. (2019). Exploring Comparability in Disclosure of Non-IFRS Performance Measures. Working Paper, https://papers.ssrn.com/sol3/papers.cfm ?abstract_id =3343965.

Cormier, D., Demaria, S., & Magnan, M. (2017). Beyond earnings : Do EBITDA reporting and governance matter for market participants ? Managerial Finance, 43(2), 193‑211.

Cormier, D., Demaria, S., & Magnan, M. (2022). The reporting of Adjusted EBITDA and stock Markets : The Case of Canada and France. Journal of Financial Reporting and Accounting, Forthcoming.

Curtis, A. B., McVay, S. E., & Whipple, B. C. (2014). The Disclosure of Non-GAAP Earnings Information in the Presence of Transitory Gains. The Accounting Review, 89(3), 933‑958.

Davis, A. K., Piger, J. M., & Sedor, L. M. (2012). Beyond the Numbers : Measuring the Information Content of Earnings Press Release Language : Content of Earnings Press Release Language. Contemporary Accounting Research, 29(3), 845‑868.

Davrinche, G. (2023). Typologie des stratégies de divulgation de résultats non-GAAP en France : une approche multifactorielle. ACCRA, 16(1), 45-77.

Dechow, P., Ge, W., & Schrand, C. (2010). Understanding earnings quality : A review of the proxies, their determinants and their consequences. Journal of Accounting and Economics, 50(2‑3), 344‑401.

Dilla, W. N., Janvrin, D. J., & Jeffrey, C. (2013). The Impact of Graphical Displays of Pro Forma Earnings Information on Professional and Nonprofessional Investors’ Earnings Judgments. Behavioral Research in Accounting, 25(1), 37‑60.

Donelson, D. C., Kartapanis, A., & Koutney, C. (2020). SEC Non-GAAP Comment Letters and Firm Disclosures. Working Paper, https://papers.ssrn.com/sol3/papers.cfm ?abstract_id =3375762.

Doyle, J. T., Jennings, J. N., & Soliman, M. T. (2013). Do managers define non-GAAP earnings to meet or beat analyst forecasts ? Journal of Accounting and Economics, 56(1), 40‑56.

Doyle, J. T., Lundholm, R. J., & Soliman, M. T. (2003). The Predictive Value of Expenses Excluded from « Pro Forma » Earnings. Review of Accounting Studies, 8, 145‑174.

EFRAG (2017). Issues Paper : Use of additional subtotals in the Statement of Financial Performance, European Financial Reporting Advisory Group, 29-30 mars 2017.

Elliott, W. B. (2006). Are Investors Influenced by Pro Forma Emphasis and Reconciliations in Earnings Announcements ? The Accounting Review, 81(1), 113‑133.

Entwistle, G. M., Feltham, G. D., & Mbagwu, C. (2006). Financial reporting regulation and the reporting of pro forma earnings. Accounting Horizons, 20(1), 39‑55.

ESMA. (2015). ESMA Guidelines on Alternative Performance Measures. European Securities and Markets Authority, 5 octobre 2015.

ESMA (2017). Questions and answers on Alternative Performance Measures. European Securities and Markets Authority, 12 juillet 2017.

ESMA. (2018). European common enforcement priorities for 2018 annual financial reports. European Securities and Markets Authority, 26 octobre 2018.

ESMA. (2019). Report On the use of Alternative Performance Measures and on the compliance with ESMA’s APM Guidelines. European Securities and Markets Authority, 20 décembre 2019.

Gomez, E. A., Heflin, F., & Wang, J. (2022). SEC Regulation and Non-GAAP Income Statements. The Accounting Review, 98(2), 149-175.

Guillamon-Saorin, E., Isidro, H., & Marques, A. (2017). Impression Management and Non-GAAP Disclosure in Earnings Announcements. Journal of Business Finance & Accounting, 44(3‑4), 448‑479.

Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure, and the capital markets : A review of the empirical disclosure literature. Journal of accounting and economics, 31(1), 405‑440.

Heflin, F., & Hsu, C. (2008). The impact of the SEC’s regulation of non-GAAP disclosures. Journal of Accounting and Economics, 46(2‑3), 349‑365.

Heflin, F., Kolev, K. S., & Whipple, B. (2022). The risk-relevance of non-GAAP earnings. Review of Accounting Studies, Forthcoming.

Henry, E., Hu, N., & Jiang, X. (2020). Relative Emphasis on Non-GAAP Earnings in Conference Calls : Determinants and Market Reaction. European Accounting Review, 29(1), 169‑197.

Hirst, D. E., & Hopkins, P. E. (1998). Comprehensive Income Reporting and Analysts’ Valuation Judgments. Journal of Accounting Research, 36, 47.

Hirst, D. E., Koonce, L., & Venkataraman, S. (2007). How Disaggregation Enhances the Credibility of Management Earnings Forecasts. Journal of Accounting Research, 45(4), 811‑837.

Hitz, J.-M. (2010). Press Release Disclosure of ‘Pro Forma’ Earnings Metrics by Large German Corporations – Empirical Evidence and Regulatory Recommendations. Accounting in Europe, 7(1), 63‑86.

Hoogervorst, H. (2015). IASB Speech : Mind the Gap (Between non-GAAP and GAAP). Korean Accounting Review International Symposium, 31 March 2015, Seoul, Korea.

Hoogervorst, H. (2018). Chairman's speech : Japan and IFRS Standards. Keindanren Hall, 29 August 2018, Tokyo, Japan.

IOSCO. (2002). Cautionary Statement Regarding Non-GAAP Results Measures. International Organization of Securities Commission, 19 mai 2002.

IOSCO (2016). Statement On Non-GAAP Financial Measures. International Organization of Securities Commission, juin 2016.

Isidro, H., & Marques, A. (2015). The Role of Institutional and Economic Factors in the Strategic Use of Non-GAAP Disclosures to Beat Earnings Benchmarks. European Accounting Review, 24(1), 95‑128.

Jana, S., & McMeeking, K. (2021). Alternative Performance Measures : Determinants of Disclosure Quality – Evidence from Germany. Accounting in Europe, 18(1), 102‑142.

Jennings, R., & Marques, A. (2011). The Joint Effects of Corporate Governance and Regulation on the Disclosure of Manager-Adjusted Non-GAAP Earnings in the US. Journal of Business Finance & Accounting, 38(3‑4), 364‑394.

Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm : Managerial behavior, agency costs and ownership structure. Journal of financial economics, 3(4), 305‑360.

Kolev, K., Marquardt, C. A., & McVay, S. E. (2008). SEC scrutiny and the evolution of non-GAAP reporting. The Accounting Review, 83(1), 157‑184.

Landsman, W. R., Miller, B. L., & Yeh, S. (2007). Implications of Components of Income Excluded from Pro Forma Earnings for Future Profitability and Equity Valuation. Journal of Business Finance & Accounting, 34(3‑4), 650‑675.

Leung, E., & Veenman, D. (2018). Non-GAAP Earnings Disclosure in Loss Firms. Journal of Accounting Research, 56(4), 1083‑1137.

Leuz, C. (2010). Different approaches to corporate reporting regulation : How jurisdictions differ and why. Accounting and Business Research, 40(3), 229‑256.

Leuz, C., & Wysocki, P. D. (2016). The Economics of Disclosure and Financial Reporting Regulation : Evidence and Suggestions for Future Research. Journal of Accounting Research, 54(2), 525‑622.

Li, F. (2008). Annual Report Readability, Current Earnings, and Earnings Persistence. Journal of Accounting and Economics, 45(2‑3), 221‑247.

Lougee, B. A., & Marquardt, C. A. (2004). Earnings informativeness and strategic disclosure : An empirical examination of “pro forma” earnings. The Accounting Review, 79(3), 769‑795.

Maines, L. A., & McDaniel, L. S. (2000). Effects of Comprehensive‐Income Characteristics on Nonprofessional Investors’ Judgments : The Role of Financial‐Statement Presentation Format. The Accounting Review, 75(2), 179‑207.

Malone, L., Tarca, A., & Wee, M. (2016). IFRS non-GAAP earnings disclosures and fair value measurement. Accounting & Finance, 56(1), 59‑97.

Marques, A. (2006). SEC interventions and the frequency and usefulness of non-GAAP financial measures. Review of Accounting Studies, 11(4), 549‑574.

Marques, A. (2010). Disclosure strategies among S&P 500 firms : Evidence on the disclosure of non-GAAP financial measures and financial statements in earnings press releases. The British Accounting Review, 42(2), 119‑131.

Merkl-Davies, D. M., & Brennan, N. M. (2017). A theoretical framework of external accounting communication : Research perspectives, traditions, and theories. Accounting, Auditing & Accountability Journal, 30(2), 433‑469.

Michailesco, C. (2017). Communiquer sur les résultats annuels  : L’usage de la tonalité par les sociétés du Cac 40. Revue Française de Gestion, 43(263), 11‑29.

Nichols, D. C., & Wahlen, J. M. (2004). How Do Earnings Numbers Relate to Stock Returns ? A Review of Classic Accounting Research with Updated Evidence. Accounting Horizons, 18(4), 263‑286.

Nichols, N. B., Gray, S. J., & Street, D. L. (2005). Pro Forma Adjustments to GAAP Earnings : Bias, Materiality, and SEC Action. Research in Accounting Regulation, 18, 29‑52.

Ohlson, J.A. (1995). Earnings, Book Values, and Dividends in Equity Valuation. Contemporary Accounting Research, 11(2), 661-687.

Petersen, M. A. (2009). Estimating Standard Errors in Finance Panel Data Sets : Comparing Approaches. Review of Financial Studies, 22(1), 435‑480.

Rainsbury, E. A. (2017). The Impact of the FMA Guidelines of Non-GAAP Earnings Disclosures. Australian Accounting Review, 27(4), 480‑493.

SEC. (2001). Cautionary Advice Regarding the Use of "Pro Forma" Financial Information in Earnings Releases. Securities and Exchange Commission, 4 décembre 2001. https://www.sec.gov/rules/other/33-8039.htm

SEC. (2003). Final Rule : Conditions for Use of Non-GAAP Financial Measures. Securities and Exchange Commission, 22 janvier 2003. https://www.sec.gov/rules/final/33-8176.htm

SEC. (2012). Investor Bulletin : American Depository Receipts. Securities and Exchange Commission, Août 2012. https://www.sec.gov/investor/alerts/adr-bulletin.pdf

Spence, M. (1973). Job Market Signaling. The Quarterly Journal of Economics, 87(3), 355‑374.

Venter, E. R., Emanuel, D., & Cahan, S. F. (2014). The Value Relevance of Mandatory Non-GAAP Earnings : The Value Relevance of Mandatory Non-GAAP Earnings. Abacus, 50(1), 1‑24.

Visani, F., Di Lascio, F. M. L., & Gardini, S. (2020). The impact of institutional and cultural factors on the use of non-GAAP financial measures. International evidence from the oil and gas industry. Journal of International Accounting, Auditing and Taxation, 40, 1-21.

Wieland, M. M., Dawkins, M. C., & Dugan, M. T. (2013). The Differential Value Relevance of S&P’s Core Earnings Versus GAAP Earnings : The Role of Stock Option Expense. Journal of Business Finance & Accounting, 40(1‑2), 55‑81.

Young, S. (2014). The drivers, consequences and policy implications of non-GAAP earnings reporting. Accounting and Business Research, 44(4), 444‑465.

Zhang, H., & Zheng, L. (2011). The valuation impact of reconciling pro forma earnings to GAAP earnings. Journal of Accounting and Economics, 51(1‑2), 186‑202.

Haut de page

Annexe

Appendix 1. Interpretation of FMA Guidelines for the Construction of the Disclosure Quality Score (DQS)

In order to ensure the coding reliability of the disclosure quality score, we carried out a careful and detailed reading of the FMA guidelines on Alternative Performance Measures. In these guidelines, each recommendation presented in bold is accompanied by an explanation specifying what is expected of firms in terms of presentation. This additional information made it easier for us to interpret each recommendation and code the items making up the score. In addition to the FMA position, we also referred to the interpretations published by ESMA, which specify how some of the guidelines transposed in France by the FMA are to be applied (ESMA, 2017).

Item 1 : Definition of NGE

According to the FMA, “issuers or persons responsible for the prospectus should disclose the definitions of all APMs used, in a clear and readable way” (FMA, 2015, p. 6). In this article, we considered this recommendation to be correctly applied if and only if an explicit definition of NGE was included within the press release. For example, the definition of an NGE within a glossary presented separately in the press release represents an explicit definition. On the other hand, implicit definitions (via a footnote) were not considered to comply with the recommendation. Proceeding in this way ensures that we have the most conservative measure possible.

Item 2 : Denomination reflecting the content and calculation method of NGE

According to the FMA, APMs “should be given meaningful labels reflecting their content and basis of calculation in order to avoid conveying misleading messages to users” (FMA, 2015, p. 6). Although the FMA does not explicitly prescribe the terminology that firms must adopt, the regulator specifies that APMs must not “convey misleading messages to users”. Several clarifications are made here :

- Firms must not use positive or over-optimistic terms, such as “guaranteed profit” or “guaranteed return on investment”.

- Firms must not use titles similar to those used to qualify the indicators defined by the accounting standards if this could lead to confusion : this is the case, for example, for titles such as “normalized earnings” or “basic earnings from operations”.

- Firms should not wrongly classify items as non-recurring or exceptional when such items are likely to occur in future periods.

Lastly, the ESMA considers that the use of terms such as “recurring income” is likely to mislead investors as to the very nature of the items excluded from the calculation of APMs. As a result, we have deemed NGE with such a denomination not to comply with item 2.

Item 3 : Reconciliation of NGE with financial statement indicators

According to the FMA, “a reconciliation of the APM to the most directly reconcilable line item, subtotal or total presented in the financial statements of the corresponding period should be disclosed, separately identifying and explaining the material reconciling items” (FMA, 2015, p. 6). This recommendation appears rather broad, given that reconciliations can take different forms depending on the level of detail provided by firms (Zhang and Zheng, 2011). For example, a reconciliation may consist of explaining the types of expenses and/or income that have been excluded, without mentioning the amounts involved. However, the regulator specifies that users must be able to identify the reconciling items and understand how the figure was calculated. Furthermore, ESMA states that firms must provide a reconciliation "in numerical form" (ESMA, 2020, p. 15).

In view of these clarifications, only reconciliations presented in numerical form that make it possible to identify the main adjustments used in calculating NGE were deemed to comply with FMA recommendations. On the other hand, incomplete reconciliations specifying only the types or amounts of adjustments used were not considered compliant. Lastly, reconciliations that did not allow income and/or expense items to be identified in the financial statements (e.g. “other non-recurring expenses”) were considered incomplete and therefore not compliant with FMA recommendations.

Item 4 : Comparative measures

According to the FMA, firms are expected to provide comparative information for all reported APMs. The regulator thus stipulates that “APMs should be accompanied by comparatives for the corresponding previous periods” (FMA, 2015, p. 7). For example, when “adjusted net income” is presented for period N, the amount of this indicator in N-1 must also be disclosed. In addition, it is important to specify that this recommendation also applies to reconciliations provided by firms (item 3). Consequently, for indicators with a reconciliation, we have ensured that comparative information is provided in order to consider item 4 as compliant.

Item 5 : Justifications for the use of NGE

According to guidance published by the FMA, firms are required to “explain the use of APMs in order to allow users to understand their relevance and reliability” (FMA, 2015, p. 7). In other words, firms are expected to explain why the disclosure of NGE, in addition to the results derived from the financial statements, is necessary. For example, the occurrence of exceptional expenses during the course of a financial year may compromise the readability of the financial statements. In such cases, management may justify the disclosure of NGE by the need to provide users with a more relevant measure.

Item 6 : Justification of the relevance of NGE

Firms are not only required to explain why they are disclosing NGE. They must also provide details on the usefulness of the indicator. Thus, according to the FMA, firms “should explain why they believe that an APM provides useful information regarding the financial position, cash-flows or financial performance” (FMA, 2015, p. 7). For example, they may justify disclosing NGE on the grounds that it : (i) constitutes a key indicator for monitoring the firm's performance and/or (ii) is used by governance bodies to assess management's ability to create shareholder value.

Item 7 : Presentation of NGE with the same emphasis as measures directly stemming from financial statements.

The FMA states that “APMs should not be displayed with more prominence, emphasis or authority than measures directly stemming from financial statements” (FMA, 2015, p. 7). The objective pursued by the regulator is to preserve a certain neutrality in corporate financial communication so as not to overly influence the users of the information.

The “same prominence” (or “same emphasis”) criterion is undoubtedly the one that entails the greatest degree of subjectivity, since it is vague and open to interpretation. In order to facilitate the correct application of this recommendation, ESMA suggests that firms take a number of factors into consideration prior to disclosing NGE. According to ESMA, the emphasis of an indicator should be assessed in relation to the following criteria :

  • The location of APMs within the press release

  • The frequency of use of APMs

  • The use of bold type or a font that highlights APMs

  • The length of the comments accompanying the disclosure of APMs

For the coding of this item, we retained respectively : (i) the location of NGE and (ii) the use of bold type and font. This choice is mainly motivated by previous literature, which attests to the effect of these two practices on investor reaction (Dilla et al., 2013 ; Elliott, 2006 ; Henry et al., 2020). On the other hand, we have left aside the criteria linked to the frequency of use and the length of the speeches accompanying NGE communication, due to the excessive amount of time such an analysis would have required.

Item 8 : Consistent communication of NGE over time

Finally, the FMA requires firms to define and calculate APMs “consistently over time” (FMA, 2015, p. 8). The purpose of the regulator is to make APMs more comparable for users. Thus, a firm deciding to disclose “adjusted operating income” in year N is required to report the same indicator in year N+1. However, the FMA recognizes that it may be necessary to change the indicator if it is no longer relevant to explain the company's performance. In such cases, management must :

  • Explain the modifications made

  • Explain why these modifications provide more reliable and relevant information

  • Provide comparative figures

Accordingly, we have considered an NGE to be consistent from one period to the next when it is the same as that disclosed the previous year, or (in the event of a change) when the above justifications are included within the press release.

Appendix 2. Example of Press Release Coding

Gemalto : résultats de l’exercice 2015

Le chiffre d’affaires annuel est en hausse de +16 % à 3,1 milliards d’euros, et le résultat des activités opérationnelles progresse de +10 % à 423 millions d’euros (383 millions d’euros en 2014)

Le segment Paiement & Identité génère un chiffre d’affaires de 1,8 milliard d’euros, représentant 58 % du chiffre d’affaires total

Les ventes de Plateformes logicielles & Services atteignent 898 millions d’euros, en hausse de +70 %

La génération de flux de trésorerie disponible s’est accélérée au cours de l’exercice, avec un montant de233 millions d’euros au second semestre

Base de préparation de l’information financière

Analyse sectorielle

Le segment Mobile désigne les activités liées aux technologies de télécommunication mobiles cellulaires, incluant Machine-to-Machine, éléments de sécurité mobiles (SIM, éléments sécurisés embarqués) et Plateformes Logicielle & Services mobiles. Le segment Paiement & Identité rend compte des activités liées aux interactions personnelles sécurisées, notamment pour les Paiements, les Programmes gouvernementaux et les Entreprises. L’acquisition de SafeNet fait partie de l’activité Entreprises.

Outre cette information sectorielle, la Société présente le chiffre d’affaires de ses segments Mobile et Paiement & Identité par type d’activité en distinguant l’activité Logiciels embarqués & Produits (E&P) et l’activité Plateformes logicielles & Services (P&S).

Taux de change courants et taux de change constants

La Société vend ses produits et services dans un très grand nombre de pays et les règlements sont en général effectués dans d’autres devises que l’euro. Les fluctuations monétaires de ces autres devises ont en particulier un impact sur la valeur en euros du chiffre d’affaires publié de la Société. Les comparaisons exprimées à taux de change constants visent à neutraliser les effets de ces fluctuations sur les résultats de la Société, en réévaluant le chiffre d’affaires de l’année précédente au taux de change moyen de l’année courante. Sauf indication contraire, les variations du chiffre d’affaires sont à taux de change constants et intègrent l’impact du programme de couverture des variations de change. Tous les autres chiffres figurant dans le présent communiqué de presse sont à taux de change courants, sauf indication contraire.

Chiffres pro forma

Suite à l’acquisition de SafeNet et afin de mieux comprendre l’évolution de ses activités, la Société présente les chiffres pro forma par segment et activité de Gemalto pour 2014 comme si SafeNet avait été consolidé sur l’ensemble de l’exercice 2014 et les variations en glissement annuel entre ces chiffres pro forma 2014 et les chiffres 2015 comme si SafeNet avait été consolidé à compter du 1er janvier 2015. La différence entre les chiffres réels et pro forma 2015 correspond à la contribution de SafeNet du 1er janvier 2015 au 7 janvier 2015, date de réalisation de l’opération. Les chiffres pro forma de SafeNet, utilisés dans le présent document, ont été convertis en euros sur la base des taux de change mensuels. Les variations du chiffre d’affaires pro forma sont à taux de change constants et s’entendent hors impact du programme de couverture des variations de change pour 2014 et 2015.

La croissance pro forma inclut la croissance organique provenant des activités de SafeNet en 2015. Cet indicateur vise à mesurer de manière fidèle la performance opérationnelle de la Société, y compris les synergies générées par l’acquisition.

Compte de résultat ajusté et résultat des activités opérationnelles non GAAP

Les états financiers consolidés ont été établis selon les normes comptables internationales (IFRS).

Afin de mieux évaluer ses performances passées et futures, la Société élabore également un compte de résultat ajusté, dans lequel le résultat des activités opérationnelles constitue la valeur de référence permettant d’évaluer l’activité et de prendre les décisions relatives à l’exploitation de la période de 2010 à 2017.

Le résultat des activités opérationnelles est un indicateur non GAAP correspondant au résultat d’exploitation IFRS retraité (i) de l’amortissement et de la dépréciation des actifs incorporels résultant d’acquisitions, (ii) des coûts liés aux restructurations et aux acquisitions, (iii) de tous les éléments de rémunération fondés sur des actions et des coûts afférents ; et (iv) des ajustements de juste valeur découlant de l’acquisition d’entreprises, qui se définissent, respectivement, comme suit :

Amortissement et dépréciation des actifs incorporels résultant d’acquisitions : charges d’amortissement et de dépréciation liées aux immobilisations incorporelles reconnues dans le cadre de l’allocation de l’excédent de la contrepartie transférée par rapport à la valeur de l’actif net acquis.

Coûts liés aux restructurations et acquisitions : (i) coûts supportés dans le cadre d’une opération de restructuration telle que définie par la norme IAS 37 (cession ou cessation d’activité, fermeture d’usine…) et coûts afférents ; (ii) charges de réorganisation correspondant aux coûts supportés au titre de la réduction des effectifs, de la consolidation de sites de production ou administratifs, de la rationalisation et de l’harmonisation du portefeuille de produits et services, ainsi que de l’intégration des systèmes informatiques, suite à un regroupement ; et (iii) coûts de transaction (comme les honoraires versés dans le cadre des processus d’acquisition).

Éléments de rémunération fondés sur des actions : (i) décote accordée aux salariés se portant acquéreurs d’actions Gemalto dans le cadre des plans Gemalto de souscription d’actions ; et (ii) amortissement de la juste valeur des options d’achat d’actions et des unités d’actions restreintes accordées par le Conseil d’administration aux salariés, ainsi que les coûts afférents.

Ajustements de juste valeur des actifs nets acquis : reprise, en compte de résultat, des ajustements de juste valeur constatés suite à un regroupement d’entreprises, conformément à la norme IFRS 3 R. Ces ajustements concernent principalement : (i) les charges d’amortissement liées à la réévaluation des travaux en cours et des produits finis comptabilisés à leur valeur de réalisation lors de l’acquisition et (ii) l’amortissement de la marge commerciale sur les soldes de produits constatés d’avance, annulée lors de l’acquisition.

Ces agrégats non-GAAP ne sont pas destinés à être utilisés isolément ni à remplacer les agrégats IFRS correspondants, ils doivent être analysés parallèlement à nos états financiers consolidés, établis conformément aux normes comptables internationales (IFRS).

Dans le compte de résultat ajusté, les charges d’exploitation correspondent à la somme des frais d’étude et recherche, des frais commerciaux et de marketing, des frais généraux et administratifs, et des autres produits (ou charges) nets.

Haut de page

Notes

1 The term GAAP refers to Generally Accepted Accounting Principles. Thus, GAAP earnings are financial indicators defined precisely by an accounting framework.

2 According to the “Autorité des Marchés Financiers” (FMA), an APM is "a financial indicator, historical or future, of performance, financial position or cash flows other than a financial indicator defined or specified in the applicable accounting framework" (FMA, 2015, p. 5). APMs therefore cover a broader panel of financial indicators that includes NGE. In this article, we focus specifically on the disclosure of NGE. We will therefore make a distinction between the terms NGE and APM to avoid confusion.

3 The Committee of European Securities Regulators (CESR) was the body responsible for assisting the European Commission in regulating financial markets. It was replaced in 2011 by the European Securities and Markets Authority (ESMA).

4 This measure of “quality” differs from the definition of Dechow et al., (2010, p. 345), who assess earnings quality for the purposes of decision-making by the user (e.g. relevance, persistence, predictive capacity, etc.).

5 According to Audit Analytics, the percentage of comment letters sent to firms concerning non-GAAP disclosure has risen from 13% in 2014 to 37% in 2017.

6 For most firms in our sample, the first disclosure affected by the FMA guidelines corresponds to the results of the fiscal year 2016 ending on December 31st (with the disclosure published in early 2017). However, for firms whose fiscal year-end does not coincide with the calendar year, we considered the first fiscal year as the one ending after July 3rd, 2016 (e.g., September 31st, 2016, March 31st, 2017).

7 Considering the impact of IFRS standards on certain disclosure strategies (Barth et al., 2012; Malone et al., 2016), we exclude this year from the study to avoid introducing a bias into our analyses.

8 . A one-sample mean test (t-test) reveals that the variable CAR has a mean statistically greater than zero (t = 3.622; p-value < 0.001). A non-parametric Wilcoxon test revealed a similar finding (z = 3.578; p-value < 0.01).

9 The variables SURPRISENG and |SURPRISENG| (r = 0.98), NG and SURPRISENG (r = 0.69), NG and |SURPRISENG| (r =0.69) are highly correlated with each other. However, this strong collinearity is not a problem, as these variables are used in separate models.

10 This result holds for all the estimates presented in Table 9 (columns 1 to 12).

11 . An American Depository Receipt (or ADR) is “a financial security representing shares of non-U.S. companies held by an American depository bank outside the United States” (SEC, 2012). The issuance of ADRs enables US investors to buy and/or sell shares in non-US firms without having to intervene in foreign markets.

12 In contrast, ADRs traded on over-the-counter (OTC) markets are not subject to any special requirements. These firms must, however, register with the SEC by means of a Form F-6.

Haut de page

Table des illustrations

Titre Table 1. Reg. G. and Reg. S-K. (item 10(e))
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-1.png
Fichier image/png, 98k
Titre Figure 1 summarizes the regulatory setting of NGE disclosure since the early 2000s.
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-2.png
Fichier image/png, 39k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-3.png
Fichier image/png, 12k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-4.png
Fichier image/png, 10k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-5.png
Fichier image/png, 20k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-6.png
Fichier image/png, 7,9k
Titre Table 2. Constitution of the Sample
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-7.png
Fichier image/png, 56k
Titre Table 3. Variables Definitions
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-8.png
Fichier image/png, 99k
Titre Table 4. Sample Distribution
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-9.png
Fichier image/png, 107k
Titre Table 5. Descriptive Statistics
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-10.png
Fichier image/png, 88k
Titre Table 6. Mean Comparisons of DQS Variables Before and After FMA Intervention
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-11.png
Fichier image/png, 92k
Titre Table 7. Correlations Matrix (Pearson)
Légende Note : This table presents the correlation matrix of the variables. All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-12.png
Fichier image/png, 99k
Titre Table 8. Hypothesis Testing
Légende Note : This table presents the results of equations (1), (2), and (3). All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-13.png
Fichier image/png, 59k
Titre Table 9. Hypothesis Testing (Quartile Analysis)
Légende Note : This table presents the results of equations (1), (2), and (3) for each quartile of the dependent variable. All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-14.png
Fichier image/png, 105k
Titre Table 10. Robustness Check (Elimination of Cross-Listed Firms)
Légende Note : This table presents the results of equations (1), (2), and (3) for each quartile of the dependent variable. All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %.
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-15.png
Fichier image/png, 74k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-16.png
Fichier image/png, 12k
Titre Table 11. Additional Analysis : Value Relevance Test
Légende Note : This table presents the results of equations (5) and (6). All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-17.png
Fichier image/png, 48k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-18.png
Fichier image/png, 18k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-19.png
Fichier image/png, 18k
Titre Table 12. Additional Analysis : FMA Intervention and Quality of Analysts’ Forecasts
Légende Note : This table presents the results of equations (5) and (6). All variables are defined in table 3. ***,**, and * represent statistical significance at respectively 1 % ; 5 % and 10 %
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-20.png
Fichier image/png, 49k
URL http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/docannexe/image/11942/img-22.png
Fichier image/png, 20k
Haut de page

Pour citer cet article

Référence électronique

Davrinche Grégoire et Dumas Guillaume, « Consequences of the Financial Markets Authority Guidelines on the Relevance of Non-GAAP Earnings »Finance Contrôle Stratégie [En ligne], 27-1 | 2024, mis en ligne le 15 mars 2024, consulté le 18 juin 2024. URL : http://0-journals-openedition-org.catalogue.libraries.london.ac.uk/fcs/11942 ; DOI : https://0-doi-org.catalogue.libraries.london.ac.uk/10.4000/fcs.11942

Haut de page

Auteurs

Davrinche Grégoire

Maître de Conférences, Université de Lille, LUMEN (ULR 4999)

Dumas Guillaume

Maître de Conférences, Université de Montpellier, MRM

Haut de page

Droits d’auteur

Le texte et les autres éléments (illustrations, fichiers annexes importés), sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search