Alternative Performance Measures: Determinants of Disclosure Quality – Evidence from Germany

ABSTRACT

Alternative performance measures (APMs) might be used to improve the information environment or strategically to mislead the market. The recently introduced European Securities and Markets Authority APM guidelines are intended to enhance corporate financial disclosures. We analyse the disclosure quality and determinants of all types of APMs in management reports of German listed firms for two financial periods. Although the quantity of APM disclosures is extensive, it differs across firms’ characteristics, and there is considerable room for improvement regarding disclosure quality. APM disclosure quality is positively associated with firm size and negatively associated with profitability. However, not all firms’ characteristics can be applied per se as universal determinants of APM disclosure quality, and a distinction must be made between different types of APMs. For example, high ownership concentration is negatively associated particularly with the quality of profitability APMs. Firms’ leverage is positively associated with the disclosure quality of non-profitability APMs.

Keywords:

• disclosure quality
• alternative performance measures
• non-GAAP measures
• pro forma financial measures
• ESMA guidelines

Subject classification codes:

• M40
• M41

Acknowledgements

The authors appreciate the valuable comments and suggestions on earlier versions of this paper received from Eduardo Damásio, Ana Marques, and other participants of the 2018 14th Workshop on European Financial Reporting (EUFIN) in Stockholm, Sweden, as well as participants of the 2019 13rd Workshop on Empirical Research in Financial Accounting in Castellón, Spain, the 2019 42nd Annual Congress of the European Accounting Association (EAA) in Paphos, Cyprus, the 2019 81st VHB annual conference in Rostock, Germany, and workshops at the University of Exeter and ESCP Business School Berlin. The constructive comments of the editor and of two anonymous reviewers are also greatly appreciated. Finally, the authors are grateful for the excellent research assistance of Steffen Schneider.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 Profitability APMs are also called ‘pro forma earnings’ or ‘non-GAAP earnings measures’ in the accounting and finance literature (e.g. Doyle et al., 2013; Jennings & Marques, 2011; Leung & Veenman, 2018; Young, 2014). We use the terms profitability/non-profitability APMs throughout the paper. Non-profitability APMs compromise the following APM categories: asset/capital structure, capital efficiency, liquidity/cash flow, stock market/valuation, and volume/growth.

2 Aripin (2010) collects and analyses 43 predefined and generalised ratios but excludes firm specific APMs, industry specific APMs and adjusted financial numbers. Agyei–Mensah (2015) does not discuss what kind of financial ratios is collected. Bini et al. (2017) focus on APMs that are presented, e.g. as ‘main performance results’ or ‘key performance indicators’.

3 Particularly, Aripin (2010) builds an index from the qualitative characteristics of the IFRS Framework to analyse APM disclosure quality. Due to strong generalisation, some of the 12 criteria derived from the IFRS characteristics are not directly related to APM disclosure quality and the qualitative disclosure requirements per se required by regulators. For example, a graphical or tabular presentation, specific disclosure locations, or a comparison with target benchmarks and industry consistency are defined as quality requirements in the conducted index. Agyei–Mensah (2015) bases his analysis on the index by Aripin (2010). Bini et al. (2017) implement their own index based on the work of Aripin (2010) in conjunction with recommendations of standard setters and regulators published until then. Consequently, Bini et al. (2017) do not cover all defined quality requirements of the new ESMA APM guidelines.

4 For example, in 2002, the IOSCO had already released a ‘Cautionary Statement Regarding Non-GAAP Results Measures’ advising issuers and investors to exert care when presenting and interpreting non-GAAP measures (IOSCO, 2002).

5 Currently, the IASB addresses the APM issue in its ‘Disclosure Initiative’ with the completed sub-project ‘Principles of Disclosure’ and in its standard-setting project ‘Primary Financial Statements’ (IASB, 2017, 2018, 2019). In its recently published Exposure Draft, the IASB intends to introduce (i) three new profit subtotals, including operating profit, (ii) a required better analysis of operating expenses, including guidance on any unusual income or expenses and (iii) a definition of ‘management performance measures’ (MPMs) with requirements for the disclosure (e.g. disclosed in a single note, reconciliation to IFRS figures). These proposals related to MPMs are in line with ESMA APM quality requirements but are not as detailed as the ESMA guidelines. For example, the IASB does not consider cash flow measures, ratios (e.g. return in equity), adjusted revenue and growth rates as meeting the definition of MPMs, and therefore only subtotals of income and expenses would be included (IASB, 2019, B80). Also, the IASB considered but rejected to define ‘EBITDA’ because of a lack of consensus with regard to the definition and calculations (IASB, 2019). Nevertheless, the IFRS proposals might aid users by enhancing the understandability of APMs and increasing compliance amongst firms in member states.

6 This distinguishes these soft law measures from technical standards (EU, 2010, Art. 16).

7 In particular, a true and fair view of the economic position has to be conveyed, opportunities and risks of future development have to be properly presented, and in all material respects it has to comply with the legal requirements. Additionally, in the absence of an applicable IFRS for the management report, in 2004 the Accounting Standards Committee of Germany (ASCG) issued German Accounting Standard (GAS) 15 on Management Reporting, superseded in 2013 by GAS 20, to complement the HGB sections. It gives specific recommendations regarding which information a management report should include. However, GAS 20 does not provide a general definition of performance measures and only contains rudimentary guidance on how to present respective financial and non-financial key performance measures (e.g. presentation of calculation, reconciliation, and discussion of changes).

8 German listed firms are subject to a two-tier enforcement system established in 2005 based on the Bilanzkontrollgesetz [Financial Statements Control Act]. The first tier involves the FREP as a government-appointed private institution. The second tier is performed by the Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin) [Federal Financial Supervisory Authority] (for further information see also Hitz & Schnack, 2019). BaFin reported to ESMA that they comply with the APM guidelines by incorporating them into their supervisory practices (ESMA, 2017b).

9 For consistency reasons, all our hypotheses are stated in the alternative form.

10 The Transparency Directive (Directive, 2004/109/EC) was issued in 2004 and developed further disclosure obligations for European security issuers in regulated markets in order to further harmonise information duties. The Directive demands issuers to publish annual reports that comprise audited financial statements, the management report and a responsibility statement by the legal representatives that the annual report conveys a true and fair view (Transparency Directive, Art. 4).

11 Both tiers differ substantially in terms of visibility to the public, media, and analyst coverage. The prime standard is intended for firms with an international focus, and firms traded in this standard are eligible for being included in one of the German stock market indices (e.g. the blue-chip market index DAX). Thus, these firms are subject to higher overall transparency requirements (e.g. analysts’ conferences at least once a year) as well as to a higher probability of being selected by the FREP for an additional enforcement (FREP, 2016b).

12 Since the ESMA guidelines gained effectiveness in July 2016, the first reporting year which is shaped by this particular regulatory setting is 2016.

13 With regard to listed firms in the Prime Standard and General Standard in 01/2017.

14 We allocate APMs into the following categories: (1) asset/capital structure (2) capital efficiency (3) liquidity/cash flow (4) profitability (5) stock market/valuation (6) volume/growth.

15 In Germany, GAS 20.P45 requires listed firms to present their internal management system, including the performance indicators used for internal management purposes. Thus, management reports typically contain a separate section that explains which performance indicators are applied for this purpose. Furthermore, several firms state which measures they regard as APMs. However, these statements were not considered decisive for the identification and inclusion of an APM. Several management reports exemplify that additional measures are disclosed which are not introduced as such by the firm. Limiting the analysis to declared measures could therefore depict a biased picture.

16 For example, for the ESMA APM quality requirement ‘Explanation on the use’: (1) Firm indicates the purpose of the APM by a ‘profitability measure’ (sub-requirement (1) purpose = 1). However, this does not sufficiently explain why a specific measure provides useful information (sub-requirement (2) why = 0), leading to an overall score for ‘Explanation on the use’ = 0. In the Appendix, the assessment of the disclosure quality for each quality requirement is provided by using a firm example.

17 For example, different APM labels for economic value added are ‘Added Value’, ‘[Firm’s name] value added’, ‘Net value added’, and ‘Value contribution’. Adjusted EBIT is also called, for example ‘EBIT before special items’, ‘EBIT pre’, ‘EBIT excluding special factors’, ‘Normalised EBIT’, ‘Recurring EBIT’, ‘EBIT I’, ‘Adjusted operating result’, ‘Adjusted operating profit’, ‘Adjusted profit from operating activities’, ‘Adjusted net operating profit’, ‘Adjusted profit from ordinary activities’, ‘Adjusted operational earnings’, ‘Operating income before extraordinary effects’, ‘Operating EBIT before extraordinary expenses’, etc. Occasionally, firms label an APM still ‘EBIT’ even if they adjusted the number disclosed as line item in the profit and loss statement.

18 We thank an anonymous referee for suggesting this analysis.

19 All p-values are two-tailed.

20 Since our sample consists of a limited number of loss firms (firm-year observations = 31), we leave further empirical testing of this relationship as an avenue for future research.

21 We chose the industry super sector Consumer Goods as our baseline industry dummy, as this industry sector shows an average disclosure quality across all sectors. Thus, there are industry sectors with a lower or higher disclosure quality (see Table 8).

22 However, the findings should be interpreted with caution for several reasons. First, as shown in the descriptive statistics, the number of sample firms in the utilities sector only amounts to six firms. However, in total this sector comprises just eight firms in Germany due to the specific characteristic of utility firms. Thus, our sample covers almost the entire sector population. Additionally, two utility firms have just been de-merged from two other sample firms in 2016, with those firms still holding major stakes. Thus, the independence of the observations may be questionable. Second, Watson et al. (2002) discuss several aspects that could explain industry variations in disclosure such as historical and bandwagon effects, public scrutiny, regulation and legitimisation. Stringent regulations are especially relevant for the utility sector.

23 We thank two anonymous referees for pointing out these arguments.

24 We also perform the random-effects regression using the log of the odds ratios of DQI as the dependent variable in order to address the possible violation of the normality assumption (Al-Shammari et al., 2008; Cooke, 1998; Inchausti, 1997). We obtain qualitatively identical results with slightly larger p-values.