Deviations from the Mandatory Adoption of IFRS in Europe? Why Non-Adoption Does Not Mean Non-Compliance

Abstract

Using Worldscope data, Pownall and Wieczynska (2018) [Deviations from the mandatory adoption of IFRS in the European Union: Implementation, enforcement, incentives, and compliance. Contemporary Accounting Research, 35(2), 1029–1066] show that there was extensive non-adoption of International Financial Reporting Standards (IFRS) by European listed firms in 2005–2012. They also suggest that much of this non-adoption is illegal. This would constitute a serious failing by European firms, auditors and regulators. However, in this note we show that there is no non-compliance with mandatory requirements to use IFRS in Germany and the UK in various years including 2012 and 2020. We also find no non-compliance in Austria and Portugal, two countries with less highly regarded enforcement. The reasons for legitimate non-adoption vary by country and by year. Our analysis is based on hand-collected data and it documents that the Worldscope database contains many errors in various fields. Therefore, we provide researchers with a hand-collected database for German and UK firms for the period 2005–2020, which can be used to supplement Worldscope. The database includes information on country of incorporation, whether consolidated financial statements are prepared, the accounting standards applied, and whether the firm is listed on an EU-regulated market. We use the database to provide some evidence on voluntary adoption of IFRS.

1. Introduction

Pownall and Wieczynska (2018, hereafter ‘PW’) investigate non-adoption of IFRS by listed firms in European Union (EU) countries for four different years (2005, 2007, 2009 and 2012). PW explain that an EU Regulation requires the use of IFRS for the consolidated statements of firms listed on EU-regulated markets. PW examine EU-listed firms as included in the Worldscope database, and they find a high level of non-adoption of IFRS. Specifically: ‘We conclude that many EU firms do not use IFRS; that some firms exploited definitions, exemptions, and deferrals to avoid adopting IFRS while some firms simply failed to comply with the regulation’ (PW, p.1029).

This is a useful reminder to us all not to exaggerate the spread of the application of IFRS. However, PW go much further, by expanding the ‘failed to comply’ part of their conclusion to suggest that most of the non-adoption was illegal: ‘the majority of non-adopters presented consolidated financial statements and were traded on EU-regulated markets’ (p.1060). Indeed, readers might infer that PW regard this widespread non-compliance as their most important finding because their paper’s title refers to ‘Deviations from the mandatory adoption of IFRS’ rather than merely to non-adoption.

PW particularly comment on German and UK listed firms, which are their largest country samples. They report average non-compliance of over 9% for firm-years in Germany and over 28% for the UK (see our Table 1). Since all such firms presented audited reports which should be subject to national monitoring and enforcement, PW’s findings (if accurate) would constitute a major indictment of the behaviour of EU firms, auditors and regulators. Indeed, PW suggest that, strictly speaking, because of poor monitoring and enforcement, ‘EU IFRS adoption was not mandatory’ (p.1030).

Table 1. PW’s ‘Mandatory IFRS’ total samples for all their years.

	All countries	Germany	UK	Austria	Portugal
Firm-years required to use IFRS, according to PW	29,652	3,870	5,976	587	350
Firm-years recorded by PW as required to use IFRS but having NoIFRS	3,027	382	1,699	45	20
Implied non-compliance	10.2%	9.9%	28.4%	7.7%	5.7%

Notes: The numbers in the first two rows are from the ‘Mandatory IFRS’ columns of Panel D of PW’s Table 2. The first row reports totals of the ‘IFRS’ and ‘NoIFRS’ columns. Using these numbers, the authors calculated the percentages in the third row.

However, we show that PW’s findings reflect problems with their data, i.e. Worldscope. We use samples of four countries for various years. For PW’s most recent year studied (2012), we use a similar sample of firms to PW’s, and we hand-collect data for it from national registries, stock exchanges and firms’ annual reports. We show that there was no such non-compliance by German and UK firms, which make up 41% of PW’s sample firms for that year and 45% overall.¹ For the UK firms, which comprise PW’s largest sample, we also check compliance for one of PW’s earlier years: 2007. We then update the study by examining the German and UK firms for 2020. In all cases, we find no non-compliance.

The literature (see Section 2.1) regards Germany and the UK as having comparatively high levels of enforcement. Thus, we were also interested to see if there was any non-compliance in countries regarded as having lower levels of enforcement. We investigate this for firms in Austria and Portugal, but again find no non-compliance.

As will be explained in Section 2, PW’s paper has already been extensively cited, including several references to the percentages of non-compliance.² The first contribution of this note is to document in detail the legal grounds on which listed European firms do not apply IFRS, thereby ruling out non-compliance as a reason, at least for the several countries and periods we examine. Next, we show how Worldscope data do not report the institutional setting of European firms in sufficient detail or quality to determine whether the firms are legally required to report under IFRS. The most frequent error is that Worldscope often incorrectly shows that a firm publishes consolidated statements. Therefore, we have created a hand-collected database for German and UK firms covered by Worldscope for the period 2005–2020, which is available for researchers. The database includes information on country of incorporation, whether consolidated financial statements are prepared, the accounting standards applied, and whether the firm is listed on an EU-regulated market. Finally, we use the database to provide some evidence on voluntary adoption of IFRS.

This note proceeds as follows. Section 2 looks at the institutional setting and the literature of relevance to this paper. In Section 3, we explain our sources of data, confirming that our data are similar to PW’s. We then discuss the various explanations of why non-adoption of IFRS does not necessarily mean non-compliance with regulations. In Section 4, we turn to our own measures of non-adoption and non-compliance with IFRS for various countries and years. In Section 5, we look in more detail at the types of error in Worldscope, using Germany in 2020 as a case study. As part of this, we present data on voluntary adoption of IFRS. Section 6 concludes.

2. Institutional Setting and Relevant Literature

In this section, we examine three background issues: the institutional setting of IFRS reporting in Europe, including enforcement quality; PW’s definitions of non-compliance, and citations to PW on this; and literature on the quality of data in financial reporting databases.

2.1 Institutional Setting

2.1.1 EU Regulation

Under Article 4 of the EU’s Regulation 1606/2002 (‘IAS Regulation’), use of EU-endorsed IFRS (hereafter ‘IFRS’) is required for the consolidated statements of EU³ firms listed on EU-regulated markets. In some EU countries, IFRS is required for firms listed on some other markets. The rules are complex, including temporary exemptions which ran until 2007. The rules vary by country because of choices allowed in the Regulation. PW explain this in detail (e.g. on page 1032). In addition, readers can consult André (2017) and country-specific papers in the same special issue of the same journal. The UK formally left the EU on 31 January 2020, but all regulations continued to apply until 31 December 2020. Even after that, the key parts of the EU Regulation on IFRS have been retained in UK legislation.⁴

As noted above, the IAS Regulation applies to firms listed on ‘EU-regulated markets’. These can be contrasted with firms listed on ‘exchange-regulated markets’ (or ‘unregulated markets’) to which the Regulation does not apply, and which are therefore not required to follow IFRS unless the particular country or exchange mandates it. Any one stock exchange can host both types of market. Byard et al. (2021) found that more than half of the initial public offerings (IPOs) in the EU in 2005/7 were made on exchange-regulated markets, and that most of the firms chose to report under local GAAPs rather than IFRS. Taking account of these points led to importantly different results from previous studies of IPO under-pricing. Pierk (2018) studied IPOs on exchange-regulated markets in five EU countries in 2005–2013. Less than 20% of the firms chose to report under IFRS. Choosing IFRS was associated with, inter alia, firm size and a subsequent move to an EU-regulated market. Hitz et al. (2020) studied ‘downlisting’ from EU-regulated to exchange-regulated markets in Germany. They found that downlisting is associated with previous censures for poor reporting and with subsequent falls in share price. Nevertheless, about half of the downlisting firms volunteered to continue using IFRS.

An important part of the regulatory context in the EU, unlike that in the USA for example, is that financial reporting is also widely required from unlisted firms. Laws also require the publication of the unconsolidated (‘stand-alone’) statements of firms, including those of parent firms and subsidiaries. Beyond the above mandated scope, IFRS is allowed (or in a few cases required) for some of this reporting, as explained in PW’s Table 1. Beuselinck et al. (2023) examine this in detail for unlisted firms.

2.1.2 Enforcement quality

Researchers have concluded that the benefits of adopting IFRS in a country rely on strong enforcement (e.g. Daske et al., 2008; Christensen et al., 2013). Some have tried to quantify enforcement quality. Hope (2003) compiled a country enforcement index comprising five elements: audit spending, judicial efficiency, rule of law, insider trading laws and shareholder protection. He calculated the index for 22 countries. However, especially in the EU, enforcement has improved since then in several countries, to coincide with the adoption of IFRS in 2005 (Hitz et al., 2012). Brown et al. (2014) note that researchers have used a range of legal system proxies to capture enforcement quality, and they suggest that the proxies should focus on factors that directly affect compliance in the field of financial reporting. For 51 countries, Brown et al. calculate, for 2002, 2005 and 2008, indices of the quality of the audit working environment and the degree of accounting enforcement activity by independent bodies. Preiato et al. (2015) apply these indices to analysts’ forecasts, using 2003–2009 data on firms from 39 countries, showing that high scores on the indices are associated with better forecasts.

The countries with the two largest samples in PW (Germany and the UK) score highly in the Brown et al. (2014) measures of enforcement quality.⁵ Amongst the weakest scores for western Europe were those for Austria and Portugal.⁶ We choose these four countries for our detailed investigations below.

Our working hypothesis is that, at least in Germany and the UK, firms which publish audited annual reports are most unlikely to fail to use IFRS if regulations require them to do so. Although weaknesses in auditing are well documented (Cahan et al., 2011; Francis, 2011), they usually relate to problems which are not immediately apparent to any observer at the date of the audit report. Thus, we would not expect to see clean audit reports if there were cases of such blatant non-compliance in countries with well-regarded enforcement regimes. We suggest that the apparent non-compliance with requirements reported by PW is caused by problems with the data. We analyse the various possible explanations in Section 3 below.

2.2 Non-Adoption and Non-Compliance According to PW, and Citations of PW’s Findings

As explained in Section 1, PW report extensive non-adoption of IFRS, and that most of it contravened regulations. PW imply⁷ that previous researchers have misled readers by stating that the EU mandates IFRS for all firms or for all listed firms, rather than only mandating it for EU firms listed on EU-regulated markets. PW offer no evidence that previous writers had referred to mandatory use of IFRS for all EU firms. The distinction is important because only a very small proportion of EU firms is listed.⁸ PW state that they investigated the ‘general assumption that all EU firms adopted IFRS’ (p.1037), and they framed their first research question as: ‘Did all or most EU firms adopt IFRS in 2005 as commonly assumed?’ (p.1037). However, we must interpret PW’s real research question as referring to listed companies because their data are restricted to listed companies.

PW at various stages discuss ‘non-compliance’ and ‘deviations from the mandatory adoption’. These terms seem to us to mean the same thing, but PW define non-compliance (p.1030, fn. 3) as follows:

Although de facto non-compliance may be argued to be a characteristic of firms which were specifically required to adopt IFRS, we focus on perceived non-compliance as we compare IFRS adoption incidence to the assumption that all EU firms or all EU firms preparing consolidated reports follow IFRS.

This definition of non-compliance appears to equate it with non-adoption. It may explain why PW’s second research question (p.1037) is ‘Why did non-IFRS users fail to comply with the requirement to adopt IFRS?’ although the research question then refers to valid types of permission for non-adoption rather than to the normal concept of failing to comply. Also, when discussing non-adoption (including legitimate non-adoption) by UK firms, their heading is ‘Non-compliance among British firms’ (p.1059).

However, in their Abstract, Introduction and Conclusions, PW say that some firms ‘simply failed to comply with the regulation’ (pp. 1029, 1032) and ‘failed to comply with mandatory IFRS adoption’ (p.1060). These expressions presumably use ‘comply’ with its normal meaning. From here on, we assume that PW intended this.

Even when discussing legitimate non-adoption of IFRS, PW suggest that firms ‘exploited definitions, exemptions, and deferrals’ (p.1029), and that firms ‘failed to adopt IFRS’ (p.1031). However, for many firms, the non-use of IFRS is both legal and rational (e.g. to avoid costs and to maintain policies over time). For example, it is rational for many German firms listed on unregulated markets to use German GAAP rather than exercising the option to use IFRS. This is because the parent company and its German subsidiaries are required by German law to prepare financial statements under German GAAP, so to volunteer to prepare IFRS consolidated statements would be an expensive exercise.

As noted in Section 1, many papers cite PW. Several of these refer to PW’s numerical findings on non-adoption.⁹ More importantly for our paper, many prior researchers specifically refer to the non-compliance alleged by PW. For example, Gordon et al. (2019, p. 10) say that PW found that ‘Some companies that should have adopted IFRS stayed with domestic GAAP’. Song and Trimble (2022, fn. 4) note that PW had found ‘as of 2012, a substantial 17% of listed EU firms still do not report under IFRS despite the mandate’; and Griffin et al. (2022, p. 37) refer to the same percentage.¹⁰ Becker et al. (2021, Section 4.1.2) say that PW ‘document that there is a substantial percentage of EU firms that are noncompliant and stick to local GAAP practices, instead of applying IFRS’; and Cereola and Dynowska (2022, pp. 4 and 6) also refer to PW’s percentages of non-compliance. Wang and Yu (2021, p. 243) repeat PW’s assertion that some firms ‘simply failed to comply with the regulation’, as do André et al. (2021, fn. 1). Downes et al. (2019, p. 45) refer to PW concerning a lack of power to enforce standards in the EU. Yamani et al. (2021, p. 529) also refer to PW as a reference for non-compliance and weak enforcement.

2.3 The Quality of Information in Databases

García Lara et al. (2006) show that the choice of database matters in empirical research. For 14 European countries, they compared the samples of firms in major databases such as Worldscope, finding important differences. Other previous research has found problems with the quality of information in such databases. For example, Daske et al. (2013, p. 500) identify many ‘inconsistencies’ in Worldscope’s ‘accounting standards followed’ (data field 07536). They hand-collected a large amount of data in order to address this problem. Nobes and Stadler (2018) provide an overview of various Worldscope data problems. They discuss cases where there is missing data in some fields for many firms, and examples of fields for which Worldscope only provides current information (unlike other fields for which it has a time series). They also discuss cases of erroneous data and misleading data.

3. Ways in Which a Firm Can Comply with Regulations Even Though Worldscope Records that it Does Not Publish IFRS Statements

3.1 Introduction and Data Requirements

In Section 2.1, we established a working hypothesis of no non-compliance (at least by German and UK firms) with the requirement to use IFRS. Of course, we (and PW) are not checking the quality of compliance, but only whether the firm (or its auditors) report the application of IFRS in the annual report.

To discover ‘deviations from the mandatory adoption of IFRS’, four types of data are needed for each firm analysed: (1) country of incorporation; (2) whether consolidated financial statements are prepared; (3) accounting standards applied; and (4) whether the firm is listed on an EU-regulated market (for either equity or debt). PW use Worldscope which only contains data on (1) to (3). Further, we suspected from the previous literature that Worldscope contains many errors, and we report on this below and further in Section 5. Therefore, we have hand-collected data on (1) to (4). Details are provided in Appendix 1.

To check that we have similar data to PW, we replicated PW’s data for 2012, using Worldscope data relating to firms’ reports for periods ending on 31 December 2012 or nearest after (see PW p.1042; hereafter ‘2012/13’), and imposing similar types of conditions as PW did (see our Appendix 2). This led to similar samples for Germany and the UK as those used by PW, as shown in the first row of our Table 2. The number of firms without IFRS statements (NoIFRS) is also similar for PW and our data, as shown in the second row. We are therefore confident that we are analysing substantially the same data as in PW.¹¹

Table 2. Replication of PW’s 2012/13 German and UK samples.

	All countries	Germany		UK
	PW^a	PW^a	Our data	PW^a	Our data
Firms with all required data	6,306	738	789^b	1,837	1,694^b
Firms recorded as NoIFRS	1,068	224	248^c	254	241^c

Notes: a = From the ‘Total’, ‘DEU’ and ‘GBR’ rows of the 2012 ‘Firm-year observations’ columns of Panel D of PW’s Table 2. The upper figures are totals of the ‘IFRS’ and ‘NoIFRS’ columns. b = See Appendix 2. c = We define NoIFRS as Worldscope’s WC07536 (accounting standards followed) being neither ‘IFRS’ (code 23) nor ‘International standards’ (code 02). There was only one case with ‘International standards’ in our data.

3.2 Explanations for Legitimate Non-Adoption of IFRS

We seek to explain why many firms are showing in Worldscope as not publishing IFRS statements. In principle, there could be three broad categories of explanation apart from non-compliance: (a) Worldscope records (correctly) that the firm does not produce consolidated statements, so it falls outside of the requirement to use IFRS, (b) Worldscope contains accounting data but the firm was not listed in the relevant year on a market which requires IFRS, or (c) there is an error of some sort in Worldscope. The first column of Table 3 expands (b) and (c), which leads to a list of six potential explanations for ‘NoIFRS’ other than non-compliance with requirements. In the sub-sections below, we give detail about these six explanations. A firm could fit into two or more of them, such as both not producing consolidated statements (because it has no subsidiaries or is otherwise exempt)¹² and not being listed on a market which requires IFRS. In such a case, we record only the first of the two explanations. That is, the explanations in Table 3 are in a hierarchy, starting from the top.

Table 3. Analysis of Worldscope’s NoIFRS firms (by explanation for Worldscope recording the lack of IFRS statements).

	[I]	[II]	[III]	[IV]	[V]	[VI]	[VII]
	Germany 2012/13	UK 2012/13	Austria 2012/13	Portugal 2012/13	UK 2007/8	Germany 2020	UK 2020
Firms with all required data	789	1,694	87	55	2,234	652	1,490
Firms recorded as NoIFRS by Worldscope	248	241	8	6	403	214	177
Explanations of why reporting is legitimate:
[1] Firm not incorporated in the country (Worldscope error)	–	2	–	–	16	1	4
[2] Firm does not provide consolidated statements – Case 1: Worldscope records this (correctly)	53	125	2	–	291	8	–
[3] Firm does not provide consolidated statements – Case 2: Worldscope error	113	77	6	4	31	122	154
[4] Firm provides IFRS statements (but Worldscope records a different GAAP)	6	6	–	–	1	3	8
[5] Firm not listed in the year (although Worldscope contains accounting data)	4	2	–	–	4	7	2
[6] Firm not listed on a market which requires IFRS	72	29	–	2	60	73	9
Deviations from mandatory adoption	0	0	0	0	0	0	0

Notes: This table lists the numbers of firms which do not publish IFRS statements according to Worldscope in the respective year, by our explanations for this. ‘Firms with all required data’ follows PW’s ‘initial sample’ (see PW’s Table 2), see Appendix 2 for details. NoIFRS is defined in footnote (c) to our Table 2. For some firms, more than one explanation applies, e.g. a firm does not publish consolidated statements and is not listed on a market which requires IFRS. The explanations are presented as a hierarchy, so (apart from Explanation 1) the total for each category may be incomplete. For [1], see an example firm recorded in Appendix 3. For [2], we define ‘does not provide consolidated statements’ as Worldscope’s WC07531 (accounting method for long term investment greater than 50%) being ‘No consolidation, cost basis (parent company only)’ (code 06) or ‘Not applicable’ (code 12). No other relevant category was present in our data; in particular, there was no case of ‘No consolidation, equity basis (parent company only)’ (code 07). Most of the ‘Not applicable’ cases relate to investment trusts in the UK. For [4], Worldscope’s WC07536 records local GAAP or US GAAP.

Explanation 1: firm not incorporated in the country, although Worldscope records otherwise

As part of following PW’s sample selection (see Appendix 2), we had already excluded firms for which Worldscope’s data field WC06026 (nation) differs from the country analysed.¹³ The aim of this was to exclude firms that are not incorporated/domiciled in the respective country. However, the data in Worldscope’s ‘nation’ may be wrong or not fit for our purpose. The latter can occur because WC06026 is a ‘current’ data field. This means that the information is not available for each firm-year, because Worldscope only provides the current information. For example, it is possible that a firm that was incorporated in the UK in 2020 had been incorporated in another country in 2012. Therefore, the current data in Worldscope is not fit for purpose when analysing 2012. Another technical problem is that Worldscope’s ‘nation’ refers to ‘the country in which the corporate office of a company is located’ (Thomson Reuters, 2015, p. 551). This is not the relevant issue for accounting requirements, which relate to country of incorporation. Therefore, we collected the latter. However, the difference only affects a small number of firms, as we confirm for Germany in Section 5.

Explanation 2: firm does not provide consolidated statements – Case 1: Worldscope records this (correctly)

As the IAS Regulation restricts the mandatory application of IFRS to consolidated statements, when searching for non-complying firms, one should exclude those that do not provide such statements. PW attempt this by excluding firms for which Worldscope records the absence of consolidated statements. This approach works reasonably well in one direction, in that there are only a few cases where Worldscope misses the existence of consolidated statements.¹⁴

However, there are also cases where Worldscope contains the opposite error, which is the one that matters for our analysis: when Worldscope records that a firm provides consolidated financial statements but it does not. Consequently, we distinguish two types of explanation involving the absence of consolidated statements: ‘Case 1’ when Worldscope correctly records that a firm does not provide consolidated statements, and ‘Case 2’ when Worldscope fails to record this but instead erroneously records that a firm provides consolidated statements. ‘Case 2’ is the subject of Explanation 3 below. The notes to Table 3 explain how we investigated these cases.

There are several legitimate explanations for a lack of consolidated statements of a listed firm: (i) no subsidiaries, (ii) no material subsidiaries, (iii) the firm is itself a subsidiary, and it is either included in an IFRS consolidation or its non-group owners do not object to the lack of consolidated statements,¹⁵ or (iv) it is a certain type of investment entity.¹⁶

Explanation 3: firm does not provide consolidated statements – Case 2: Worldscope erroneously records that it does

PW report (p.1058) that they checked the quality of Wordscope data for some fields but they do not mention checking whether the firms really did produce consolidated statements. However, we checked, by hand-collecting data from firms’ annual reports, which we mainly access from the national registries.¹⁷ For many firms, information was only available in the national language. As Table 3 shows, we discovered that Worldscope contains a large number of errors in this field.

Explanation 4: firm provides IFRS statements, but Worldscope records a different GAAP

A further potential explanation for the data incorrectly implying non-compliance with the requirement to produce IFRS statements is that the database incorrectly records the GAAP applied. PW report (p.1058) that they found low error rates on this issue when they checked Worldscope’s data for 10 percent of firms, and we concur. Nevertheless, we found a few cases where Worldscope recorded the use of local GAAP or US GAAP even though the firm published IFRS statements. These cases are mostly errors in Worldscope, though some firms¹⁸ do indeed publish US GAAP statements (as recorded in Worldscope) but they also publish IFRS statements, thereby complying with the legal requirements.¹⁹ In these cases, Worldscope’s data is true but not fair.

Explanation 5: firm not listed in the year although Worldscope contains accounting data

This explanation applies where Worldscope contains accounting data for a firm for a particular year but the firm was not listed in that year.

Explanation 6: firm not listed on a market which requires IFRS

This final explanation concerns a firm not being listed on an EU market which falls within the EU’s definition of ‘regulated’ (or which otherwise requires the use of IFRS). It is complicated to investigate. PW were aware of the general issue, and they noted (p.1053) that the UK’s Alternative Investment Market (AIM) did not require IFRS until 2007. However, there are several other less-than-fully regulated markets in Germany and the UK which did not require IFRS in the various years investigated here.²⁰ Worldscope does not give enough detail to determine whether or not a firm’s listing is on a market which requires IFRS. To some extent, PW adjusted for this by using a list provided by the European Securities and Markets Authority (ESMA), but that list included only 4,885 of PW’s initial 8,107 firms. Our approach to identifying the firms listed on these ‘unregulated’ markets (for all firms remaining after all the above types of exclusion), was to check four sources: lists provided by the various unregulated markets, firms’ annual reports, firms’ websites, and ad hoc news reports (which in Germany often include listing information). Consistent with the first sentence of this paragraph, EU firms that are only listed on a US stock exchange are not required to use IFRS.

4. Empirical Analyses by Country and by Year

4.1 Non-Compliance in Germany and the UK in 2012/13

As noted above, PW measure non-adoption and non-compliance²¹ for four different years. We first focus on PW’s most recent year (2012/13) and on German and UK firms which, as noted, comprise 41% of PW’s sample.²² Our Table 2 shows the sample sizes and the amount of non-adoption of IFRS for 2012/13 according to Worldscope. As the ‘PW’ columns in Table 2 show, Germany and the UK make up 45% of the non-adopting firms in their sample for 2012/13 (478 firms out of 1,068). For the separate countries, PW do not report how much of this involves non-compliance with mandatory use of IFRS, but they report that non-compliance was 8.2% for European countries taken together in 2012/13.²³ For their four sample years taken together, PW report that Germany and the UK were home to 69% of the alleged non-compliance (2,081 non-compliant firm-years out of 3,027; see the second row of our Table 1).

We examined the six potential explanations for why this does not involve non-compliance, taking the full 2012/13 samples of ‘NoIFRS’ listed firms for Germany and the UK from the second row of Table 2. In Appendix 3, we give examples of firms from each country which fit each of the six explanations. Explanations 2 and 6 were partially adjusted for in PW’s data but the other four were not. As may be seen from Table 3 (Columns [I] and [II]), in total the six explanations cover all the firms for which Worldscope records the lack of IFRS statements. We thus conclude that non-compliance with requirements to use IFRS did not exist for German and UK firms in 2012/13.

4.2 Non-Compliance in Austria and Portugal in 2012/13

As discussed in Section 2.1, we are interested to see if there was non-compliance with regulations in countries regarded as having lower levels of enforcement, and we choose Austria and Portugal. Table 1 shows the number of firm-years in PW’s data for those two countries, with the second row showing the alleged amount of non-compliance.

We repeated, for these two countries, the exercise explained above. That is, we started with the Worldscope data for 2012/13 for the two countries. The second row of Table 3 (Columns [III] and [IV]) shows the numbers of Austrian and Portuguese firms not using IFRS for their reports. Our number of NoIFRS firms for Austria (8) and Portugal (6) is very similar to PW’s (7 and 5, respectively). There were legitimate reasons for non-adoption in all cases, as shown in the rest of Table 3. In Appendix 3, we show the names of all the firms of the two countries, with the explanations. Once more, even for these two countries which score lower in previous compliance research, we confirm that there were no cases of non-compliance with the requirement to use IFRS.

4.3 Non-Compliance in the UK in 2007/8 and in Germany and the UK in 2020

It might be expected that compliance with a regulation would be at its weakest in the early years of its being in force. To check whether PW’s claims of non-compliance might have had some foundation earlier than in their most recent year, we checked compliance for one of their earlier years: 2007/8.²⁴ We did this for the UK, for which PW had the largest sample, the largest degree of alleged non-compliance, and for which they especially commented on non-compliance. Another reason for choosing to study the UK rather than Germany (or both) for this year is that 2007/8 was not an early year of IFRS adoption for many German firms because German law had allowed adoption from 1998,²⁵ whereas IFRS was not allowed for statutory reporting in the UK until 2005.

On the other hand, it is possible that firms and auditors become complacent over time, so we also up-dated our analysis by looking at the 2020 annual reports of all the German and UK listed firms in Worldscope’s files. Specifically, 2020 refers to all accounting periods ending between 6 January 2020 and 5 January 2021.

Table 3 (Columns [V] to [VII]) reports our findings relating to 2007/8 and 2020, using the same headings of ‘explanation’ as in the sections above. Once more, despite the very large numbers of firms involved, we find no cases of non-compliance with regulations in either country in either the earlier year or the later year. As may be seen, the most common explanation for apparent non-compliance for both German and UK firms in 2020 (and for all four countries in 2012/13) was one type of error in Worldscope: it recorded the existence of consolidated statements when there were none. In the UK, such firms were mostly investment trusts. The prevalence of this error has increased over time. For both Germany and the UK, there were also many cases of the firm not being listed on an EU-regulated market.

5. Our Database and What it Reveals About Worldscope Errors and Voluntary Adoptions of IFRS

5.1 Our Database

Having discovered many errors in Worldscope, we created a database for German and UK firms covered by Worldscope for the period 2005–2020. Our database contains data for each firm on the four matters discussed in Section 3.1: country of incorporation, consolidated or unconsolidated statements, the accounting standards applied, and whether listed on an EU-regulated market. Additionally, we collected data on the primary listing, e.g. Prime Standard at the Frankfurt Stock Exchange.²⁶ See Appendix 1 for detailed information about our database. The database is available to readers of this journal.

In the sub-section below, we report on various aspects of our data relating to Germany in 2020. First, a comparison of our data with Worldscope’s reveals its errors in more detail than discussed above. Secondly, we report further on compliance with regulations and on voluntary adoptions of IFRS.

5.2 Findings from the German Data for 2020

Because of its hierarchy approach, Table 3 does not reveal all of the errors in Worldscope. Table 4 compares Wordscope’s data on Germany with our data. This shows that Worldscope contains many errors, especially (as expected from Section 4) regarding whether a firm prepares consolidated financial statements. We conclude that Worldscope is no longer useful in helping to identify which firms provide only unconsolidated financial statements. Worldscope is also not useful for identifying the regulatory status of the stock market on which a firm is listed, because Worldscope does not include data on that issue. Even Table 4 somewhat understates the amount of error in Worldscope. For example, on ‘Accounting standards’, Worldscope shows that 477 firms used IFRS, whereas we count 471 (‘IFRS’ and ‘IFRS + Local’). However, Worldscope misrecorded more than six firms: it recorded eight firms as ‘IFRS’ although they used Local GAAP and three firms as ‘Local GAAP’ although they used IFRS. Therefore, for this field, Worldscope is wrong in 11 out of 697 firms (1.6%).²⁷ This percentage of error might be acceptable for some research questions, though not for detecting whether there are any ‘deviations from mandatory adoption’.

Table 4. Comparison of Worldscope data and our data: Germany 2020.

	Worldscope	Our database
Firms with data for 2020 in Worldscope	697	697
Country of incorporation:
Germany	677	673
Other	20	24
Consolidated financial statements?
Yes	687	546
No	10	150
NA (firm did not publish annual report)	–	1
Accounting standards:
IFRS	477	461
IFRS + Local	–	10
Local GAAP	216	223
US GAAP	2	–
US GAAP + IFRS	–	2
OTHER	2	–
NA (firm did not publish annual report)	–	1
Listing status:
EU-regulated market	NI	395
EU unregulated market	NI	226
Non-EU listed	NI	8
Not listed but capital-market oriented (IFRS required)	NI	10
Not listed and not capital-market oriented	NI	58
Primary listing:
EU-regulated market:
Frankfurt Prime Standard	NI	281
Frankfurt General Standard	NI	88
Regulated market at other stock exchange	NI	26
EU unregulated market:
Frankfurt Scale (premium segment of open market)	NI	43
Frankfurt Basic Board (basic open market)	NI	55
Premium segment of open market at other stock exchange	NI	72
Basic open market at other stock exchange	NI	56

Notes: We define ‘Firms with data for 2020 in Worldscope’ as firms with an entry in Worldscope’s data field WC05350 (fiscal period end date) in the period from 6 January 2020 – 5 January 2021, inclusive. For the column ‘Worldscope’, we use the following Worldscope data fields: WC06026 (nation) for ‘Country of incorporation’, WC07531 (accounting method for long term investment greater than 50%) for ‘Consolidated financial statements?’ and WC07536 (accounting standards followed) for ‘Accounting standards’. Our database records two accounting standards (and consolidation information) in the following cases: (1) A firm prepares consolidated financial statements or (if the firm does not prepare consolidated financial statements) unconsolidated financial statements using GAAP that is different from the GAAP used for its statutory financial statements; (2) Worldscope’s data do not correspond to the consolidated financial statements although such statements exist. For both cases, the first accounting standards (and consolidation information) in our database correspond to the data recorded in Worldscope; see Appendix A1.3 for more details. NA = not available. NI = no information.

Using our database, we can also report on two other matters, in the context of German firms in 2020. First, we can confirm that there was no non-compliance among the few firms which Worldscope incorrectly identified as filing IFRS statements (that is, we found that the firms did not use IFRS, but they were not required to use IFRS). Secondly, we can report on the amount of voluntary adoption of IFRS by German firms in 2020. Table 5 shows the analysis. Panel A of Table 5 reports that 39% (i.e. 50 out of 127) of the firms that prepared consolidated financial statements and were listed on an EU unregulated market voluntarily used IFRS. Additionally, Panel B of Table 5 reports that 9 firms voluntarily prepared unconsolidated IFRS financial statements alongside their statutory German GAAP financial statements. For German firms, this is surprising, given that local GAAP statements must be prepared for tax purposes.²⁸ Taken together with our above findings about compliance with requirements to use IFRS, this further undermines the impression given by PW of European reluctance to use IFRS (see our Section 2.2).

Table 5. Analysis of German firms in 2020.

Listing status	IFRS	IFRS + Local	Local	US GAAP + IFRS	Total
Panel A: Consolidated financial statements
EU-regulated market	352	–	–	–	352
EU unregulated market	50†	–	77	–	127
Non-EU listed	4^a	–	–	1†	5
Not listed but capital-market oriented	10	–	–	–	10
Not listed and not capital-market oriented	26^b	–	7	–	33
Total	442	-	84	1	527
Panel B: Unconsolidated financial statements
EU-regulated market	–	6†	24	–	30
EU unregulated market	–	2†	92	–	94
Non-EU listed	–	–	–	–	-
Not listed but capital-market oriented	–	–	–	–	-
Not listed and not capital-market oriented	–	1†	20	–	21
Total	-	9	136	-	145

Notes: This table analyses the German firms from Table 4, i.e. those incorporated in Germany in 2020 (673 firms) excluding the one firm that did not publish an annual report. Panel A reports statistics of listing status and accounting standards for firms preparing consolidated financial statements. Panel B reports the respective statistics for firms which only prepare unconsolidated financial statements. † denotes voluntary IFRS adopters. a = These firms provide IFRS financial statements for their US listing (instead of providing US GAAP statements), so we do not consider them to be voluntary IFRS adopters. b = Not all of these firms should be considered as voluntary IFRS adopters because some firms had their IPO after 2020 and needed to prepare IFRS financial statements for their prospectus.

6. Conclusions

PW usefully remind us that there is extensive non-adoption of IFRS by listed EU firms. They go further by suggesting that most of this non-adoption involved non-compliance with requirements because: ‘the majority of non-adopters presented consolidated financial statements and were traded on EU-regulated markets’ (p.1060). To us, such obvious illegality in published audited financial statements seemed implausible, and we formed a hypothesis of no non-compliance, which we then tried to falsify. First, we looked at the two largest country samples of PW’s most recent year: German and UK firms in 2012/13. These firms comprise 41% of PW’s sample and 45% of the non-adopters in that year; and those countries’ firms comprise 69% of the alleged non-compliance over PW’s whole period. We examined the annual reports and the listing status of the firms, finding that none of the non-adopters fits PW’s above description, so that none of the firms were contravening the requirements.²⁹

We then tried more exacting tests, by looking at compliance in 2012/13 in two countries with less well-regarded enforcement regimes (Austria and Portugal) and then at compliance in an earlier year of IFRS adoption in the country with the largest sample and the largest amount of alleged non-compliance (the UK in 2007/8). We then updated the examination of German and UK firms to 2020. In all cases, we found no non-compliance. An interesting feature of our results is that the importance of the various explanations for legal non-adoption varies by country and by year. Generally, the most important reason is that the firm does not prepare consolidated statements, but there are also many firms on unregulated markets in some of our countries. Future research could investigate which reasons allowing non-adoption apply in other countries and other years.

Our finding of full compliance with requirements is relevant for readers of PW’s paper entitled ‘Deviations from the mandatory adoption of IFRS … ’ because we have shown that, for four countries (including the two largest country samples) and for three different years, there was no deviation from mandatory adoption. We thus find no support for PW’s claim that ‘the majority of non-adopters’ (p.1060) failed to comply with requirements or for their suggestion that IFRS is not strictly mandatory in the EU because of poor monitoring and enforcement (p.1030).

Our findings further imply that PW’s discussions of potential problems with enforcement regimes (e.g. pp. 1045, 1053) are not relevant for this issue because there has been no lack of enforcement. PW note (p.1036) that, of 197 enforcement cases published by ESMA, there were no cases about lack of adoption of IFRS. They also find that ‘for mandatory adopters none of the enforcement variables are significantly related to non-adoption’ (p.1054). Eventually, PW conclude that ‘enforcement has only a small role to play’ (p.1057). These points are not surprising: weak enforcement does not explain the non-adoption of IFRS because (for all our samples) the non-adoption is not non-compliance.

PW are in the mainstream in using a database such as Worldscope, and they are alert to potential data problems. They suggest (p.1058) that Worldscope data of 2009 onwards ‘are quite accurate and appropriate for use in research addressing questions based on accounting standard choice’. However, such confidence should not be placed in all fields. Most importantly, we discovered a high rate of error in Worldscope’s field that records the existence of consolidated statements. Researchers who use it should beware. There are other sources that might provide reliable data and, thus, serve as alternatives to manual data collection (e.g. Capital IQ, Orbis, Compustat Global). Future research could evaluate the reliability of these.

As part of our work, we provide researchers with a hand-collected database for German and UK firms covered by Worldscope for the period 2005–2020. The database includes information on country of incorporation, whether consolidated financial statements are prepared, the accounting standards applied, and whether the firm is listed on an EU-regulated market. We encourage the accounting community to build on this by contributing to and coordinating a joint data collection effort. This would add further countries. The database could be used to check the reliability of the commercial sources. In addition to use in studies of non-adoption, the database should help more generally by distinguishing between listings on regulated and unregulated markets.

Supplemental Data and Research Materials

Our database will be publicly available in the following GitHub repository: https://github.com/christian-stadler/ifrs-database. The data used in the paper is available under the following DOI: 10.5281/zenodo.7837457.

Disclosure statement

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

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 See the ‘initial sample’ firms in Panel A of PW’s Table 2 in the column headed ‘2005–12 percent’.

2 Google Scholar reported 50 citations (to the published and pre-published versions of the paper) on 13 June 2022. Section 2.2 of our paper looks at the papers which refer to non-compliance.

3 Strictly speaking, the scope is the somewhat larger group of countries called the European Economic Area, but we will refer to the EU throughout.

4 See Amendments to UK and Republic of Ireland Accounting Standards. UK Exit from the European Union, Financial Reporting Council, December 2020; available at: https://www.frc.org.uk/getattachment/8214be0e-d10e-440c-9406-5be7d049aefb/Amendments-to-UK-and-RoI-accounting-standards-UK-exit-from-the-EU-(Dec-2020).pdf (accessed 16.7.2022).

5 Brown et al.’s Table 5 shows the UK with the highest score (54) of any IFRS-adopting country for 2008. The four large western European members of the EU, including Germany (at 44), have very similar high scores.

6 Austria scores 27 and Portugal 29 for 2008.

7 See the first two sentences of PW’s paper (p.1030).

8 Let us take the examples of Germany and the UK. Public companies, i.e. public limited companies (PLCs) in the UK and most commonly Aktiengesellschaften (AGs) in Germany, are allowed to be listed but do not need to be. In Germany, there were about 11,000 AGs and 864,000 private companies (GmbHs) in 2018 (source: data on ‘Turnover tax (assessments) by legal forms’ on the website of the Federal Statistical Office of Germany). In the UK, there were about 7,000 PLCs and 4,022,000 private companies in 2018 (Companies House, 2018).

9 Such as, Chen et al. (2015, p. 80), Enomoto et al. (2018, p. 190), and Downes et al. (2018, p. 383).

10 To be more exact, PW (p.1031) imply that 87% (of these 17%) did not comply, because the remaining 13% did not produce consolidated statements.

11 We downloaded our data on 29 September 2021. We are most grateful to Maria Wieczynska for replying to our request for information, and telling us that the PW data was downloaded in 2013 (e-mail from M. Wieczynska to the authors on 30 July 2019).

12 As shown in Table 3, this is a prevalent UK explanation. Apart from the lack of subsidiaries, further exemptions in the UK at the time were set out in FRS 2 Accounting for Subsidiary Undertakings, para. 21.

13 Incidentally, for the UK, this excludes several firms incorporated in Guernsey and Jersey, which were and are not part of the UK or of the EU.

14 Taking the example of German firms for 2012/13, we find four such firms which actually do provide consolidated statements.

15 Issue (iii) is regulated by IFRS 10, para. 4.

16 See IFRS 10, para.s 4B and 31.

17 The Bundesanzeiger for Germany and Companies House for the UK.

18 For 2012/13, the examples are Fresenius Medical Care (German sample) and Carnival (UK sample).

19 As a somewhat different example, ITN Nanovation (a German firm) initially published 2012 consolidated financial statements under local GAAP, explaining that a change in corporate structure and financial difficulties meant that preparing IFRS statements was an ‘intolerable economic burden’, with a note on this by the auditors. However, one year later they published an IFRS report for 2012. This could be regarded as late filing rather than lack of an IFRS report.

20 For Germany, this applies if a firm was only listed on Open Markets (Freiverkehr), which existed in Frankfurt, Berlin, Bremen, Dusseldorf, Hamburg, Hannover, Munich and Stuttgart. For the UK, examples are the ICAP Securities & Derivatives Exchange (ISDX) and Aquis Exchange.

21 In the normal sense of the term (i.e. in this case, departure from the requirements of EU and national regulations, p.1061).

22 45% overall, and 41% of the firms for 2012/13; see Panel A of PW’s Table 2. For the UK, many of these firms have year-ends in 2013, such as 31 March 2013.

23 PW refer to ‘above 8 percent’ (p.1043); and this can be derived from the 2012 ‘Mandatory’ numbers of Panel C of their Table 2.

24 We chose 2007/8 rather than 2005/6 because (as PW explain) an important exchange-regulated segment required the use of IFRS from 2007, and this makes it easier to compare 2007/8 with later years.

25 Under the Kapitalaufnahmeerleichterungsgesetz of 1998. A few German firms adopted IFRS even earlier by means of dual reporting.

26 As part of this work, we investigated Worldscope’s coverage of German firms listed in the Frankfurt Stock Exchange’s Prime Standard, General Standard and Scale on 31/12/2020. We find that Worldscope provides full coverage.

27 We do not consider the two cases where Worldscope recorded ‘OTHER’ as errors, because these relate to firms incorporated in Switzerland which used Swiss GAAP. Given that Worldscope includes these firms in its Germany list, it would not be appropriate to record ‘Local standards’.

28 For filing purposes in Germany, some large firms are allowed to deposit IFRS statements (HGB § 325 2(a)). For filing and tax purposes in the UK, IFRS is allowed.

29 With the caveat relating to one German firm in footnote 19. Also, in preparing our database, we found one further problem in a year not examined by PW. The 2009 financial statements of a German firm, PA Power Automation AG, did not apply IFRS when they should have done. The annual report (p.29) explains that the firm had moved to an unregulated market before the report was published, and that the expense of reporting under IFRS was therefore not justified. The auditor’s report notes the non-compliance.

Appendix 1.

Our Database

This Appendix provides detailed information about our database.

A1.1. Aims

The primary aim of our database is to provide accurate data to assess whether a firm covered by Worldscope should prepare financial statements under IFRS. This includes: (1) country of incorporation; (2) whether consolidated financial statements are prepared; (3) accounting standards applied; and (4) whether the firm is listed on an EU-regulated market (for either equity or debt).

The secondary aim of our database is to correct errors/problems in Worldscope.

A1.2. Coverage

Our database covers the following countries, firms and years:

Countries:

Germany and the UK.

A1.2.1. Firms:

Germany: All firms which Worldscope included in its ‘Germany’ list (Worldscope ‘Series’ WSCOPEBD) on 29 September 2021.

UK: All firms which Worldscope included in its ‘United Kingdom’ list (Worldscope ‘Series’ WSCOPEUK) on 29 September 2021 for which Worldscope also recorded ‘United Kingdom’ in its data field WC06026 (nation).

A1.2.2. Years:

2005–2020. Specifically, all accounting periods ending between 15 December 2005 and 5 January 2021 with data in Worldscope, i.e. an entry in Worldscope’s data field WC05350 (fiscal period end date).

We collect the earliest and the most recent available years in the sample period for all firms. We also collect interim years, which include (but are not limited to) the following:

1. Firm-years used in our empirical analysis in this paper (see Table 3).

2. Firm-years of changes: For any data we collect on an annual basis, if there is a difference between the earliest and the most recent available year, we collect data for the year of the change and the prior year.

3. Firm-years where the time-series of Worldscope’s data on consolidation (data field WC07531) or accounting standards (data field WC07536) suggest a change: If Worldscope correctly identifies a change, we collect the data for the year of the change and the prior year.

4. Firm-years where the time-series of the labels of the filings in the national registry indicates a possible change in consolidation (e.g. ‘Group of companies’ accounts’ versus ‘Full accounts’ in the UK registry): If there is a change, we collect the data for the year of the change and the prior year.

5. For all remaining firm-years, we infer that there is no change in the data between the firm-years we collected. For each data category, our database specifies which firm-years are inferred.

A1.3. Data Definitions and Sources

Table

Data	Scores	Sources	Collected for
CountryWorldscopeList	–	Worldscope country lists, e.g. Germany if included in Worldscope 'Series' WSCOPEBD	All years
CompanyNameOfficial	–	1 = national registry (default) 2 = annual report 3 = prospectus (recorded as shown in national registry) 4 = company website 11 = prospectus or other pre-IPO report 98 = as most recent available year	Most recent available year (which is applied to all other years)
ISIN	–	1 = primary exchange (default) 2 = annual report 3 = ad hoc news (precise source provided in notes field) 4 = company website 8 = other (details provided in notes field) 98 = as most recent available year	Most recent available year (which is applied to all other years)
WorldscopePermanentID	–	Worldscope data field WC06105 (to connect our database to Worldscope)	All years
AccountingPeriodEndDate	–	Worldscope data field WC05350 (to connect our database to Worldscope); we identify a few cases where Worldscope recorded a wrong date (this is documented in CompanyNameOfficialNotes)	All years
CountryOfIncorporation	–	1 = filing in national registry as domestic firm (default) 2 = annual report 4 = company website 9 = assumption (details provided in notes field) 11 = prospectus or other pre-IPO report 22 = annual report of later year 97 = as most recent available year unless evidence of change (for Germany only) 99 = inferred	Earliest available year; most recent available year; change year (including prior year); additional years; intermediate years (with no change) inferred
Consolidation	c = consolidated financial statements u = unconsolidated financial statements c + c etc (if two accounting standards are recorded) NA = not available	1 = annual report policy note or audit report (default) 9 = assumption (details provided in notes field) 11 = prospectus or other pre-IPO report 22 = annual report of later year 99 = inferred	Earliest available year; most recent available year; change year (including prior year); additional years; intermediate years (with no change) inferred
AccountingStandards	IFRS Local UK GAAP (for non-UK firms) US GAAP US GAAP + IFRS etc (for two accounting standards) NA = not available	1 = annual report policy note or audit report (default) 4 = company website 9 = assumption (details provided in notes field) 11 = prospectus or other pre-IPO report 22 = annual report of later year 99 = inferred	Earliest available year; most recent available year; change year (including prior year); additional years; intermediate years (with no change) inferred
ListingStatus	0 = not listed 1 = EU-regulated market 2 = EU unregulated market (except AIM) 3 = not listed but capital market oriented 9 = non-EU listed (e.g. NASDAQ) 22 = EU unregulated market – AIM	1 = primary exchange (default) 2 = annual report 3 = ad hoc news (precise source provided in notes) 4 = company website 8 = other (details provided in notes field) 21 = annual report of earlier year 22 = annual report of later year 99 = inferred	Earliest available year; most recent available year; change year (including prior year); additional years; intermediate years (with no change) inferred
ListingPrimary	e.g. FRA Prime Standard (for Frankfurt Stock Exchange Prime Standard)	1 = primary exchange (default) 2 = annual report 3 = ad hoc news (precise source provided in notes) 4 = company website	2020

Notes:

1. Each data category (except CountryWorldscopeList, WorldscopePermanentID and AccountingPeriodEndDate) has three fields: main data, source and notes; for example, AccountingStandards, AccountingStandardsSource and AccountingStandardsNotes. For some firm-years, the notes fields include information that explain the scoring or provide details (e.g. page number in the annual report where the relevant information has been obtained).

2. We access annual reports from the following sources: national registries (Bundesanzeiger for Germany and Companies House for the UK), company websites, Refinitiv Eikon, HV-Info.de, EDGAR, SEDAR and Investegate.

3. inferred = We did not collect the data but inferred that there is no change between the firm-years we collected.

4. Our database records two accounting standards (and consolidation information) in the following two cases: (1) A firm prepares consolidated financial statements or (if the firm does not prepare consolidated financial statements) unconsolidated financial statements using GAAP that is different from the GAAP used for its statutory financial statements; most commonly, this applies to firms that prepare US GAAP financial statements for US listing purposes and statutory financial statements using IFRS or Local GAAP; it also applies to firms that prepare unconsolidated IFRS financial statements for investor information purposes and statutory unconsolidated financial statements using Local GAAP (e.g. in Germany). (2) Worldscope’s data do not correspond to the consolidated financial statements although such statements exist. For both cases, the first accounting standards (and consolidation information) in our database correspond to the data recorded in Worldscope (which may contain errors, see A1.5 below). Therefore, our database includes for example ‘Local + IFRS’ and ‘IFRS + Local’; or ‘Local + Local’. For (1), generally we only report two accounting standards (and consolidation information) if Worldscope (for any year of the firm) reports GAAP that is different from the GAAP required for statutory financial statements. Therefore, there may be more firms that prepare US GAAP financial statements, but it is outside the scope of our database to fully document the use of US GAAP (see the aims in A1.1 above).

A1.4. Scoring Details

General

1. Before recording data that differs from Worldscope, we check that total assets from our data source matches total assets recorded in Worldscope. If the presentation currency in our data source differs from the local currency (i.e. € in Germany and £ in the UK), we convert the reported amount into the local currency using Worldscope’s data field WC18214 (exchange rate used in translating balance sheet). This is because Worldscope shows total assets of all firms in a national list (e.g. WSCOPEBD for Germany and WSCOPEUK for the UK, see A.1.2) in the local currency; the national lists are based on Worldscope’s data field WC06027 (nation code), which ‘represents the Country of Primary Listing for each set of financials the company files’ (Thomson Reuters, 2015, p.551). By ‘match’ we do not necessarily mean that total assets from the annual report and total assets in Worldscope are the same; this is because Worldscope excludes certain items from its reported total assets, including: deferred tax assets, reinsurance assets, negative goodwill and deficit not covered by equity.

2. If Worldscope’s fiscal period end date (data field WC05350) has an entry but total assets (data field WC02999) is missing, we only record data if we can find an annual report and the accounting standards are in line with Worldscope or (when Worldscope does not report accounting standards) if the accounting standards are consistent with the other years of the firm.

3. The German registry only contains data from 2006, so we use HV-Info.de to access earlier annual reports as a default.

4. For firms incorporated in Germany and the UK, we state when the annual report used is not from the national registry (in CompanyNameOfficialNotes, ConsolidationNotes or AccountingStandardsNotes). In Germany, this can occur for example when a firm prepares consolidated financial statements on a voluntary basis. We do not provide this information for the German 2005 firm-years because the German registry does not include 2005 data.

CompanyNameOfficial

1. If the most recent available year of a firm in our database ends before a firm’s initial public offering (IPO), then we record the IPO name from the prospectus. If the firm is included in the national registry (of Germany or the UK), then we record the IPO name from the prospectus as shown in the national registry.

2. For some firms that changed their name, the UK registry (Companies House) fails to record the official name used in the annual report. In such cases, we use the annual report as the data source and mention this in CompanyNameOfficialNotes.

3. Some prior names and (for the UK) company numbers are provided in CompanyNameOfficialNotes, especially when this is helpful to find the company in the national registry. For example, in the UK, this applies when a firm is subject to a scheme of arrangement that involves a reverse takeover because the financial statements of what is treated as the same firm in Worldscope are under two different company numbers in the national registry.

CountryOfIncorporation

1. When a firm files annual reports in a national registry, we record that they are incorporated in the respective country, except if the following evidence indicates the contrary: in the UK, if the company number starts with FC; in Germany, if the legal form is different from any German legal form (e.g. Aktiengesellschaft).

2. For all firm-years, including pre-IPO firm-years, we record the country of incorporation that applies at the accounting period end date.

Consolidation

1. We score ‘consolidated’ if the firm’s annual report states that consolidated financial statements or group financial statements have been prepared. The latter includes cases in which no subsidiaries are consolidated but subsidiaries, joint ventures or associated companies are included at fair value or using the equity method.

2. When the prior period of an annual report is the only source that generates a match with the total assets recorded in Worldscope, we carefully investigate whether the consolidation information in the annual report also applies to the prior year and provide details in ConsolidationNotes. This is because there are examples where the annual report states e.g. ‘Consolidated statement of financial position’ but the prior year data are unconsolidated.

3. When the source is a prospectus that shows consolidated financial statements and covers several years, we score ‘consolidated’ even though it is possible that earlier years are not consolidated.

AccountingStandards

1. Local refers to the national GAAP of a firm’s CountryOfIncorporation. For example, if a UK firm uses UK GAAP, we score Local; but if a Jersey firm uses UK GAAP, we score UK GAAP.

2. AccountingStandardNotes states when a firm’s accounting period began on a date before the cut-off date that requires use of IFRS. This is for example relevant for AIM-listed firms, who did not have to use IFRS for accounting periods beginning before 1 January 2007.

ListingStatus

1. We have a separate category for the UK’s Alternative Investment Market (AIM) because UK-incorporated AIM-listed firms have to use IFRS for consolidated financial statements for accounting periods beginning on or after 1 January 2007.

2. The primary data source is information from the national stock exchanges. Whenever available, we use Excel documents published on the websites of the stock exchanges.

3. When using ad hoc reports, the date of the data source generally differs from the accounting period end date. To ensure that the data apply on the accounting period end date, we consult (whenever possible) one ad hoc report before and one ad hoc report after the accounting period end date; we provide a note if this is not possible.

4. The main sources of ad hoc reports are DGAP for Germany and Investegate for the UK.

5. ListingStatusNotes provides information that explain the scoring, e.g. IPO dates.

ListingPrimary

1. For Germany, we assume that a listing in the Prime or General Standard of the Frankfurt Stock Exchange is a firm’s primary listing. Data for this is available from the Frankfurt Stock Exchange. However, for other firms this information must be manually collected. It is not possible to use data from regional stock exchanges because a firm can be listed on several stock exchanges, and it is not possible to know which is the primary listing.

2. For the UK, we assume that a listing on the London Stock Exchange (Main Market or AIM) is a firm’s primary listing.

A1.5. Using Our Database

Our database can be connected to Worldscope via the data fields WorldscopePermamentID and AccountingPeriodEndDate.

Our data field CountryOfIncorporation differs from Worldscope’s ‘nation’ (data field WC06026) because we focus on country of incorporation (which determines the GAAP that a firm is required to use) whereas Worldscope focuses on the location of the corporate office of the firm; and we provide annual data whereas Worldscope only provides ‘current data’, i.e. only the current information. There is no data field in Worldscope that is similar to our ListingStatus.

Worldscope provides data on consolidation (data field WC07531), but we document that it contains many errors (see Tables 3 and 4). Worldscope also provides data on accounting standards (data field WC07536), but we document that it contains some errors. For these two data fields, our data can be used to correct Worldscope errors in the following way:

1. If our database reports one GAAP and our data on consolidation and/or accounting standards differ from Worldscope: This indicates that Worldscope’s data is wrong and should be replaced with our data.

2. If our database reports two accounting standards and the first accounting standards and/or the first consolidation information differs from Worldscope: This indicates that Worldscope’s data are wrong and should be replaced with our data.

3. In both cases, we are confident that the data fields in Worldscope for the respective firm-year are based on the financial statements we have used to collect our data because we have checked that total assets from our data source matches Worldscope.

If our database reports two accounting standards, the second accounting standards and/or the second consolidation information should not be used to correct data in Worldscope because Worldscope would not have used the respective financial statements to populate their data fields. It mainly provides information on statutory financial statements.

For using our database beyond 2020: Our 2020 data can be used as a starting point to check whether Worldscope continues to include errors for those firms where our data on consolidation and accounting standards differ from Worldscope.

Appendix 2.

Sample Selection Relating to 2012/13

Table

	Germany	UK
Firms in Worldscope as of 29 September 2021 (active and inactive)	1,910	5,878
Observations of firms with data for 2012/13 in Worldscope	854	2,024
Less:
- Duplicates	3	3
- Worldscope’s ‘nation’ differs from the respective country	17	309
- Total assets is zero	1	1
- Missing SEDOL number	28	10
- Missing data on accounting standards followed	6	4
- Missing total assets data	10	3
Firms with all required data	789	1,694

Notes: This appendix shows our sample selection relating to 2012/13, which follows PW’s ‘initial sample’ (see PW’s Table 2). We define ‘Observations of firms with data for 2012/13 in Worldscope’ as observations with an entry in Worldscope’s data field WC05350 (fiscal period end date) in the period from 31 December 2012 – 30 December 2013, inclusive. Duplicates arise for a firm which published two reports in that period because of a short accounting period; in such cases we keep the more recent observation (as PW do). With the aim of excluding firms that are not incorporated/domiciled in the respective country, we exclude firms for which Worldscope’s data field WC06026 (nation) differs from the respective country; out of the 17 (309) firms excluded from the Germany (UK) sample, there were 12 (39) firms for which Worldscope’s ‘nation’ was a different EU country. For the other restrictions, we use Worldscope’s WC02999 (total assets), WC06006 (SEDOL number) and WC07536 (accounting standards followed).

Appendix 3.

Explanations for Worldscope Recording the Lack of IFRS Statements in 2012/13, with Example Firms

Table

	Explanation	Example German firms	Example UK firms	All Austrian firms	All Portuguese firms
[1]	Firm not incorporated in the country (Worldscope error)	–	KNOT Offshore Partners (incorporated in the Marshall Islands)	–	–
[2]	Firm does not provide consolidated statements – Case 1: Worldscope records this (correctly)	Deutsche Grundstücks-auktionen Hamburger Getreide-Lagerhaus	Aberdeen Japan Investment Trust; Henderson Eurotrust	AB Effectenbeteiligungen; Stadlauer Malzfabrik	–
[3]	Firm does not provide consolidated statements – Case 2: Worldscope error	Bremer Straßenbahn; Eisen- und Hüttenwerke	John Lewis of Hungerford; Schroder Income Growth Fund	A-TEC Industries; Athos Immobilien; Burgenland Holding; Josef Manner & Comp.; Maschinenfabrik Heid; Unternehmens Invest	Companhia Industrial Produtora de Antibióticos; Companhia Portuguesa de Amidos; Litho Formas Portuguesa; Sporting Clube de Braga – Futebol
[4]	Firm provides IFRS statements (but Worldscope records a different GAAP)	Alexanderwerk; Fresenius Medical Care (IFRS and US GAAP statements)	Betfair; Carnival (IFRS and US GAAP statements)	–	–
[5]	Firm not listed in the year (although Worldscope contains accounting data)	Knorr Bremse; Veritas	Nexus Infrastructure; RFC 2012	–	–
[6]	Firm not listed on a market which requires IFRS	Deutsche Rohstoff (Open Markets Frankfurt, Berlin, Dusseldorf and Stuttgart); MPC Münchmeyer Petersen Capital (Open Markets Frankfurt, Berlin, Bremen, Dusseldorf, Hannover, Munich and Stuttgart)	DXS International (ISDX); Williams Grand Prix Holdings (Open Market Frankfurt)	–	Conduril Engenharia; Sonagi