The persistence of international accounting differences as measured on transition to IFRS

Abstract

The international accounting classification literature emphasises the importance of understanding how institutional factors shape accounting regulations and practices. With the mandatory adoption of International Financial Reporting Standards (IFRS) in the European Union and Australia in 2005, our empirical study examines whether three international accounting classification systems relating to equity financing, law and culture still had merit as measured on transition to IFRS and explore whether they are effective in grouping accounting systems. Using IFRS as the yardstick, we find statistically significant differences in the measurement of shareholders’ equity as between strong (Class A) versus weak (Class B) equity financing systems, common law versus code law systems and cultural systems based on ‘Anglo’, ‘Nordic’ and ‘More Developed Latin’ cultural groups. With regard to the measurement of net income, however, we find statistically significant differences only in respect of strong (Class A) versus weak (Class B) equity financing systems. Our findings demonstrate that traditional international accounting system differences still persisted at the time of IFRS adoption even after long periods of harmonisation and growing international accounting convergence.

Keywords:

• international accounting
• IFRS
• accounting systems
• accounting harmonisation
• international accounting classification

Notes

1 Nobes (1998, 2006) suggests that the strong link between financial accounting and taxation appears because of the lack of competing purposes for accounting, that is, the lack of information demands of outside shareholders provides larger room for tax rules to have an influence on financial reporting.

2 In addition, Nobes (1998) suggests a factor concerning colonial heritage.

3 (1) Professionalism versus Statutory Control; (2) Uniformity versus Flexibility; (3) Conservatism versus Optimism; (4) Secrecy versus Transparency.

4 When testing the hypotheses mentioned in the text, we will use an index of comparability, where shareholders’ equity (or net income) according to national GAAP is compared to shareholders’ equity (or net income) according to IFRS, using IFRS as the benchmark. This is further described in the methodology section.

5 IFRS 1 allows for keeping national GAAP values in the opening IFRS balance sheet in selected areas instead of applying the general rule of retrospective application (cf. Cormier et al. 2009). This lowers the probability of detecting national GAAP differences by studying the reconciliations from national GAAP to IFRS. An example: a company values equity at 80 according to national GAAP. The value according to IFRS is 100, but the first-time adoption rule allows the company to keep the national GAAP number of 80 under IFRS. Then there will not appear to be any difference in measurement of shareholders’ equity according to national GAAP versus IFRS. For a real-world illustration, see Nobes's (2006, p. 242) example concerning goodwill and business combinations.

6 We refer here to the general mechanism of income shifting across periods due to accounting conservatism, acknowledged already by Paton and Littleton (1940, p. 128). Further, there is an element of management behaviour involved, as conservative accounting practices tend to offer timing opportunities with regard to the realisation of profits and the creation of reserves. The effects of income shifting from good years to bad years may thus be reinforced by management incentives to smooth earnings over time. Income smoothing is a general phenomenon and the flexibility of companies to smooth income will vary because of differences in the nature of their operations (Trueman and Titman 1988), however, empirical studies tend to show that earnings are in fact smoother in Continental Europe compared to Anglo-American countries (Leuz et al. 2003, p. 513).

7 The reference to ‘2004–2005’ rather than just 2004 is because companies with financial years that deviate from calendar years will have a first-adoption year beginning during 2005 (e.g. 1 April 2005–31 March 2006) with a corresponding transition year (e.g. 1 April 2004–31 March 2005).

8 We acknowledge that GNP growth for a country is a somewhat crude indicator of whether the year was generally a good or a bad year for the listed companies of that country. Anecdotally, Daimler-Benz measured its net income more conservatively under German GAAP compared to US GAAP during 1991 and 1992 and less conservatively during 1993. The corresponding GNP growth numbers for Germany were +5.1% in 1991, +1.9% in 1992 and −1.0% in 1993 (Source: The World Bank).

9 Denmark is the smallest country included. According to the market capitalisation statistics from the Federation of European Securities Exchanges, Denmark was the 11th largest stock market in Europe per 31 December 2004. Excluding Switzerland, which did not require IFRS from 2005, both Belgium and Finland ranked before Denmark. However, we have not been able to find a usable top-100 sample of companies in these countries.

10 According to the Financial Statements Law Reform Act (Bilanzrechtsreformgesetz) implemented in December 2004, German firms who applied US GAAP (and German firms who only had debt securities listed) were allowed to defer the mandatory adoption of IFRS until 2007.

11 The first step of the interquartile-range method is to identify the interquartile range (IQR). Next, the first (ql) and third (q3) quartiles are determined. Finally, all observations lying below the ql −1.5IQR or above q3 +1.5IQR range are considered outliers and excluded from the data set.