Changing from German GAAP to IFRS or US GAAP: A Survey of German Companies

ABSTRACT

Since 1993 an increasing number of listed German companies have been publishing their consolidated financial statements in accordance with either IFRS or US GAAP. In 1998 this was approved as a substitute for the consolidated German GAAP financial statements of listed companies (§292a HGB). Our study surveys the motives that led these companies to opt for international reporting systems (IFRS or US GAAP) rather than German GAAP and considers whether these objectives have been achieved. Rather surprisingly, we find that even though companies state that their overall expectations have been met to a satisfactory degree, a detailed analysis shows that several of the ex-ante objectives have not been achieved from an ex-post point of view. Additionally, we use logistic regression analysis to show that companies choosing IFRS rather than US GAAP and vice versa differ distinctly in the objectives they pursue with their choice of international GAAP.

Notes

1. The findings of the studies are heterogeneous with respect to different market segments and with respect to the differentiation between IFRS and US GAAP; tests about the achievement of both objectives and a statistical discrimination to link objectives and the reporting system chosen are not provided by these studies.

2. Nevertheless, if there had been a tendency towards one type of international GAAP this would introduce a bias into our analysis.

3. Note that the Neuer Markt was terminated in 2003 due to several financial fraud cases that strongly reduced the investors' trust in this segment. The remaining companies listed in the Neuer Markt were included into other capital-market segments.

4. In this context, Pellens (2001, p. 89) mentions the IPO of Deutsche Telekom with a volume of about €10 billion in 1996 that could not be supplied by the capital market in Germany alone.

5. In 2000, 219 companies were listed in the Amtlicher Handel. Fifty-five of these companies used either IFRS or US GAAP, and 54 of them were included in the DAX100. The other 46 DAX100 companies were still using German GAAP.

6. The companies that did only plan to switch to an international reporting system were excluded in the latter part of the analysis.

7. The size of a company was determined by its market capitalization. CDAX or Composite DAX is a stock index comprising all companies listed on the German stock market.

8. The coefficient of variation (CV) describes the homogeneity of the answers – the higher the CV the more heterogeneous the opinion on a specific aspect.

9. The Mann–Whitney U-test is designed to compare arithmetic means of two independent samples even if the items are not normally distributed. A significant Z-value shows statistical differences of the arithmetic means (Toutenburg, 2000, p. 172).

10. The Wilcoxon test is used to compare the rank of an item in two independent samples (Toutenburg, 2000, p. 182).

11. The number of companies in the analysis from here on is reduced by 11 as these had not yet adopted a different system at the cut-off time of the survey.

12. This study carried out between November 1996 and February 1997 was answered by 56 companies from a sample of the 200 biggest companies (size criterion: amount of turnover). Multiple answers were possible.

13. Only 79 questionnaires of the survey could be analysed with respect to this research question.

14. The requirement for the test is the existence of at least 25 cases in each dependent variable. The condition is met using 40 observations of IFRS users and 39 observations of US GAAP users. Segmentation between DAX100 and Neuer Markt therefore is not possible. The Durbin–Watson statistic and the multicollinearity test deny the existence of autocorrelation and multicollinearity. For more details on logistic regressions see Krafft (1997, p. 265).