Portal de Periódicos da CAPES

The effect of revised IAS 14 on segment reporting by IAS companies

2004; Routledge; Volume: 13; Issue: 2 Linguagem: Inglês

10.1080/0963818032000138206

1468-4497

Jenice Prather‐Kinsey, Gary K. Meek,

Accounting Theory and Financial Reporting

Abstract International Accounting Standard (IAS) 14 on segment reporting was revised in 1997. IAS 14R substantially changed segment reporting requirements in response to numerous criticisms of the original standard. The objective of this study is to determine how IAS 14R affected the segment disclosure practices of companies claiming to comply with IAS. This paper examines the following questions: (1) What items of information are disclosed under IAS 14 and IAS 14R? Was there a gain or a loss of information disclosed for business and geographic segments with the implementation of IAS 14R? (2) Has the number of business and geographic segments reported by companies changed with the implementation of IAS 14R? (3) Are companies disclosing the items required by IAS 14R? (4) Are companies' segment reporting practices related to size, country of domicile, industry, international listing status, and having a then-Big Five auditor? We find that the impact of IAS 14R is mixed. Companies are responding to IAS 14R, but not wholly embracing it. Our findings suggest that companies audited by a Big Five (now Big Four) firm and, to a lesser extent, companies that are larger, listed on multiple stock exchanges, and from Switzerland have greater compliance with IAS 14R than other companies in our study. Gary K. Meek, Oscar S. Gellein/Deloitte & Touche Professor, School of Accounting, College of Business Administration, Oklahoma State University, Stillwater, OK 74078-4011, USA. E-mail: gmeek@okstate.edu Acknowledgements The authors gratefully acknowledge the helpful comments received at the 25th Annual Congress of the European Accounting Association, and the Kansas State University Central States Accounting Research Workshop. Special thanks are due to Sid Gray, Don Herrmann, Donna Street, Wayne Thomas, Nancy Weatherholt, the two reviewers and Kari Lukka (editor). The authors also thank Ron Howren for his indispensable work in our data analysis. Notes Gary K. Meek, Oscar S. Gellein/Deloitte & Touche Professor, School of Accounting, College of Business Administration, Oklahoma State University, Stillwater, OK 74078-4011, USA. E-mail: gmeek@okstate.edu 1.A summary of the criticisms is contained in Pacter (Citation1994) and Street et al. (Citation2000, pp. 259–285). 2.The Canadian standard, contained in CICA Handbook Section 1701 and the US standard, SFAS No. 131, were also issued in 1997. Some differences between IAS 14R and SFAS No. 131 are: (a) IAS 14R requires the same accounting policies for segment reporting as used in the consolidated financial statements, whereas SFAS No. 131 requires the accounting policies used for internal reporting purposes. (b) IAS 14R identifies segments by reference to the dominant source of risk and return, while SFAS No. 131 bases reportable segments on the company's internal reporting system. Both standards require the designation of primary and secondary bases of segmentation, with more disclosures required for the primary segments. However, IAS 14R requires that the primary segments must be either business or geographical segments with the other one treated as the secondary segment. There is no similar requirement under SFAS No. 131: reportable segments could consist of a combination of business and geographical segments. (c) IAS 14R requires disclosure of liabilities by primary segments and capital expenditures by geographical segments, whereas SFAS No. 131 does not. 3.The introduction to IAS 14R (pages 3–5) details the major changes from the original IAS 14. 4.While we are unable to directly compare our sample to that of Street and Nichols (Citation2002), we can do so indirectly. The Street and Nichols (Citation2002) sample starts with 279 companies identified in Street and Gray (Citation2001) and, after eliminating certain companies, ends up with 210 companies. Street and Gray (Citation2001, pp. 63–68) list the 279 companies in their sample. We note that 66 of the 211 companies in our sample (roughly one-third) are not in Street and Gray (Citation2001) and, therefore, not in Street and Nichols (Citation2002). By comparing our Table 3 (below) to Table 2 in Street and Gray (Citation2001), we note that we have proportionally more German, fewer Chinese and more 'other' Pacific Rim (e.g. Australia and New Zealand) companies. 5.The IASC made no claims as to the accuracy of these lists and in our experience they represent something of a 'moving target'. There were 414 companies on the list in August 1997, 817 in September 1999, and 207 in May 2001. 6.We emphasize that the sample consists of companies claiming to comply with IAS. One purpose of this study is to see if companies claiming compliance actually do comply with IAS. 7.Our sampling procedure is similar to that used by other studies focusing on IAS reporting, such as Street and Gray (Citation2001). As noted in Table 2, we obtained 330 annual reports from the 584 requested, a response rate of 57%. By comparison, Street and Gray (Citation2001) received 67% of the annual reports they requested. Every attempt was made to obtain as many annual reports as possible and to get as representative a sample as possible. In the end, we can only base our study on the annual reports received. 8.The sample contains no UK or US companies. Only a handful of UK and US firms follow IAS and most of them are either financial institutions or non-profit organizations, which we deleted from our sample. During the 1997–2000 time period, the IASC lists that we obtained identified only four UK firms (two are financial institutions and one is a non-profit organization). The lists identified only six US firms (three are financial institutions and one is a non-profit organization). The remaining (three) UK and US firms either did not claim IAS compliance or did not respond to our request for an annual report. 9.One way to look at the pattern is to compare disclosures based on companies' designation of primary and secondary basis of segmentation after implementing IAS 14R. Recall that under IAS 14R, a company designates a primary basis of segmentation – either groups of related products and services or geography – determined by the dominant source and nature of the company's risk and returns. Its secondary basis of segmentation is then the other one. This is how Street and Nichols (Citation2002) look at their two-year patterns. The approach requires an assumption that a company's primary and secondary bases of segmentation in 1999 were also their bases of segmentation in 1998 and 1997. Instead, we look at disclosure patterns according to line of business and geographic segments. Later in the paper we look at compliance with IAS 14R where compliance is measured using primary and secondary bases of segmentation. 10.We lose nine companies because they had no segment information for 1997 and/or 1999. In other words, they were not in compliance with IAS 14 (1997) and/or IAS 14R (1999). Thus, Table 6 is based on 57 companies instead of 66. 11.The first year that IAS 14R was mandatory was 1999. 12.'One-segment' companies are those companies that (a) operate in one line of business and designate business segments as their primary basis of segmentation or (b) operate in one geographic region and designate geography as their primary basis of segmentation. Fourteen of the 134 companies in our 1999 sample are one-segment companies. Thus, Table 7 is based on 120 companies. 13.Ninety-six of the 120 companies designated products and services as their primary basis of segmentation and 24 designated geography as the primary basis of segmentation. 14.Street and Gray (Citation2001, pp. 11–18) discuss other variables that have been hypothesized to influence financial reporting practices. They note that leverage (gearing) is consistently not associated with disclosure in prior international accounting literature. There is also little empirical support for profitability (e.g. return on assets) or extent of multinational operations (e.g. foreign sales divided by total sales). We exclude these variables in our analysis, though we did test them and found that they are not statistically significant. A few other variables such as size of the home equity market and dispersion of stock ownership have been rarely considered in the empirical literature. 15.A company was considered to have disclosed a reconciliation (Item 8) if it provided a reconciliation of any one of the four listed in Table 7. This introduces some error in our measure of the extent of disclosure, but it is unclear to us how the error might bias the results. For example, companies reconciling only one of the items get the same score as companies reconciling three or four of the items, understating a company's extent of disclosure. On the other hand, we felt that counting each reconciliation item as a possible disclosure would overstate the overall extent of disclosure. Our approach seemed to us to be a reasonable compromise to the situation. 16.Measuring the extent of disclosure this way assumes that all items are equally important, which we admit is unlikely to be true. However, any proposed weighting scheme is likely to be as subject to criticism as an unweighted scheme. Also, the eight items we focus on reflect what seems to us to be the most important potential segment disclosures required by IAS 14R. 17.The F-statistic measures the marginal contribution of an independent variable to the overall regression results, assuming that the other independent variables are already included in the regression equation (Kmenta, Citation1971, pp. 369–391). 18.Our R 2's are lower than those reported in some other disclosure studies. (For example, Street and Gray, Citation2001, report R 2's of 27%.) However, we focus on a single standard, IAS 14R, while these other studies have looked at total disclosures across many standards and categories. Thus, we would expect our R 2's to be lower. We reiterate that the regressions are statistically significant. 19.These results may be compared to Street and Nichols (Citation2002) who find an increase in the amount of primary and secondary segment information disclosed and a marginal increase in the number of business segments reported. They also report non-compliance with IAS 14R, but our results show lower compliance than theirs.