On the Relationship between Voluntary Disclosure, Earnings Smoothing and the Value-Relevance of Earnings: The Case of Switzerland

Abstract

This paper examines whether voluntary disclosure by Swiss firms constrains the use of discretionary accruals to smooth earnings, and explores the effect of voluntary disclosure on the value relevance of earnings. We focus on Swiss firms because Switzerland's financial reporting system provides managers with extensive discretion in corporate disclosure, and there are important variations in the level of information provided in their annual reports. We consider that managers can choose two different ways to voluntarily convey information, either through the quality and quantity of annual report disclosure or, through compliance with International Accounting Standards (IAS)/International Financial Reporting Standards (IFRS) or US Generally Accepted Accounting Principles (GAAP). Relying on a simultaneous equations approach, our results suggest that Swiss firms use discretionary accruals to smooth earnings. However, this relation is reduced for firms that voluntarily disclose more information in their annual report or comply with IAS/IFRS or US GAAP. Moreover, we show that discretionary accruals of high disclosers or of firms voluntarily complying with IAS/IFRS or US GAAP receive a lower valuation weight.

Acknowledgements

The authors wish to thank the Social Sciences and Humanities Research Council of Canada, Fonds québécois de la recherche sur la société et la culture (FQRSC), Corporate Reporting Chair (UQAM) and the Lawrence Bloomberg Chair in Accountancy (Concordia) for their financial support. Comments from workshop participants at Université Laval and participants at the 27th Annual Congress of the European Accounting Association and the 2005 American Accounting Association Annual Meeting are also greatly appreciated.

Notes

1. Firms listed on the SWX are required to comply with either IAS/IFRS or US GAAP since 2005, with the stated objective to increase corporate transparency.

2. Last reported audited EPS is followed in order by management forecast of EPS, last reported unaudited EPS, reporting a profit, analysts' consensus forecast and finally, the lead analyst forecast.

3. Collins and Hribar (2002) show that the use of the balance sheet approach may lead to a systematic bias in estimating discretionary accruals. They suggest the use of the difference between net income and cash flow from operations as the correct measure of total accruals.

4. A reduced form of the smoothing model including ETD, DISCORE and the residual from the disclosure model, or ETD, IASUS and the residual from the disclosure model is estimated using OLS.

5. The disclosure threshold for blockholdings is 5% in Switzerland. Consequently, the three ownership structure variables correspond to the sum of blocks greater than 5%.

6. We use the Breush–Pagan test to test for the presence of heteroskedasticity. The test is significant for all three models: disclosure model, p < 0.050; smoothing model, p < 0.000; value-relevance model, p < 0.000.

7. Results are similar when we use a log transformation for trading volume instead of a binary variable (coefficient = 0.735; t = 5.910).

8. To assess the robustness of our results, we replicate the analysis using the two following recomputed dependent variables. First, consistent with Lang and Lundholm (1993), Botosan (1997) and Hail (2002), we compute rank scores calculated by dividing the rank of a firm's DISCORE by the number of firms included in the sample. This procedure allows for the study of the economic determinants of corporate disclosure quality by comparing firms with one another, instead of on an arbitrary base of 20 points. Second, we standardize the disclosure score by the year median. Our results remain virtually similar in both cases. Overall, we conclude that Swiss firms' disclosure strategy is associated with the level of information costs faced by investors.

9.