The Challenge of Assessing the Market Value of Private Companies Using a Standardised Combination Method for Tax Purposes – Lessons to be Learnt from Past Experience

Abstract

When taxes on capital or wealth are levied, in most countries companies have to be assessed in terms of their market value (MV). Estimating the MV of private companies for tax purposes is a challenging task for tax authorities as MVs are not available. In this study, I empirically analyse to what extent an accounting-based tax valuation method for private companies, a simplified residual income model, succeeds in matching the MV. I refer to the mandatory Standardised Combination Model that is a special case of methods commonly used in several countries. In the absence of market prices for private companies, I use a sample of small German public companies as a proxy. I validate this approach using a sensitivity analysis that involves matching the sample of public companies with that of private companies. The results imply that the mandatory Standardised Combination Model leads to a severe unequal treatment not only between public and private companies but also among private companies across and within industries. Furthermore, I simulate the effects of variation in the key parameters and highlight their impact on the approximation quality of the Standardised Combination Method. The findings are relevant to tax reform discussions as well as to tax policy-makers and practitioners in many countries.

Acknowledgements

I am grateful to the anonymous reviewers for their useful suggestions on improving this paper. I also thank the participants of the 2008 American Accounting Association Annual Congress in Anaheim, the 2006 arqus Annual Meeting in Magdeburg, the 2008 European Accounting Association Annual Congress in Rotterdam, and to the University of Hannover for their valuable comments. I owe my gratitude to Urska Kosi, Rainer Niemann, Hollis A. Skaife, Caren Sureth, David Veenman, and Dennis Voeller for their very helpful feedback. I acknowledge the German Research Foundation (DFG) for the financial support.

Notes

1 A combination of different value estimates is also frequently used in US IRS tax-related appraisal cases (Yee, 2004; Chen et al., 2010).

2 See, for example, Hand and Landsmann (1998), Frankel and Lee (1998), Penman and Sougiannis (1998), and Dechow et al. (1999).

3 For reasons of simplification, it is rounded to 0.68 (R 100 para. 2 Inheritance Tax Code Directives).

4 See Section 2.

5 The effects remain qualitatively unchanged when variables are scaled by total assets or sales.

6 See Schanz and Schanz (2010) and Pfaff and Schröer (1996, p. 964).

7 See Cheng and McNamara (2000, p. 363) for the validation of this procedure.

8 See Yoo (2006, p. 113), Liu et al. (2002, p. 145, 2007, p. 60), and Bhojraj and Lee (2002, p. 416).

9 See Harris et al. (1994, p. 191), Pfaff and Schröer (1996, p. 964), and Eberhartinger (1999, p. 95).

10 Company-year observations after the adoption are eliminated, since only local GAAP accounts may enter the sample. Thus, companies that adopt IFRS remain in the sample but only with their observations before the adoption. Eleven companies adopted IFRS in 2004. Even though it is a voluntary adoption, it may also be interpreted as an early mandatory adoption. Nevertheless, the distinction has no effect on the results.

11 Harris et al. (1994, pp. 196–197) and Lo (2003, pp. 202–203) apply a four- to two-digit SIC code; Cravens and Oliver (2000, p. 530) match based on the three-digit SIC code.