Discussion of ‘Making Deferred Taxes Relevant’

Abstract

In this discussion of Brouwer and Naarding's article ‘Making Deferred Taxes Relevant’, which is published in this issue of Accounting in Europe, I question several aspects of their proposal to change the tax accounting standard. I argue that a quest for more value relevance of individual balance sheet items is not a good guideline for accounting standard setting. The distinction between book-first and tax-first temporary differences may be helpful for some analytical purposes, but it is not sufficiently robust to serve as a basis for an accounting standard. However, I agree with the authors that the efforts to improve IAS 12 should not be abandoned.

Keywords:

• IFRS
• deferred tax
• value relevance

Acknowledgements

I have received helpful comments from Andrew Lennard and from the authors of the article discussed to previous versions of the paper.

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1 Compare here the distinction between relative and incremental association studies in Holthausen and Watts (2001).

2 That tax accruals increase the association between accounting numbers and security prices is a long-established result in accounting research (e.g. Beaver & Dukes, 1972, 1973; Chaney & Jeter, 1994).

3 B&N also refer to Guenther and Sansing (2000, 2004) and Dotan (2003). Although these authors discuss the value relevance implications of the book-first tax-first split of temporary differences, it is not evident that they advocate the split introduced in the accounting standards.

4 B&N present these definitions as a synthesis of the literature they refer to (Dotan, 2003; Guenther & Sansing, 2000, 2004); however, none of the cited sources use a wording that resembles that of B&N.

5 To be even more precise, by accounting income in this context must be meant profit before tax. If deferred tax income expense were included, there will be no separation between book-first and tax-first.

6 Laux (2013) lists the components according to their classification in the income statement. Under IFRS, common disclosure practice is to analyse temporary differences by class of asset and liability (cf. IAS 12.81g). Thus, normal reporting would not allow a separation between temporary differences on fixed assets caused by depreciation and impairment losses, respectively.

7 The recently enacted US tax reform (‘Tax Cuts and Jobs Act’), which will permit immediate expensing for tax purposes of some long-lived assets, will provide a lot of new cases of pure tax-first temporary differences.

8 For US firms, Poterba, Rao, and Seidman (2007) report large aggregate DTLs related to PPE for all years 1993–2004.

9 The method, therefore, contradicts the claim that PPE temporary differences do not represent future tax cash flows. The proposed DTA of B&N is precisely the present value of the tax cash flows created by tax depreciation.

10 Laux (2013) and Amir and Sougiannis (1999) find it value relevant, while Amir, Kirschenheiter, and Willard (1997) fail to find value relevance.

11 An even simpler way of seeing the inadequacies of the proposed approach is by substituting the tax depreciation of Example 4 by immediate expensing for tax purposes. Under IAS 12, a DTL will arise upon acquisition (the ‘initial recognition exemption’ will not apply, cf. paragraph 15b ii), which will smooth the tax expense over the lifetime of the asset. Under B&N's approach, the tax expense will every year be equal to the current tax expense (there is no tax amortization benefit), which is zero in year one and 250 in all subsequent years.

12 On discounting, B&N further suggest the use of a pre-tax discount rate, for example, for the present value measurement of future deductions resulting from tax depreciations. To discount tax flows (which are not taxable) with a pre-tax rate seems illogical. However, they do not justify this choice other than with a general reference to literature, and I will not discuss it further here. On the use of pre-tax and post-tax rates for accounting purposes, see Kvaal (2007).

13 On this note, I refer to the discussion about whether goodwill is an asset in the basis for conclusions of IFRS 3 (BC313–318).

14 IAS 12 actually says something of that kind in the third paragraph of its present ‘Objective’, but the content of that text is limited to the classification of tax in profit or loss, other comprehensive income or equity.