Write-offs and Profitability in Private Firms: Disentangling the Impact of Tax-Minimisation Incentives

Abstract

Private firms are likely to use the financial reporting process more for other objectives, such as tax savings, than for communicating performance. However, observing firms choosing accounting policies for tax-minimisation purposes is not straightforward due to (i) tax and non-tax costs of reporting lower income (ii) accounting policies that result in lower reported income and no tax savings but generate non-tax benefits (iii) preparers' multiple incentives and (iv) econometric issues. We observe a large sample of 20,505 private firms writing off assets in two separate regimes, one that generates tax savings and one that does not. Firms significantly decrease, but continue to use, write-offs after the adverse change in tax treatment of write-offs. The exogenous tax change should not affect other reporting incentives. This allows us to disentangle the tax-minimisation incentive from other (un-observable) incentives, including debt contracting, dividends and employee relations that contribute to the observed anomalous positive relationship between write-offs and profitability. We show that for private firms (i) obtaining tax savings is important overall (ii) non-tax costs and benefits are probably also important and (iii) earnings informativeness for future cash flows increases after the adverse tax legislation change.

Acknowledgements

The authors wish to thank Bisnode, d.o.o., a commercial database provider, for providing the data on ownership of companies and data on auditing in electronic form. The authors gratefully acknowledge the two anonymous reviewers for their helpful suggestions and comments on earlier versions of the paper. The authors also thank William Rees, participants at the 31st EAA annual congress in Rotterdam and participants at the research seminar at the Faculty of Economics, University of Ljubljana, for their suggestions. This research is part of the INTACCT programme – The European IFRS Revolution (Contract No. MRTN-CT-2006-035850).

Notes

Asset write-offs in this study should be understood as partial decreases in the assets' book values. This is different from Francis et al. (1996), who use the term write-off to refer to both complete and partial downward asset revaluations.

It should be noted that write-off operating expenses have no direct effect on current-period cash flows from operations; they directly affect the firm's tax base and, thus, total cash flows.

Garrod et al. (2007) show that firms are aware of this possibility.

Unfortunately, this hypothesis cannot be tested in our empirical setting, as the dividend-payout data is not available.

However, the amendment of the Corporate Income Tax Act 2 did not affect annual depreciation/amortisation rates prescribed for tax reporting. In both years, firms apply linear depreciation/amortisation methods and maximum rates allowed for tax purposes (e.g. 5% for buildings, 25% for equipment, 50% for computer equipment).

This does affect empirical estimations as any un-modelled factor causes the omitted variable bias.

For a more detailed background about Slovenian private firms, see Garrod et al. (2008) and Pusnik and Tajnikar (2008).

A firm is defined as small if it fulfils two of the following three criteria: average number of employees in the last fiscal year does not exceed 50, net turnover is less than EUR 4.17 million and total assets at the end of the fiscal year do not exceed EUR 2.09 million (Amendment F of Companies Act, 2001, paragraph 52; valid during the sample period in this paper).

The system of data collection resembles the Belgian system (e.g. Deloof and Jegers, 1999; Vander Bauwhede et al., 2003), except that it is broader in scope as it applies to all firms. A predecessor of the Agency was a government body governing inter-company payments and financial reporting.

We thank an anonymous reviewer for pointing out this issue.

The CA and FA subsets cannot be directly compared because there is some overlap between the two. Specifically, companies that write off both types of assets belong to both subsets.

A caveat applies in using interaction terms in logit and probit models (see Ai and Norton, 2003), although not explicitly in tobit models.

Other approaches have been developed when there are multiple EIV variables in the model. See Cready et al. (2001) for an accounting-based application of Klepper and Leamer's solution of a multiple EIV problem (1984).

Similar to Burgstahler et al. (2006), we must estimate cash flows from operations using the balance-sheet approach, as firms in our sample do not report cash flow statements. The accrual component of earnings is calculated as ACC _t  = (Δtotal current assets − Δcash) − (Δtotal current liabilities − Δshort-term debt) − depreciation/amortisation, where Δ denotes the change over the last fiscal year.

Given the audit-market structure in Slovenia we substitute Big-4 dummy with a Big-5 dummy and results remain essentially the same (Zaman Groff and Valentincic, 2011).

This is not to say that these firms were not using other means of accounting discretion to achieve specific financial reporting outcomes.

These firms could still have used write-offs to achieve objectives of the financial process other than reporting firm performance, but it is unlikely that any such incentive would have changed as a result of a switch in the tax treatment of write-offs.

Additional information

Notes on contributors

Urska Kosi

Paper accepted by Salvador Carmona.

Aljosa Valentincic

Paper accepted by Salvador Carmona.