Reconciling Competing Reporting Objectives Through Deferred Tax Accounts: Evidence on Private Italian Firms

Abstract

This paper contributes to the growing literature on earnings management in private firms by focusing on deferred taxes. This accounting treatment requires sophisticated use of accruals that provides the chance to manage earnings and net assets without affecting the tax payable. We argue that in a setting with high book-tax conformity, the small room that allegedly exists to recognise deferred taxes remains a comfortable avenue to reach reporting objectives that a tax-minimisation strategy may preclude. We use a sample of private firms operating in a credit- and tax-driven environment such as Italy to test this expectation. Our results show that private firms use deferred taxes to extract multiple financial reporting benefits that may facilitate debt contracting: smoothing earnings over time, meeting/beating historical earnings, avoiding reporting accounting losses, and managing leverage. Tax loss carryforwards are the source of deferred tax assets where the exercise of discretion becomes more critical.

Keywords:

• private firms
• deferred taxes
• earnings smoothing
• book-tax conformity
• tax loss carryforwards

JEL:

• M400
• M410
• M480

Acknowledgements

I thank Andrei Filip and Araceli Mora (Editors in Chief), Urska Kosi (Associate Editor) and the two anonymous referees for their helpful and constructive comments. I also thank Stephen Zeff, Christopher Nobes, Niclas Hellman and other participants in the 14th Workshop on European Financial Reporting (Stocholm, 2018), Alberto Quagli and Francesco Avallone and other participants in the IX Financial Reporting Workshop (Bologna, 2018) for their encouraging comments on a previous draft of this paper. I am also grateful to Gianluigi Roberto, Francesco Vallascas, Laura Mulas and Viviana Ecca for their thoughtful suggestions. I gratefully acknowledge financial support from the Fondazione di Sardegna.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 The Pearson correlation matrix for the explanatory variables in our regression model (1) is presented in Appendix 3. Few cases register a high correlation term of magnitude. DHE appears significantly and negatively correlated with ΔTLCF (– 0.42) and positively correlated with EBD (0.49). We are not surprised by these correlations, as we have defined these variables as changes in earnings and components of earnings. To further alleviate the risk of multicollinearity, variance inflation factors (VIF) were calculated for all regressions. Because P<HE_S registers the highest VIF, equalling 2.9, multicollinearity is not a major issue in our model.

2 The variance of the difference between the observed and expected number of observations for interval i is Npi (1 − pi) + (1/4) N (pi − 1 + pi + 1) (2 − pi − 1 − pi + 1). Under the formula originally employed by Burgstahler and Dichev (1997, note 6)—where the first term in the last parentheses is 1 instead of 2—we obtained similar values, only slightly higher.

3 We measure the economic significance following Mitton’s (2022) suggestions: for indicator explanatory variables, we measure the change in the dependent variable as a percentage of its standard deviation, associated with a change from zero to one in the explanatory variable; for continuous variables, we measure the change in the dependent variable as a percentage of its standard deviation, associated with a one-standard-deviation change in the explanatory variable (based on the estimated regression coefficient). These standardised measures of economic significance avoid spurious inflation and abnormal high significance when the dependent variable is continuous and includes negative values.

4 The untabulated standard deviation of the dependent variable, ΔDTA, equals 0.0034 and 0.0038 for suspect firm-years reaching prior year earnings and zero earnings, respectively.

Additional information

Funding

This work was supported by Fondazione di Sardegna.