Does accounting for uncertain tax benefits provide information about the relation between book-tax differences and earnings persistence?

Purpose – This study aims to use research setting provided by the implementation of Financial Accounting Standards Board Interpretation 48 (FIN48) to help develop a further understanding of large positive book–tax differences (LPBTD) and their relationship with earnings persistence. Extant literature indicates that the tax information provided in financial statements, such as large book–tax differences, is useful for detecting earnings management and signals less persistent future earnings. However, more information is needed about the causes of large book–tax differences and their abilities to signal the differences in earnings persistence (Blaylock et al. , 2012). Design/methodology/approach – In the first step, temporary book–tax differences are ranked by quintiles based on the approach in Hanlon’s (2005) study and the highest quintile in the sample observations are designated as large positive temporary book–tax differences (LPBTD). In the second step, differences in the persistence of earnings for high tax-planning firms as measured by UTB_NonETR are searched for. In further testing, an ordered logistic model and the Vuong (1989) test are applied to compare both the incremental and the relative ability of UTB_NonETR and Cash-ETR to explain the ranking order of temporary book–tax differences. Findings – The negative relation between temporary differences and earnings persistence is moderated by the level of tax planning as measured by UTB_NonETRs. More specifically, the persistence of earnings appears to be higher for firm-years with large UTB_NonETRs. When comparing the relative power of UTB_NonETR with Cash-ETR, the results indicate that UTB_NonETR is incrementally useful for explaining the ranking orders of temporary book–tax differences. However, it appears that neither UTB_NonETR nor Cash-ETR is relatively more useful over another under the Vuong (1989) test. Originality/value – First, the part of UTB, if recognized, that would not affect earnings (UTB_NonETR) is used as an empirical proxy and its usefulness is tested in the context of book–tax differences and the persistence of earnings. Second, new evidence is provided supporting the predictions, as in Ayers’ et al. (2010) and Blaylock et al. ’s (2012) studies, that the level of tax planning will attenuate the negative association between large book–tax differences and earnings quality. Third, the findings can contribute to the post-implementation review of FIN48 (Financial Accounting Foundation, 2012) supporting the argument that FIN48 can provide decision-useful information for financial statement users.