TY - UNPB

T1 - On the valuation of tax expense

AU - Gao, Zhan

AU - Givoly, Dan

AU - Laux, Rick

PY - 2015

Y1 - 2015

N2 - A growing body of research documents that tax expense is positively associated with contemporaneous stock returns, suggesting that an increase in tax expense is “good.” However, it is difficult to reconcile opposing explanations for the anomalous association because prior studies employ a variety of tax expense measures. We first demonstrate that, controlling for pretax income, the tax expense measures used in prior studies can be interpreted as a test of whether book-tax differences provide incremental information to the market. We then devise a specification which clearly distinguishes between temporary and permanent differences and examine the value relevance of each component. We find that temporary differences can be either positively or negatively associated with contemporaneous returns, depending on if they are calculated using current or deferred tax expense, respectively. We further demonstrate that the positive association between returns and permanent differences documented in prior research only exists when there are extreme changes in permanent differences which are likely caused by unusual events.

AB - A growing body of research documents that tax expense is positively associated with contemporaneous stock returns, suggesting that an increase in tax expense is “good.” However, it is difficult to reconcile opposing explanations for the anomalous association because prior studies employ a variety of tax expense measures. We first demonstrate that, controlling for pretax income, the tax expense measures used in prior studies can be interpreted as a test of whether book-tax differences provide incremental information to the market. We then devise a specification which clearly distinguishes between temporary and permanent differences and examine the value relevance of each component. We find that temporary differences can be either positively or negatively associated with contemporaneous returns, depending on if they are calculated using current or deferred tax expense, respectively. We further demonstrate that the positive association between returns and permanent differences documented in prior research only exists when there are extreme changes in permanent differences which are likely caused by unusual events.

M3 - Working paper

BT - On the valuation of tax expense

ER -