This study investigates the determinants and effects of voluntary book-taxdifference (BTD) disclosures in earnings releases. Unlike prior studies, I find noevidence that managers are more likely to voluntarily disclose BTD information whenfirms have low earnings quality. I also find that managers are more likely to discloseBTD information when firms have large negative but not large positive BTDs. BecauseBTDs are particularly informative when earnings quality is low and when book incomesignificantly exceeds taxable income (i.e., large positive BTDs), these results suggest thatmanagers selectively disclose BTD information in earnings releases. Interestingly, I alsofind that managers are more willing to disclose BTD information when tax avoidanceactivities are high. This result suggests that managers are willing to bear some taxrelateddisclosure costs to reassure investors that BTDs are not due to aggressivefinancial reporting. Prior research provides evidence of a systematic association between BTDscomputed using required 10-K tax disclosures and future forecast errors and stockreturns. I provide evidence that voluntary BTD disclosures attenuate the associationbetween BTDs and future forecast errors. I also provide limited evidence that voluntaryBTD disclosures attenuate the association between BTDs and future stock returns. Theseresults suggest that voluntary BTD disclosures help analysts and investors impound BTDinformation into earnings forecasts and stock prices.
