Conservatism and Book-Tax Differences

Abstract

I investigate the relation between conservatism and book-tax differences (BTDs), where “conservatism” refers to conservatism in the financial statement context (as opposed to the taxable income context). Specifically, “conservatism” refers to conditional and unconditional conservatism per Basu (1997). The primary objective of my study is to further the use of BTDs in financial statement analysis. I find that, on average, the conditional and unconditional financial statement conservatism of firm-years with large positive BTDs (book income is greater than taxable income) is similar to that of other sample firm-years, while the conditional and unconditional conservatism in taxable income of firm-years with large positive BTDs is greater than that of other sample firm-years. Thus, contrary to teachings in popular accounting textbooks, my main findings suggest that large positive BTDs do not, on average, reveal information about a firm's relative level of financial statement conservatism. I additionally find that, on average, the unconditional financial statement conservatism among firm-years with large negative BTDs (LNBTDs, taxable income is greater than book income) is greater than that of other sample firm-years, and, at the same time, the conditional and unconditional conservatism in taxable income among firm-years with large negative BTDs is less than that of other sample firm-years. Because the increased financial statement conservatism among firm-years with LNBTDs is unconditional in nature, and is not conditional in nature, the increased conservatism among firm-years in this subsample does not translate into higher earnings quality.