The Persistence and Value Relevance of Earnings From Tax Savings

Abstract

This dissertation examines the persistence and value relevance of earnings attributable to tax savings and the extent to which this persistence and value relevance differs from those of nontax earnings. After controlling for factors previously shown to be systematically associated with the tax component of earnings, results show that tax savings are significant and statistically persistent but statistically less persistent than earnings from nontax sources. Results also reveal that the persistence of tax savings changes across tax regimes whereas earnings from nontax sources remain relatively unchanged. Contextual analysis shows that (1) the persistence of tax savings is largely driven by firms in the pharmaceutical, oil and gas, financial services, insurance and real estate industries, (2) the persistence of tax savings is increasing in the R&D tax credit and (3) this persistence is increasing in settings where the ratio of foreign over domestic earnings is increasing. Additionally, the persistence and value relevance of tax savings is increasing for positive tax savings, implying a market reward (penalty) for lower (higher) tax savings (reported effective tax rates). When I compare the results from my valuation tests with those from my persistence tests, I find that tax savings are sometimes not persistent but value relevant and sometimes persistent but not value relevant whereas the persistence and value relevance of nontax earnings are always consistent. These findings are consistent with managerial opportunistic behavior, a market that suspects managerial opportunistic behavior or a stock market that does not understand fully the persistence of tax savings relative to nontax savings. Results from the Mishkin (1983) test show that the market appears to significantly overestimate both the persistence of tax savings and nontax earnings, implying that securities are mispriced. This potential mispricing appears to be more severe for tax savings, implying that, on average, the market does not appear to understand fully the persistence and value relevance of the tax component of earnings. Finally, this study reconciles some of the mixed results of prior research and carries significant implications for policy makers, firm management, market participants and accounting researchers.