We examine the magnitude and sources of difference between income for tax and financial reporting purposes using publicly available data from1988 to 1998. We find evidence that the book-tax income spread has generally increased over time, but that a relatively small set of variables are able to explain this increase. We also find that these same variables explain a large percentage of the variation in the book-tax spread across firms. While neither supporting, nor disproving, the existence and growth in tax sheltering behavior, the results do suggest that financial statement-based measures of income have become less representative of firms' taxable income.

This paper examines the ability of financial statement measures of average and marginal tax rates (MTR) to capture tax attributes utilizing firm-level tax and financial data. The results suggest commonly used average tax rate measures provide little insight about statutory tax burdens, and may introduce substantial bias into analyses of tax incidence. Financial statement-based proxies for MTR, particularly those based on simulation methods, are found to perform well in estimating current year tax rates. Both current year and present value MTR are found to be highly correlated with an easily constructed binary proxy of firms' tax status. r

No abstract

We hypothesize that when confronted with a loss, investors price earnings conditional on the likelihood of the firm's return to profitability. We argue such pricing is consistent with the abandonment option hypothesis as described by Hayn (1995) and show both the pricing of losses and their characteristics vary as a function of their expected reversal. We document a more pronounced stock price response to transitory losses (i.e., losses likely to reverse), consistent with investors assessing the likelihood of exercising the abandonment option to be smaller. However, we also find evidence consistent with investors pricing persistent losses (i.e., losses not likely to reverse) negatively, a result inconsistent with the abandonment option hypothesis. Further analysis shows investors price the components of losses differently depending on the likelihood of reversal. Aggregate accruals explain the pricing of persistent losses while aggregate cash flows explain the pricing of transitory losses. The result for persistent losses relates to the presence of an increasingly larger R&D component: investors reward firms that make larger R&D outlays with larger returns. One consequence of the growing R&D component in persistent losses is that they have become a weaker indicator of the likelihood of exercising the abandonment option.

No abstract