An empirical analysis of the valuation allowance for deferred tax assets under SFAS No. 109, 'Accounting for Income Taxes'

Positive accounting literature has provided empirical evidence regarding firms' characteristics that cause their managers to make particular accounting decisions. The purpose of this study is to extend this line of research to the area of accounting for income taxes. The research is designed to identify accounting choice variables that influence managers' decisions to change the level of the valuation allowance. Specifically, the study presents an empirical examination of whether managerial discretion over the valuation allowance appears to be used for earnings management purposes.; This study uses various univariate tests and multivariate estimation procedures to determine which factors influence managers' decisions to change the level of the valuation allowance. The analysis includes investigation of the relationships among observed valuation allowance changes and closeness to debt constraints, management compensation agreements, political sensitivity, "big bath" behavior, and income smoothing. Only firms that change the valuation allowance are included in the empirical analysis.; Results of both the univariate tests and multivariate estimation procedures provide no evidence of a relationship between binding debt covenants and observed valuation allowance changes. The results for the bonus plan hypothesis are mixed, implying that other factors used in awarding management bonuses alleviate the potential earnings management incentive of the annual bonus plans. While the findings provide no support for the political costs hypothesis, there is some support for the economic argument that large firms have more sources of income (e.g., carrybacks, tax-planning strategies) cited in SFAS No. 109 to conclude that a valuation allowance is not required. The income smoothing hypothesis is not supported by the univariate or multivariate tests. This study found that firms attempt to take a "big bath" rather than smooth their income in years when their earnings are below a normal earnings level. Overall, the results show that a good deal of variation in the valuation allowance is explained by the "big bath" motivation.