Accrual Reversal Effect and Conditional Conservatism

Abstract

This paper examines the relationship between two salient features embedded in the modern financial accounting information system: accrual reversal and accounting conservatism. This relationship is analyzed using a moral hazard model in a single-period setting and two types of two-period models: pooling and separating. When the effect of accrual reversal is considered in the long term, accounting conservatism as an information bias was found to be an optimal choice for the principal in a two-separating-period setting, particularly when business risk is high and/or the informativeness of the accounting information system is low; however, accounting conservatism could never be used as an information bias under a single-period or two-pooling-period settings, even with limited liability conditions. These findings indicate that accrual reversal could be considered a driving force for conditional accounting conservatism, but not for unconditional conservatism. Moreover, accrual reversal may provide an explanation for the seemingly contradictory behavior of accounting standard-setting bodies that introduced conditional conservatism, although expressing negative attitudes toward accounting conservatism.