This paper examines individual investors’ short- and long-term trading reactions to the 1986 Tax Reform Act's (TRA 86) capital gain tax rate increase. Consistent with a tax-induced trading model, we document a December 1986 change in year-end trading patterns. TRA 86 also had long-term effects on individual investors’ trading decisions. The results suggest that individual investors were less (more) willing to sell accrued gain (loss) stocks after TRA 86. Our ability to document predicted trading behavior is in part due to our trading metric, which captures individual investors’ selling activity better than the volume-based metrics used in prior research.

This paper presents a model of optimal tax-motivated intertemporal income shifting given a quadratic cost function that relates the costs associated with shifting income to the amount of income shifted. By formally modeling the income-shifting decision, we: (1) show how parameter estimates of the income-shifting cost function can be extracted from a linear regression where a proxy for income shifted is the dependent variable, (2) provide insight into prior tax-motivated income-shifting research, and (3) clarify the interpretation of independent variables that capture the interaction between tax incentives and nontax costs. We then provide an empirical application of our method for quantifying the costs to shift federal taxable income by investigating the income-shifting behavior of firms in the property and casualty (P&C) insurance industry following the Tax Reform Act of 1986. Our results suggest that the parameters of the cost function are negatively related to firm size, the cost to shift a significant amount of income is nontrivial, and the marginal cost to shift income increases as more income is shifted.

Measurement error in unexpected earnings is recognized as a source of bias in examinations of the relation between earnings and returns. Reverse regression procedures are commonly used as a means of coping with this bias. This study examines the properties of reverse regression procedures in multi-interacted variable settings with a specific focus on the earnings response coefficient (ERC) analysis of Collins and Kothari (J. Account. Econom. 11 (1989) 143.). It shows that both conventional reverse regression techniques and novel techniques employed by Collins and Kothari are not robust. It also demonstrates how reverse regression techniques can be successfully employed in such settings using non-interacted-variable designs.