Content area
Full Text
This research empirically examines whether firms' capital expenditure patterns are consistent with the timing patterns that minimize taxes. Specifically, if marginal tax rates remain constant over time, the present value of investment-related tax shields is greater if a depreciable asset is placed in service in the current tax year rather than in a subsequent tax year. As a result, firms can reduce the net cost of any capital expenditures that can be flexibly timed around the turn of the tax year if they are made late in the current year (year t) rather than early in the following year (year t + 1). Thus, we predict that, ceteris paribus,(1) firms make more capital expenditures in the fourth quarter of year t than in the first quarter of year t + 1.
Moreover, MacKie-Mason (1990) finds that a firm does not attempt to maximize its current year's tax shields unless it is paying taxes at a high rate. Based on this evidence and given the large first-year tax benefits of an investment in a depreciable asset, we predict that high-tax firms make more of their turn-of-the-year capital expenditures late in the current tax year than low-tax firms.
Lastly, fixed assets often serve as collateral for debt financing, especially for high-tax firms (Dhaliwal et al. 1992). As a result, high-tax firms should issue long-term debt in a pattern that coincides with the timing of their capital expenditures. That is, high-tax firms should issue more long-term debt in the fourth quarter of year t than in the first quarter of year t + 1.
We find support for these predictions based on tests of the observed quarterly timing of firms' capital expenditures and long-term debt issuance from 1979 through 1989. In other words, the timing of tax shield recognition to minimize taxes is documented in this study. This supports the contention that in the hierarchy of responses to tax rules, "[s]tanding at the top of the hierarchy, the most clearly responsive to tax incentives, is the timing of economic transactions" (Slemrod 1990, 6).
Our results are important for several reasons. First, Scholes et al. (1992, 4) find that "while income-shifting occurs, it does not shift uniformly across income and expense items." The items tested in Scholes et...