Portfolio Optimization Subject to Tax Bracket Constraints: A Linear Programming Approach.

Abstract Conventional investment and tax-planning wisdom holds that the marginal tax rate is the critical variable in planning for the next dollar of taxable investment. The traditional approach to maximize after-tax returns is to invest incremental investment dollars in the asset yielding the highest after-tax return. In this paper we develop and implement a linear programming model to serve as an alternative to traditional analysis. Given a regime characterized by progressive tax rates, we argue that as an investor's taxable income exceeds the upper boundary of a tax bracket, optimal strategy results from a comparison of the after-tax return from an intra-bracket investment shift relative to the after-tax return from the traditional, or inter-bracket, approach. Our analysis suggests that tax planners may wish to jointly consider intra-bracket and inter-bracket allocations of investment dollars.