Tax-Aware Portfolio Construction via Convex Optimization

We describe an optimization-based tax-aware portfolio construction method that adds tax liability to standard Markowitz-based construction. Our produces a trade list specifies the number of shares buy each asset and sell from lot held. To avoid wash sales (in which some realized capital losses are disallowed), we assume monthly cannot simultaneously same asset. The problem is not convex, but it becomes convex when specify, for asset, whether or it. It can be solved using mixed-integer optimization methods at cost very long solve times instances. present custom relaxation borrows curvature risk model. This provide good approximation true liability, while greatly enhancing computational tractability. requires solution only two problems: first determines second generates final list. In our numerical experiments, almost always solves nonconvex optimality, does not, close optimal. Backtests show performance indistinguishable obtained globally optimal solution, with significantly reduced effort.