Multi-period portfolio optimization using model predictive control with mean-variance and risk parity frameworks

• Propose model predictive control with risk-parity objective transaction control. Derive a successive convex algorithm to efficiently solve the MPC. MPC strategies beat benchmarks in out-of-sample tests various asset universes. Multi-period planning provides advantage via over single-period planning. We employ for multi-period portfolio optimization problem. In addition mean-variance objective, we construct whose allocation is given by and provide program that 30 times faster robust solutions experiments. comprehensive comparison of models regard horizon, parameter estimation, as well function choice. Computational results on multi-asset universe show perform better than their single period counterparts period, 2006–2020, presence market impact costs. The risk-adjusted performance both formulations fix-mix benchmarks, achieve Sharpe ratio 0.64 0.97, respectively. also include different universes (Fama French industry portfolios) alternative estimation methods (Bayes-Stein Black-Litterman) consistent findings.