MULTI-PERIOD MEAN–VARIANCE PORTFOLIO OPTIMIZATION BASED ON MONTE-CARLO SIMULATION

• Main Author: Naufal Daffa Andarwan, Muhammad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/74348
• Institution: Institut Teknologi Bandung

Mean-variance optimization or usually known as the Markowitz portfolio model was a cornerstone of modern portfolio theory. This model helped investors create mathematically optimal portfolios. Generally, the Markowitz portfolio model provided investors with an optimal portfolio for a single time period. In this final project, we would extend this model to multiple periods, allowing investors to perform asset re-balancing more than once. The writer would explore portfolio optimization cases with and without constraints. The multi-period problem in this final project was modeled with multi-stage strategy developed by (Cong, 2016), with the addition of the Karush-Kuhn-Tucker (KKT) method for constrained cases. The portfolio optimization cases considered were unconstrained and bounded leverage constrained cases. The numerical simulation would l be performed using the Monte Carlo method. The results obtained for the unconstrained and non-periodic contribution optimization problems were consistent with the pre-commitment analytic solution demonstrating the equivalence of the two methods. For the constrained and with periodic contribution cases, the additional Kurish-Kuhn-Tucker method was consistent with the benchmark result in Cong and Oosterle (2016). The highest efficient frontier was achieved in the unconstrained and then bounded leverage cases.