An optimal portfolio mix for stable and turbulent times: A study on selected Philippine assets using modified Markowitz mean-variance model for the years 2001-2010

• Main Authors: Apolonio, David Renan T., Serrano, Jose Leandro D., Uy, Samuel Johnson Y.
• Format: text
• Language: English
• Published: Animo Repository 2012
• Subjects: Portfolio management > Philippines
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/9690
• Institution: De La Salle University

The study in 1953, Harry Markowitz introduced the mean-variance optimization model. This study addresses the problems of the original model by segmenting the data and introducing different risk regimes. The study adopted a modified Markowitz model that was introduced by Chow, et al. Multivariate outliers were identified in order to segment the data into stable and turbulent regimes. The researchers used six asset classes for a period of ten years. These asset classes were thus used to estimate a portfolio from the full sample returns in order to give the optimal portfolio for stable periods and also to estimate a portfolio from the outlier-sample returns to give the optimal portfolio for turbulent periods. Furthermore, the researchers blend the portfolio estimated from the full-sample with the portfolio estimated from the outlier-sample in order to minimize the trade-off between return and risk while incorporating the probability of occurrence for each period as well as the investor's risk aversion. In all of the portfolios, it was observed that majority of the assets consisted mostly of foreign exchange. Thus, a BSP constraint was applied to foreign exchange to further diversify the assets and make the portfolio more realistic.