Figure - uploaded by Fernando Suarez
Content may be subject to copyright.
Source publication
We propose a new approach to portfolio optimization that utilizes a unique combination of synthetic data generation and a CVaR-constraint. We formulate the portfolio optimization problem as an asset allocation problem in which each asset class is accessed through a passive (index) fund. The asset-class weights are determined by solving an optimizat...
Contexts in source publication
Context 1
... following example will help to assess the merits of our approach vis-` a-vis other alternative asset allocation schemes. Consider the case of an investor who has access to ten asset classes (a diverse assortment of stocks, bonds and commodities) based on the indices described in Table 1. We further assume that the investor has a medium-to long-term horizon and that he/she will be rebalancing his/her portfolio (recalculating the asset allocation weights) once a year, which for simplicity we assume that is done at the beginning of the calendar year (January). ...Context 2
... further assume that the investor has a medium-to long-term horizon and that he/she will be rebalancing his/her portfolio (recalculating the asset allocation weights) once a year, which for simplicity we assume that is done at the beginning of the calendar year (January). We consider the period January 2003-June 2022, a time span for which we have gathered daily returns data corresponding to all the indices listed in Table 1. Finally, we assume that the investor will rely on a 5-year lookback period to, first, generate synthetic returns data (via the Modified CTGAN approach outlined in the previous section), and then, would rely on the linear optimization framework described in (5) to determine the asset allocation weights. ...Context 3
... following example will help to assess the merits of our approach vis-` a-vis other alternative asset allocation schemes. Consider the case of an investor who has access to ten asset classes (a diverse assortment of stocks, bonds and commodities) based on the indices described in Table 1. We further assume that the investor has a medium-to long-term horizon and that he/she will be rebalancing his/her portfolio (recalculating the asset allocation weights) once a year, which for simplicity we assume that is done at the beginning of the calendar year (January). ...Context 4
... further assume that the investor has a medium-to long-term horizon and that he/she will be rebalancing his/her portfolio (recalculating the asset allocation weights) once a year, which for simplicity we assume that is done at the beginning of the calendar year (January). We consider the period January 2003-June 2022, a time span for which we have gathered daily returns data corresponding to all the indices listed in Table 1. Finally, we assume that the investor will rely on a 5-year lookback period to, first, generate synthetic returns data (via the Modified CTGAN approach outlined in the previous section), and then, would rely on the linear optimization framework described in (5) to determine the asset allocation weights. ...