Article ID: | iaor20021544 |
Country: | United States |
Volume: | 48 |
Issue: | 3 |
Start Page Number: | 185 |
End Page Number: | 200 |
Publication Date: | Apr 2001 |
Journal: | Naval Research Logistics |
Authors: | Michalowski Wojtek, Ogryczak Wodzimierz |
Keywords: | programming: linear |
A mathematical model of portfolio optimization is usually represented as a bicriteria optimization problem where a reasonable tradeoff between expected rate of return and risk is sought. In a classical Markowitz model, the risk is measured by a variance, thus resulting in a quadratic programming model. As an alternative, the MAD model was developed by Konno and Yamazaki, where risk is measured by (mean) absolute deviation instead of a variance. The MAD model is computationally attractive, since it is easily transformed into a linear programming problem. An extension to the MAD model proposed in this paper allows us to measure risk using downside deviations, with the ability to penalize larger downside deviations. Hence, it provides for better modeling of risk averse preferences. The resulting