An analytical derivation of properly efficient sets in multi-objective portfolio selection (Q7002498)

The paper deals with the multi-objective portfolio selection and the corresponding multi-objective optimization problem. The goal is to show that the sets of properly and weakly efficient solutions of the optimization problem with quadratic and linear objectives, say\N\[\N\min \{ x^T\Sigma_i x\}, \, i=1,2,\N\]\N\[\N\max \{ c_i^Tx \} , \, i=1, \dots ,l ,\N\]\N\[\N\text{s.t. }A^T x = b ,\N\]\Ncan be characterized in terms of the weighted-sums method\N\[\N\min\{ \lambda_1X^T \Sigma_1 x + \lambda_2X^T \Sigma_2 x - \sum_{i=1}^l \lambda_{i+2} c_i^T x \}\N\]\N\[\N\text{s.t. }A^T x = b, \, \lambda_i \geq 0.\N\]\NThis is performed under the assumption that the matrices $\Sigma_1$ and $\Sigma_2$ are positive definite and the rank of the matrix $(A,c_1, \dots , c_l\}$ equals rank$(A) +l$.