The Portfolio Selection Problem is amongst the most studied issues in finance. ealistic portfolio selection under short selling is recently receiving more at- tention amongst scholars, but often the computational phase is not discussed properly in terms of strategies used and models at hand; furthermore most approaches are aimed in determining single risk/return points. The aim of this paper is instead to draw out the whole efficient frontier in pres- ence of realistic short selling constraints. We extend our previous metaheuristic approach for Portfolio Selection to the case of short sellings and we propose a new set of benchmark instances, constructed from real-world market data. Our solver favorably compares with a Mixed Quadratic Programming formulation of the problem solved with IBM ILOG CPLEX 12.2.
Local Search for Constrained Financial Portfolio Selection Problems with Short Sellings / L. Di Gaspero; G. di Tollo; A. Roli; A. Schaerf. - STAMPA. - 6683:(2011), pp. 450-453. (Intervento presentato al convegno LION 5 tenutosi a Roma nel Jan 17-21, 2011) [10.1007/978-3-642-25566-3_34].
Local Search for Constrained Financial Portfolio Selection Problems with Short Sellings
ROLI, ANDREA;
2011
Abstract
The Portfolio Selection Problem is amongst the most studied issues in finance. ealistic portfolio selection under short selling is recently receiving more at- tention amongst scholars, but often the computational phase is not discussed properly in terms of strategies used and models at hand; furthermore most approaches are aimed in determining single risk/return points. The aim of this paper is instead to draw out the whole efficient frontier in pres- ence of realistic short selling constraints. We extend our previous metaheuristic approach for Portfolio Selection to the case of short sellings and we propose a new set of benchmark instances, constructed from real-world market data. Our solver favorably compares with a Mixed Quadratic Programming formulation of the problem solved with IBM ILOG CPLEX 12.2.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.