A Comparative Study of Multi-objective Evolutionary Algorithms to Optimize the Selection of Investment Portfolios with Cardinality Constraints [chapter]

Feijoo E. Colomine Duran, Carlos Cotta, Antonio J. Fernández-Leiva
2012 Lecture Notes in Computer Science  
We consider the problem of selecting investment components according to two partially opposed measures: the portfolio performance and its risk. We approach this within Markowitz's model, considering the case of mutual funds market in Europe until July 2010. Comparisons were made on three multi-objective evolutionary algorithms, namely NSGA-II, SPEA2 and IBEA. Two well-known performance measures are considered for this purpose: hypervolume and R2 indicator. The comparative analysis also includes
more » ... an assessment of the financial efficiency of the investment portfolio selected according to Sharpe's index, which is a measure of performance/risk. The experimental results hint at the superiority of the indicator-based evolutionary algorithm.
doi:10.1007/978-3-642-29178-4_17 fatcat:6dhmauexafac5lvmwyixhcadwq