Evaluating Fund Performance by Compromise Programming with Linear-Quadratic Composite Metric: An Actual Case on The CaixaBank in Spain

This paper proposes an additive measure on the basis of compromise programming to evaluate fund performance from multiple criteria, of which the most usual are profitability and risk. This proposal is motivated by the fact that compromise programming is a sound decision support model to obtain scores of alternatives by minimizing weighted distances to an ideal point, the weights reflecting the investor's preferences for the criteria. To define the distance objective function, the linear-quadratic composite metric is used, which combines advantages of linear and non-linear objective functions. A critical advantage of compromise programming for fund performance evaluation is that the model can be extended to more than two financial criteria while other measures currently used (either ratio-based or leverage-based measures) only consider two criteria, say, profitability and risk. In the application, three investor's profiles are defined, which involve different weighting systems and lead to different fund rankings. These rankings are compared with domination relationships, the latter formulating if a fund is dominated or non-dominated by convex combinations of other funds. Numerical tables are provided with data, computational process and results, which are analysed. Copyright © 2012 John Wiley & Sons, Ltd.