A new behavioral model for portfolio selection using the Half-Full/Half-Empty approach

In this paper our goal is to compare, in the context of portfolio selection, the so-called behavioral models (Prospect Theory [1] and Half-Full/Half-Empty model (HF-HE) [2]) with the traditional portfolio approaches (Minimum- variance portfolio [3], Mean-absolute deviation portfolio [4] and portfolio equally weighted [5]) to incorporate investors' perception of risk into the choices of optimal portfolios. In decision-making under risk problems we analyze theoretical and prac- tical aspects of the HF-HE model which allows modeling the cognitive distor- tions of individuals based on the simple concept provided by the HF-HE glass metaphor: a DM is optimistic when evaluating the outcomes of a generic lottery above its expected value or pessimistic when evaluating the outcomes of a generic lottery below its expected value. Out-of-sample analysis is carried out on four equity markets (two from the US, one from the UK and one from Europe) where, for testing and validation, we compare the performance obtained by the HF-HE model with that of the PT and the standard portfolio selection models. We are able to demonstrate through this research work, how the superi- ority of the HF-HE model is due to the possibility of inuencing the results of the analysis by calibrating the model parameters. Indeed, the HF-HE model has proved to be very sensitive to the value of the DM's degree of optimism/pessimism.