An interaction-based combined portfolio strategy with applications to stock markets

Each portfolio model has its own advantages. Behavioral portfolio models can depict investor psychology, but it is not easy to obtain the preference parameters. We propose an interaction-based combined (InCo) portfolio that determines the proportions of assets based on investor preferences. We take the existing models as alternatives and innovatively develop a visualization tool to present the portfolio performance. Hesitant fuzzy set theory is used to describe investor evaluations and to then determine the proportions of alternatives. The out-of-sample performance of four strategies is tested on four datasets. The results show that InCo portfolio performs better in various markets.