Normalized Expected Utility-Entropy investment decision model and its application in stock selection

Abstract

We first introduce the normalized Expected Utility-Entropy (EU-E) decision model, which is a weighted linear average of normalized expected utility and information entropy. Based on the normalized EU-E decision model, we establish a normalized EU-E investment decision model. Then we apply the model to stock selection when we invest in the 40 sample stocks of Shenzhen component index. It has concluded that portfolios of 4 stocks selected by normalized EU-E model with larger tradeoff coefficient λ are more efficient than that of those selected with smaller tradeoff coefficient λ with relative general utility function. Thus, this has demonstrated that we should not only take the expected utility of a risky action itself into account but also the information entropy to measure the uncertainty of the state of nature, which further verified the usefulness of the information entropy.