Mean-semi-entropy portfolio adjusting model with transaction costs

Considering the financial markets and investor's desire to reallocate wealth among the existing holding assets, in this study, we propose a mean-semi-entropy portfolio adjusting model, in which semi-entropy is employed to measure the downside uncertainty of portfolio. Transaction costs that are induced in the adjusting process are taken into account in model formulation to better trade off between risk and return.The calculation of semi-entropy is simplified by approximately converting it into fitting functions, and numerical analyses demonstrate that a polynomial one is the best approximation after comparing the fitting performance and complexity. To solve the model, the return of risky assets are captured by triangular fuzzy variables. By introducing a risk-averse factor, the proposed mean-semi-entropy portfolio adjusting model is transformed into a deterministic programming program which can be easily solved. Finally, numerical experiments with real data are provided to illustrate the effectiveness of the proposed model and results show that the adjusted portfolio is distributive which is required by investors in practice.