.png){kind=icon}
.png){kind=icon}
Return
Logarithmic-exponential utility for portfolio optimization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.eswa.2025.130439.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Expected utility maximization is a useful framework for incorporating the risk preferences of investors into portfolio optimization. However, designing suitable and practical utility functions remains challenging due to the diversity of risk preferences among different investors. In this paper, we propose a novel logarithmic-exponential (log-exp) utility function that incorporates decreasing absolute risk-aversion and increasing absolute risk-aversion into a single parameter value. It generalizes most commonly used utility functions that can only reflect a single absolute risk-aversion attitude. We also propose a log-exp utility-based sparse portfolio optimization model with reweighted ℓ1 regularization, and design a solving algorithm that achieves linear convergence to the global optimum. Extensive experiments on seven benchmark data sets show that the proposed method performs well in both return and risk control.
Journal
IF:
7.5
Papers:
2.9W
Citations:
10.2W