Abstract
When making investment decisions, securities investors are most concerned about the benefits they can obtain and the risks they must bear. In order to solve the limitations of Markowitz's classic securities portfolio model to measure investment risk based on the return and variance of securities, this paper analyzes the nature and characteristics of risk, based on the concept of thermodynamic entropy, and focuses on demonstrating entropy. After quantifying the rationality and superiority of securities investment risks, a securities investment portfolio model under entropy risk is constructed. This model carries out a brand-new optimization modeling of traditional securities portfolio models from the perspective of entropy and opens up new research perspectives. Afterwards, in view of the main defects of the model, the model was revised and optimized in the incomplete financial market in combination with the maximum entropy of information theory, and it was concluded that investors can reduce or resist risks by controlling the decentralized structure of the investment portfolio. Finally, based on our country's national conditions, some attention problems and solutions are given in the application of our country's securities industry.