Financial Portfolio Model Based On Particle Swarm Optimization

Markowitz advanced his financial portfolio model in 1952 and also set up the beginning of ration analysis to financial investment. Now portfolio optimization problem turns to be more complex because of the higher demand proposed by the investors and the deeper research carried on by the scholars. Intelligence computation, such as GA and artificial neural network is introduced to solve this problem while traditional mathematics methods are in the mess.This thesis we suggest some improvements based on Markowitz's average-variance model according to the reality of China's stock market. We used a Hybrid Particle Swarm Optimization Algorithm joined crossover operator to solve this portfolio problem, which proved to more efficiently than GA by studying the stocks in Shanghai market. We also find that a portfolio with more assets has more probability to balance return and risk.According to the utility function, we present a multistage decision-making model that incorporates Particle Swarm Optimization (PSO) Algorithm into multi-stage portfolio optimization system, so that we can capture the variation of the market more quickly. Our system is demonstrated to be more efficiently by optimizing the allocation of cash and various stocks in Shanghai market and compared with other invest strategies. This model provides a realistic method to control the risk of different markets in multi-stages.