Research On Portfolio Application Based On Improved Artificial Bee Colony Algorithm

A portfolio is the allocation of assets in a way that satisfies the trade-off between risk and return.Markowitz proposed the mean-variance model of portfolio,which changed the descriptive qualitative analysis to scientific and rigorous quantitative analysis,and could obtain the distribution of random variables according to historical data.With the development of economy,financial investment has become an important part of China’s market economy.Therefore,the research and analysis of investment portfolio has important theoretical significance and practical value.In the process of studying the portfolio problem,more and more complex models are put forward,and simple mathematical methods are difficult to solve,while artificial bee colony algorithm is widely used in optimization problems because of its few parameters and simple structure,so this paper chooses artificial bee colony algorithm to solve the portfolio model.The main contents and innovations of this paper are as follows:(1)The mean-variance-skewness model of a single target is established.Considering the skewness distribution of return rate and the risk preference of investors,on the basis of the mean-variance theory proposed by Markowitz,the existing portfolio model is improved,and the mean-variance-skewness coefficient and risk aversion coefficient are introduced to establish a single-objective mean-variance-skewness model.In the mean-variance-skewness model,the skewness distribution is described by the third-order central moment,and investors are divided into three types: risk preference,risk neutrality and risk aversion,which can more accurately present the actual situation of the securities market.When solving the model,the penalty function method is used to transform the model,which makes the solving process easier.(2)An improved artificial colony algorithm based on slime mold and Marine predators is proposed.In order to solve the portfolio model better,an improved artificial bee colony algorithm is proposed.The slime mold foraging strategy was introduced in the search process of the leading bees,and the leading bees generated positive and negative feedback according to the concentration of food source,generated adaptive weights,adjusted the search mode,and searched in the direction of the optimal solution,which accelerated the convergence speed.In the reconnaissance bee stage,the aggregation effect of fish devices in the Marine predator algorithm is introduced to avoid optimization stagnation.The simulation results of 15 benchmark test functions show that the improved artificial bee colony algorithm has faster convergence speed,higher convergence accuracy and better stability.(3)The mean-variance-skewness model was optimized by basic artificial bee colony algorithm and improved artificial bee colony algorithm respectively and applied to portfolio application.Five stocks with good momentum,including petrochina,Yaling Pharmaceutical,Ningde Times,Bank of Chengdu and Luzhou Lajiao,were selected to calculate the 100-day return rate of each stock with daily closing price.Then,the basic artificial bee colony algorithm and improved artificial bee colony algorithm were used to solve the mean-variantbias model,and the optimal portfolio was obtained.The empirical analysis shows that the improved artificial bee colony algorithm can select the portfolio with higher expected rate of return,smaller variance and higher skewness,which is beneficial for investors to obtain greater benefits in the securities market.