Research On The Optimization Of Investment Portfolio Based On Improved Particle Swarm Algorithm

Investors want to gain as much as possible and face as little risk as possible,but high returns are often accompanied by high risks,especially stock investments known for high returns and high risks,and people try to reduce risk by building stock portfolios.However,no matter how many stocks are included in the portfolio,no loss can be avoided in the downward trend of the economy.The emergence of options brings a glimmer of vitality to many shareholders,because options not only have the high leverage of general derivatives,but also have the function of protection.In February 2015,the first non-commodity option in China-Shanghai Stock Exchange 50 etf option was listed and circulated.Because of the immature capital market in China,there are many restrictions on the trading of 50 etf options in Shanghai Stock Exchange,and the delivery mode is not the traditional price spread delivery,but the physical delivery.The cost of buying and selling the real goods will even bring loss to the profits.Therefore,it is necessary to verify whether the 50 etf option of Shanghai Stock Exchange can be as effective as the option in theory.As an artificial intelligence optimization algorithm,Particle Swarm Optimization(PSO)is more and more widely used in solving various problems because of its fast convergence rate,easy realization and no requirement of differentiability,derivability and continuity of the optimized function.But the portfolio problem is more and more complex,and the performance requirements for solving algorithms are getting higher and higher,so it is necessary and meaningful to improve the particle swarm optimization algorithm.The main work is done as follows:(1)Combined with the actual transaction cost of Chinaundefineds capital market and the limitation of the number of securities,this paper puts forward some constraints,such as considering the transaction cost and the limitation of the minimum trading unit of securities,and integrates into the mean-CvaR model of 50 etf options in Shanghai Stock Exchange.(2)Combine the two optimization ideas based on complex PSO and dynamic factor PSO,proposes a complex PSO algorithm based on adaptive weight dynamic factor,and uses the benchmark function to simulate and test the improved algorithm.The simulation results show that the improved algorithm has the best solution result and the fastest convergence rate,that is,the improved algorithm has better performance.(3)Select 8 stock from Shanghai 50 etf components and 50 etf call and put options of Shanghai Stock Exchange to build a portfolio,and the improved algorithm and the two algorithms referenced in the paper are coded and adjusted,so that the algorithm is suitable for the solution of discrete problems such as portfolio,and the capital market is on the rise by using the three adjusted algorithms,respectively.The investment portfolio of the added option is solved under the three states of descent and steady oscillation.The validity of options in this market state is judged by the number of options in the solution result,and the performance of the algorithm in portfolio solution is judged by the adaptive value algorithm corresponding to the solution results.The empirical results show that when the capital market is on the rise,options can fully play the role of leverage,that is,the portfolio of options can obtain high returns,and when the capital market is in a downward trend,the addition of options can reduce the loss.In this time,options play a full role in safeguarding,that is,under these two trends,options are effective.The algorithm combined with two improved ideas gives full play to the advantages of the two ideas.No matter solving the continuity problem or the discrete problem such as portfolio,we can always get the minimum adaptive value,that is,relative to the other two algorithms.The performance is optimal.