An Empirical Study Of Index Replication Based On Optimized Genetic Algorithm

Index replication refers to the behavior of constructing a portfolio of assets in real time to track the rate of return of a benchmark index in the underlying market.It belongs to the category of passive investment.Because of its benchmark market income,it has many advantages such as relatively small risk,low transaction rate and strong liquidity.With the overall downturn in the A-share market in 2018 and the frequent occurrence of black swan events,the overall performance of stock-based active funds has been sluggish,and even a large number of star products have caused huge losses to investors.Nearly 60%of active equity funds underperformed the market index.This series of situations has sounded the alarm for investors,making them aware of the safety and reliability of indexed investments,and the market's attention is getting hotter.However,relying on the complete replication of the index constituents one by one is subject to many obstacles and cannot be well indexed.In this case,investors try to use the optimal planning replication method to transform the index replication problem into the optimal programming problem.Solve which constituents are selected to characterize the index and how many constituents to choose to build the tracking combination.When the method of exponential replication is determined,it is extremely critical to use which numerical optimization solution technique to solve the optimal value.This paper should be produced in this context,the purpose is to clarify that the optimal replication method based on the optimization genetic algorithm can be used for index replication.Secondly,the adaptive genetic algorithm is superior to the classical genetic algorithm in solving the stochastic numerical optimization.This method can be a good guide for daily indexed investment research.In the research process of the thesis,theoretical analysis and empirical analysis are combined.Firstly,based on the literature review method,the relevant research results in the current academic circles are summarized,and the methodological methods related to index replication are sorted out to clarify the relevant methodological contexts.Based on the previous studies,the research is expanded.Secondly,the insufficiency of the tracking error deviation index is proposed,and the simulation experiment is carried out to prove the problem,and then the solution is proposed to guide the subsequent empirical research.Finally,the empirical part is based on the benchmark index of investment feasibility,designing the control group according to the idea of control variables,and using the combination of classical genetic algorithm as the control group.In the empirical aspect,the genetic algorithm is described in detail,and combined with the exponential replication optimization to solve this specific problem,the key parameters are analyzed in detail and the specific role is explained.Based on the classical genetic algorithm,an optimization scheme of adaptive genetic algorithm is proposed.The crossover operator and mutation operator of classical genetic algorithm are adaptively updated,and the algorithm logic framework is give.The optimized genetic algorithm is used for the tracking solution of the optimal planning replication method and the excellent tracking effect is obtained.It can be seen that the crossover operator and the mutation operator in the genetic process are set to adaptive update.Compared with the fixed value,there may be Help the algorithm to better converge to solve the global optimal solution.The whole backtesting period is from 2009/01/01 to 2019/01/01,based on the 10-year long-cycle backtesting,and the survivor's biased display is also avoided.From the empirical backtesting results,the solution based on the optimized genetic algorithm is obtained.The tracking combination is reduced by 5%on the tracking error compared to the tracking combination solved by the classical genetic algorithm.From the perspective of the empirical results of the thesis,the exponential replication investment method based on the optimal genetic algorithm for solving the optimal planning replication method proposed in this paper is feasible.The empirical results show that this method can be used to conduct related research on investment.The half-year adjustment frequency,the small number of stock positions,and the low tracking error also make the combination of the optimized genetic algorithm and the optimal planning replication method worthy of public expectation.