基于遗传算法的投资组合模型及实证研究
Abstract
Portfoliobasedonuncertainenvironment,anyincomewillbeaccompaniedby
certainrisks,inordertospreadagreatriskandgetamorcstableinvestmentincome,canbeinavarietyofsecuritiesindifferentproportiontocarryoutinvestment,thekeyofportfolioisbasedonthepreferencesofinvestors,makeabanlancebetweenincomeandrisk,identifyavarietyofsecuritiesintheportfolioofproportion,SOcangetasatisfiedportfolio.
Thegeneticalgorithmisakindofsearchingmethodwhichsimulatesthenatural
evolution.Itssimpleandeasytoimplement,especiallyitdoesn’tneedthespecialfieldknowledge,butsimpleusesfitnessfunctionasanassessmenttoinstructthesearchprocess,SOithasbeenusedinverybroadfields.Nowthegeneticalgorithmhasgotalotoffruitsandscholarsbegintopayattentiontoit.Thispapergivesabriefintroductiontothebasicconceptionandtheoriesofthegeneticalgorithm.Anditalsoprovidesadetailedexplanationonthecomponentsofthegeneticalgorithm,withtheemphasisonthestrategiesofselection,crossover,mutationandthebasictheorem--schematheorem.
Thispapercombinesportfolioandgeneticalgorithm,andbringupaportfolio
basedongeneticalgorithm.Throughstudyofsomeportfoliomodel,consideringthepractrcalityofmoderportfolioselectiontheory,weproposeaconstrainedportfolioselectionmodelbasedonthenoshortsale,transactioncostsandminimunlotconstraints.Duetothesecomplexconstrainsourproposedproblembecomesamixedintegeroptimizationproblemandtraditionalalgorithmcan’tsolveitefficiently.Thus,allimprovedgeneticalgorithmisdesignedtosolveourproposedproblem.
Atlast,fromChina’Ssecuritiesmarketasevidence,Canbeseetheportfolioand
algorithmthatweproposedisfeasibilityandvalidity.
Keywords:portfolioinvestment,geneticalgorithm,investmentconstrain