欢迎来到天天文库
浏览记录
ID:34908540
大小:381.53 KB
页数:4页
时间:2019-03-13
《Developing Portfolio Optimizaton Models in Matlab》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、UsingtheDowJonesIndustrialAverageasDevelopingPortfolioabenchmark,wewillimplementaportfoliooptimizationmethodologybasedoncapitalOptimizationModelsassetpricingandmean-varianceanalysis.Ourgoalsaretouseconsistent,repeatablestepsandtoconstructrealistic,optimalportfol
2、iosthatarestableovertime.ByBOBTAyLOREstimatingAssetReturnMomentsPortfoliooptimizationwasfirstdevelopedinthe1950s,butThemeanandcovarianceofassetreturnsareprimaryinputsforportfoliooptimization.anumberofpracticalandtheoreticalproblemshavelimitedEstimatingthesemomen
3、tsinvolvesthreetasks:acquiringdata,dealingwithmissingitsusebyinvestmentmanagers.Forexample,itisoftendata,andsettingupasuitablebenchmark.difficulttoobtainsufficienthigh-qualityhistoricaldataforAcquiringDatathoroughanalysis.Inaddition,theefficientfrontierwhereWeus
4、eMATLABandDatafeedToolboxtoobtainreturndataforstocksandmar-optimalportfolioslietendstoshiftovertime,quicklymakingketindexes.Inourexampleweacquiremonthlytotalreturndataon44blue-chiptheseportfoliossuboptimal.stocksandtheDowJonesIndustrialAver-Moderndataanalysistoo
5、ls,suchasMATLAB®andage(DJIA)fromYahoo!Finance.DealingwithMissingDataFinancialToolbox,canovercomethesechallenges.Unfortunately,historicalfinancialdataisoftenmessyandincomplete.WeusetheFinancialToolboxfunctionecmnmletodealwithdatasetsthathavemissingvalues(represen
6、tedasNaNsinMATLAB).ThisfunctionusesallavailabledatatoobtainbestestimatesforassetreturnmomentsinthepresenceofNaNs—anicealternativetotheusualadhocapproaches.SettingupaBenchmarkWeuseamarketindexasourbenchmark,sincemarketreturnisthemaindriverofassetreturnsincapitala
7、ssetpricing.Byremovingmarketreturnsfromthedatawecanfocusonnon-marketreturnsandrisks.InourexamplewesubtractthereturnoftheDJIAfromindividualassetreturns.UsingClassicMean-VarianceAnalysisInmean-varianceanalysis,expectedreturnisplottedagainstrisk(thestandarddeviatio
8、nofassetreturns)foragivenportfolio.Wegener-Figure1.Theefficientfrontier.aterandomcombinationsofportfolioweightstoproduceascatterplotoftheexpectedreturnandriskforeachp
此文档下载收益归作者所有