MS-智能手机交易技术、投资者行为和共同基金表现（英）_市场营销策划_2024年市场报告-3月第4周.docx

informsMANAGEMENTSCIENCEArticlesInAdvance,pp.1-20ISSN0025-1909(print),ISSN1526-5501(online)SmartphoneTradingTechnology,InvestorBehavior,andMutualFundPerformancePgAJ4>SaJSlq8=B.AwosnWUOSdJO1.E寸：EZIE"寸ZoZIPjoNOeUOSGZOCZ一Z/二Z-Aq8JOSWJOJ=UIO±PaPEO-UMOaXiaoCenaaMaysBusinessSchl,TexasA&MUniversity,CollegeStation,Texas77843Contact:xcenmays.tamu.edu,https:/orcid.org/0000-0003-3915-5889(XC)Keywords:smartphonetechnologyinvestmentmutualfundfintechinvestorbehaviorAbstract.Usingproprietaryindividual-leveltradingdataaroundanaturalexperimentthereleaseofasmartphonetradingappbyalargeinvestmentadvisorthisstudyinvestigateshowsmartphonetradingtechnologyaffectsretailinvestorbehaviorandmutualfundperformance.Appadoptionbyretailinvestorsleadstoanincreaseininvestorattentionandtradingvolume.Appadopters"flowsbecomemoresensitivetoshort-termfundreturnsandmarketsentiment,resultinginhigheraggregateflowvolumeamongadopters.Thefundsmoreexposedtotheshockexperienceagreaterdeclineinabnormalreturns,likelyattributabletohigherfundflowvolumeandliquiditycosts.Asaresult,bothadoptersandnonadoptersexperienceadeclineintheirmutualfundinvestmentreturns.History:Acceptedby1.inWilliamCong,finance.Funding:TheW.EdwardsDemingCenterandtheEugene1.angEntrepreneurshipCenteratColumbiaUniVerSityprovidedfundingtosupportthisresearch.SupplementalMaterial:Theonlineappendicesanddatafilesareavailableatmnsc.2021.02099.Received:June26,2021Revised:June21.2022;March28,2023Accepted:May29,2023PublishedOnlineinArticlesinAdvance:November28,2023Copyright:©2023INFORMStheonehand,traditionalpersonalmputer(PC)tradingalreadyprovideshigh-speedexecutionWithlowtransactioncosts.TheincrementalcostreductionassociatedwithgoingmobilemaynotsignificantlyaffectinvestortradingdecisionsifthetradingappsprovidethesameinformationcontentasPCplatforms.Howzzgoingmobile,zaffectsinvestortradingdecisionsandmutualfundperformanceisanopenquestion.OnOntheotherhand,theconvenientaccessandintuitiveinterfaceofthesmartphoneappsmayleadtoincreasedusageofatradingplatformandheightenedinvestorattention.Thisgreaterexposuretomarketinformationmayamplifytradingpatternsofmutualfundinvestors(e.g.,flowperformancesensitivity),whichcouldresultinhigherflowvolumeandliquiditycostsatthemutualfundlevel(Edelen1999zCovalandStafford2007).Toinvestigatetheeffectofsmartphonetradingoninvestorbehaviorandfundperformance,IexplorethereleaseofasmartphonetradingappbyoneofChina'slargestinvestmentadvisorsasashocktoinvestors/adoptionofsmartphonetradingtechnology.TheappprovidesinvestorswithamoreflexiblewaytoaccessmarketinformationandexecutetradesmparedwiththeexistingonlinetradingplatformforPCs.Istartbyusingadifference-in-differences(DID)frameworktocomparethebehaviorchangesofappadopters(investorswhoswitchedfromPCtradingtotheapp)andmatchednonadopters(investorswhocontinuedtouseonlythetraditionalonlineplatform)followingthereleaseoftheapp.ThisapproachenablesmetoisolatePgAJ4>SaJSlq8IW.AwoUsnWUOSJodJO工E寸：EZIE"寸ZoZq2rt工OZcoSGZOCZIZ二Z-Aq=PUPEO-UMOatheincrementaleffectOfgoingmobilewhilecontrollingforinvestorheterogeneityandsomekeytradingplatformfeatures(e.g.ztransactionfeesandexecutionspeeds).Myanalysisindicatesthatfollowingthereleaseoftheappzadoptersdemonstrateasignificantincreaseinattention,asmeasuredbythenumberofloginstothetradingplatform,aswellasintradingvolume,relativetononadopterswithsimilarpredeterminedcharacteristics.Interestingly,goingmobilealsoaffectsadopters/responsestoshort-termmarketsignals.Forexample,relativetononadoptersaftertheshock,theadopters'flowdisplayssignificantlyhighersensitivitytofundreturnsinthepreviousweek,whichhasnotbeenpreviouslydocumented.Additionally,adopters'flowsensitivitytomarketsentimentincreasescomparedwiththatofnonadopters.Furtheranalysesbasedonthecross-individualvariationinincrementalattentionsuggestthatincreasedinvestorattentionlikelycontributestotheobservedchangesintradingbehavior.Thepreviousanalysishasonecaveatinvestorsmayself-selectintoadoptingtheapp,leadingtoendogeneityconcerns.Toaddressthisissue,Iusethevariationintreatmentintensityaroundtheborderbetweentwoprovincesdifferentiallyexposedtothetechnologyshock.Thevariationinexposurearisesfromtheheterogeneityinmobilenetworkqualityandsmartphonepenetrationratesacrossprovinces.4Specifically,Icompareinvestorslivinginpairsofneighboringprefectures3thatshareaborderbutbelongtodifferentprovinceswithvaryingsmartphonepenetrationrates.Priortotheshock,theseinvestorshadsimilarobservablecharacteristics,exceptfortheexposuretothetechnologyshock.Aftertheshock,theseinvestors'tradingbehaviordiverges,lendingsupporttothecausalinterpretationoftheapp,simpactoninvestorattention,tradingvolumes,and(weekly)flowsensitivitytofundperformanceandmarketsentiment.Onenaturalquestioniswhetherthechangesininvestorbehaviorcanhaveasignificantimpactonthemutualfundsasawhole.Onepossibilityisthattheincreasedtradingcomovementamongindividualinvestorsleadstoahighervolumeoffundflows,whichinturncouldresultingreaterliquiditycostsformutualfunds.Toinvestigatethetechnologyzsimpactonfundflows,Ifirstcompareadopterswithnonadoptersintermsoftheaggregateflowsaroundthetimeoftheappzsrelease.Ifindthattheabsolutevalueofaggregateflowsfromadoptersmorethandoubles,andadopters'incrementalflowsaresignificantlyhigherthanthoseofnonadopters.Second,Iexplorethedifferencesintechnologyexposureacrossmutualfunds.Icreatethreeproxiesforsuchexposurebasedoninvestors7appadoptionrates,predeterminedinvestor-typecomposition,andpredeterminedgeographicdistributionofinvestors.Myanalysisrevealsasignificantincreaseintheabsolutevalueoffundflowsaftertheap,srelease,andthisincreaseismorepronouncedinthemutualfundswithhightechnologyexposure.Thesefindingssuggestthattheapp,sreleasehasanamplifyingeffectonthevolumeofflowatthefundlevel.Tounderstandwhethertheincreaseinfundflowvolumeresultsinlowerfundperformance,IcomparefundswithdifferenttechnologyexposurelevelsintermsOfthechangeinabnormalreturnsaroundtheshock.Theanalysisrevealsthatthefundsavailableontheappexperienceasignificantdeclineinabnormalreturnsaftertheshock.Regardlessofthetechnologyexposuremeasures,thereductioninabnormalreturnsisparticularlypronouncedamongthehigh-exposurefunds,consistentwithacausalinterpretationofthefundperformanceeffect.Theconcurrentincreaseinfundflowvolumeanddecreaseinreturnssuggeststhattheheightenedliquiditycostsarisingfromincreasedappflowslikelycontributetothedeclineinfundperformance.Tofurthersupporttheliquiditycostmechanism,Idocumentanegativeassociationbetweentheabsolutevalueoffundflowsandthefundreturnsduringthesubsequentweekinthepostshockperiod.Inaddition,thetreatmenteffectsofthetechnologyonfundreturnsbecomeweakeraftercontrollingfortheabsolutevalueofaggregatefundflowsinthepreviousweek,suggestingthattheheightenedvolumeoffundflowpartlyexplainsthedeteriorationinfundperformance.Overall,theseresultsareconsistentwiththeliquiditycostmechanism.6Thispapermakesseveralcontributionstotheliterature.First,itisthefirststudy,tomyknowledge,thatexaminestheimpactofsmartphonetradingtechnologyonmutualfunds.Evidenceonhowsmartphonetradingaffectssecuritypricesorfundperformanceremainsscarce,likelyduetodataconstraints.Thisstudyusesproprietarydatathatprovidesindividual-levelloginandtradingrecords,enablingmetoidentifythetradesofappusersandinvestigatetheaggregateeffectsofsmartphonetradingtechnologyonthepricingofmutualfundshares.Also,theDIDdesignemployedinthisstudyallowsmetoconnectincreasedinvestorattentionandchangesintradingbehaviortothefundperformanceeffect.Second,thisstudyntributestotheliteratureonhowtechnologyaffectsinvestors7tradingbehavior.PriorstudieshavedocumentedthatinvestorstrademorefrequentlyaftertransitioningfromtelephonetradingtoanInternetplatform(BarberandOdean2002,Choietal.2002).However,whatwelearnedabouttheshiftfromtelephonetoonlinetradingdoesnotnecessarilycarryovertosmartphonetrading.GiventhattheInternetalreadyprovideshigh-speedexecutionwithlowtradingcosts,itisunclearwhetherthemigrationfromPCstosmartphoneshasmarginaleffects(1.iaoetal.2021).Theresultsofthisstudysuggestthatthemigrationtosmartphonetradingsignificantlyaffectsinvestors/PgAJ4>SaJSlq8=B.AwosnWUOSdJO1.E寸：EZIE"寸ZoZIPjoNOeUOSGZOCZ一Z/二Z-Aq8JOSWJOJ=UIO±PaPEO-UMOatradingbehaviorandamplifiesinvestors'responsetoshort-termmarkettrendsandfundreturns.SeveralrecentstudiesalsoaddressthebroadthemeofsmartphonetradingandFinTech.Welch(2022)focusesonthecollectivestockinvestmentofRobinhoodusersandfindstheseusers'portfoliostilttowardhigh-volumestocksandlargefirms.Usingsimilardata,Barberetal.(2022)documenthowappnotifications,suchasdisplayingyztopmovers/7affectinvestortradingdecisions.Additionally,Oziketal.(2021)documentanincreaseinretailtradingduringtheCOVID-19pandemiclockdown.Usingtransaction-leveldatazKaldaetal.(2021)showthatinvestorspurchasemoreIottery-Iikeassetsonsmartphonesthanonnonsmartphoneplatforms.Inantemporanusstudy,Hongetal.(2021)focusondigitalpaymenttechnology'sspillovereffectonriskyinvestmentparticipation.D7Acuntoetal.(2021)studytheeffectofadoptingaFinTechappthatprovideshouseholdswithpeers'spendinginformation.Incomparison,thisstudydepartsfromtheliteraturebyexamininghowsmartphonetradingtechnologyaffectsmutualfundinvestors7responsetomarketinformationandhowtheinvestorbehaviorchangesaffectmutualfundperformance.Also,thegranulardataandempiricalsettingusedinthisstudyenablemetoconductageographicvariationanalysistoshedlightoncausalinferences.Theremainderofthispaperisorganizedasfollows:Section2discussestheinstitutionalbackground;Section3detailstheresearchdesignandtheempiricalresultsonhowsmartphonetradingtechnologyaffectsinvestorbehavior;Section4analyzesthetechnology'seffectonfundflowandperformanceanditsimplicationsonindividualwealth;andSection5concludes.2. InstitutionalBackgroundandData2.1.EmpiricalSettingThisstudyfocusesonthemutualfundindustryandexploresthereleaseofasmartphonetradingappbyoneofthelargestmutualfundmanagementcompaniesinChina.ThemutualfundindustryinChinaoperatessimilarlytothatintheUnitedStates,withopen-endedinvestmentfundsthatareprofessionallymanagedandplfundsfromnumerousinvestors.Thefocalcompany,theAdvisor,managedmorethan300billionRMB(approximately$43billionUSD)inassetsbytheendof2017through65differentmutualfunds,includingequity,bond,hybrid,andmoneymarketfunds.About73%oftheholdingsinthesefundswerepurchasedthroughtheAdvisor,sdirectsalesplatform,whichallowsinvestorstotradethecompany'smutualfundsdirectlywithoutinvolvinganexternaldistributor.7ThesmartphonetradingappisanintegralpartoftheAdvisor'sdirectsalesplatform,anditenablesinvestorstotradeallmutualfundsprovidedbytheAdvisor,butnootherassets,throughtheapp.8ThisappisdifferentfrommoregenerictradingappsprovidedbybrokeragefirmslikeRobinhoodbecauseitexclusivelyservesthemutualfundinvestorsoftheAdvisor,anditdoesnotprovideaccesstomutualfundsinotherfundfamiliesornonmutualfundassets.Beforethelaunchofthesmartphoneappzinvestorshadtwomethodstoaccessthedirectsalesplatform.TheprimarymethodinvolvesvisitingtheofficialwebsitethroughanInternetbrowser,mostlythroughPCs.Similartothesmartphoneapp,theofficialwebsiteprovidesaccessexclusivelytothemutualfundsmanagedbytheAdvisorandnootherassets.Avisittothewebsiteallowsinvestorstotrademutualfundsharesandaccessinformationsuchasmarketupdates,fundprofiles,andfundperformance.Thesecondmethodisthroughawebbasedclientterminaldesignedforfeaturephones,whichinvestorsrarelyuseduetoitsslowandunstableconnection.Evenbeforethereleaseoftheapp,investorswidelyadoptedthedirectsalesplatform.Bytradingfundsharesthroughdirectsales,investorsincurlowertransactionfeesthanwhenusingexternaldistributionchannelssuchasbanksandsecuritiescompanies.Thismakesthedirectsaleschannelparticularlyattractivetoinvestorswhoaresensitivetotransactioncosts.9InSeptember2013ztheAdvisormadeamajorupgradetothedirectsalesplatformwiththelaunchofasmartphoneapp.However,tradingthroughtheofficialwebsiteremainedunchanged.Theappofferscomprehensiveservicesandisasignificantimprovementoverthewebbasedclientterminalinseveralways.First,theappisfasterandmorereliablethantheclientterminalasitworksintandemwithlocaldevicesratherthanrelyingonaremoteserver.Second,theapphasamoreintuitiveanduser-friendlyinterface,allowinguserstoconvenientlymonitorportfolios,placeorders,andobtaininformation.Finally,theappstreamlineswiretransfersbetweentradingandbankaccounts,improvingefficiencyforinvestors.Thedesirablefeaturesoftheappquicklyenticedmanyinvestorstoadoptit.AsshowninFigure1,thegrowthinthenumberofactiveusers10ontheplatformsignificantlyacceleratedfollowingthelaunchofthesmartphoneapp.Withinthreemonthsofitsdebut,thenumberofactiveusersmorethantripled,primarilyduetotheexpansionofappusersafterthetechnologyshock.1,Theappandthewebsitehavemanysimilarities.Forexample,bothallowinvestorstotradeonlymutualfundsmanagedbytheAdvisorandtransfermoneyinandoutoftheinvestmentaccounts.Bothalsoprovideinformationsuchasassetbalances,portfolioperformance,transactionhistory,fundprofiles,andmarketperformance.However,therearetwokeydifferencesbetweentheappandthewebsite.First,theappprovidesmoreconvenientaccesstothedirectsalesplatform,allowinguserstoaccessinformationandtrademutualfundsharesfromP2J2S3JSlqoc=B<AIUOUSnWUOSJodJOH-EFEzlB<寸ZOZIPJEWOZUO-ggzocz一Z/二Z-AqonJOsu三OJU一(UOJJpoPEOWMOFigure1.(Coloronline)NumberofActiveUsersontheDirectSalesPlatformNotes.Thefiguredisplaysmonthlyactiveusersofthedirectsalesplatformbeforeandafterthereleaseofthesmartphonetradingapp.Thehorizontalaxisrepresentsthemonthrelativetotheapp,srelease.Theverticalaxisrepresentsthenumberofinvestorswhologgedintotheplatform(throughtheofficialwebsite,theclientterminal,orthesmartphonetradingapp)atleastonceduringagivenmonthandthenumberOfuserswhoprimarilyloginthroughtheappzwebsite,andclientterminal,respectively.EBSn>woJBqEnNanylocationcoveredbyamobilenetwork.Theconveniencemayaffectinvestors'tradingbehaviorbyincreasinginvestors'attentiontofundinvestment.Second,theapppresentsinformationinadifferentformatthanthewebsiteduetothesmallerscreensofsmartphones.Priorstudiesfindtheinformationpresentationformatmayaffectinvestors*behavior.Forinstance,DzAcuntoandRossi(2021)showdifferencesinthegraphicalfeaturesofinformationcanchangepeople,sinterpretationoftheinformationandmaybeinterpretedasrobo-advising.DzAcuntoetal.(2021)findgraphicsthatshowpeers'andusers'spendingcanaffectusers'spendingdecisions.OnlineAppendixApresentsadetailedmparisonoftheinformationcontentanddisplayoftheappandthewebsite.Overall,theinformationcontentavailableontheappiscomparabletothatonthewebsite.Becauseofthesmallerscreensizeofsmartphones,theapptypicallyprovidesless-detailedinformationoneachscreenandinsteadusesbuttonsthatdirectuserstoadditionalscreenswithmoreinformation.Theappmayalsousenotificationfunctionstodeliverinformation.12Thesedisparities,combinedwiththeconvenienceofaccessinginformationandtrading,couldcontributetobehaviorchangesofappadopters.Theintroductionofthesmartphoneappoffersanaturalempiricalsettingtostudytheimpactofsmartphonetradingtechnologyonbothinvestorbehaviorandmutualfun