股票市场与经济增长文献综述及外文文献资料文档格式.docx

股票市场与经济增长文献综述及外文文献资料文档格式.docx

《股票市场与经济增长文献综述及外文文献资料文档格式.docx》由会员分享，可在线阅读，更多相关《股票市场与经济增长文献综述及外文文献资料文档格式.docx（14页珍藏版）》请在冰豆网上搜索。

ImpactofStockMarketStructureonIntertradeTimeandPriceDynamics

作者:

Ivanov,Plamen;

Yuen,Ainslie;

Perakakis,Pandelis

期刊:

InternationalReviewofFinancialAnalysis（Volume92014）:

207-219

年份:

2015

ImpactofStockMarketStructureonIntertradeTimeandPriceDynamics

Introduction

Theimpactofmarketstructureandassociatedrulesofoperationonmarketefficiencyandstockpriceformationhaveattractedconsiderablepublicattention[1].DevelopmentsontheNewYorkStockExchange（NYSE）[1],[2],haveraisedtheprofileofthemarketoperatingmechanism,the"

marketstructure"

employedbyastockmarket.Thishasalsobeenofconcerntothoseinvolvedinstockmarketregulation,onbehalfofinvestors[1],[3],sinceoptimizingmarketstructureresultsinmoreeffectivelyfunctioningmarkets[4]andincreasescompetitivenessformarketshareinlistedstocks[5].ThetwomajorstockmarketsintheU.S.,theNYSEandtheNationalAssociationofSecuritiesDealersAutomatedQuotationSystem（NASDAQ）NationalMarkethaveverydifferentstructures[6],[7],andthereiscontinuingcontroversyoverwhetherreporteddifferencesinstockpricebehaviorareduetodifferencesinmarketstructureorcompanycharacteristics[8].ComparativestudiesoftheNYSEandNASDAQhaveprimarilyfocusedonstockpricestoprovideevidencethatmarketorganizationalstructureaffectsthepriceformationprocess[4],[9],[10].IthasbeenshownthatstocksregisteredontheNASDAQmaybecharacterizedbyalargerbid-askspread[11]andhigherpricevolatility[4],[9],[10].However,thisisoftenattributedtothemarketcapitalization,growthrateorthenatureofthecompanieslistedontheNASDAQ[8].Empiricalstudieshavealsoemphasizedthedominantroleandimpactoftradingvolumeonprices[12],[13];

sincetradedvolumeisdeterminedbyinvestorsitisdifficulttoisolatetheeffectsofmarketstructureonpriceformation.Astheinfluenceofmarketstructureonstockpricesmaybeobscuredbyexogenousfactorssuchasdemandandsupply[12],[13],wehypothesizethatmodulationoftheflowoftransactionsduetomarketoperationsmaycarryastrongerimprintoftheinternalmarketmechanism.

HereweanalysetimesbetweenconsecutivetransactionsforadiversegroupofstocksregisteredontheNYSEandNASDAQmarkets,andwerelatethedynamicalpropertiesoftheintertradetimeswiththoseofthecorrespondingpricefluctuations.Tounderstandhowmarketstructuremayaffectstockprices,westudytheinformationcontainedinthetimesbetweenconsecutivestocktransactions.Asmarket-specificoperationsmaymodulatetheflowoftransactions,wehypothesizethatdynamicalfeaturesoftransactiontimingreflecttheunderlyingmarketmechanism.Specifically,weaskifstocksofcompanieswithdiversecharacteristicsregisteredonagivenmarketexhibitcommonfeaturesintheirtransactiontiming,whichmaybeassociatedwiththeparticularmarketstructure.Further,weinvestigatehowthedynamicalpropertiesoftransactiontimingoverarangeoftimescalesrelatetostockpricedynamicsandwhethermarketstructureaffectsthetemporalorganisationofpricefluctuations.

Toprobehowmarketstructureinfluencesthetradingofstocks,weconsiderthetwomajorU.S.stockmarkets,theNYSEandtheNASDAQ.AlltransactionsontheNYSEofagivenstockarecentralisedandarecontrolledbyasinglehumanoperatorcalleda"

specialist"

whoseprimaryroleistomatchtogetherpublicbuyandsellordersonthebasisofprice,inanauction-likesetting[6].TheNYSEspecialistisunderobligationtomaintainbothpricecontinuityanda"

fairandorderlymarket"

[6],aswellastointervene,usinghisownfirm'

sinventoryofavailablestock,toprovideliquidityintheeventofanorderimbalance,thuspreventingsharpchangesinthestockprice[6].TheNYSEregulationsallowforconsiderableflexibilitywithinthespecialist'

soperations[2].

Incontrast,tradingontheNASDAQisdecentralised,withtradinginagivenstockmanagedbyanumberofdealerscalled"

marketmakers"

.Thesemarketmakersmaintainastockinventory,postingtheirbestpricesatwhichtheyarepreparedtoimmediatelybuyandsellstock[7].Marketmakerscompetewitheachotherfororders,sointheorycompetitionensuresthatinvestorsgetthebestprices.Alternatively,anordercanbeplacedintoanAlternativeTradingSystem（ATS）,operatedbyNASDmembersorNASD-memberaffiliatesanddesignedtoallowtwosubscriberstomeetdirectlyonthesystemundertheregulationofathirdparty.ThemostcommonlyusedformofATSistheElectronicCommunicationNetwork（ECN）,afacilitythatmatchescustomerbuyandsellordersdirectlythroughacomputernetwork.

Athirdalternative,incasetheorderplacedisverysmall,istoentertheorderintotheSmallOrderExecutionSystem（SOES）,whichisanelectronicnetworkdesignedtoallowfastautomaticrouting,executionandreportingofordersof500sharesorless.Ordersareautomaticallyroutedtomarketmakerswhosequotesarecurrentlyidenticaltothehighestbid（buy）andthelowestoffer（sell）prices.ParticipationintheSOESsystemwasmademandatory[7]afterthemarketcrashofOctober1987,asoneofthereportedproblemsontheNASDAQduringthecrashwastheinabilitytoreachmarketmakersbythephoneduringperiodsofrapidpricemovement.

Tosummarizethedifferencesbetweenthetwomarketstructures,eachmarketmakerontheNASDAQmaintainshisowninventoryofstockinordertobuyandsell[7].Incomparison,theNYSEspecialistrarelyuseshisownfirm'

sinventory:

suchtransactionsinvolvelessthan15%oftradingvolume[14].AlthoughseveralregionalexchangesmaytradeNYSElistedstocks,priceformationhasprimarilybeenattributedtoNYSEtrading[15].Incontrast,theNASDAQmarketreliesoncompetitionbetweenmultipledealersforpublicorderstofacilitatethepriceformationprocess[11].Moreover,asubstantialfractionofsharevolumeontheNASDAQisnothandledbydealers,butistradedelectronicallyvianetworksforsmallpublicordersandforinstitutionalinvestors[7].SuchfragmentationoftheNASDAQstockmarkethasbeenassociatedwithhigherpricevolatility[4].

HereweasktowhatextentsuchstructuralandoperationaldifferencesbetweentheNYSEandNASDAQmarketsaffecttheflowoftransactions.Itisdifficulttoanswerwhetherdifferencesinintertradetimesareduetoindividualcompanycharacteristicsorexternalmarketinfluences（Fig.1）.Twoempiricalstudieshaveconsideredonlyasinglecompanystockoverashortperiodofafewmonths[16],[17].Studieswhichconsideredalargergroupofstockseitherdidnotfindcommonfeaturesintheintertradetimes[18],[19]ordidnotcomparebetweenmarkets[20]-[22].TheonlycomparativestudyconsideredasingleNYSEandasingleParisstock,findingsomedifferencesintheirintertradetimes,butthosemaywellbeduetoadifferentcultureoftrading[23].Toprobeforevidenceoftheimpactofmarketstructureonthetradingofstocks,weemployconceptsandmethodsfromstatisticalphysicstoinvestigatethecorrelationpropertiesoftransactiontimingfordiversecompanies,overtimescalesrangingfromsecondsuptoayear.

Figure1.Relationshipbetweenstockpriceandtradingactivity.

RepresentativeexampleoftimeseriesderivedfromtheTradesandQuotes（TAQ）databasefortransactionsofstockinCompaqComputerCorp.（CPQ）registeredontheNYSE.（a）PriceofCPQstockoverathreeweekperiodfrom20Feb.-8Mar.1996（42606trades）.On1Mar.1996Compaqreportedthatitwouldcutproductpricesinordertomeetsalestargets,leadingtoadropinthestockprice.（b）Intertradetimes（ITT）ofCPQstockoverthesameperiod.Dataexhibitcomplexfluctuations,adailypatternoftradingactivity（withshortITTattheopenandcloseofatradingdayandlongerITTinbetween）,andhighlyheterogeneousstructure,asseenintheflurryoftradesfollowingthepricedrop.TherelaxationtimeoftheITTresponsefollowingthepricedropextendsoverseveraldays,suggestingthatinformationmaybecontainedinthetemporalstructureoftradingactivity.Dataincludetransactionsoccurringbetween9.30amand4pmEST,excludingweekendsandholidays.

Data

WeexamineonehundredstockslistedontheNYSE,fromelevenindustrysectors:

Technology-Hardware（5）,Semiconductors

（2）,Pharmaceutical&

MedicalEquipment（10）,Financial（8）,Automotive（9）,Defense/Aerospace（9）,Mining,Metals&

SteelWorks（8）,Chemicals&

Plastics（7）,Retail&

Food（17）,Petroleum,Gas&

HeavyMachinery（10）,TelephoneServiceProviders（7）,Electric&

PowerServices（8）.Westudythetimeintervalsbetweensuccessivestocktrades,overaperiodoffouryears-4Jan.1993to31Dec.1996-asrecordedintheTradesandQuotes（TAQ）databasefromtheNYSE（Table1）.

Table1.CharacteristicsofonehundredNYSEstocksstudiedovertheperiod4Jan.1993-31Dec.1996.

WealsoanalyseonehundredNASDAQstocksfromfourteenindustrysectors:

Technology-Hardware（28）,Technology-Software（16）,Semiconductors（7）,Pharmaceutical,Biotechnology&

MedicalEquipment（12）,Financial（5）,Automotive

（1）,SteelWorks

（1）,Chemicals

（1）,Retail&

Food（16）,Petroleum,Gas&

HeavyMachinery

（2）,Telephone&

CableTelevisionServiceProviders（5）,Services

（2）,Transportation（3）,ElectricalApparatus

（1）.WestudythetimeintervalsbetweensuccessivestocktradesasrecordedintheTAQdatabase,fortwenty-ninecompaniesovertheperiod4Jan.1993-31Dec.1996,andseventyonecompaniesovertheperiod3Jan.1994-30Nov.1995（markedwith（*）inTable2）.Forbothmarkets,weselectcompanieswithaveragemarketcapitalisationsrangingoverthreedecades