Savings Investment and Capital Mobility within ChinaWord下载.docx
- 文档编号:6679303
- 上传时间:2023-05-07
- 格式:DOCX
- 页数:24
- 大小:72.36KB
Savings Investment and Capital Mobility within ChinaWord下载.docx
《Savings Investment and Capital Mobility within ChinaWord下载.docx》由会员分享,可在线阅读,更多相关《Savings Investment and Capital Mobility within ChinaWord下载.docx(24页珍藏版)》请在冰点文库上搜索。
Preliminaryversion:
July,2008
JELClassification:
O16;
R1
Keywords:
Capitalmobility;
Marketintegration;
Non-stationarypanel;
Allocationefficiency
Abstract:
ThispaperaddressesthecapitalmobilityamongregionswithinChina.Usingarangeofpanelestimatorswhichdealwiththenon-stationarityoftimeseriescomponents,individualheterogeneityandcommonunobservedfactors,weshowthatthesavingsandinvestment(bothexpressedasratiostoGDP)arepositivelycorrelatedforasampleof28Chineseprovincesovertheperiod1978to2006.AccordingtotheFeldstein-Horioka’sargument(1980,EconomicJournal(90),pp.314-329),suchacorrelationcanbeinterpretedasevidenceoflowcapitalmobility.Inaddition,bymeansofGrangercausalitytests,wefailtoprovideevidenceinfavorofthehypothesisofefficientcapitalallocationinChineseprovinces.Tosumup,ourfindingssuggestthatthecapitalcannotgoanywhereinsearchofthebestreturn-risktrade-offwithinChinaand,inturn,appearbeinefficientlyconstrainedinsideaprovince.Thus,furtherreformsaimedatremovingtheinterregionalcapitalmobilitybarriersaresolelyneeded.
1.Introduction
Thereisageneralagreementthatdespitethesuccessfulintegrationintotheglobaleconomysince1978,China’sdomesticmarketremainsfragmentedintomanyregionalblocs,bothintermsofgoodsandfactorsofproduction.Indeed,numerousanecdotescouldbementionedtosupportthisview,whereasthemostrepresentativestorymaybethecommodity‘wars’betweenregionswithinChina,suchas‘cottonwar’,‘silkcocoonwar’and‘teawar’,whicheruptedinthelate1980saswellastherecent‘taxicabwar’observedintheearly2000,(seeWorldBank(1994,chapter2)andWedeman(2003)).
Inparticular,sincethebeginningof1990s,theissueofChina’sdomesticmarketintegrationhasreceivedmoreattentionofeconomicscholars.Agrowingliteraturehasproceededalongfivemajorlines,suchaspriceconvergence(WorldBank(1994),Young(2000),andFanandWei(2006)),productionstructuresimilarity(WorldBank(1994),Young(2000)andBaietal.(2004)),synchronizationofthebusinesscycle(Tang(1998),Xu(2002)andXuandVoon(2003)),internaltradelink(WorldBank(1994),Naughton(2003)andPoncet(2005))andlabormigration(Wei(1997),Poncet(2004)andLiangandMa(2004)).However,toourknowledge,thetopicofcapitalmarketintegrationinsideChinaisrarelydiscussed,withthenotableexceptionofBoyreau-DebrayandWei(2004,hereafterBDW).TheauthorsfindthatsavingsandinvestmentinsideChineseprovincesarecloselylinked.AccordingtoFeldstein-Horioka(1980,hereafterFH),suchresultscouldbeinterpretedasevidenceoflowcapitalmobility.Inaddition,usinganotherapproachbasedonrisk-sharingliterature,theauthorbearoutfurtherthecapitalimmobilitywithinChina.
FollowingBDW(2004),weattempttoreexaminethecapitalmobilityacrossChineseregions,whichcanbeseenasanimportantindicatorofdomesticfinancialmarketintegration.Ourstudyhasthreedistinguishingfeatures:
First,tomeasurethedegreeofcapitalmobility,weuseasampleincluding28Chineseprovinces(orprovince-levelautonomousregionsormunicipalities)observedovertheperiodof1978to2006(insteadof1978to2001inBDW).Thankstothelargertimedimension,wegainabetterunderstandingoftimeseriespropertiesofsavingsandinvestmentdata.Second,todealwithprovince-specificheterogeneityandcommonunobservedfactors,whichwillbediscussedindetailshortly,asetofpanelestimatorshasbeenusedintheFHregressions.Third,bymeansofGrangertests,weinvestigatealsothecausalrelationshipbetweeninvestmentandincomegrowthratesandtempttoassesstheefficiencyofChina’sfinancialmarket..
Thereminderofthepaperisorganizedasfollows:
Section2presentsbrieflytheFHapproachaswellassomerelatedalternativeinterpretations.Section3examinesthestationarityofsavingsandinvestmentratesanddetectstheexistenceofcointegratingrelationshipbetweenthetwoseries.Section4showstheresultsofFHregressionsfromavarietyofpanelmodels.AshortcomparisonbetweenChina’scapitalmobilityandevidencefromothercountriesisalsogiven.Section5discussesfurthertheefficiencyofcapitalallocationinChina.Thelastsectionconcludesthepaper.
2.FHapproachandalternativeinterpretations
ThebasicideaofFHapproachisresumedas‘withperfectworldcapitalmobility,thereshouldbenorelationbetweendomesticsavinganddomesticinvestment:
savingineachcountryrespondstotheworldwideopportunitiesforinvestmentwhileinvestmentinthatcountryisfinancedbytheworldwidepoolofcapital’(FH(1980,pp.317)).Toverifythishypothesis,FHstudiedthefollowingmodel:
whereIandSdenotetheratiosofdomesticinvestmentandnationalsavingstoGDP,respectively.Fromasampleof16OECDeconomiesover1960-1974,FHshowedthattheparameterofsavings,β,whichissometimescalled‘savings-retentionratio’or‘FHcoefficient’,isclosetounity.Accordingtothem,thoseresultsimplythattheinternationalcapitalmarketisweaklyintegratedandnearly100%ofincrementalsavingsareretainedinsidenationalborders.Theirfindingshavebeenconfirmedbysubsequentresearch.Infact,asstressedbyTesar(1991)whoprovidesacomprehensivesurveyofFHliterature,thepositivesavings-investmentcorrelationisbothashort-runandlong-runphenomenonandisnotrestrictedtoaparticularsampleofcountries.
However,giventheconsiderableimprovementininternationalfinancialmarketliberalizationoverrecentdecades,thesignificantassociationbetweensavingsandinvestmentrateshasbeenknownasthe‘FHpuzzle’.TheimplicationsofFHtestforcapitalmobilityhavebeenquestionedbynumerousfollowingcontributions,whicharguethatthesavingsretentionratioestimatesshedlittlelightoncapitalmobility.Forourpurposes,threealternativeviewsareparticularlynoteworthy.First,thelong-runassociationbetweensavingsandinvestmentcanbeinterpretedasadirectimplicationoftheintertemporalbudgetconstraint.Sincethecumulatedbalanceofpaymentsorregionaldebtcannotexplode,thesavingsandinvestmentseriesmustbecointegratedwithaunitcoefficientwhentimeseriesdimensiongoestoinfinity(seeCoakleyetal.(1996),Jansen(1998)andTaylor(2002)).Second,itseemsthatthesavings-investmentlinkisheterogeneousacrossindividuals.Inparticular,itmaybedistortedbyeconomicsizeofacountry.Ontheonehand,Harberger(1980)conjecturesthatascountriesbecomelarger,theybecomemorediversifiedandtheneedtoborrowfromabroadintheeventofashockdeclines.Ontheotherhand,asarguedbyDooleyetal.,(1987)andBaxterandCrucini(1993),ifacountryislargeenoughtoaffecttheworldinterestrate,aninitialincreaseinitssavings,willlowertheworldinterestrateandconsequentlyboostinvestmentinthatcountry.Third,someexogenousdisturbancescouldalsogenerateco-movementsinsavingsandinvestment,whichareunrelatedtothecapitalmobility.Forinstance,Obstfeld(1986)arguedthatapositiveproductivityshockwouldcausethecurrentwagetoriseabovethepermanentwage.Inthiscase,householdstendtoincreasesavingstosmooththeirconsumptionovertime.Iftheshockispersistent,higherproductivity(duetohigherwage)willstimulateinvestmentinfutureperiods.Inaddition,asTesar(1991)discusses,suchapositivesavings-investmentassociationwillbespreadtoagroupofcountriesifexogenousshocksarecorrelatedamongthem.
Totackletheaforementioned‘noises’thatrendtheFHtestirresponsivetocapitalmobility,wefirstinvestigatethelong-runrelationshipbetweensavingsandinvestmenttodetectweathertheintertemporalbudgetconstraintappliesinoursample.Next,todifferentdegree,weallowfortheprovince-specificheterogeneitythatisprobablycausedby‘sizeeffects’.Finally,weassumethatexogenousdisturbancesbehaveasunobservedcommonfactorsaffectingsimultaneouslyallofcrosssectionalunits(provinces)inthesample.Variousapproacheshavebeenusedtodealwithsuchlatentfactors.
3.Stationarityandcointegrationtestsforsavingsandinvestmentseries
BeforeestimatingtheFHcoefficient,thispartinvestigatesthetimeseriespropertiesofsavingsandinvestmentdata.First,weconducttheADF(augmentedDickey-Fuller)-GLS(GeneralizedLeastSquare)unitroottestsforindividualseries.ItisknownthatthisapproachdevelopedbyElliottetal.(1996)hassignificantlygreaterpowerthantheordinaryADFtest.AscanbeseeninTable1,withafewexceptions,theADF-GLSstatisticsareinsignificantat10%formostofseries,suggestingthatbothsavingsandinvestmentratesseemtobenon-stationary(orintegrated)processes.
Moreover,thestationaritypropertiesarefurtherexaminedbyperformingIm,PesaranandShin(2003,hereafterIPS)’spanelunitroottest,whichstartsfromanADFregressionasfollows:
whereα,tandprepresentindividual-specificconstants,timetrendandlaglength,respectively.Thenullhypothesisisthateachseriescontainsaunitroot(ρi=1foralli),whereasthealternativehypothesisisthatonlyafractionofseriescontainsaunitroot(ρi<
1forsomei).Thus,considerableheterogeneity(intermsofautoregressiveparameter,interceptaswellasdeterministictrend)isallowedhere.Moreinterestingly,theIPSprocedurecanalsocontrolforthecrosssectionaldependencebyincludingtimedummiesinequation
(2),orequivalently,bytakingdeviationsfromcros
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- Savings Investment and Capital Mobility within China
![提示](https://static.bingdoc.com/images/bang_tan.gif)
链接地址:https://www.bingdoc.com/p-6679303.html