1、1本科毕业论文外文翻译外文文献译文标题金融发展，对外贸易和经济增长以印度为例资料来源INTERNATIONALJOURNALOFSOCIALECONOMICS2007,349作者SALIHTURANKATIRCIOGLU；NESLIHANKAHYALAR；HASRETBENAR本研究利用19652004印度的年度数据调查了可能的关于协整关系和因果关系方向之间的金融发展，国际贸易和经济增长关系。结果表明，在印度金融发展，国际贸易和实际收入的增长存在着长期稳定的均衡关系。格兰杰因果检验表明，增长的实际收入导致国际贸易部门增长，即出口和进口部门。因此，存在单向因果关系，即从实际收入的增长对国际贸易的
2、推动。另一方面，双向因果关系已获得的实际收入的增长和经济发展的措施，即广义货币，国内信贷，这是类似的结果，DEMETRIADES和LUITEL（1996），他们还调查三者之间的双向因果关系，金融发展与印度经济增长。然而，结论与SIVASUBRAMANIAN（2003）发现共同指出，使用广义货币量作为金融业发展的措施有单向因果关系，金融发展实际促进国内生产总值增长。此外，这项研究表明，印度出口导致印度金融发展（包括广义货币量，国内信贷）改变。但由于进口，有单向的因果关系，即广义货币对进口和进口对国内信用的作用。如果对结果进行总结，研究结果表明，供给和需求追随假设不能被推断为印度经济本身。此外，1
3、9652004的出口和IMPORTLED假说不能再次被推断为是印度经济基础上的样本期。但需要注意的是，高盛等（2002）在调查出口导向型经济增长假说的基础上对印度沿海主要地区的区域进行了分析。许多宏观经济时间序列可能包含单位根的发展并带动了理论的非平稳时间序列分析。实证研究表明，存在非平稳的时间序列的考虑可能会导致虚假回归结果和无效的结论，运用格兰杰因果关系。TODA和菲利普（1993）使格兰杰因果关系的处理方法在I（1）变量系统，因果关系的长期关系在协整的非平稳时间序列时可以推断。存在一个协整关系的基础形式的向量误差修正（VEC）技术规格。此外，标准格兰杰或模拟试验由于省略了误差修正条款，测
4、试提供了无效的因果关系的信息（DOYLE，2001）。简单的格兰杰因果关系检验协整向量变得不恰当时，得到了这一系列。根据格兰杰定理，结果表明与协整的纠错申述方程（4）和（5）。这些都是必要的，以增加简单的格兰杰因果关系检验和误差修正机制，从适当的残差的协整关系检验因果关系2DLNYTC0XKI1BIDLNYT2IXKI1AIDLNXT2IPIECTT21UT4DLNXTC0XKI1GIDLNXT2IXKI16IDLNYT2IHIECTT211T5这是最通用的模型与趋势；TM是模型的趋势；T是最大限制模型没有趋势。括号中的数字是滞后长度用于检验（由ACI设置为最大3确定）删除序列相关的残差。当使
5、用PP试验，括号中的数字代表NEWEYWEST带（由BARTLETTKERNEL确定）。，和表示拒绝零假设的15和10PERCENTLEVELS，分别。国内生产总值国内生产总值，M2代表货币和准货币作为国内生产总值的百分数，EXP代表国内信贷银行部门提供的国内生产总值的百分数，EXP为货物和服务出口和进口货物和服务站。测试单位根已在EVIEWS51中对ADF和PP进行了单位根检验。在必要的条件下进行数据的平稳性检验，我们进行了约翰森的协整检验（约翰森，1988；约翰森和JUSELIUS，1990），这是非常敏感的选择滞后长度（CHANG，2002），探索从任何可能的长期关系的因素考虑。本文首先
6、审查国际贸易与经济增长之间可能的联系，并共同整合金融发展；其次，在测试这三个变量因果关系的基础上，对印度经济进行以供给，需求跟进和贸易为主导的假说。这项研究的结果可能会给有趣的结论三文学的原因值得注意的是，也有研究探索之间的关系经济增长和金融业在印度的发展。阿加瓦尔（2000）审查了在印度金融部门的改革和表明，重要的是要考虑的脆弱性，印度经济金融危机由于高的经常账户赤字，巨额财政赤字和缓慢增长的出口。该研究的结论和SIVASUBRAMANIAN（2003）调查之间的因果关系，金融发展与经济增长在印度使用的3因果关系分析。他们发现，期间的1970年至1999年金融业的发展作为衡量立方米/国内生产
7、总值导致国内生产总值增长。本研究DEMETRIADES和LUITEL（1996）调查之间的关系，金融发展，经济增长和金融部门控制印度。他们发现金融发展与印度经济增长有双向因果关系。他们还指出，通过改变银行的行为，政策影响金融发展，也影响到经济增长，金融深化和金融部门政策的影响。另一方面，以甫洛娃（2004）调查，在印度贸易自由化对于公司的生产力的影响，其中发现，贸易自由化（尤其是降低关税）增加生产力公司。这项研究还声称，在印度生产力和盈利的公司可能会导致经济福利的改善更密集的私有化的努力。本研究的BAJPAI（2001）表明，在印度每年因为结构变化的工业，金融和贸易等降低保护水平，放开价格，并
8、继续在银行部门的改革有潜在的增长78的可能。萨克斯等人（2002）表明，在印度沿海地区如泰米尔纳德邦，马哈拉施特拉和古吉拉特因为地域的经济绩效而出口增长，关键是通过增加出口到沿海地区，因此当地生产率大大提高。本研究的BAJPAI（2002）指出，与印度在1991开始的经济改革相比，民间投资的角色获得了很大的意义。国家一级的数据，对外商直接投资审批建议，相对快速增长的国家已经吸引了更多的外国直接投资的水平。这意味着，长期均衡关系之间存在着的这些对它也是有用的，说一个以上的协整向量获得了LNGDP和LNEXP，LNGDP和LNIMP，和LNM2和LNIMP而对其他变量的协整与最多一个协整向量。本研
9、究调查了可能的协整关系和因果关系的方向之间的金融发展，国际贸易和经济增长在印度用年度数据，涵盖的时间19652004。结果表明，在印度金融发展，国际贸易和实际收入的增长存在着长期稳定的均衡关系。格兰杰因果检验表明，增长的实际收入增长导致国际贸易部门，即出口和进口。因此，存在单向因果关系，从实际收入的增长对国际贸易的增长。另一方面，双向因果关系已获得的实际收入的增长和经济发展的措施，即M2，国内信贷。这是类似的结果，DEMETRIADES和LUITEL（1996），他们还调查之间的双向因果关系，金融发展与印度经济增长。然而，结论和SIVASUBRAMANIAN（2003）发现，使用M2作为金融业
10、发展的措施有单向因果关系，金融发展实际国内生产总值增长。4外文文献原文TITLEFINANCIALDEVELOPMENT,TRADEANDGROWTHTRIANGLETHECASEOFINDIAMATERIALSOURCEINTERNATIONALJOURNALOFSOCIALECONOMICS2007,349AUTHORSALIHTURANKATIRCIOGLU；NESLIHANKAHYALAR；HASRETBENARABSTRACTPURPOSETHISPAPERAIMSTOINVESTIGATETHEPOSSIBLECOINTEGRATIONANDTHEDIRECTIONOFCAUSAL
11、ITYBETWEENFINANCIALDEVELOPMENT,INTERNATIONALTRADEANDECONOMICGROWTHININDIADESIGN/METHODOLOGY/APPROACHANNUALDATACOVERINGTHE19652004PERIODHAVEBEENUSEDTOINVESTIGATECOINTEGRATIONANDGRANGERCAUSALITYTESTSBETWEENFINANCIALDEVELOPMENT,INTERNATIONALTRADE,ANDGROWTHAFTEREMPLOYINGUNITROOTTESTSTOSEEIFTHEVARIABLESU
12、NDERCONSIDERATIONARESTATIONARYKEYWORDSFINANCIALMANAGEMENT,TRADE,ECONOMICGROWTH,INDIAPAPERTYPERESEARCHPAPER1INTRODUCTIONTHEFUNDAMENTALQUESTIONINTHERELEVANTEMPIRICALLITERATUREISDOESFINANCIALDEVELOPMENTORTRADECAUSESECONOMICGROWTHORISFINANCIALDEVELOPMENTORINCREASEINTRADEANENGINEOFGROWTHFORANECONOMYONECR
13、UCIALFACTORTHATHASBEGUNTORECEIVECONSIDERABLEATTENTIONMORERECENTLYISTHEROLEOFFINANCIALMARKETANDBANKINGSECTORINTHEDEVELOPMENTOFGROWTHPROCESSTHENEXUSBETWEENECONOMICGROWTHANDFINANCIALDEVELOPMENTHASBEENCONDUCTEDONNUMBEROFDIVERGENTLINESAFTERTHEEXTENSIVESTUDIESINTHISFIELD,ITISNOWWELLRECOGNIZEDTHATFINANCIAL
14、DEVELOPMENTISACRUCIALFACTORFORECONOMICGROWTHCALDERONANDLIU,2003ASITISANECESSARYCONDITIONFORACHIEVINGAHIGHRATEOFECONOMICGROWTHCHANG,2002ANDHASASTRONGPOSITIVERELATIONSHIPWITHECONOMICGROWTHMAZURANDALEXANDER,2001HOWEVER,DEGREGORIOANDGUIDOTTI1995POINTOUTTHATFINANCIALDEVELOPMENTSIGNIFICANTLYREDUCESECONOMI
15、CGROWTHFORCOUNTRIESESPECIALLYINLATINAMERICAEXPERIENCINGRELATIVELYHIGHINFLATIONRATESALTHOUGH,THEDIRECTIONOFCAUSALITYBETWEENFINANCIALDEVELOPMENTANDECONOMICGROWTHISINATTENTIONOFTHERESEARCHERSINTHERELEVANTLITERATURE,THISCAUSALRELATIONSHIPGENERALLYREMAINSUNCLEARCALDERONANDLIU,2003PATRICK1966DEVELOPEDTWOH
16、YPOTHESESTESTINGTHEPOSSIBLEDIRECTIONSOFCAUSALITYBETWEENFINANCIALDEVELOPMENTANDECONOMICGROWTH,THATIS,THESUPPLYLEADINGHYPOTHESIS,WHEREITPOSITSACAUSALRELATIONSHIPFROMFINANCIALDEVELOPMENTTOECONOMICGROWTH,ANDTHEDEMANDFOLLOWINGHYPOTHESIS,WHEREITPOSTULATESACAUSALRELATIONSHIPFROMECONOMICGROWTHTOFINANCIAL5DE
17、VELOPMENTINTHEEMPIRICALLITERATURE,MCKINNON1973,KINGANDLEVINE1993,NEUSSERANDKUGLER1998ANDLEVINEETAL2000SUPPORTTHESUPPLYLEADINGHYPOTHESISWHILEGURLEYANDSHAW1967,GOLDSMITH1969ANDJUNG1986SUPPORTTHEDEMANDFOLLOWINGHYPOTHESISEMPIRICALSTUDIESOFTHETRADELEDGROWTHTLGHYPOTHESISFAILTOPRODUCECONCLUSIVEFINDINGSGILE
18、SANDWILLIAMS,1999DEME,2002THENEWTRADETHEORYHASCONTRIBUTEDTOTHETHEORETICALRELATIONSHIPBETWEENEXPORTSANDGROWTHREGARDINGEFFECTSONTECHNICALEFFICIENCYDOYLE,2001RIVERABATIZANDROMER1991SHOWTHATEXPANSIONOFINTERNATIONALTRADEINCREASESGROWTHBYINCREASINGTHENUMBEROFSPECIALIZEDPRODUCTIONINPUTSHOWEVER,THISOUTCOMEI
19、SAMBIGUOUSWHENTHEREISIMPERFECTCOMPETITIONANDINCREASINGRETURNSTOSCALEDOYLE,2001KRUGMAN1979,DIXITANDNORMAN1980ANDLANCASTER1980SHOWECONOMIESOFSCALEASAMAJORCAUSEOFINTERNATIONALTRADE,HINTINGTHEVALIDITYOFTHEGROWTHLEDEXPORTSHYPOTHESIS31DATADATAUSEDINTHISPAPERFORTHEINDIANECONOMYAREANNUALFIGURESCOVERINGTHEPE
20、RIOD19652004THEVARIABLESOFTHESTUDYAREMEASUREDASFOLLOWSREALGROSSDOMESTICPRODUCTGDPAT1995CONSTANTUSPRICESLNGDP,THEFIRSTFINANCIALDEVELOPMENTMEASURE1ISTHERATIOOFBROADMONEYM2TONOMINALGDP,NAMELYLNM2ANDTHESECONDFINANCIALDEVELOPMENTMEASUREISTHERATIOOFDOMESTICCREDITTONOMINALGDP,NAMELY,LNDCTHEREAREMANYSTUDIES
21、INTHELITERATUREWHICHUSESTHEPROXYFORTRADEOPENNESSASTHERATIOOFTRADEOFGOODSANDSERVICESINCLUDINGEXPORTSANDIMPORTSRELATIVETOGDPANDTHEREALSOMANYSTUDIESWHICHUSESEXPORTSANDIMPORTSSEPARATELYTOCONSIDERINDIVIDUALEFFECTSTHISSTUDYWILLUSEREALEXPORTSOFGOODSANDSERVICESLNEXPANDREALIMPORTSOFGOODSANDSERVICESLNMPWHEREB
22、OTHAREAT1995CONSTANTUSPRICESTOCAPTUREINDIVIDUALRELATIONSHIPSWITHOTHERVARIABLESOFTHESTUDYALLOFTHEVARIABLESINTHESTUDYAREATTHEIRNATURALLOGARITHMDATAWEREGATHEREDFROMWORLDBANKDATABASEFORWORLDDEVELOPMENTINDICATORSWORLDBANK,200532METHODOLOGYTHEAUGMENTEDDICKEYFULLERADFANDPHILLIPSPERRONPP2UNITROOTTESTSAREEMP
23、LOYEDTOTESTTHEINTEGRATIONLEVELANDTHEPOSSIBLECOINTEGRATIONAMONGTHEVARIABLESDICKEYANDFULLER,1981PHILLIPSANDPERRON,1988THEPPPROCEDURES,WHICHCOMPUTEARESIDUALVARIANCETHATISROBUSTTOAUTOCORRELATION,AREAPPLIEDTOTESTFORUNITROOTSASANALTERNATIVETOADFUNITROOTTESTUNLESSTHERESEARCHERKNOWSTHEACTUALDATAGENERATINGPR
24、OCESS,THEREISAQUESTIONCONCERNINGWHETHERITISMOSTAPPROPRIATETOINCLUDECONSTANTTERMANDTRENDFACTORINTHEUNITROOTPROCESSENDERS,1995ITMIGHTSEEMREASONABLETOTESTTHEEXISTENCEOFAUNITROOTINTHESERIESUSINGTHEMOSTGENERALOF6THEMODELSTHATISWHEREYISTHESERIESTTIMETRENDFACTORACONSTANTTERMDRIFT1TGAUSSIANWHITENOISEANDPTHE
25、LAGORDERTHENUMBEROFLAGS“P”INTHEDEPENDENTVARIABLEWASCHOSENBYTHEAKAIKEINFORMATIONCRITERIAAICTOENSURETHATTHEERRORSAREWHITENOISEONEPROBLEMWITHTHEPRESENCEOFTHEADDITIONALESTIMATEDPARAMETERSISTHATITREDUCESDEGREESOFFREEDOMANDTHEPOWEROFTHETESTONTHEOTHERHAND,THERESEARCHERMAYFAILTOREJECTTHENULLHYPOTHESISOFAUNI
26、TROOTG0BECAUSEOFAMISSPECIFICATIONCONCERNINGTHEDETERMINISTICPARTOFTHEREGRESSIONTHEREFORE,DOLDADOETAL1990ALSOSUGGESTSTARTINGFROMTHEMOSTGENERALMODELTOTESTFORAUNITROOTWHENTHEFORMOFTHEDATAGENERATINGPROCESSISUNKNOWNTHEGENERALPRINCIPLEISTOCHOOSEASPECIFICATIONTHATISAPLAUSIBLEDESCRIPTIONOFTHEDATAUNDERBOTHTHE
27、NULLANDALTERNATIVEHYPOTHESESHAMILTON,1994IFTHEINTERCEPTORTIMETRENDISINAPPROPRIATELYOMITTED,THEPOWEROFTHETESTCANGOTOZEROCAMPBELLANDPERRON,1991“REDUCEDPOWERMEANSTHATTHERESEARCHERWILLCONCLUDETHATTHEPROCESSCONTAINSAUNITROOTWHEN,INFACT,NONEISPRESENT”ENDERS,1995,P255ALINEARCOMBINATIONOFINTEGRATEDVARIABLES
28、ARESAIDTOBECOINTEGRATEDIFTHEVARIABLESARESTATIONARYMANYECONOMICMODELSENTAILSUCHCOINTEGRATINGRELATIONSHIPSENDERS,1995FINANCIALDEVELOPMENT,TRADEANDGROWTHAFTERTHEORDEROFINTEGRATIONISDETERMINED,COINTEGRATIONBETWEENTHEVARIABLESSHOULDBETESTEDTOIDENTIFYANYLONGRUNRELATIONSHIPJOHANSENTRACETESTISUSEDFORTHECOIN
29、TEGRATIONTESTINTHISPAPERCHEUNGANDLAI1993MENTIONTHATTHETRACETESTISMOREROBUSTTHANTHEMAXIMUMEIGENVALUETESTFORCOINTEGRATIONTHEJOHANSENTRACETESTATTEMPTSTODETERMINETHENUMBEROFCOINTEGRATINGVECTORSAMONGVARIABLESTHERESHOULDBEATLEASTONECOINTEGRATINGVECTORFORAPOSSIBLECOINTEGRATIONTHEJOHANSEN1988ANDJOHANSENANDJ
30、USELIUS1990APPROACHALLOWSTHEESTIMATINGOFALLPOSSIBLECOINTEGRATINGVECTORSBETWEENTHESETOFVARIABLESANDITISTHEMOSTRELIABLETESTTOAVOIDTHEPROBLEMSWHICHSTEMSFROMENGLEANDGRANGER1987PROCEDURE3THISPROCEDURECANBEEXPRESSEDINTHEFOLLOWINGVARMODELXTP1XT21PKXT2KMETFORT1T2WHEREXT,XT21,XT2KAREVECTORSOFCURRENTANDLAGGED
31、VALUESOFPVARIABLESWHICHAREI1INTHEMODELP1,PKAREMATRICESOFCOEFFICIENTSWITHPXPDIMENSIONSMISANINTERCEPTVECTOR4ANDETISAVECTOROFRANDOMERRORSTHENUMBEROFLAGGEDVALUES,INPRACTICE,ISDETERMINEDINSUCHAWAYTHATERROR7TERMSARENOTSIGNIFICANTLYAUTOCORRELATEDTHERANKOFPISTHENUMBEROFCOINTEGRATINGRELATIONSHIPSIERWHICHISDE
32、TERMINEDBYTESTINGWHETHERITSEIGENVALUESLIARESTATISTICALLYDIFFERENTFROMZEROJOHANSEN1988ANDJOHANSENANDJUSELIUS1990PROPOSETHATUSINGTHEEIGENVALUESOFPORDEREDFROMTHELARGESTTOTHESMALLESTISFORCOMPUTATIONOFTRACESTATISTICS5THETRACESTATISTICLTRACEISCOMPUTEDBYTHEFOLLOWINGFORMULA6LTRACE2TXLN12LIIR1N213ANDTHEHYPOT
33、HESESAREH0R0H1R1H0R1H1R2H0R2H1R3THEFINDINGTHATMANYMACROTIMESERIESMAYCONTAINAUNITROOTHASSPURREDTHEDEVELOPMENTOFTHETHEORYOFNONSTATIONARYTIMESERIESANALYSISEMPIRICALSTUDIESHAVESHOWNTHATTHEEXISTENCEOFNONSTATIONARITYINTHETIMESERIESCONSIDEREDCANLEADTOSPURIOUSREGRESSIONRESULTSANDINVALIDATETHECONCLUSIONSREAC
34、HEDUSINGGRANGERCAUSALITYTODAANDPHILLIPS1993HAVELEDTHEMETHODSTODEALWITHGRANGERCAUSALITYINI1SYSTEMSOFVARIABLESACAUSALLONGRUNRELATIONSHIPBETWEENNONSTATIONARYTIMESERIESWHENTHEYARECOINTEGRATEDCOULDBEINFERREDTHEREFORE,IFCOINTEGRATIONANALYSISISOMITTED,CAUSALITYTESTSPRESENTEVIDENCEOFSIMULTANEOUSCORRELATIONS
35、RATHERTHANCAUSALRELATIONSBETWEENVARIABLESTHEPRESENCEOFACOINTEGRATINGRELATIONFORMSTHEBASISOFTHEVECTORERRORCORRECTIONVECSPECIFICATIONADDITIONALLY,STANDARDGRANGERORSIMSTESTSMAYPROVIDEINVALIDCAUSALINFORMATIONDUETOTHEOMISSIONOFERRORCORRECTIONTERMSFROMTHETESTSDOYLE,2001THESIMPLEGRANGERSCAUSALITYTESTBECOME
36、SINAPPROPRIATEWHENCOINTEGRATINGVECTORSAREOBTAINEDINTHESERIESACCORDINGTOGRANGERSREPRESENTATIONTHEOREM,THERESULTSOFCOINTEGRATIONIMPLYTHATXANDYHAVETHEFOLLOWINGERRORCORRECTIONREPRESENTATIONSINEQUATIONS4AND5THESEARENECESSARYTOAUGMENTTHESIMPLEGRANGERCAUSALITYTESTWITHTHEERRORCORRECTIONMECHANISMECM,DERIVEDF
37、ROMTHERESIDUALSOFTHEAPPROPRIATECOINTEGRATIONRELATIONSHIPTOTESTFORCAUSALITYDLNYTC0XKI1BIDLNYT2IXKI1AIDLNXT2IPIECTT21UT48DLNXTC0XKI1GIDLNXT2IXKI16IDLNYT2IHIECTT211T5WHEREYANDXARETHEVARIABLESUNDERCONSIDERATION,ANDRIISTHEADJUSTMENTCOEFFICIENTWHILEECTT21EXPRESSESTHEERRORCORRECTIONTERMOFGROWTHEQUATION,DIN
38、DICATESFIRSTDIFFERENCEOPERATORINEQUATION4,XGRANGERCAUSESYIFAIANDRIARESIGNIFICANTLYDIFFERENTFROMZEROINEQUATION5,YGRANGERCAUSESXIF6IANDHIARESIGNIFICANTLYDIFFERENTFROMZEROFSTATISTICISUSEDTOTESTTHEJOINTNULLHYPOTHESISOFAI,6I0,ANDTTESTISEMPLOYEDTOESTIMATETHESIGNIFICANCEOFTHEERRORCOEFFICIENT3RESULTSTABLEIG
39、IVESADFANDPPTESTRESULTSFORUNITROOT,WHICHPROVETHATALLTHEVARIABLESAREINTEGRATEDOFORDERONETHATISI1THISINDICATESTHATTHEFIRSTDIFFERENCESOFLNGDP,LNM2,LNDC,LNEXPANDLNIMPARESTATIONARYINTHEINDIANCASEFORTHISSAMPLEPERIODSTATISTICSLNGDPLAGLNM2LAGLNDCLAGLNEXPLAGLNIMPLAGLEVELSTTADF2148022551220132138021580TMADF19
40、30205112112104501150TADF104202461178160703850TTPP21184220922129321507206122TMPP40970250210431663849934TPP111614690198412773849924FIRSTDIFFERENCETTADF2534324311242102798026340TMADF2692024441242202753025660TADF2060222870225612052322641TTPP288562372424172288312277612TMPP26901238242420227375256459TPP220
41、9422851236932462224511NOTESTTREPRESENTSTHEMOSTGENERALMODELWITHADRIFTANDTRENDTMISTHEMODELWITHADRIFTANDWITHOUTTRENDTISTHEMOSTRESTRICTEDMODELWITHOUTADRIFTANDTRENDNUMBERSINBRACKETSARELAGLENGTHSUSEDINADFTESTASDETERMINEDBYAICSETTOMAXIMUM3TOREMOVESERIALCORRELATIONINTHERESIDUALSWHENUSINGPPTEST,NUMBERSINBRAC
42、KETSREPRESENTNEWEYWESTBANDWITHASDETERMINEDBYBARTLETTKERNEL,ANDDENOTEREJECTIONOFTHENULLHYPOTHESISATTHE1,5AND10PERCENTLEVELS,RESPECTIVELYGDPSTANDSFORGROSSDOMESTICPRODUCT,M2STANDSFORMONEYANDQUASIMONEYASPERCENTOFGDP,ANDDCSTANDSFORDOMESTICCREDITPROVIDEDBYBANKINGSECTORASPERCENTOFGDP,EXPSTANDSFOREXPORTSOFG
43、OODSANDSERVICESANDIMPSTANDSFORIMPORTSOFGOODSANDSERVICESTESTSFORUNITROOTSHAVEBEENCARRIEDOUTINEVIEWS51TABLEIADFANDPPTESTSFORUNITROOTFINANCIALDEVELOPMENT,TRADEANDGROWTHHAVINGESTABLISHEDTHENECESSARYCONDITIONSFORTHESTATIONARITYOFDATAUNDERINSPECTION,WECONDUCTJOHANSENSCOINTEGRATIONTESTJOHANSEN,1988JOHANSEN
44、ANDJUSELIUS,1990,WHICHISVERYSENSITIVETOTHECHOICEOFLAGLENGTHCHANG,2002,TOEXPLOREANYPOSSIBLELONGRUNRELATIONSHIPAMONGTHEVARIABLESUNDERCONSIDERATIONWEEMPLOYBOTHAKAIKEANDSCHWARTZCRITERIATOSELECTTHENUMBEROFLAGSINTHECOINTEGRATIONTESTWHERETHETWOCRITERIASUGGESTAVARMODELWITH1LAGTHERESULTSSHOWINGNUMBEROFCOINTE
45、GRATINGVECTORSAREREPORTEDINTABLEIITHATPRESENTSONLYTHETRACETESTRESULTSASSUGGESTEDBYCHEUNGANDLAI19937JOHANSENTESTRESULTSSHOWTHATEVERYPAIRINTABLEIIISCOINTEGRATEDWITHEACHOTHERTHISMEANSTHATLONGRUNEQUILIBRIUMRELATIONSHIPEXISTSBETWEENTHESEPAIRSITISALSOUSEFULTOMENTIONTHATMORETHANONECOINTEGRATINGVECTORHASBEE
46、NOBTAINEDBETWEENLNGDPANDLNEXP,LNGDPANDLNIMP,ANDLNM2ANDLNIMPWHEREASOTHERPAIROFTHEVARIABLESARECOINTEGRATEDWITHATMOSTONECOINTEGRATINGVECTORNOTESRDENOTESTHENUMBEROFCOINTEGRATINGVECTORSAKAIKEINFORMATIONCRITERIONAICANDSCHWARTZCRITERIASCWEREUSEDTOSELECTTHENUMBEROFLAGSREQUIREDINTHECOINTEGRATIONTESTBOTHGAVET
47、HESAMELEVELOFLAGORDERTABLEIICOINTEGRATIONTESTSUSINGTHEJOHANSEN1988ANDJOHANSENANDJUSELIUS1990APPROACHSINCE,COINTEGRATIONRELATIONSHIPISFOUNDBETWEENTHEVARIABLESUNDERINSPECTION,ANECMMODELSHOULDBECONSTRUCTEDTODETERMINETHEDIRECTIONOFTHECAUSALITYGRANGER1988MENTIONSTHATTHERESHOULDBEATLEASTONEDIRECTIONOFCAUS
48、ALITYAMONGTHEVARIABLESIFTHEYARE10COINTEGRATEDTHECAUSALITYMODELISEXPRESSEDASANERRORCORRECTIONMODELASINEQUATIONS4AND5SINCETHEVARIABLESARECOINTEGRATEDTABLEIIIREPORTSTHEFSTATISTICSANDTSTATISTICSFORERRORCORRECTIONTERMCONSTRUCTEDUNDERTHENULLHYPOTHESISOFNONCAUSALITYREJECTIONOFTHENULLHYPOTHESISIMPLIESTHATTH
49、ECORRESPONDINGVARIABLEGRANGERCAUSESTHEDEPENDENTVARIABLETHEGRANGERCAUSALITYTESTRESULTSTABLEIIISUGGESTTHATUNIDIRECTIONALCAUSALITYRUNSFROMGDPTOEXPORTS,FROMGDPTOIMPORTS,FROMEXPORTSTOM2,FROMM2TOIMPORTS,FROMEXPORTSTODOMESTICCREDITS,FROMIMPORTSTODOMESTICCREDITS,ANDFROMEXPORTSTOIMPORTSBIDIRECTIONALCAUSALITYHASALSOBEENOBTAINEDBETWEENGDPANDM2,ANDBETWEENGDPANDDOMESTICCREDITSFINALLY,NODIRECTIONOFCAUSALITYHASBEENOBTAINEDBETWEENM2ANDDOMESTICCREDITSIFTHESERESULTSARETOBESUMMARIZED,THESUPPLYLEADING,THEDEMANDFOLLOWING,EXPORTLEDANDIMPORTLEDHYPOTH
