purposejjcao.com

BOPCOM-05/61EighteenthMeetingoftheIMFCommitteeonBalanceofPaymentsStatisticsWashington,D.
C.
,June27–July1,2005RoundTrippingDIRECTINVESTMENTTECHNICALEXPERTGROUP(DITEG)OUTCOMEPAPER(DITEG)#13December23,20041.
Topic:RoundTripping2.
Issues:SeeDITEGIssuePaper#13byHongKong,China3.
Recommendations:(i)ThegroupexpressedahighlevelofinterestintheconceptofroundtrippingandagreedthatitisaninterestingdimensionofFDI.
(ii)ThegroupexpressedsomeconcernsaboutthemethodologypresentlybeingusedbyHongKongChina,aswellasconcernsaboutthedifficultyofeffectivelyidentifyingandcoveringallroundtrippingFDIflows.
Italsorecognizedtheneedformoreworktobedoneondevelopingthemethodologyandconcepts.
(iii)Thegroupagreedthatitwouldbeusefultoextendtheanalysisofroundtrippingtocoveradditionalsituations,suchaswherethetwoendsofthechainofroundtrippingflowsarenotinthesamecountrybutratherinthesamegeographicalregionoreconomiczone.
(iv)ThegrouprecognizedthatmanyoftheissuesdiscussedunderroundtrippingareconnectedtootherFDIconceptualissues,suchasUltimateBeneficialOwners(UBOs),andthatitwouldbenecessarytoconsidertheseissuesinacoherentmannerinthenewstandards.
(v)Thegroupagreedthat,ifitisdecidedthatdataonroundtrippingistobeincludedinthebalanceofpayments(BOP)/internationalinvestmentposition(IIP)frameworkinthenewmanual,thedatashouldbetreatedassupplementalBOP/IIPpresentationsratherthanasmandatory,standardcomponents.
4.
RejectedAlternatives:None.
5.
QuestionsfortheIMFCommitteeonBalanceofPayments(theCommittee)andtheOECDWorkshopinInternationalInvestmentStatistics(WIIS)(i)DotheCommitteeandtheWIISconsiderthatdataonroundtrippingshouldbeincludedintheframeworkoftheBOP/IIPstatisticsatthisstage,giventheexperimentalnatureofthedata(ii)Iftheanswerto(i)is"Yes",dotheCommitteeandtheWIISconsiderthatanypresentationofdataonroundtrippingintheBOP/IIPstatisticsshouldbesupplementalratherthanastandardormandatoryreportingrequirementIMFCOMMITTEEONBALANCEOFPAYMENTSSTATISTICSANDOECDWORKSHOPONINTERNATIONALINVESTMENTSTATISTICSDIRECTINVESTMENTTECHNICALEXPERTGROUP(DITEG)ISSUESPAPER(DITEG)#13ROUNDTRIPPINGPreparedbyCensusandStatisticsDepartmentHongKong,ChinaSeptember2004-2-DIRECTINVESTMENTTECHNICALEXPERTGROUPISSUESPAPER(DITEG)#13:ROUNDTRIPPINGI.
Currentinternationalstandardsforthestatisticaltreatmentoftheissue1.
Theterm"roundtripping"isnotexplicitlydiscussedintheBPM5andOECDBenchmarkDefinitionofForeignDirectInvestment(BenchmarkDefinition).
ItisdefinedintheAnnotatedOutline(AO)fortherevisionofBPM5asthechannellingbydirectinvestorsoflocalfundstoSPEsabroadandthesubsequentreturnofthefundstothelocaleconomyintheformofdirectinvestment.
2.
Para.
365oftheBPM5andpara.
39oftheBenchmarkDefinitionstatethatSPEsbeincludedasdirectinvestmententerprisesiftheymeetthecriteriaofdirectinvestment,exceptforSPEswithasolepurposeofservinginafinancialintermediarycapacity.
AlthoughSPEs,typicallysetupinoffshorefinancialcentresortaxhaveneconomies,mayhavedifferentstructuresorpurposes,theyareanintegralpartofthestructureofthedirectinvestmentnetwork.
Assuch,roundtrippingfundsflowedwithinthedirectinvestmentnetworkarecurrentlyrecordedasFDItransactionsonagrossbasis,i.
e.
asdirectinvestmentabroadforthelocalfundschannelledtoSPEsabroad,andasdirectinvestmentinthereportingeconomyforthesubsequentreturnofthefundstothelocaleconomy.
II.
Concerns/shortcomingsofthecurrenttreatment3.
FDIinflowtosomeeconomieshasincreasedsignificantlyinthepastseveralyears.
AshareofthetotalFDIinflowtotheseeconomies(1)isexpectedtobeofthenatureofroundtrippingfunds,giventhatinthoseeconomies,investmentincentivesareavailabletoforeigninvestorsbutnotdomesticones.
Inadditiontotaxandfiscaladvantagesthatareprovidedtoforeigninvestors,thereareotherincentivesforroundtripping,suchassafetyandriskmanagementofcapital,accessingbetterfinancialservices,etc.
Someelaborationsontheseincentivesaregivenbelow:(a)TaxandfiscaladvantagesSomeeconomiesprovidepreferentialpoliciestoattractforeigndirectinvestment,includinglowtaxation,favourablelanduserights,convenientadministrativesupport,etc.
Sinceitisnotalwayseasyforlocalenterprisestofindforeigninvestorswhoarewillingtoinvestinthem,theymayfirstlychannelcapitalabroadwhichisthendisguisedasforeigncapitalforlocalinvestmenttotakeadvantageofthepreferentialtreatmentsonlyavailabletoforeigninvestors.
(1)Forinstance,40%oftotalFDIinflowtoHongKongbetween1998and2002wasrelatedtoroundtripping.
-3-(b)PropertyrightprotectionInfrastructureforpropertyrightprotectioninsomeeconomiesisnotwellestablished.
Therefore,theenterprisesintheseeconomiesmayhavethemotivationtoparktheirwealthinaffiliatedenterprisessetupinoverseaseconomieshavingbetterlegalandinstitutionalsettingsforpropertyrightprotection.
Besides,someinvestorsmayprefertokeeptheiridentitiesanonymousbyinvestingthroughcompaniessetupinoffshorefinancialcentres.
CapitalwillthenbebroughtbacktothehosteconomiesintheformofFDIifthereareprofitableinvestmentopportunities.
(c)ExpectationsonexchangecontrolandexchangerateSomeeconomieshavecontroloncapitalaccountandexchangerate.
Expectationsonchangesinexchangecontrolandexchangeratemaygenerateroundtrippingforlargerflexibilityinforeignexchangemanagement.
(d)AccessingbetterfinancialservicesFinancialmarketsofsomeeconomiesarenotwelldeveloped.
Enterprisesoftheseeconomieshavetoaccessoverseasfinancialmarketsforobtainingbetterfinancialservices,suchaslistingofcompaniesinoverseasstockmarkets.
ThefundsraisedwillbebroughtbacktothehosteconomiesintheformofFDI.
Roundtrippingmayoccuraspartofthisprocess.
4.
AlthoughthecurrenttreatmentofrecordingofroundtrippingfundsunderFDIisinconformitywithrecommendationsoftheBPM5andBenchmarkDefinition,somearguethattheseroundtrippingfundsleadtoanoverstatementofthetruemagnitudeofFDI.
InordertobetterreflecttheabilityoftheseeconomiestoattractFDIotherthanroundtripping,andtoenhanceinternationalcomparability,itisusefultocompileasupplementarysetofFDIstatisticsthatexcludesroundtrippingfunds.
5.
WhileageneraldescriptionofroundtrippingactivityisgivenintheAO,morespecificguidelinesareneededinthenextBoPmanualforbothcompilersanduserstobetterunderstandtheactivity,giventhatitisoftendifficulttofollowcloselytheintricateenterprisestructureofdirectinvestmentnetworkandthefundsflowingwithinthenetwork.
6.
Fromtheperspectiveofthehosteconomy,thefollowingtwotypesofFDIflowsareconsideredasroundtripping:(a)domesticinvestmentdisguisedasforeigninvestmentthroughnon-residentSPE,e.
g.
inFigure1(a),CompanyAinthehosteconomyprovidesFDIfundstoanon-residentSPE(CompanyB)forinvestingbackinCompanyCinthehosteconomy.
(b)channellingofFDIfundsthroughlocalSPE,e.
g.
inFigure1(b),CompanyA'ineconomyXchannelsFDIfundstoCompanyC'inthesameeconomythroughaSPEinthehosteconomy(CompanyB').
-4-7.
Sincethisroundtrippingofcapitalfundsusuallyhavelittleeconomicsignificancetothehosteconomy,itisusefultoexcludethemforthepurposeofcompilingasupplementarysetofFDIstatisticsforthehosteconomytobetterreflectitsabilityinattractingFDI.
8.
TocompilethesupplementarysetofFDIstatisticsforthehosteconomy,onebasicapproachistoidentifyandascertainthetrueroundtrippingnatureofFDIfundsreceivedfromandprovidedbycompaniesinthehosteconomy.
Forinstance,incaseofFigure1(a),CompanyCcanbeaskedtoconfirmwhethertheFDIfundsreceivedfromCompanyBareinfactlocalfundsprovidedbyanothercompanyinthehosteconomy,i.
e.
CompanyA.
Moreover,CompanyAcanbeaskedtoconfirmwhethertheFDIfundsprovidedtoCompanyBwouldbe,andatwhattime,channelledbacktoanothercompanyinthehosteconomy,i.
e.
CompanyC.
ForthecaseofFigure1(b),CompanyB'canbeaskedtoconfirmwhethertheFDIfundsreceivedfromCompanyA'ineconomyXwouldbe,andatwhattime,channelledbacktoCompanyC'ineconomyX.
9.
Nevertheless,itisdifficulttoascertainthetruenatureofroundtrippingfundssimplybysurveyingthecompanies.
Becauseoftheinherentunderlyingincentivesforroundtripping(e.
g.
todisguisethecapitalflowsfortakingadvantageofpreferentialtreatmentsonlyavailabletoforeigninvestors),somecompaniesactuallyinvolvedinroundtrippingactivitiesmaynotbewillingtodisclosethetruenatureofFDIflows.
ForthetypeofroundtrippingshowninFigure1(b),theremaybeanadditionaldifficultyintryingtocontactandobtaininformationfromCompanyB'sinceitisaSPE(e.
g.
abrass-platecompany).
III.
Possiblealternativetreatments10.
ToovercometheinherentdifficultiesofdirectlyobtainingcomprehensiveinformationfromcompaniesonthetrueroundtrippingnatureofFDIfunds,compilerscanconsideridentifyingsomerecognisablestructuresofdirectinvestmentgroupstoassistinthecompilationofroundtrippingstatistics.
Sinceitmaybedifficulttofollowcloselyacomplicatedgroupstructureandthefundsflowingwithinthegroup,focusofthesurveyworkcouldfirstlybeplacedonthoserelativelysimplegroupstructuresthatareseentobeconducivetoroundtripping.
Formosteconomies,thisapproachwouldendupcapturingmostBCAHostEconomyFigure1(a)Figure1(b)EconomyXHostEconomyB'A'C'RestoftheWorld-5-oftheroundtrippingflows,whileenhancingtheeffectivenessofthesurveyapparatusasthemaintoolfordatacollection.
AnexampleofasimplegroupstructurethatisconducivetoroundtrippingisshowninFigure2.
11.
InFigure2,CompaniesA,BandCaremembersofadirectinvestmentgroup,whereCompaniesAandCareresidentsofthesamehosteconomy,andCompanyBisanon-residentSPE.
Thissimplegroupstructurewillbeconducivetoroundtrippingifthefollowingtwoconditionsaremet:(a)CompanyBiswholly-ownedbyCompanyA(i.
e.
the100%arrowshowninthediagram);and(b)CompanyBdoesnotownanycompanyotherthanCompanyC.
12.
Underthissimplestructureofdirectinvestmentgroup,itislikelythatFDIfundsreceivedbyCompanyCfromCompanyBareinfactprovidedbyCompanyA,i.
e.
roundtripping.
ThetwoconditionsgivenaboveprecludethepossibilitythatthefundsreceivedbyCompanyCareinfactprovidedbyotherdirectinvestorsordirectinvestmententerprisesofCompanyB.
13.
Ofcourse,itispossibleforCompanyBtoborrowfundsfromunaffiliatedcompaniesandchannelthefundstoCompanyCintheformofFDI.
ItisthereforeessentialtofindoutfromCompanyCaboutthesourceofFDIfundsprovidedbyCompanyB,i.
e.
whetherthefundscomefromCompanyA,orareborrowedfromothercompanies.
Forthepurposeofcompilingroundtrippingstatistics,intheeventthatCompanyCfailstofurtherclarifyandascertainthenatureandsourceofFDIfundsprovidedbyCompanyB,giventhestructureofthisdirectinvestmentgroup,itcanbeimputedthatthefundsareactuallyprovidedbyCompanyA,i.
e.
roundtripping.
BCAHostEconomyFigure2RestoftheWorld100%$x$y≥n%Remark:n%isthethresholdsignifyingadirectinvestmentrelationship.
-6-Timeofrecording14.
InFigure2,itisobviousthatCompanyAmayprovidecapitalfundstoCompanyBatatimedifferentfromthatwhenthefundsarechannelledbackfromCompanyBtoCompanyC.
AtthetimewhenfundsflowoutfromCompanyAtotheSPEabroad,onecannotbecertainthattheywillactuallyflowbacktothehosteconomy,notwithstandingtheobviousintentionbehindandthelikelihood.
Itisthereforereasonabletoconsiderthesefundsasroundtrippingfundsonlywhentheyactuallyflowbacktothehosteconomy.
Onlyatthetimewhenthisroundtrippingnatureofthefundsisascertained,bothoutflowandinflowofroundtrippingfundscanberecorded(i.
e.
atthetimewhenthefundsflowbacktothehosteconomy).
15.
Iftheflowsoffundsinoppositedirections(outflowandinflow)inaroundtrippingactivitytakeplaceintwodifferentreferenceperiods,saytheoutflowoccursinperiodt–1andtheinflowofthesameamountoccursinperiodt,anoutflowwillneedtobeimputedandrecordedtogetherwiththeinflowinperiodt,giventheruleofrecordingstatedabove.
16.
Thereisamajorlimitationforthisruleofrecording.
IncompilingthesupplementarysetofFDIstatisticsthatexcludesroundtrippingfunds,itispossiblethat,foragivenreferenceperiod,thetotalFDIoutflowtoaparticulareconomyissmallerthantheimputedoutflowofroundtrippingfundstothateconomy,givinganegativeFDIoutflowtothateconomyafterexcludingroundtrippingfunds.
Thiswillbedifficulttointerpret.
Selectingalongerreferenceperiod(e.
g.
oneyear)andadoptingabroadergeographicalbreakdown(e.
g.
aregion)willreducethechanceofrecordingsuchanegativeoutflowofFDIotherthanroundtripping.
Treatmentofdifferentflowsinoppositedirections17.
Thedefinitionofroundtrippingimpliesthatthetotalsizesoftheflowsoffundsinoppositedirections(outflowandtheninflow)inroundtrippingshouldbeidentical.
Inpractice,foragivenreferenceperiod,itispossiblethattheyaredifferent.
18.
ForthecaseshowninFigure2,theoutflowof$xmaybegreaterthan,equalto,orevensmallerthantheinflowof$yforareferenceperiod.
SincefundsflowingouttotheSPEabroadwillonlybeconsideredasroundtrippingfundswhentheyactuallyflowbacktothehosteconomy,$yshouldbetakenasthesizeoftherecordedroundtrippingfundsinbothdirections.
Evenif$xisgreaterthan$y,only$ywillberecordedasroundtrippinginthereferenceperiod,andtheresidual(i.
e.
$x–$y)wouldbetakenasreturningtothehosteconomyinalaterperiod.
Ontheotherhand,if$xissmallerthan$y,thegap(i.
e.
$y–$x)wouldbetakenashavingbeenprovidedtoCompanyBbyCompanyAinanearlierperiod.
Directinvestmentcomponentsofroundtrippingfunds19.
Roundtrippingfundsmaycoveranycomponentsofdirectinvestmentcapital,namely,equitycapital,reinvestedearning,orothercapital.
Itispossiblethatthecomponentsofdirectinvestmentcapitalaredifferentfortheoppositeflowsinroundtripping.
Forinstance,CompanyAmayprovideothercapitaltoCompanyBwhichinturnusesthefundstoincreaseitsholdingsofequitycapitalofCompanyC.
-7-Otherstructuresofdirectinvestmentgroupsconducivetoroundtripping20.
ThesimplegroupstructureshowninFigure2canbeeasilyextendedtocoverothermorecomplicatedgroupstructuresthatareconducivetoroundtrippingandareusefulforidentifyingpotentialroundtrippingactivitiesinsurveyworkforthepurposeofcompilingroundtrippingstatistics.
TwoexamplesofthesestructuresaregiveninFigures3and4below.
B1A1AFigure4HostEconomyCB2B3RestoftheWorldA2≥n%≥n%≥n%100%100%100%Remark:n%isthethresholdsignifyingadirectinvestmentrelationship.
B3CA100%≥n%B2B1100%100%HostEconomyFigure3RestoftheWorldRemark:n%isthethresholdsignifyingadirectinvestmentrelationship.
-8-21.
Insurveywork,forastructureofdirectinvestmentgrouptobeseenasconducivetoroundtripping(suchasthatshowninFigure3or4),theownershipstructureofthegroupshouldbesetupinawaythatprecludethepossibilitythatfundsreceivedbyCompanyCareinfactprovidedbyotherdirectinvestorsordirectinvestmententerprisesofCompanyB3.
Forthepurposeofcompilingroundtrippingstatistics,intheeventthatCompanyCfailstofurtherclarifyandascertainthenatureandsourceofFDIfundsprovidedbyCompanyB3,itcanstillbeimputedthatthefundsareactuallyprovidedbyCompanyA,i.
e.
roundtripping.
IV.
Pointsfordiscussion(1)Shouldbothtypesofroundtripping(i.
e.
domesticinvestmentdisguisedasforeigninvestmentthroughnon-residentSPEs,andchannellingofFDIfundsthroughlocalSPEs)beexcludedincompilingthesupplementarysetofFDIstatisticsforthehosteconomy(2)IncompilingthesupplementarystatisticsofFDIotherthanroundtripping,shouldsurveyworkbefocusedonlyonthosestructuresofdirectinvestmentgroupsthatareseentobeconducivetoroundtripping,giventhatitisdifficulttodirectlyobtainfrommostcompaniescomprehensiveinformationonthetrueroundtrippingnatureofFDIfunds(3)IsthesuggestedruleofrecordingforroundtrippingfundsappropriateIsthereanymethodtoovercomethelimitationofthepossibilityofrecordingnegativeFDIoutflowtoaneconomyafterexcludingroundtripping(4)ShouldroundtrippingfundscoverallcomponentsofdirectinvestmentcapitalIsitappropriatetoallowtherecordingofdifferentcomponentsofdirectinvestmentcapitalfortheoppositeFDIflows(outflowandinflow)inroundtripping(5)ArethereotherstructuresofdirectinvestmentgroupsthatarealsoseentobeconducivetoroundtrippingReferencesBalanceofPaymentManual,fifthedition,IMF,1993BenchmarkDefinitionofForeignDirectInvestment,thirdedition,OECD,1996AnnotatedOutlinefortheRevisionofBPM5,IMF,April2004