[PDF] Dominant-currency pricing and the global output spillovers from US PDF ecb.wp2308~ead44a2c81.en.pdf

currency Underlying this prediction is a change in an economy's net exports ( 2019) provide extensive empirical evidence based on global trade data for 2500 country pairs American Economic Journal: Macroeconomics 7, 44–76

Previous PDF

Next PDF

[PDF] Foreign exchange

5 mar 2021 · 1) PPP data from OECD, based in Local Currency per USD 2) Current 2019 2020 2021 World: Higher financial stress supports USD

[PDF] 2019 FX Outlook: Peak Dollar - ING Think - ING Bank

30 nov 2018 · As 2019 progresses, expect a bearish dollar narrative to develop US rates should be coming off their highs by the end of the year and as US growth converges lower on the rest of the world, expect investors to rotate out of US asset markets

[PDF] Canadian Dollar - CIBC

CANADIAN DOLLAR: STORM FRONT COMING OUTLOOK AND PORTFOLIO IMPLICATIONS By Vincent Lépine, Éric Morin, and Michael Sager1 June 2019

[PDF] Forecasting Dollar Real Exchange Rates and the Role of Real

∗We thank seminar participants at the 2019 MEG meeting for useful comments using the ratio of the root mean squared prediction error (RRMSPE) criteria If underlying driving forces of the American economy are more closely connected

[PDF] Forecasting the US Dollar-Korean Won Exchange Rate: A Factor

ican real activity factors are employed, whereas American nominal/financial market factors help improve short-run prediction accuracy Third March 2019

Exchange Rates Change by Using Economic Data - ScienceDirect

Keywords: Artificial Neural Networks; Crisis Forecast; Currency and Gold Rate; There are many studies that can affect and predict gold prices and exchange rates in Turkey, Turkey American interest data and interest data were used as external factors Guler Tepecik/ Procedia Computer Science 00 (2019) 000– 000

[PDF] Dominant-currency pricing and the global output spillovers from US

[PDF] Dollar and Exports - Bank for International Settlements

American University Hyun Song Shin of trade due to Gopinath and Stein (2017 ) and Gopinath et al (2019) These papers that are dollar-funded, consistent with the prediction that the exports of intermediate goods are more sensitive to a

[PDF] The Quanto Theory of Exchange Rates - personallseacuk

American Economic Review 2019, 109(3): 810–843 correlation between the S&P 500 index and the dollar/euro exchange rate If the euro appreciates against

Working Paper Series

Dominant-currency pricing and the

global output spillovers from

US dollar appreciation

Georgios Georgiadis, Ben Schumann

Disclaimer: This

paper should not be reported as representing the views of the European Central Bank (ECB). The views expressed are those of the authors and do not necessarily reflect those of the ECB.

No 2308 / August 2019

Abstract

Dierentexp ort-pricingcurrencyparadigmsha ve dierentimplicationsfor ahost of issuesthat arecritical forp olicymakers such asbusinesscycleco-move ment, optimal monetaryp olicy,optimumcurre ncyareas andinternationalmonetary policyco-ordination. Unfortunately, theliteraturehas notreac hedaconsensus on whichpricingparadigm best describes thedata.Againstthis background,we test forthe empiricalrelev anceof dominant-currencypricing(DCP). Specically , werst setup astructural three-country NewKeynesian dynamicsto chasticgen- eral equilibriummodelwhic hnestsDCP, producer-currencypricing (PCP)and local-currencypricing (LCP).In themo del,und er DCPtheoutputspillo versfrom shocksthatappreciate theUS dollarm ultilaterallyd eclin ewith aneconomy's export-importUSdollar pricingshare dierential, i.e.the dierenceb etween the shareof aneconom y'se xportsand importsthatarepricedinthe dominant currency.Underlying thisprediction isa chan gein aneconom y'snetexports in responsetom ultilateralc hangesintheUS dollarexchangerate that arises becauseof dierencesin theexten tto which exportsandimp ortsar epricedin the dominantcurrency. Wethenconfron tthispredictionof DCPwith thedata in asample ofup to46 advanced ande merging marketeconomiesfor thetime periodfrom1995 to2018. Specically ,con trollingfor othercross-bordertrans- mission channels,we documentthat consistentwiththepredictionfrom DCPthe output spilloversfromUSdollar appreciationcorrelate negatively withrecipien t economies' export-importUSdollarin voicing sharedieren tials.Wedocument that these ndingsarerobustto consideringUS demand,US monetaryp olicyand exogenous exchangerate shocksas atriggerofUSdollarap preciation,as well as to accountingfortherole ofcommo dity tradein USdollar invoicing. Keywords:Dominant-currencypricing, USsho cks, spillovers. JEL-Classication: F42,E52, C50.ECB Working Paper Series No 2308 / August 20191

Non-technicalsummary

In theclassic Mundellianin ternationalmacroeconomicmodelit isassumedthatexp ortprices are stickyinthe producer's currency(pro ducer-currencypricing;PCP).An alternative sug- gested morerecen tlyistoassume thatexp ortprices arestic kyin theimporter'scur ren cy (local-currencypricing; LCP).And even morerecen tly,ithas bee nsuggestedasyetanother alternativethat export pricesaresticky ina dominant currencyregardlessoftheir destina- tion, forexample inUS dollar(dominan t-currencypricing; DCP).Imp ortantly,the dynamic relationships betweenkeymacroeconomic variablesinresp onsetodomesticand USshocks are ratherdieren tunderDCP, PCPand LCP.For example,in contrast towidespreadintu- itions shapedby theimplicitassumptionof PCP, changes inan economy's exchangerated o not aectits termsof tradeunder DCP, andhence play littleoreve nno rolein expenditure switching.Also, underDCP pass-throughof changes inan economy's exchangerateagainst the dominantcurrencyishigh whilethat ofc hangesin exchange ratesagainst non-dominant currencies issm all.Andwhileam ultilateraldepreciation ofan economy's exchangerate has alimited expansionaryimpact onits exports, expenditure switchin gstilloccurs through imports.And nally, theUSdollar exchange rateis drivingglobaltrade ratherthanonly transactions inwhic htheUSis inv olved. Becausedieren texportpricingparadigmshave dierentimplications fora variet yof issuesincludingbusinesscycleco-m ov ement andrisk sharing, optimalmon etarypolicy,opt imumcurrenc yareas,andinternationalmonetary pol- icy co-ordination,it iscritical foracademics andp olicymakers tokn ow whichparadigmis relevantinthe data. In thispaper wecontrib utetotheliterature byprovidingevidencefor theempirical relevance of DCPfor alarge shareof globaltrade. Inparticular, basedon astructural model we rst showthatunder DCPthe outputspillo vers froma multilateral USdollarappreciation decline withan economy's export-importUS dollarpricingsharedierential,i.e. thedier- ence betweentheshareofan econom y'sexp ortsand imports thataresubjecttoDCP .The intuitionis thatas anec onom y'sshare ofexportsthatissubjectto DCPrises relative tothe correspondingshare ofimp orts,appreciation oftheUSdollar willelicit strongerexp enditure switchingin theec onom y'stradingpartnersandhenceproducea largerdecline inits exports than inits imports, whichimp liesthatitsnet exportsandthusoutput willfall morestron gly. Weexploit thisprediction fromthe model inorder totest fortheempiricalrelev anceof DCP in thedata. Inparticular, because ofthe lackofdata oneconomies' export-imp ortUSdol- lar pricingshare dierentials, wetest thejoint hypothesisthat thedata arecharacterised bypartial DCPand thatthe export inv oicingcurrency coincideswiththepricing currency. More concretely,we testempiricallywhethereconomies' export-imp ortUS dollarin voicing share dierentialsarenegativ elyr elated tocross-countrydierencesintheestimatedoutput spilloversfromUS dollarap preciation. Specically,we rstestimatethespillo vers fromUS dollarappreciation usingtwo-country

ECB Working Paper Series No 2308 / August 20192

VARm odelsforupto 46adv ancedandemergingmark eteconomies forthe timeperiod from

1995 to2018. Thet wo-coun tryVARmodelsconsistofthe USandoneeconom yatthetime.

Weiden tifyapositiv eUS demand,contractionarymonetarypolicy andexogenous exchange rate shocksthatappreciatethe USdollar usingstandard signrestrictions. Ina secondstep, werun cross-sectional regressionsinorderto explainthe cross-country dierences inthe estimated outputspillo versfromUSdol larappreciation.Sp ecically, weregressthe output spilloversestimatedfrom theV ARmo delson theexport-import USdollarinv oicingshare dierential.In theregressions we alsocon trolforcross-country dierences inthesusceptibility to otherspil lovertransmissionchannelsthatare relevan tinthe structuralmo del:Economies ' susceptibilityto bilateral USandmultilateral demandc hannelsof spilloversre ected by their overallopennessto tradeaswell astheir bilateraltrade integration withtheUS;and dierences inthe strengthof expenditure switching conditionalontradepricing re ected byeconomies' exch angerate exibilityvis-a-vis the USdollar.Ourndings areconsistent with thepredictions ofpartial DCPfrom thetheoretical model (andthe assumptionthat the currencyof pricingcoincides withthe currencyof inv oicingof exports): Economiesthat exhibit largerexp ort-importUSdollarinv oicingshare dierenti alsexhibitsmaller|less positiveormore negative |output spilloversfrom USdoll arappreciation. Wedocumentth at these ndingsare robustto arange ofalte rnativ esp ecicationsof theempiricalframework bothregarding theV ARmo delandthecross-s ectionalregressions,including inparticular controllingfor theimp ortanc eofcommoditytrade.Finally ,w ealsoconsiderthespillovers from Euroinsteadof USdollar appreciationand documen tthat thendings provide less supportfor partialDCP ,whi chisc onsistent withwhat onewouldexpectgiventhemuchless prominentrole ofthe Euroin globaltrade.

ECB Working Paper Series No 2308 / August 20193

1 Introduction

In theclassic Mundellianin ternational macroeconomicmodelitis assumedthatexport prices are stickyinthe producer's currency(pro ducer-currencypricing;PCPMundell, 1963;Flem- ing, 1962;Obstfeld andRogo, 1995).An alternative suggestedmore recently istoassume that exportpricesare sticky inthe importer'scurrency(lo cal-currencypricing; LCPB etts and Devereux,1996,2000; Engel,2000; Devereux andEngel, 2003).And evenmorerece ntly , it hasb eensuggestedasy etanother alternative thatexportprices arestic kyinadominan t currency regardlessof theirdestination, forexample inUS dollar(dominan t-currencypricing; DCP Casaset al.,2016; Boze tal., 2017;Gopinath etal.,2019).Imp ortantly ,the dynamic relationships betweenkeymacroeconomic variablesinresp onsetodomesticand USshocks are ratherdieren tunderDCP, PCPand LCP.For example,in contrast towidespreadintu- itions shapedby theimplicitassumptionof PCP, changes inan economy's exchangerated o not aectits termsof tradeunder DCP, andhence play littleoreve nno rolein expenditure switching.Also, underDCP pass-throughof changes inan economy's exchangerateagainst the dominantcurrencyishigh whilethat ofc hangesin exchange ratesagainst non-dominant currencies issm all.Andwhileam ultilateraldepreciation ofan economy's exchangerate has alimited expansionaryimpact onits exports, expenditure switchin gstilloccurs through imports.And nally, theUSdollar exchange rateis drivingglobaltrade ratherthanonly transactions inwhic htheUSis inv olved. Becausedieren texportpricingparadigmshave dierentimplications fora variet yof issuesincludingbusinesscycleco-m ov ement andrisk sharing, optimalmon etarypolicy,opt imumcurrenc yareas,andinternationalmonetary pol- icy co-ordination,it iscritical foracademics andp olicymakers tokn ow whichparadigmis relevantinthe data. Against thisbackground, Casasetal.(2016) asw ellas Bozet al.(2017) provide evidence suggesting thatDCP tsthe datab etterthan PCPor LCP.

1Specically,Casaset al.(2016)

analyse customsdata fromColom biaon exportsandimp ortsat thermlev eland nd that |consisten twiththep redictionsof DCP| pass-throughofchanges inthe USdollar exchangerate top esoimp ortandexport pricesislargerelativ eto thatofchanges inbilateral exchangerates, regardlessof theorigin ofimp ortsor thedestination ofexp orts.Regarding quantities,Casas etal. (2016)nd thatfollo winga depreciationof thep esoagainsttheUS dollar Colombianexport quantitiestodollar destinationsd onot fall,that exportstonon- dollar destinationsdofall, andthat imports fromb othdollar andnon-d ollareconomies fall. As regardsexc hangeratepass-throughtoprices, Casaset al.(2016) documen tthat when quantitiesresp ondtherelevan texc hangerateisthepeso-dollar andnottheb ilateralexc hange rate. Andnally ,Casasetal. (2016)nd thatColom bia'sterms-of-trade |when adjusted for commodityprices|are ratherstable. Bozet al.(2017) provide similarevidence basedon an annualpaneldataset ofbilat eral import andexportpricesand quantitiesfor55 economies.1 The papersby Casasetal.(2016)as well asBozetal. (2017)ha ve been combine din toGopinathetal. (2019).ECB Working Paper Series No 2308 / August 20194 In thispap erwecon tributeto theliterature by providingnewevidence fortheempir ical relevanceof DCPand by addressingsome importantshortcomings ofexisting work. Speci- cally,w erstshow thatin astandardstructural open-econom ymo delunder DCPtheoutput spilloversfromsho cks thatappreciatetheUSdollarm ultilaterally declinewith aneconom y's export-importDCPshare dierential, i.e.the dierencebetw eenthe shareof aneconom y's exportsand imports thataresubject toDCP .The intuitionisthat asan economy'ssh areof exportsthat issub jectto DCPrisesrel ative tothe corresponding shareofimports,am ulti- lateral appreciationof theUS dollare lic its strongerexpenditureswitchingintheeconom y's trading partnersand hencepro ducesa largerdeclineinits exports, which impliesthat its net exportswillf allmore strongly.We exploitthis predictionfromthemo delinorde rto test for theem piricalrelevanceofDCP inthedata.In particular,b ecauseof thelac kof dataon economies' export-importUSdollarpricing sh aredieren tials ,wetestthe jointhypothesis that thedata arec haracterised byDCPandthatthe exportinvoicingcurrency coincideswith thepricingcurrency.More concretely, wetest empiricallywhetherecon omies' export-imp ort US dollarin voicingsharedierentialsare negatively relatedtocross-country dierencesin the estimatedoutput spillov ersfromshocks thatappreciatetheUS dollarmultilaterally.We nd thatthe dataare consistent withthe hypothesisthat alarge shareof globaltradeis subjecttoDCP: Outputspillo vers from shocksthatappreciatethe USdollarmultilaterally are statisticallysignican tlyandnegatively relatedto economies'export-import USdollar invoicingsharedieren tials. In moredetail, we rstconsiderathree-coun tryNew Keynesiandynamic stoc hasticgeneral equilibrium modelfortheUSand there stof thew orld.W esplit therestofth ew orld in tworegionsofequal sizein orderto account forthe role ofDCP fornon-US, third- countrytrade. Importan tly,themodelweset upn eststhecases ofPCP,LCP andDCP as wellascom binationsthereof. Wethe nexplore theoutputspillov ersfromUSdemand, monetary policyandunco ve redinterestrateparity(UIP)sho cksthat appreciatethe US dollar multilaterally,andinparticular how thesev aryacross dierentassumptionsregarding the currencyin which exportsarepriced. Inparticular, incase ofPCP theoutputspillov ers from USdollar appreciationto therest ofthe world arep osit ive. Specically,US dollar appreciation increasesth erestofthe world's netexp ortsvis- a-vistheUS, asit reduces(raises) the restof thew orld'simp ortsfrom(exports to)theUSthrough expenditure switching.In contrast,in caseof LCPthe outputspill ov ersfrom USdollar appreciationtotherestof the worldare muted.Specically ,asboth USandrestofthew orldexp ortpricesarestic ky in theimp orter'scurrency,US dollarappreciationdoes nottrigger signican texp enditure switching.Most importan tly,however,both incaseofPCPandLCPthereis noexpenditure switchingin tradeb etw eentheeconomiesintherest oftheworld astheir exchange rates depreciate symmetricallyagainst theUS dollar. In caseoffullDCP allexp ortpricesarestic kyin USdollar regardlessofthedestination and origin, whichim pliesthatUSdollarappreciationdo eselicit expenditure swi tching intrade

ECB Working Paper Series No 2308 / August 20195

betweentheeconomiesin therest ofthe world. Specically,assume forsimplicitythattrade of economiesin therest ofthe world isexclusiv elyin tra-regional,i.e.therest ofthe world doesnot trade withtheUS.Because pricesof intra-regional imports in therest ofthe world are stickyinUSdollar, USdollar appreciationtriggers expenditure switching aw ay from importsfrom othereconomies inthe restof thew orldto wards domestically-prod uced goods. As amirror image,economies' exports toother economiesintherest ofthe world pricedin US dollarac cordinglydeclineduetothese expenditure switching eects.Most importan tly, however,becauseUS dollarappreciationreduces intra-regional exportsandimp ortsth rough expenditureswitc hingtotheexact samedegree inall economiesin therest ofth ew orld, intra-regionalnet exports andhenceGDPare unchanged. In contrasttothe caseof fullDCP ,in caseof partialDCP |that isth es haresofan economy'sexp ortsandimports thatare subjecttoDCP bein gdieren tingeneral| the output spilloversfromUSdollar appreciationthat stemfrom intra-regional tradein theres t of thew orldareingeneral dierent fromzero. For example,incaseof aneconom yin the rest ofthe world featuringap ositiveexport-imp ortDCP sharedierential|i.e. alarger share ofexp ortsthanimports be ingpricedin USdollar|USdollarappreciationreducesits importsfrom othereconomies inthe restof thew orldb yless throughexp enditureswitc hing than itreduces itsexp ortsto othereconomiesin therest ofthe world. Asare sult,the output spilloversfromUS dollarappreciation thatarise throughin tra-regionaltrade aresmaller | less positiveormorenegativ e| foran economywitha positiv eexp ort-import DCPshare dierential.The model thusimplies atestablepredictionregardi ngthemagnitude ofoutput spilloversfromUS dollarappreciation underpartial DCP:The outputspillo vers fromUS dollar appreciationdecline witheconomies' non-USexp ort-import DCPshare dierentials. Wedo cumentthatthisprediction frompartialDCPis qualitatively robustto changes in modelparametrisation, considering avariety ofs hocksthatinduce USdollar appreciation suchas positiv eUSdemand,contractionaryUS mone tary policyandUIP shoc ks,aswellasto the inclusionof additionalm odel featuressuchastradein intermediate inputsforproduction. Due tothe lack ofdataonexp ortan dimp ortpricing, weconsiderthe jointhypothesisof partial DCPb eingempiricallyrelevan tand thatthecurrencyofexport pricingcoincides with the currencyof inv oicing.Thepredictionfromthisjoin th ypothesis isthat outputspillo vers from USdollar appreciationare negatively relatedto economies'non-USexport-imp ortUS dollarinvoicingshare dierential. Weth enconfrontthisprediction ofthemodel withthe data.Sp ecically, we rstestimatethe spilloversfromUS dollarappreciation usingt wo-coun tryV ARmo delsforupto46advanced and emergingmark eteconomiesforthe timep eriod from1995 to2018. Thetwo-countryV AR modelsconsist ofth eUS andonenon-USeconom yat thetime. We identify USdemand,US monetary policyandUIP shoc ksusing model-consis tent signrestrictions.Inasecondstep, werun cr oss-sectionalregressionsofthecross-countrydierences inthe spillov ersfrom US dollar appreciationoneconomies'non-US export-imp ortUS dollarin voicingsharedieren-

ECB Working Paper Series No 2308 / August 20196

tial. Inthe regressionsw ealso control forcross-coun trydierencesinthe susceptibility to other spillovertransmissionchannels thatare relevantin thestructural model: Economies' susceptibilityto bilateralUS andm ultilateraldemand channels ofspillovers re ectedb ytheir overallopenness totradeasw ellas theirbilateral tradein tegrationwiththeUS; anddif- ferences inthe strengthof expenditure swit chingconditionalon tradepricingre ectedby economies' exchangerate exibilit yvis- a-vistheUSdollar. Ourndingsareconsisten twith the predictionsof partialDCP fromthe theoreticalmo deland theh ypothesised coincidence of thecurrency ofpricing withthe currencyof inv oicingof exports: Economiesthat exhibita larger non-USexp ort-importUSdollarinv oicingshare dierential exhibitsmaller|lesspos- itiveor morenegativ e| outputspillovers from USdollar appreciation.We documentth at these ndingsare robustto arange ofalte rnativ esp ecicationsof theempiricalframework bothregarding thet wo-coun tryVARmodelsandthecross-sectional regression,inparticular controllingfor thep ossiblerole ofcommodit ytrade inUS dollarinvoicingandidentication using externalinstrumen tsinsteadofsign restrictions.Fin ally, we alsoconsider thespillovers from multilateralEuroinstead ofUS dollarappr eciation anddo cument thatthendings provideless support forpartialDCP, which isconsisten twithwhatone wouldexpe ctgiv en the muchlessprominen trole oftheEuroinglobaltrade. Our paperisrelated toexisting literature.Dev ereuxet al.(2007) discussth e\USdollar standard" asa combination ofPCPandLCP ina tw o-country model forthe USand therest of thew orld.Cookand Devereux(2006)and Goldberg andTille(2009)are earlycon tributions discussing therole ofDCP forthird-coun trytrade. Inparticular, CookandDe vereux (2006) explore theeects ofDCP ina partialequilibrium three-country NewKeynesian model and showthat itis criticalin orderto account forthe muted dynamicsof SouthEastAsianexp orts against thebac kgroundofthesizable currencydepreciations duringthe Asiannancial crisis. Goldbergand Tille(2009) explorew elfareand international monetarypolicyco-ordination under DCPin astatic three-country generalequilibrium model. Morerecen tly, Gopinath et al.(2019) considera rich three-country NewKeynesianDSGEmo delthatnestsPCP , LCP andDCP ,andde riv etestablepredictionsinordertoassesstheempiricalrelev anceof DCP.In particular,Gopinath etal. (2019)sho wthat underDCP tradeprices andquantities respondto theUS dollarrather thanto bilateralexc hangerates. Furthermore, Gopinathet al. (2019) provideextensive empiricalevidencebasedon globaltrade datafor 2500coun tryp airs as wellascustoms datafor Colomb iathat isconsisten twiththesepredictionsfromDCP . 2 Finally,Zhang (2018)sets upa tw o-perio dthree-coun trygeneralequilibriummodelandshows that underDCP exchange ratesofeconomieswith largershares ofthe consumptionbask et priced inUS dollarexhibit smaller depreciationsand largerinterestr aterises inresp onseto contractionaryUS monetaryp olicysho cks.Zhang(2018) thenprovidesempiricalevide nc e using high-frequencyestimates ofthe nancialmark eteects ofUS monetarypolicysho cks for aset ofadv ancedeconomies withdev elope dnancial marketsthatareconsistentwith2 Similar evidenceisfound by Chenet al.(2018) usinghighly disaggregatedcustoms datafor the UK.

ECB Working Paper Series No 2308 / August 20197

the predictionsof DCP. 3 Wecon tributeto thisexisting literatureb ypro vidingnew evidencefortheempirical rel- evanceof DCPin several dimensionsand byaddressingsome importan tshortcomings of existing work.Moresp ecically, relativetoGopinathetal. (2019),wecon tributeto the literature byassessingthe empiricalrelev anceof DCPtes tingforits predictionsfor out- putrather thantrade priceor quantit yspillo versfromchangesin theUSdollarexchange rate. Moreover,weconsider exogenouschangesin theUS dollare xchange raterather than reduced-form regressions,whic hismor econsisten twiththeexperiments inthetheoretical modelsfrom which thetestable predictionsof DCParederivedandwhic hr ed uces therisk of omittedv ariablebiasinthe empiricalanalysis. Relative toZhang (2018),w econ tribute to theliterature by assessingtheempiricalrelev anceof DCPtesting fori tspredictionsfor output spilloversfromUSshoc ksat thebusiness cyclefrequency ratherthan thoseon asset prices atthe dailyfrequency ,whic hdoes notindicatehow persistenttheestimatedeects are athorizons relevan tformacroeconomicmodels andp olicymakers.Moreov er,relativeto Zhang (2018)w ealsoincludein ouranalysis abroad rangeof emergingmarket ec onomies instead offo cusingonadvanced economieswith developednancial markets. Thisis par- ticularly informativeforassessingtheempiricalrelev anceof DCPas USdollar inv oicingofquotesdbs_dbs4.pdfusesText_8

[PDF] [PDF] Dominant-currency pricing and the global output spillovers from US