重构韩国银行全球预测模型：框架与分析

1HyukjinYoon*·SeungryeolJeong** 2*Financial Modeling Team,Office of Economic Modeling&Policy Analysis,Bank of Korea(Tel:02-759-5285,e-mail:hyukjin.yoon@bok.or.kr)**PolicyAnalysisTeam,MonetaryPolicyDepartment,BankofKorea(Tel:02-759-4806,e-mail:srj@bok.or.kr)▪Disclaimer:The views expressed herein are those of the authors and donotnecessarily reflect the official views of the Bank of Korea.When r-eportingor citing this paper,the authors’names should be always e-xplicitlystated.▪We would like to express our sincere appreciation to Jeongik Lee,generalmanagerofthe Officeof EconomicModelingandPolicy Analysis;ByeonghoBae,general manager of the Human Resources Development Institute;andMingyuSon,head of the Financial Modeling Team,for their significantcontributionto the preparation of this paper.Any errors in this paper aresolelythe responsibilityof theauthors.ContentUltimately,our findings highlight the need to continue diversifying export markets inresponseto the evolving global trade environment.They also underscore the importanceofclosely monitoring the U.S.dollar and financial conditions,given their significant influ-enceon both thedomestic and global economies.Ⅰ.IntroductionⅡ.BOK-GPMOverview1.BasicFrameworkofModel2.TransmissionChannelsofExternalShocks3.ModelEstimationProcessⅢ.ResultsoftheEstimatedModel1.GoodnessofFit2.Impulse-ResponseAnalysisⅣ.SummaryandImplications 3businesscycles,including the pandemic peri-od,andinvolvedre-estimatingallparameters.Second,we incorporated theemergingAsia region,one of the key eco-nomicpartners for Korea,into the model.Thisadditionreflectsthegrowingim-portanceof emerging markets since the2008global financial crisis.Third,we en-hancethe exchange rate channel to reflectthedollar’s dominant pricing role in globaltrade.By adding the U.S.dollar index,themodelcan now better capture how changesinthe dollar’s value,often driven by U.S.monetarypolicy,spill over to other econo-mies2).Fourth,the financial channel is en-hancedto better model cross-country link-ages,using credit spreads as the key trans-missionmechanism.Overall,the redevelopedBOK-GPMshows a strong fit for key macro-economicvariables like GDP growth andpromisesto be a versatile tool for analyzingthetransmission ofdiverse external shocks.Thispaper proceeds as follows.SectionⅡdescribes the model's structure.SectionⅢevaluates its performance and analyzesthetransmission of key external shocks us-ingimpulse-response functions.The finalsectionconcludes by summarizing the resultsandtheirpolicy implications.1)TheBOK‐GPMisasemi‐structuralmodel:itmaintains theoreticalcoherence byadaptingkey behavioralequationsderived from a DSGE framework to real‐world conditions,while enhancing data fit throughvariablesandBayesianestimationofprincipalparameters.For a detailed description of the original development of the BOK-GPM,see Kang et al.2)Whereas the legacy model accounted solely for real effective exchange rate movements,theredevelopedBOK-GPM further incorporates the impact of dollar index fluctuations,reflecting the Ⅰ.IntroductionSincethe 1980s,the rapid expansion ofglobaltrade and capital flows has strength-enedboth real-financial linkages and thecross-countrysynchronization of economicvariables.More recently,global trade un-certaintyhasintensified,COVID-19pandemic and a trend towardtradefragmentation.These global dynamicsareparticularly consequential for a small,openeconomy like Korea,which is highlydependenton external demand.A preciseunderstandingof how external shocks affectthedomestic real and financial sectors isthereforeessential for improving the accu-racyofeconomicanalysis.This requires continuously refiningexistinganalytical frameworks to incorporatethelatest international research,particularlyonthe economic impact of shifting tradepatternsand the global spillovers of U.S.monetarypolicy.Therefore,this paper firstpresentsthe redevelopment of the Bank ofKorea'sGlobal Projection Model(BOK-GPM)1),whichwas originally built in 2014 based ontheIMF's global modeling framework.Itthenuses the new model to analyze thetransmissionof keyexternal shocks.Thisredevelopment of the BOK-GPM fo-cusedon four key enhancements.First,theestimationsample is extended from its origi-nal2000-2010 period to cover data through2023.This allows the model to better reflecttheevolving global trade environment andtrend‐cycledecomposition(2014).dollar’suniqe roleininternationaltrade. drivenbytheforecastsandpolicyofmacroeconomic Ⅱ.BOK-GPMOverview1.BasicFrameworkofModelTheBOK-GPM is a multi-country modelcomposedofsix economies:UnitedStates,China,the Euro Area,Japan,andemerging Asia.The Euro Area includes19member countries,while the emergingAsiaregioncomprisesIndia,Malaysia,thePhilippines,Singapore,andThailand3).Collectively,these economies ac-countfor approximately 70%of global GDPasof 2023.The five foreign economies alsorepresent63%of Korea’s total trade.A com-parisonwith 2014 data shows that while thetradeshares of China and Japan have de-clined,the U.S.share h