ContentslistsavailableatScienceDirect
Journal
of
Financial
Stability
journal homepage:www.elsevier.com/locate/jfstabil
The
influence
of
government
intervention
on
the
trajectory
of
bank
performance
during
the
global
financial
crisis:
A
comparative
study
among
Asian
economies
夽
Cherng
G.
Ding
a,∗,
Chiu-Hui
Wu
a,
Pao-Long
Chang
baInstituteofBusinessandManagement,NationalChiaoTungUniversity,Taipei,Taiwan bDepartmentofBusinessAdministration,FengChiaUniversity,Taichung,Taiwan
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received25July2011
Receivedinrevisedform12April2012 Accepted7November2012
Available online 16 November 2012 JELclassification: C1 G2 H8 Keywords: Bankperformance Globalfinancialcrisis Governmentintervention Piecewiselatenttrajectorymodel
a
b
s
t
r
a
c
t
Theglobalfinancialcrisisthatstartedfrom2007onwardsspreadaroundtheworldandimpactedthe per-formanceofbanksinmajoreconomies.Manygovernmentshaveusedavarietyofinterventionpolicies torecovertheirfinancialsystems.Byexaminingthedynamicchangesinbankperformancebeforeand aftergovernmentintervention,thisstudydemonstratestheuseofthepiecewiselatenttrajectorymodel. WeusedthedatacollectedfromBloombergforbanksoffivemajorAsianeconomies,Japan,SouthKorea, HongKong,SingaporeandTaiwan,overtheeleven-quarterperiodfromthe4thquarterof2007tothe 2ndquarterof2010onsixfinancialperformanceindicatorsreflectingsolvency,creditriskand profitabil-ity.Thechangepatternsofbankperformancebefore/aftergovernmentinterventionduringtheglobal financialcrisishavebeencomparedamongthefiveeconomies.Ourempiricalresultsindicatethat,on average,thebankperformanceintermsofsolvency,creditrisk,andprofitabilityimprovesafter govern-mentintervention.Moreover,theinfluenceofgovernmentinterventiononbankperformancedepends ontheevaluativefinancialindicator,theeconomy,andwhetherbanksareinternationalized.SouthKorea andHongKonghavebeenidentifiedtobetheeconomieswithstrongerbankperformanceafter govern-mentintervention.PoliciesdemonstratedusefulinSouthKoreaandHongKonghavebeensummarized anddiscussed.
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
In1929,theU.S.financialsystemcollapsed,thestockmarket crashedandinlessthantwoweeks,morethan300billionUSD worthofwealthdisappeared.Thiseconomiccrisisquicklyspread toothereconomiesalso,resultingintheGreatDepression.Nearly 80yearslater,anotherfinancialdisasterthatbeganin2007because oftheU.S.sub-primemortgagecrisisresultedinasnowballeffect thathascontinuedtohurtfinancialmarketssincethen.In2008, MerrillLynch wasacquiredby BankofAmerica, AIG(American InternationalGroup)reportedhuge losses,theU.S.government tookoverthemanagementofFannieMaeandFreddieMac,and LehmanBrothersHoldingsInc.andWashingtonMutual,thelargest
夽 TheauthorsaregratefultoMr.PheeBoonKang,Mr.LawrenceTong,Professor Cheng-YingWu,ProfessorIftekharHasan,theManagingEditor,andananonymous refereefortheirconstructivecommentsandsuggestions.Thisworkwaspartially supportedbygrantNSC100-2410-H-009-008-MY2fromtheNationalScience Coun-cilofTaiwan,R.O.C.
∗ Correspondingauthor.Tel.:+886223494932;fax:+886223494922. E-mailaddresses:cding@mail.nctu.edu.tw(C.G.Ding),
jessiech.wu@yahoo.com.tw(C.-H.Wu),plchang@fcu.edu.tw(P.-L.Chang).
savings andloaninstitution inAmerica,announced bankruptcy, makingtheFederalReserveBoardprovideemergencyrelief.
Theglobalfinancialcrisisin2008wasachainreactionofcredit riskinherentinengineeredfinancialinstruments.Itwasthe com-binationofthreefinancialproducts,ResidentialMortgageBacked Securities (RMBSs), CollateralizedDebt Obligations (CDOs), and CreditDefaultSwaps(CDSs),inadditiontosomeothermajor prod-ucts,thatcausedthesub-primemortgagecrisiswhichlaterspread acrossmarkets.Sincemanyfinancialproductsareinternationaland liquid,theAmericanfinancialcrisishadacontagioneffectonall economiesaroundtheworld.Thecrisispushedtheworldeconomy intoasevererecession.Italsoledtoshrinkageofthejobmarket, lackoffundsincreditmarkets,soaringofcommercialinterestrates, anddownturnintherealestatemarket.Americanfinancial insti-tutionsreportedsignificantlossesinthefourthquarterof2008 whileEuropeencounteredshortfallsinbothdomesticandexternal demands.
TheglobalfinancialcrisisalsoimpactedJapanandthe“Four AsianDragons”(SouthKorea,HongKong,SingaporeandTaiwan), whichhadbeenrapidlygrowingexport-orientedeconomiessince 1970,untiltheywerehitbytheAsianfinancialcrisisof1997.They wereinevitablyaffectedbythisglobalfinancialcrisisaswell.Asian banks’earningssufferedfromthecrisisandtheirexistingfinancial 1572-3089/$–seefrontmatter © 2012 Elsevier B.V. All rights reserved.
problemswereworsened.Theywerefacedwithmanyunexpected challenges.
Atthemostcriticalmomentduringthefinancialcrisis,nineU.S. financialinstitutionsthatwereonthevergeofcollapsingreceived government’sfinancialaid.Theyhadacombinedtotallossofclose to100billionUSDin2008,forcingtheU.S.governmenttoinject 175billionUSDinto theTroubledAsset Relief Program(TARP), makingthis a pioneering actof governmentintervention in an efforttostaveoffanevenmoremassivefinancialcrisis.Most Euro-pean bailout programs combined government guaranteed debt issuanceprogramswithdirectequityinjections.National govern-mentsaroundtheworldweretryingtosavestockmarketsfrom collapsing,inanefforttoresolvetheworseningfinancialcrisis. Issuesregardingregulationinresponsetothefinancialcrisishave beendiscussed(e.g.,Goodhart,2008;Krimminger,2008;Praetand Nguyen,2008).
Nowthatwehavefacedaglobalfinancialcrisisnotseenina cen-tury,weneedtoassesswhethertheAsiangovernmentsintervened properly. Sincethe bankindustry, withhighfinancialleverage, iscloselyconnectedwiththefinancialsystem,andbank perfor-mancereflectsdirectlyfinancialstability,thebankindustryshould receiveparticularattention.Hasbankperformanceimprovedafter interventionduringtheglobalfinancialcrisis?Thefocusofthis studyistoexaminewhethergovernmentinterventioncanexert influenceonthetrajectoryofbankperformance,andtocompare theeffectsofinterventionamongdifferenteconomies.Specifically, therearethreeresearchpurposesofthisstudy.First,weanalyzethe changesinbankperformanceovertimeforfiveAsianeconomies (Japan,SouthKorea,HongKong,SingaporeandTaiwan)beforeand aftergovernmentinterventionduringthefinancialcrisis.Thesefive economies arechosenbecauseoftheirsimilaritiesinopenness, geographyandthepreponderanceoftheirfinancialmarkets. Sec-ond,wecomparethedifferencesamongtheseeconomies.Third,we summarizeremedialdevicesthathavebeendemonstrateduseful anddiscusssomepolicyimplications.
Thispaperis,tothebestofourknowledge,thefirststudythat appliesthepiecewiselatenttrajectorymodeltoanalyzethechange inbankperformanceovertime,beforeandaftergovernment inter-vention.Themodelisusefulfordescribingthechangeoftrajectory thatoccursafteraparticularimportantpointintimeand conduct-ingcomparisons(Flora,2008).Section2reviewsrelatedliterature. Section 3 presents the method.Section 4 shows the empirical resultsanddiscussessomepolicyimplications,andSection5gives theconclusion.
2. Relatedliteratureandhypotheses 2.1. Financialperformanceindicatorsforbanks
Mishkin(1999)indicatesthatabankisafinancial intermedi-aryand makesa profitthroughtransactionalprocesses,suchas thesellingofdebt(acceptscustomers’deposits)andpurchasingof properties(extendingloanstocustomers).WhentheU.S.Financial SupervisoryAuthorityevaluatesthefinancialpositionofabank,it usestheCAMELstandard,whichisaninternationallyrecognized ratingsystemforfinancialinstitutions,andusestheindicatorsfor capitaladequacy,creditrisk,management,earningsandliquidity. Solvencyisakeyfactortofinancialmarketefficiencyand stabil-ity(Barrelletal.,2010;Raddatz,2006);intheeventofthemarket worsening,thecrisisisthenattributedtoalackofliquidityand insolvency.
Stojanovic et al. (2008) analyze whether banks underwrit-ing mortgages increase their risk, by using CAMEL financial and economic variables regarding capital adequacy, credit risk,
profitability,liquidityrisk,andsensitivity.Arena(2008)findthat bankfundamentalssignificantlyaffectthelikelihoodofbank fail-uresforthebanksinEastAsiaand LatinAmerica,basedonthe indicatorsof assetquality, solvency,liquidity, return onassets, andinterestratesanddeposits.Zopounidis(1998)usesthe princi-palcomponentfactoranalysistoconductdimensionreductionfor financialindicators.Theresultingfactorsincludedsolvency, invest-mentactivity,capitalstructure,andprofitability.Raveh(2000)uses theco-plotmethod,atwodimensionalgraphicdisplaytechnique, toevaluateabank’sperformancebyanalyzingliquidity,investment activity,capitalstructure,andprofitabilityofthebank.
2.2. Effectsofgovernmentintervention
WhenPresidentReaganassumedthePresidencyin1981,the U.S.economywascaughtinanextremelyunusualdilemma com-prisingsevereinflationandrisingunemployment.Itwasinthis environmentthattheeconomicpoliciesthatcametobeknown as“Reaganomics”wereborn.Atthetime,itwasconsideredtobe astrategythatactivelyfacedrealityandquicklyadoptedcounter measurestoaddressthedilemma.Ononehand,thegovernment adoptedafiscalpolicyofreducingtaxesandtaxrateswhile increas-inggovernmentspending(suchasasubstantialincreaseinmilitary spending);ontheotherhand,theFederalReserveadoptedapolicy ofreducinginterestratestoincreasemoneysupplyfor encourag-inginvestmentandstimulatingeconomicdevelopment.Asaresult, therewassignificantimprovementintermsofinflationandthe unemploymentrateandfinancialmarketswereonceagainactive, implyingthatgovernmentinterventioncanlifteconomyand finan-cialmarkets.
Researchonchanges inperformance ofthevarious typesof bankshasshowndifferencesbetweenpublicsectorand private banksintheIndiancommercialbankingsystem,indicatingthatthe averageefficiencywashigheramongpublicsectorbanksbecause oftheirhighercredibilityandcompetitivenesssupportedby gov-ernment(Bhattacharyyaetal.,1997).Ongenaetal.(2003)analyze theNorwegianbankingcrisisduringtheperiod1988–1991and findthatbothNorwayandJapanstartedfacingthefinancial cri-sisroughly atthesametime.However,becauseofintervention bythegovernment,Norwayovercameitsfinancialcrisisinfour yearswhileJapancontinuedtosuffer.Inthe1980s,Japanfaceda highlyriskymortgagemarket,endlessproblemswiththe finan-cialinstitutions,andrelentless appreciationof theYenbecause offoreignpressure. Aftera coupleof failedrescueattempts by theMinistryofFinance,Japansubsequentlyexperienceda finan-cialdisaster.DuringtheAsianfinancialcrisis(1997–1999),capital injectionsbytheJapanesegovernmentweremeanttosavethe banks.However,theresultwasineffectiveduetoinsufficient fund-ing.HoshiandKashyap(2010)havedescribedindetailthethree phasesoftheJapanesecrisis.Theyalsoexaminemajorresponses bytheJapanesegovernmenttothefinancialcrisisandshowthe successorfailureofvariousgovernmentinterventions.They con-cludethatgovernmentinterventionsareonlypartiallysuccessful inrecapitalizingthebanks.TheJapanesegovernmentintervention seemstobeunsatisfactory.ThedragtheJapaneseeconomy has experiencedmightbebecauseofimproperhandlingofbanks dur-ingtheentiredecadein the1990s.Drawingsomelessonsfrom this,theTaiwanesegovernmentstartedafinancialrevolutionof sortsin2001whenitaggressivelystartedeliminatingidle bank-ingaccountsandre-establishingactivefinancialmotivations.By jointeffortsofallconcernedparties,allidlebankingaccountswere closedandmorefundswerereleasedintothemarket,and non-performingloansratiowassignificantlyreduced.Thiswastaken asasuccessfuleffectofthe1stfinancialreform,butthehugecost waspaidbythewholesociety(Hsiaoetal.,2010).SouthKoreaalso
experiencedtheAsianfinancialcrisisandtheSouthKorean govern-mentproposedavarietyofpoliciesandfinancialreforms,which ledtotherevivalofitseconomyafterthecrisis.Thefinancialcrisis promptedthegovernmenttoinitiateanumberofreformsinthe areasoffinance,corporategovernance,andthelabormarket, giv-ingSouthKoreaabrandnewopportunitytocreateacompetitive economicenvironment(LeeandLee,2008).
Pasquariello(2007)conductsanempiricalinvestigationofthe impactofcentralbankinterventionontheprocessofprice for-mationinforeignexchangemarketsandexploitsinsightsfromthe analysisofmarketliquiditytoexplaintheeffectivenessofthe inter-vention.However,duetothelargeamountof foreignexchange trades,directinterventionbyanindividualcentralbankmay pro-ducelimitedeffectsunlesscentralbanksagreewitheachotherto interveneincurrencymarketsviaswaparrangements(Cukierman, 2011).
Goodhart(2008)summarizestheaspectsconcerningfinancial regulationtocopewithfinancialcrises.Theyinclude:(1)deposit insurance;(2)banksolvencyregimes(‘promptcorrectiveaction’); (3)moneymarketoperationsbycentralbanks;(4)liquidityrisk management; (5) procyclicality in capital adequacy ratios; (6) boundariesofregulation,conduits,SIVsandreputationalrisk;(7) crisismanagement.However,itshouldbecautioustousepolicies sincenotallpoliciesareeffective.Theeffectsofgovernment inter-ventionvarywitheconomiesbecausethefocusofrescueprograms proposedbydifferenteconomiesmightbequitedifferent.
Accordingtoananalysisofperformanceofdomesticand for-eignbanksbySabi(1996),themoreinternationalizedthebanks are,themoretheyareaffectedbyafinancialcrisis.Mostofthe local banksare less globalized.They have limited involvement inand interfacewiththeglobalmarkets.Also,theirbondsand securitiestradingcapacityissmaller,andhence,thelimited inter-nationalizationhasresultedinlessdirectimpactsfromtheglobal financialcrisis.Foreignbanksweremoreefficientthandomestic banks,butthishigherefficiencywasunabletoproducehigher pro-fits(SturmandWilliams,2004).Cukierman(2011)mentionsthat whenfacingacrisis,largerbanksnotonlydonotinternalize sys-temicrisks,butalsoexpecttobebailedoutbygovernment,whichis the“too-big-to-fail”problem.Sohn(2010)indicatesthatthe “too-big-to-fail”expectationmakesthemarketignorethepossibility thatfournationwidebanksinSouthKoreamightbeforcedtoclose. However,largerbanksarebetterabletodiversifytheirloan portfo-lios,reducingtheirassetrisk,andinternationalizedbankscanhave accesstoparentfunding(Arena,2008;CalomirisandMason,2001). Accordingtotheliteraturereviewgivenabove,weexpectthat H1. Onaverage,thetrajectoryofbankperformanceintermsof
solvency,creditrisk,andprofitabilityimprovesaftergovernment interventionduringtheglobalfinancialcrisis.
H2. Theeffectsofgovernmentinterventiononthetrajectoryof
bankperformanceduringtheglobalfinancialcrisisdependonthe economyandwhetherbanksareinternationalized.
3. Methods 3.1. Data
Sixfinancialindicatorsforassessingbankperformanceareused inthisstudy.Theyincludetotaldeposits/totalassets(abbreviated asDAinthisarticle),totaldeposits/totalfunding(abbreviatedas DF),totalallowanceforbaddebt/totalloans(abbreviatedasABDL), loanlossreserves/non-performingloans(abbreviatedasLRNPL), returnonassets(abbreviatedasROA)andgrowthrateofassets (abbreviatedasGRA).DAandDFaresolvency-relatedindicators.
Table1
Samplecharacteristics.
Internationalization Japan South Korea Hong Kong Singapore Taiwan With 5 3 7 3 10 Without 5 3 1 0 5
Thevaluesinthecellsarethenumbersofbanks.Thetotalnumberofbanksis42.
ABDLandLRNPLareindicatorsrelatedtocreditrisk.ROAandGRA
areprofitability-relatedindicators.
The financial data used for this study are obtained from
Bloomberg’s (acomprehensivefinancial databasefor individual
banksacrosstheworld)quarterlyreportsduringtheperiodfrom
the4thquarterof2007tothe2ndquarterof2010,inwhichthe
2008financialcrisisoccurred.Bloombergdoesnotreportfull
finan-cialinformation ofbanksinTaiwan.Asa result,balancesheets
and financialstatementsweregatheredseparatelyfromreports
producedbyfinancialregulatoryagenciesinTaiwan.Byfollowing
similarstudies(e.g.,HoshiandKashyap,2010;Rotheli,2010),we
selectedlargebanksonly.Thereasonisthattheysufferedmuch morefromtheglobalfinancialcrisisandbenefitedmoreby gov-ernmentintervention,andthereforetheyarerepresentative.The sampleconsistsof42bankswithcompletefinancialdata,selected fromJapan,SouthKorea,HongKong,SingaporeandTaiwanamong theworld’stop1000banks.10banksarefromJapan,6fromSouth Korea, 8banksarein HongKong, 3banksin Singaporeand 15 banksinTaiwan.Ofthese,28areinternationalizedbanksand14are non-internationalizedbanks.Samplecharacteristicsaredisplayed inTable1.
3.2. Thepiecewiselatenttrajectorymodel
Inordertomeettheresearchobjectivesandtobeinlinewith thenatureofthedata,thepiecewiselatenttrajectorymodel(e.g., BollenandCurran,2006,Section4.3;Flora,2008;Lietal.,2001) isusedinthisstudy.Themodelissuitableforsituationswherea fixedtransitionpointduringthetimeperiodhasbeendetermined anddifferentlineartrajectoriesneedtobefittedbeforeandafter thattransition.Inthisstudy,governmentinterventionisa partic-ulareventduringtheperiodoftheglobalfinancialcrisis.Sinceour purposeistoexamineifthechangepatternsofbankperformance beforeandafterintervention(atransitionpoint)aredifferent,in ordertoseetheeffectsofintervention,thepiecewiselatent tra-jectorymodelisappropriate.Themodelcannotonlycharacterize intra-bank(within-bank)changesinperformanceovertime but alsotestinter-bank(between-bank)differencesinchanges.
Thepiecewiselatenttrajectorymodelinthisstudyisgivenby xitk =˛ik+1tˇ1ik+2tˇ2ik+εitk, i=1,2,...,42;
t=1,2,...,11; k=1,2,...,6, (1) wherexitkisthevalueofthetrajectoryvariablexk(thekth
finan-cialperformanceindicator)forbankiattimet,˛ikistherandom
intercept and ˇ1ik and ˇ2ik are the random slopes before and
aftergovernmentinterventionofthetrajectoryforbanki associ-atedwiththekthfinancialindicator,1tand2tarefixedloading
coefficients,representingthepassageoftimebeforeandafter inter-vention,andεitkisthecorrespondingerror,reflectingthedeparture
fromthetrajectory.Sincethetransitionpoint(thetimepointof governmentintervention)isatt=4(the3rdquarterof2008),both 14 and 24 are setas zero. 1t and 2t are specified as [11,
12,13,...,1,11]=[−3−2−100000000]and[21,22,23,...,
2,11]=[00001234567].ThemodelinEq.(1)isreferredtoasthe
level-1modelandεitkthelevel-1errors.Itisassumedthatlevel-1 errorshavezeromeanandidenticalcovariancestructureεforall
Table2
Themeantrajectoryofbankperformanceforeachcombinationofthelevelsofeconomyandinternationalization.
Economy Withinternationalization Withoutinternationalization
Japan 00+01+(10+11)1t+(20+21)2t 00+01+05+(10+11+15)1t+(20+21+25)2t
SouthKorea 00+02+(10+12)1t+(20+22)2t 00+02+05+(10+12+15)1t+(20+22+25)2t
HongKong 00+03+(10+13)1t+(20+23)2t 00+03+05+(10+13+15)1t+(20+23+25)2t
Singapore 00+04+(10+14)1t+(20+24)2t 00+04+05+(10+14+15)1t+(20+24+25)2t
Taiwan 00+101t+202t 00+05+(10+15)1t+(20+25)2t
banks.AcommonlyusedstructureforεisAR(1)(thefirst-order
autoregressive)(e.g.,Littelletal.,2006,p.175),givenby
ε=
⎡
⎢
⎢
⎢
⎣
2 ε1 ε1ε2 ··· ε1ε11 ε1ε2 2 ε2 ··· ε2ε11 . . . . . . . .. ... ε1ε11 ε2ε11 ··· 2 ε11⎤
⎥
⎥
⎥
⎦
=2ε⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
1 2 ··· 10 1 ··· 9 2 1 . . . 8 . . . . . . . . . . .. . . . 10 9 8 ··· 1⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
, (2) where2εisthecommonvarianceandisthelag-1autocorrelation
coefficient.
Therandominterceptandrandomslopes(alsocalledgrowth factors)inthelevel-1modelcanbeexplainedbytimeinvariant pre-dictors.Themodelisreferredtoasthelevel-2model.Inthisstudy, twotimeinvariantpredictorsareused.Oneistheeconomy,andthe otherisinternationalization.Sincefiveeconomies(Japan,South Korea, HongKong,Singapore and Taiwan) aretobecompared, fourdummyvariables,DEcon1,DEcon2,DEcon3andDEcon4are createdforthem,definedasfollows:DEcon1=1forJapanand=0 forothereconomies;DEcon2=1forSouthKoreaand=0forother economies;DEcon3=1forHongKongand=0forothereconomies; andDEcon4=1forSingaporeand=0forothereconomies.Thereare twolevelsforinternationalization(with/without),andonedummy variableDIntiscreated.DInt=1forbankswithout international-ization;DInt=0forbankswithinternationalization.Therefore,the level-2modelisgivenby(subscriptkisomittedforsimplicitysince eachindicatoristreatedindividually):
˛i=00+01DEcon1i+02DEcon2i+03DEcon3i+04DEcon4i+05DInti+0i,
ˇ1i=10+11DEcon1i+12DEcon2i+13DEcon3i+14DEcon4i+15DInti+1i,
ˇ2i=20+21DEcon1i+22DEcon2i+23DEcon3i+24DEcon4i+25DInti+2i,
(3)
wheregamma’sarefixedgrowthparameterstobeestimatedand 0i,1iand2iarelevel-2errors,withzeromeansandthe
covari-ancematrixof =
⎡
⎢
⎣
2 0 01 02 01 2 1 12 02 12 2 2⎤
⎥
⎦
. (4)Combininglevel-1andlevel-2modelsleadsto(subscriptkis omit-ted):
xit =˛i+1tˇ1i+2tˇ2i+εit
=00+01DEcon1i+02DEcon2i+03DEcon3i+04DEcon4i+05DInti+101t+
11DEcon1i1t+12DEcon2i1t+13DEcon3i1t+14DEcon4i1t+15DInti1t+
202t+21DEcon1i2t+22DEcon2i2t+23DEcon3i2t+24DEcon4i2t+
25DInti2t+(0i+1i1t+2i2t+εit).
(5)
Thecombinedmodelgivenaboveinvolvescross-levelinteractions (11,...,15,21,...,25).Whentheirestimatesaresignificant,
furthercomparativeanalysisofthepiecewiselineartrajectories needstobeconducted.Tofacilitatetheinterpretationofgrowth
parametersinEq.(5)andsubsequentcomparison,wesummarize themeantrajectoryfor eachcombination oflevelsof economy andinternationalizationinTable2.Accordingtothetable,10and
20represent,respectively,themeanslopesoftrajectorybefore
andafterinterventionforinternationalizedbanksinTaiwan.11,
12, 13,and 14 represent,respectively,thedifferences ofthe
meanslopesoftrajectorybeforeinterventionbetweenJapan,South Korea, Hong Kong, Singapore, and Taiwan,regardless of inter-nationalized or non-internationalizedbanks. 21,22, 23, and
24represent,respectively,thoseafterintervention.15denotes
the difference of the mean slopes of trajectory between non-internationalizedandinternationalizedbanksbeforeintervention, regardlessoftheeconomy.25denotesthatafterintervention.
4. Resultsanddiscussion 4.1. Testresults
Mainresultsoftheanalysisbythepiecewiselatenttrajectory modelfor thesix financialperformance indicatorsarereported in Table 3. Analysis is conducted with SAS PROC MIXED (SAS InstituteInc.,2010).Itappearsthatsomeoftheestimatesof cross-levelinteractionsaresignificant.Inotherwords,theinfluenceof governmentinterventiononthetrajectoryofbankperformance dependsontheevaluativefinancialindicator, theeconomy,and whetherbanksareinternationalized.Furthercomparativeanalysis
ofthepiecewiselineartrajectoriesfordifferentgroupsisneeded toobtainedmoreinsightfulresults.
To see how the mean slopes of trajectory before and after intervention differ between internationalized and non-internationalizedbanks,refertotheresultsassociatedwithDInt*T1 andDInt*T2inTable3.Non-Internationalizedbanksperformworse thaninternationalized banksinABDLbeforegovernment inter-ventionbutbetterafterintervention(thedifferencesofthemean slopes are 6.924 and −0.4908, respectively, both significant at the0.001level).Non-internationalized banksperformbetterin GRAafterintervention(thedifferenceofthemeanslopes=1.1912, p<0.05).Nosignificantdifferenceinthemeanslopesoftrajectory
ofsolvencywasfound,regardlessofbeforeorafterintervention. Since internationalized banks involve financial markets world-wide,theyreceivemoreimpactsfromtheglobalfinancialcrisis. However,theyhave moresourcesof fundingsupport,andmay
Table3
Mainresultsoftheanalysisbythepiecewiselatenttrajectorymodelforthesixfinancialindicators.
Solvency Creditrisk Profitability
DA DF ABDL LRNPL ROA GRA
Intercept 77.1407*** 82.5699*** 0.1046 73.6129*** 0.2881** 6.6524* DEcon1 −12.0087* −7.5608 −1.4289* 10.6806 0.1308 1.0985 DEcon2 −23.5542*** −19.0132*** −1.6280* 88.4864*** 0.6004*** 12.3588*** DEcon3 −8.9866 2.3150 0.1484 −14.2073 0.1240 3.1258 DEcon4 −18.0160* −1.2345 0.1070 19.1078 0.7122*** −0.3226 DInt 2.6969 3.1427 2.7846*** −2.2430 −0.2017* −5.8718* T1 0.1532 0.0217 −2.3063* −2.0386 −0.0506 0.4532 DEcon1*T1 −0.6560 −0.7645 −1.2392 −2.1322 0.0122 1.0634 DEcon2*T1 −1.1131* −1.1297 −2.1251 1.0314 −0.0281 0.7190 DEcon3*T1 −0.0990 0.3571 1.4892 −0.0105 −0.1480** −3.4275** DEcon4*T1 0.2243 0.5134 2.3488 −2.4086 −0.0089 −2.7505 DInt*T1 0.0378 0.2507 6.9240*** 2.7077 0.0022 −1.3784 T2 0.6482** −0.8170** 0.1619* 5.6225*** −0.0059 −0.5480 DEcon1*T2 −0.3718 1.0067** 0.1002 −9.6325*** −0.1186*** −0.7483 DEcon2*T2 0.0215 1.7000*** 0.1596 −11.4448*** −0.0089 −1.5645* DEcon3*T2 −0.0743 1.1385** −0.1566 −2.7545 0.0528 0.4074 DEcon4*T2 −0.1419 1.1462* −0.1750 −6.1213 −0.0057 0.6689 DInt*T2 0.0785 0.1330 −0.4908*** 1.0050 0.0095 1.1912*
DA=totaldeposits/totalassets,DF=totaldeposits/totalfunding,ABDL=totalallowanceforbaddebt/totalloan,LRNPL=loanlossreserves/non-performingloans,ROA=return onassets,andGRA=growthrateofassets.DummyvariableDInt=0forbankswithinternationalizationand1forbankswithoutinternationalization;Dummyvariable DEcon1=1forJapanand0otherwise;DEcon2=1forSouthKoreaand0otherwise;DEcon3=1forHongKongand0otherwise;DEcon4=1forSingaporeand0otherwise.T1 andT2represent,respectively,1tand2t.ThepiecewiselatenttrajectorymodelisgiveninEq.(5).Thevaluesinthetableareestimatesofthegrowthparameters.
*p<0.05. **p<0.01. ***p<0.001.
be“too-big-to-fail”. Executives of largeinternationalized finan-cialinstitutionsexpectgovernmentswillrescuetheirinstitutions whenacrisisoccursbecausegovernmentsneedtoinstill confi-dencebypromisingtoinjectadditionalcapitalwhennecessaryin ordertoprotectthestabilityofthefinancialsystem.Ourresults indicatethatalthoughnon-internationalizedbanksperformworse increditriskbeforegovernmentintervention,theirrecoveryspeed isfasterafterintervention.Itmaytakesomemoretimefor inter-nationalizedbankstoimprovetheircreditriskandrecovertheir profitability.
Table4displaysthemeanslopesoftrajectoryofperformance intermsofthesixfinancialindicatorsacrossthefiveeconomies and internationalized and non-internationalized banks before
Table4
Comparisonbetweenthemeanslopesoftrajectoryofperformancebeforeandafter governmentinterventionacrossthefiveeconomiesandinternationalizedand non-internationalizedbanks.
Financialperformanceindicator Governmentintervention
Before After After−before
Solvency DA −0.1567 0.5742*** 0.7309* DF −0.0576 0.2478 0.3054 Creditrisk ABDL 1.2504 −0.0979* −1.3483 LRNPL −1.3887 0.1344 1.5231 Profitability ROA −0.0841*** −0.0172 0.0669* GRA −1.1151* −0.1998 0.9153
DA=total deposits/total assets, DF=total deposits/total funding, ABDL=total allowance for bad debt/total loan, LRNPL=loan loss reserves/non-performing loans, ROA=return on assets, and GRA=growth rate of assets. The cor-responding statistical null hypotheses for examining the mean slopes of trajectoryof performancebefore andafter intervention andtheir differences are given, respectively, by H0(before): 10+(11+12+13+14)/5+15/2=0,
H0(after): 20+(21+22+23+24)/5+25/2=0, and H0(after−before):
[20+(21+22+23+24)/5+25/2]−[10+(11+12+13+14)/5+15/2]=0. *p<0.05.
***p<0.001.
andaftergovernmentinterventionaswellastheirdifferences.By Table2,thestatisticalnullhypotheses forexaminingthemean slopesoftrajectoryofperformanceintermsofafinancialindicator before and after intervention and their difference are given, respectively,byH0(before):10+(11+12+13+14)/5+15/2=0,
H0(after):20+(21+22+23+24)/5+25/2=0, and
H0(after−before):[20+(21+22+23+24)/5+25/2]−[10+(11
+12+13+14)/5+15/2]=0.Onaverage,theslopesoftrajectory
ofbankperformanceintermsofDA,ABDL,ROA,andGRAshow improvement after intervention. Although the tests associated with (after−before) for ABDL and GRA are both insignificant, themeanslopeassociatedwithABDLchangesfrominsignificant before intervention to significantly negative (−0.0979, p<0.05) after intervention and that associated with GRA changes from significantly negative (−1.1151, p<0.05) beforeintervention to insignificantafterintervention.Theformerindicatesasignificant tendencytoimproveincreditrisk,andthelattershowsa signif-icant alleviation of the decliningprofitability after government intervention.Therefore,H1issupported.
We summarize the mean slopes of trajectory of bank per-formanceintermsofthesixindicatorsbefore/aftergovernment interventionfordifferentcombinationsofeconomies(Japan,South Korea,HongKong,SingaporeandTaiwan)andinternationalization (with and without, denoted by Int0 and Int1, respectively) in Table 5. In addition, the mean slopes before/after government interventionarecomparedandtestedforthesignificanceofthe differences.InSouthKorea,themeanslopesoftrajectoryof perfor-manceintermsofDAandDFaresignificantlygreater(reflecting the improvement of solvency) after intervention than before for internationalized banks (with the estimates of 1.6297 and 1.991) and for non-internationalizedbanks(with theestimates of1.6704and1.8733).InTaiwan,themeanslopeoftrajectoryof DAisinsignificantbeforeinterventionbutbecomessignificantly positive after intervention;however, that of DF is insignificant beforeinterventionbutbecomessignificantlynegativeafter inter-vention,regardlessofinternationalizedornon-internationalized banks. No significantdifference wasfound in othereconomies for internationalized or non-internationalized banks.It appears
Table5
Themeanslopesoftrajectoryofperformancebefore/aftergovernmentinterventionbyeconomyandinternationalization.
Indicator Japan SouthKorea HongKong Singapore Taiwan
Int1 Int0 Int1 Int0 Int1 Int0 Int1 Int0 Int1 Int0
Solvency DA Before −0.4651 −0.5028 −0.9222* −0.9599* 0.0919 0.0542 0.4153 0.3775 0.1910 0.1532 After 0.3550 0.2765 0.7483* 0.6698 0.6524 0.5739 0.5848 0.5063 0.7267* 0.6482** After−before 0.8200 0.7793 1.6704* 1.6297* 0.5604 0.5197 0.1695 0.1288 0.5358 0.4950 DF Before −0.4920 0.5250 −0.8572 −1.1080 0.6296 0.3788 0.7859 0.5351 0.2725 0.0217 After 0.3228 0.3364 0.0160* 0.8830* 0.4545 0.3215 0.4623 0.3292 −0.6839* −0.817** After−before 0.8148 0.9325 1.8733* 1.9910* −0.1750 −0.0573 −0.3236 −0.2059 −0.9564 −0.8387 Creditrisk ABDL Before 3.3785** −3.5455** 2.4926 −4.4314** 6.1069*** −0.8171 6.9665*** 0.0425 4.6177*** −2.3067* After −0.2288** 0.2621** −0.1693 0.3215** −0.4855*** 0.0053 −0.5039*** −0.0131 −0.3289*** 0.1619* After−before −3.6072** 3.8076** −2.6620 4.7529** −6.5925*** 0.8224 −7.4704*** −0.0556 −4.9466*** 2.4682* LRNPL Before −1.4631 −4.1708 1.7006 −1.0071 0.6586 −2.0491 −1.7395 −4.4472 0.6692 −2.0386 After −3.0050 −4.0101* −4.8173* −5.8223* 3.8730 2.8680 0.5062 −0.4989 6.6275*** 5.6225*** After−before −1.5420 0.1607 −6.5179 −4.8152 3.2144 4.9171 2.2456 3.9483 5.9583 7.6610* Profitability ROA Before −0.0362 −0.0384 −0.0765 −0.0787 −0.1964*** −0.1986*** −0.0573 −0.0595 −0.0484 −0.0506 After −0.1150*** −0.1245*** −0.0053 −0.0148 0.0565 0.0469 −0.0021 −0.0117 0.0036 −0.0059 After−before −0.0788 −0.0861 0.0712 0.0638 0.2529** 0.2455*** 0.0552 0.0478 0.0521 0.0447 GRA Before 0.1382 1.5166 −0.2062* 1.1722 −4.3526*** −2.9742** −3.6757* −2.2973 −0.9252 0.4532 After −0.1052 −1.2964* −0.9214 −2.1126** 1.0505 −0.1407 1.3120 0.1208 0.6431 −0.5480 After−before −0.2434 −2.8130* −0.7153 −3.2848* 5.4031** 2.8336* 4.9877* 2.4181 1.5683 −1.0013 DA=totaldeposits/totalassets,DF=totaldeposits/totalfunding,ABDL=totalallowanceforbaddebt/totalloan,LRNPL=loanlossreserves/non-performingloans,ROA=return onassets,andGRA=growthrateofassets.BankswithandwithoutinternationalizationarerepresentedbyInt0andInt1,respectively.
*p<0.05. **p<0.01. ***p<0.001.
that, on average, South Korean banks benefit by government interventionmorethanothereconomiesinsolvency.
ThemeanslopesoftrajectoryofABDLafterinterventionbecome significantlylessthanbefore(reflectingsignificantimprovementin creditrisk)fornon-internationalizedbanksinalleconomiesexcept South Korea,butbecomesignificantly greater(reflecting signif-icant deterioration) for internationalized banksin Japan,South Korea,andTaiwan.Onlynon-internationalizedbanksbenefitsby governmentinterventioninABDL.Moreover,LRNPLshows signifi-cantimprovementforinternationalizedandnon-internationalized banksinSouthKoreanbutsignificantdeteriorationfor internation-alizedandnon-internationalizedbanksinTaiwan.
ThemeanslopesoftrajectoryofROAandGRAaresignificantly greater(reflectingtheimprovementofprofitability)after interven-tionthanbeforeforinternationalized andnon-internationalized banks in Hong Kong. The mean slope of trajectory of GRA is significantly greater after intervention than before for non-internationalizedbanksinSingapore.Theperformanceintermsof ROAofinternationalizedandnon-internationalizedbanksinJapan becomesworseafterintervention.
Theresultsindicatethatgovernmentinterventionmay signif-icantlyimprovebankperformance,dependingontheevaluative financialindicator,theeconomy,andwhetherbanksare interna-tionalized.Therefore,H2issupported.
4.2. Comparisonsamongeconomies
Pairwisecomparisonsof themeanslopesoflineartrajectory ofperformanceintermsofthesixfinancialindicatorsreflecting
solvency,creditriskandprofitabilityamongeconomies,beforeand afterintervention,basedontheresultsshowninTable5,are sum-marizedinTable6.Thosebeforeandaftergovernmentintervention are given in the upper and lower triangular matrices, respec-tively.Infact,theresultscanalsobeobtainedbasedonthemean trajectories shown in Table 2. For example, the differences of themean slopes of trajectory between Japanand South Korea beforeandafterinterventionaregivenbyˆ11−ˆ12andˆ21−ˆ22,
respectively.ThosebetweenJapanandHongKongaregivenby ˆ
11−ˆ13 andˆ21−ˆ23.For DA,thesignificantdifferencesexist
only between South Korea and other economies before inter-vention. For DF, thesignificant differences exist only between Taiwanand othersafterintervention.ThebanksinTaiwan per-form,on average, theworst in DF. For ABDLand LRNPL, none ofthedifferencesamongthefiveeconomiesissignificantbefore intervention.ThebanksinJapanand SouthKoreaperformbest in LRNPL after intervention. The banks in Hong Kong perform betterthanthoseinJapanandSouthKoreainABDL.Itis inter-esting that the pairwise differences among economies before and afterintervention mostly showopposite directions inROA and GRA, though some of them are insignificant. The results reflecttheeffectsof governmentinterventiononthedifference ofprofitabilityamongeconomies.Japanperformstheworstafter interventionin ROA.Governmentintervention isineffective for Japan,consistentwithHoshi and Kashyap(2010), whoanalyze various intervention and bailout programs conducted by the Japanesegovernmentatvariousstagesofthefinancialcrisisand conclude that intervention by Japanese government lacks effi-ciency.
Table6
Pairwisecomparisonsofthemeanslopesoftrajectoryamongeconomiesbeforeandaftergovernmentintervention.
Japan SouthKorea HongKong Singapore Taiwan
(a)Solvency–DA Japan – 0.4571 −0.5570 −0.8804 −0.6560 SouthKorea −0.3933 – −1.0141* −1.3375* −1.1131** HongKong −0.2974 0.0959 – −0.3233 −0.0990 Singapore −0.2299 0.1634 0.0675 – 0.2243 Taiwan −0.3718 0.0215 −0.0744 −0.1419 – (b)Solvency–DF Japan – 0.3652 −1.1215 −1.2779 −0.7645 SouthKorea −0.6932 – −1.4868 −1.6431 −1.1297 HongKong −0.1317 0.5615 – −0.1563 0.3571 Singapore −0.1395 0.5538 −0.0077 – 0.5134 Taiwan 1.0067** 1.7000*** 1.1385** 1.1462* –
(c)Creditrisk–ABDL
Japan – 0.8858 −2.7285 −3.5880 −1.2392 SouthKorea −0.0594 – −3.6143 −4.4738 −2.1251 HongKong 0.2568* 0.3162** – −0.8596 1.4892 Singapore 0.2752 0.3346 0.0184 – 2.3488 Taiwan 0.1002 0.1596 −0.1566 −0.1750 – (d)CreditRisk–LRNPL Japan – −3.1636 −2.1217 0.2764 −2.1322 SouthKorea 1.8122 – 1.0419 3.4400 1.0314 HongKong −6.8781* −8.6903** – 2.3981 −0.0105 Singapore −3.5112 −5.3234 3.3668 – −2.4086 Taiwan −9.6325*** −11.4448*** −2.7545 −6.1213 –
(e)Profitability–ROA
Japan – 0.0403 0.1602** 0.0211 0.0122 SouthKorea −0.1097* – 0.1199* −0.0192 −0.0281 HongKong −0.1714*** −0.0617 – −0.1391* −0.1480** Singapore −0.1129* −0.0031 0.0586 – −0.0089 Taiwan −0.1186*** −0.0089 0.0528 −0.0057 – (f)Profitability–GRA Japan – 0.3444 4.4908*** 3.8139** 1.0634 SouthKorea 0.8162 – 4.1465** 3.4695 0.7190 HongKong −1.1557 −1.9719* – −0.6769 −3.4275** Singapore −1.4172 −2.2334 −0.2615 – −2.7505 Taiwan −0.7483 −1.5645* 0.4074 0.6689 –
DA=totaldeposits/totalassets,DF=totaldeposits/totalfunding,ABDL=totalallowanceforbaddebt/totalloan,LRNPL=loanlossreserves/non-performingloans,ROA=return onassets,andGRA=growthrateofassets.Thepairwisedifferencesbeforeandaftergovernmentinterventionaregivenintheupperandlowertriangularmatrices,respectively. Forexample,thedifferencesofthemeanslopesoftrajectoryofDAbetweenJapanandSouthKoreabeforeandafterinterventionare0.4571and−0.3933,respectively.Those betweenJapanandHongKongbeforeandafterinterventionare−0.5570and−0.2974,respectively.
*p<0.05. **p<0.01. ***p<0.001.
4.3. Policyimplications
Theeconomieswithbetterbankperformanceintermsofeach financialindicatorareidentifiedbasedonthetwoconditions.One isthatthemeanslope oftrajectoryofbankperformance inthe economysignificantlyimprovesafterinterventionforboth inter-nationalizedandnon-internationalizedbanks,andtheotheristhat themeanslopeoftrajectoryafterinterventionfortheeconomyis notworsethanthatforanyothereconomy.Twoconditionsneed tobebothmet.Theeconomiessatisfyingthefirstconditioncanbe foundbyTable5,andthosesatisfyingthesecondconditioncanbe obtainedbyTable6.
Accordingto Tables 5 and 6,the banksin South Korea and TaiwanperformedbetterinDAafterinterventionthanbeforeand noneof themean slopesof trajectory of DA afterintervention amongthefiveeconomiesshowedsignificantdifference.Therefore, SouthKoreaandTaiwanarethosewithstrongerbankperformance inDAafterintervention.ForDF,SouthKoreaistheonlyeconomy satisfyingthetwoconditions.ByconsideringDAandDFtogether, SouthKoreaistheeconomywiththebestbankperformanceafter intervention.SouthKoreaisalsotheeconomywiththebestbank performanceintermsofLRNPL,butnoeconomycanbeidentified
forABDL.HongKonghasbeenshowninasimilarwaytobethe onlyeconomywhosebanksperformedbetterinROAandGRAafter governmentintervention.TheresultsaresummarizedinTable7.
Theexistenceofvariationinthechangepatternofbank per-formanceamongeconomiesmaybeattributabletodifferencesin governmentinterventionpoliciesduringtheglobalfinancial cri-sis.Inresponsetotheimpact,Asianeconomieshaveimplemented aseriesofpoliciestostabilizefinancialmarketsandtorevitalize theireconomies.IthasbeenshowninTable7thatthebanksin SouthKoreaperformedthebestinsolvencyandcreditrisk and thoseinHongKongperformedthebestinprofitabilityafter gov-ernmentintervention.Itisdesirabletofindeffectivemeansused bySouthKoreaandHongKong.Throughliteraturereview,the poli-ciesadoptedbythetwoeconomieshavealsobeensummarizedin Table7.Theyarebrieflydiscussedasfollows.
ThepoliciesusedtoimprovesolvencyinSouthKoreainclude guaranteeondeposits(CentralBankoftheRepublicofChina,2009; KimandRhee,2009;Lee,2009),interestratecut(KimandKim, 2012a),anddepreciationofdomesticcurrency(Kawai,2010;Kim and Kim,2012a;Lee,2009).Blanketguarantee ondeposits and theincreaseofinsuranceamountondepositswereusedinSouth Korea.Thetargetratedecreasedsharplyfrom5.25%inSeptember
Table7
Economieswithstrongerbankperformanceaftergovernmentinterventionandthepoliciesused.
Financialindicator Economywithstronger bankperformance
Policy
DA,DF(solvency) SouthKorea 1.Guaranteeondeposits(CentralBankoftheRepublicofChina,2009;KimandRhee,2009;Lee,2009) 2.Interestratecut(KimandKim,2012a)
3.Depreciationofdomesticcurrency(Kawai,2010;KimandKim,2012a;Lee,2009) LRNPL(creditrisk) SouthKorea 1.Bankrecapitalization(DiamondandRajan,2011;KimandRhee,2009;ParkandLee,2009)
2.Creditguarantee(CentralBankoftheRepublicofChina,2009;KimandRhee,2009;Lee,2009) ROA,GRA(profitability) HongKong 1.Investmentencouragement(Chenetal.,2011;ZhangandTong,2009)
2.Consumerandbusinessfinancing(ZhangandTong,2009) 3.Expansionofpublicexpenditures(ZhangandTong,2009)
DA=totaldeposits/totalassets,DF=totaldeposits/totalfunding,ABDL=totalallowanceforbaddebt/totalloan,LRNPL=loanlossreserves/non-performingloans,ROA=return onassets,andGRA=growthrateofassets.
2008to3%bytheendoftheyear,andfurtherdecreasedto2% inFebruary2009.Theamountofliquidityincreasedrapidly.The loanstocommercialbanksbytheBankofKoreajumpedfrom6000 to8000trillionKRWbeforeNovember2008,to21121trillionKRW byDecember2008,andfurtherto33055trillionKRWbyMarch 2009.TheKoreanwondepreciatedabout30%innominaleffective termsbetweenSeptember2008andMarch2009(−7%forJapan, 8.8%forSingapore,−0.1%forHongKong,and9.5%forTaiwan)with themaximumdepreciationrateof67.8%occurringonMarch2, 2009,whereasforeignreserveaccumulationincreasedfrom200.6 billionUSDbytheendof2008to270billionUSDbytheendof 2009.Dollardepreciationhasboostedbanks’competitivenessin solvency.
ThepoliciesusedtoimprovecreditriskinSouthKoreainclude bankrecapitalization(DiamondandRajan,2011;KimandRhee, 2009;ParkandLee,2009)andcreditguarantee(CentralBankofthe RepublicofChina,2009;KimandRhee,2009;Lee,2009).Disposalof non-performingassetsandrecapitalizationweretiedtogether.The SouthKoreangovernmentestablishedtheBankRecapitalization Fund,andusedittoaccumulatecapitaltoprepareforrecessionand apossibleincreaseinbanklosses.Liquiditywasboostedby inject-ing20trillionwonintheeconomy,mainlybywayofBankofKorea’s loans and institutional and privateinvestment purchases guar-anteedbytheKoreaCreditGuaranteeFund.SouthKoreanbanks increasedBIScapitaladequacyratioover12%andcorecapitaltier 1ratioover9%.
ThepoliciesusedtoimproveprofitabilityofbanksinHongKong includeinvestmentencouragement(Chenetal.,2011;Zhangand Tong,2009)andconsumerandbusinessfinancing(ZhangandTong, 2009).Freetradeagreementsweresignedandtaxbreakswereused toattractforeigninvestmentandtofacilitatebuildingaregional financialcenter.Infrastructureprojectsboostedinvestmentinthe shortrun.HongKonggovernment’scommitmentwasincreasedto 100billionHKDwiththeprovisionofupto70%guaranteeto partici-patinglendinginstitutionsgrantingloanstoHongKongcompanies. Governmentscanplayanimportantroleinstabilizing finan-cialsystems.Policymakersmaylearntheremedialdevicesthat havebeendemonstratedusefulinSouthKorea andHongKong. Well-designedpolicies,withappropriatetiming,scopeandmeans shouldbeimplementedinearlystagesofcrisestoachievebetter effects.TheMonetaryStabilizationBond(MSB)isauniquepolicy tooldevisedbytheKoreangovernmenttocontrolexcess liquid-itytomanageexport-driveneconomicgrowth.TheMSBspread performsasignalingroleinthetransmissionofexternalshocks todomesticfinancialmarkets(Kimand Kim,2012b).Moreover, closeinternationalcooperationamongtheauthoritiesof individ-ualcountriesisneededtofindbettersolutionstotheproblemswe arefacing(Hildebrand,2008;McDowell,2012).Specifically,such areasastransparency,valuationsissues,riskmanagement,market
infrastructure,ratingagencies,andliquiditymanagementcouldbe prioritiesforaction(Ackermann,2008).
5. Conclusion
Inthisarticle,howthetrajectoryofbankperformanceis influ-encedby government intervention isexamined and compared. FiveAsianeconomies(Japan,SouthKorea,HongKong,Singapore andTaiwan) arecovered,over theperiod 2007–2010,in which theglobalfinancialcrisisoccurred.Thepiecewiselatenttrajectory modelwasusedfortheanalysis.
Ourempiricalresultsindicatethat,duringtheglobalfinancial crisis,thelineartrajectoryofbankperformanceaftergovernment intervention(inthe3rdquarterof2008)mightimprove,depending ontheperformanceindicator,theeconomy,andwhetherbanksare internationalized.Itwouldseemreasonabletorelyonshort-term perspectiveofgovernmentinterventiontocopewiththefinancial crisis.Thecrisiscanbejust anotheropportunitytoimprovethe currentfinancialsystem.Wehavelearnedsomefinancialreforms toreducepotentiallossesinfacingpossiblefinancialcrisisinthe future.
Thestandard eventstudy methodology maybe usedin the futuretogainmoreinsightabouttheeffectsofgovernment inter-ventiononbankperformance.In additiontothebankindustry, otherindustriesmaybenefitfromgovernmentintervention dur-ingthefinancialcrisis.Thepiecewiselatenttrajectorymodelcan beused again toexamine the changepatterns of performance before/aftergovernmentinterventionforotherindustriesto bet-terassesstheeffectivenessandefficiencyofthepoliciesmade.We suggestthatthepiecewiselatenttrajectorymodelbeusedin situ-ationswherethecomparisonsofthechangepatternsbefore/after interventionareofinterest.
References
Ackermann,J.,2008.Thesubprimecrisisanditsconsequences.JournalofFinancial Stability4,329–337.
Arena,M.,2008.Bankfailuresandbankfundamentals:acomparativeanalysisof LatinAmericaandEastAsiaduringtheninetiesusingbank-leveldata.Journal ofBankingandFinance32,299–310.
Barrell,R.,Davis,E.P.,Karim,D.,Liadze,I.,2010.Bankregulation,propertypricesand earlywarningsystemsforbankingcrisesinOECDeconomies.JournalofBanking andFinance34,2255–2264.
Bhattacharyya,A.,Lovell,C.A.K.,Sahay,P.,1997.Theimpactofliberalizationonthe productionefficiencyofIndiacommercialbanks.EuropeanJournalof Opera-tionalResearch98,332–345.
Bollen,K.A.,Curran,P.J.,2006.LatentCurveModels:AStructuralEquation Perspec-tive.Wiley,Hoboken,NJ.
Calomiris,C.W.,Mason,J.R.,2001,May.CausesofU.S.BankDistressduringthe Depression.FederalReserveBankofChicago,pp.530–554.
CentralBankoftheRepublicofChina,2009,May.FinancialStabilityReport(Issue No.3).CentralBankoftheRepublicofChina.
Chen,H.Y.,Lee,C.F.,Tai,T.,Wang,K.,2011.Fiscalandmonetarypoliciesinreaction tothefinancialtsunamibytheTaiwanesegovernment.ReviewofPacificBasin FinancialMarketsandPolicies14,153–169.
Cukierman,A.,2011.Reflectionsonthecrisisandonitslessonsforregulatoryreform andforcentralbankpolicies.JournalofFinancialStability7,26–37.
Diamond,D.,Rajan,R.,2011.Fearoffiresales,illiquidityseeking,andcreditfreezes. TheQuarterlyJournalofEconomics126,557–591.
Flora,D.B.,2008.Specifyingpiecewiselatenttrajectorymodelsforlongitudinaldata. StructuralEquationModeling15,513–533.
Goodhart,C.A.E.,2008.Theregulatoryresponsetothefinancialcrisis.Journalof FinancialStability4,351–358.
Hildebrand,P.M.,2008.Thesub-primecrisis:acentralbanker’sperspective.Journal ofFinancialStability4,313–320.
Hoshi,T.,Kashyap,A.K.,2010.WilltheU.S.bankrecapitalizationsucceed?Eight lessonsfromJapan.JournalofFinancialEconomies97,398–417.
Hsiao,H.C.,Chang,H.,Cianci,A.M.,Huang,L.H.,2010.Firstfinancialrestructuring andoperatingefficiency:evidencefromTaiwanesecommercialbanks.Journal ofBankingandFinance34,1461–1471.
Kawai,M.,2010.Reformoftheinternationalfinancialarchitecture:anAsian per-spective.TheSingaporeEconomicReview55,207–242.
Kim,B.H.,Kim,S.,2012a.TransmissionoftheglobalfinancialcrisistoKorea.Journal ofPolicyModeling,http://dx.doi.org/10.1016/j.jpolmod.2012.01.005. Kim,I.J.,Rhee,Y.,2009.GlobalfinancialcrisisandtheKoreaneconomy.SeoulJournal
ofEconomics22,145–179.
Kim,J.,Kim,Y.C.,2012b.Financialcrisisandatransmissionmechanismofexternal shocks:thesignalingroleoftheKoreanMonetaryStabilizationBond.Journalof FinancialStability,http://dx.doi.org/10.1016/j.jfs.2012.06.002.
Krimminger,M.H.,2008.Theresolutionofcross-borderbanks:issuesfordeposit insurersandproposalsforcooperation.JournalofFinancialStability4,376–390. Lee,C.P.,2009.Notablefeaturesof2008financialcrisisanddirectionsofadjustment
bytheKoreaneconomy.SeoulJournalofEconomics22,181–215.
Lee,K.,Lee,C.H.,2008.Themiracletocrisisandthemirageofthepostcrisisreform inKorea:assessmentaftertenyears.JournalofAsianEconomies19,425–437. Li,F.,Duncan,T.E.,Duncan,S.C.,Hops,H.,2001.Piecewisegrowthmixturemodeling
ofadolescentalcoholusedata.StructuralEquationModeling8,175–204. Littell,R.C.,Milliken,G.A.,Stroup,W.W.,Wolfinger,R.D.,Schabenberger,O.,2006.
SASforMixedModels,Seconded.SASInstitute,Cary,NC.
McDowell,D.,2012.TheUSas‘SovereignInternationalLast-ResortLender’:theFed’s currencyswapprogrammeduringthegreatpanicof2007–09.NewPolitical Economy17,157–178.
Mishkin,F.S.,1999.LessonsfromtheAsiancrisis.JournalofInternationalMoney andFinance18,709–723.
Ongena,S.,Smith,D.C.,Michalsen,D.,2003.Firmsandtheirdistressed banks: lessonsfromtheNorwegianbankingcrisis.JournalofFinancialEconomies67, 81–112.
Park,J.H.,Lee,S.,2009.Globalfinancialcrisis:background,prospects,anditsimpacts onKorea.SeoulJournalofEconomics22,77–94.
Pasquariello,P.,2007.Informativetradingorjustnoise?Ananalysisofcentralbank interventions.JournalofFinancialMarkets10,107–143.
Praet,P.,Nguyen,G.,2008.Overviewofrecentpolicyinitiativesinresponsetothe crisis.JournalofFinancialStability4,368–375.
Raddatz,C.,2006.Liquidityneedsandvulnerabilitytofinancialunderdevelopment. JournalofFinancialEconomies80,677–722.
Raveh,A.,2000.TheGreekbankingsystem:reanalysisofperformance.European JournalofOperationalResearch120,525–534.
Rotheli,T.F.,2010.Causesofthefinancialcrisis:riskmisperception,policymistakes, andbanks’boundedrationality.JournalofSocio-Economics39,119–126. Sabi,M.,1996.Comparativeanalysisofforeignanddomesticbankoperationsin
Hungary.JournalofComparativeEconomies22,179–188.
SASInstituteInc.,2010.SAS/STATUser’sGuide(SAS9.2).SASInstituteInc.,Cary,NC. Sohn,W.,2010.Marketresponsetobankrelationships:evidencefromSouthKorea
bankreform.JournalofBankingandFinance34,2042–2055.
Stojanovic,D.,Vaughan,M.D.,Yeager,T.J.,2008.Dofederalhomeloanbank mem-bershipandadvancesincreasebankrisk-taking?JournalofBankingandFinance 32,680–698.
Sturm,J.E.,Williams,B.,2004.Foreignbankentry,deregulationandbankefficiency: lessonsfromtheAustralianexperience.JournalofBankingandFinance28, 1775–1799.
Zhang,Y.,Tong,S.T.,2009.HongKong’sEconomyintheFinancialCrisis,EAI Back-groundBriefNo.448.NationalUniversityofSingapore,Singapore.
Zopounidis,C.,1998.OperationalToolsintheManagementofFinancialRisk.Kluwer AcademicPublishersGroups,NewYork.