Multilateral aspects of - the capital flow management measures - PowerPoint PPT Presentation

Multilateral aspects of - the capital flow management measures Pornpinun Chantapacdepong Monetary Policy Group Bank of Thailand The views expressed in this presentation are those of the authors and do not necessarily represent those of the Bank of Thailand.

Motivati ation on • There was little e conce ncern about t the spillo lover er effects ects of capital controls … It is widely agreed that capital controls have no significant effects on aggregate volume of capital inflows as it can only shift the composition and lengthen maturity of flows. What if the control successfully reduce the volume of certain types of flows into a country, but simply shift the challenges of large inflows into other countries.

What this paper does… • Exami mine cross market t linkages within EAEs: by measu surin ing g correla elation ions s of volatilit ility/shoc /shocks s in stock returns ns and in net foreig eign n equity ity flows. • Identif tifies s the mechanism sm of the spread ad of turmoil moil across s countr trie ies in the region. Volatil tility ity spillo lover er (pure e contagion tagion vs vs inter erde depe pende ndence ce) • Analyze ze whether r unilate tera ral l movements nts create tes s side effects cts to neighbori ring countr tries. s. 3

What this paper does not do… • Prove e whether her the coordinate dinate action on is superior erior to unilate ateral ral capital al flow measures • Asses ess effecti ctivenes eness of capita tal l flow measur ures es vis-à-vi vis its objecti ctive. e. 4

Organ anizati ization n of study • Challeng llenge from volatile tile capital tal flows • Eviden dence ce of volatil ility ity spillo lover er in EAEs • Multilate tilateral ral effects cts of the Capital tal flow manag agement ement measur ures es (CFMs) 5

‘Challenges’ to EAEs from m volati atile le capital al flows 6

Inflows are mainly for portfolio investment… Net private capital flows, 1990-2010 (% of GDP, 4 quarters MA) Source: Pradhan et. Al. (2011) Note: NIEs of Asia includes Hong Kong, Korea, Singapore, and Taiwan

…with the trend of larger inflows going forward Centr ntral l bank k balance sheets ts (% of GDP) Source: Credit Suisse (29 February 2012) 8

Unfor ortu tunate ately, , EAEs s have financ ncia ial markets s with h limited ted absor orpti ptive capac acit ity. Billions USD Stock market size Bond market size FX market size Thailand 218 180 7.4 Malaysia 90 183 7.3 Indonesia 130 98 3.4 Philippines 27 55 5.0 Singapore 282 115 265.9 Korea 1,627 1,085 43.8 Japan 4,280 11,521 312.3 USA 30,455 25,064 904.4 Source: WDI and BIS Note: Stock market, total value at end-2010; Domestic debt securities, amount outstanding at end-2009; FX, average daily turnover in April 2010 9

How connec ected ted are these se volati atile le capital al flows in EAEs? and by which h mechani anism? sm? 10

Data a and methodolo dology • The time varying cross s market t correlatio lation coefficie cients nts was calcul ulate ated d by DCC GARCH model l (Engle, 2002 2002) • Event t studie ies s for multilate ateral al impacts ts of the announ unce cemen ment t of capital al flows measur ures s (CFMs) in EAEs during ng 2009 2009-2010 2010. • Using stock market t data: 1) Equity y price (1992 1992-prese sent) nt) 2) Net foreign equity ty flows (2006 2006-pr present) nt) 11

Define ine pure e contagio agion n versus us interde depende pendenc nce • This study y follows definiti ition of Forbes and Rigobon (2002 2002) and Masson (1998 1998, , 1999 1999). ). • Contag agio ion: n: crisis is was triggered d and spread ad by investor’s psychological behaviour or panic movements ts rather r than being induced d by economic ic fundamental amental links. . • Interde depe pende dence ce/spil spillo lovers rs: : a crisis is in a countr try y affects cts the fundam amental ntal of the neighbour urs s through trade de or financial cial linkages. s. 12

How to test t for existenc stence e of contagion agion? • The examinati ation simply y tests s if this volatili ility ty transm nsmissi ission on changes s significantl cantly y after r the shocks s or crisis is. • Contag agio ion is defined d as a significant icant increase ase in cross-mar arket t linkages s (co-movemen ment) t) after r shock. • Interde depe pende dence ce or spillovers: : two markets ts show a high degree of co-movemen ment t during period d of stabili ility/ ty/the the co-movement ment does not increase ase significan icantl tly y after r shock. 13

Crisis in one country could coordinate investor expectations. The co-movement in price would exist because of correlation in memories rather than fundamental. During ing the crisis sis period iod, Compar mparing ing the two episode sodes s implied correlation for the of calm periods iods: the degree equity return among ASEAN of financial interdependence markets increase sharply in has increased from the pre- both the 1997 and 2008 crisis, Asian crisis (1992-1996) to however, the latter event incurs the pre-US crisis period. a more dramatic rise 14

Condit nditio ional nal correlation elation coefficien icient t of stock returns ns (Thailand ailand as a crisis sis originator ginator in 1997 1997) 15

Condit nditio ional nal correlation elation coefficient icient of stock returns ns (Thailand ailand as a crisis sis origina ginator tor in 1997 1997) 16

Condit nditio ional nal correlation elation coefficient icient of stock returns ns (US US as a crisi sis s origina ginator tor in 2008 2008) 17

Condit nditio ional nal correlation elation coefficien icient t of stock returns ns (US US as a crisi sis s origina ginator tor in 2008 2008) 18

Impacts of shock originated within EAEs could be more severe than those created outside the region • Recentl ntly , there’s significant increase in the pair -wise conditional correlation among Asian countries itself. • Coefficients are even stronger than in comparison with the US. • One could link this phenomenon to the increased financial integration, which has intensified contagion effects across markets. 19

Policy cy respo ponse nse of each country ntry in EAEs? 20

Capit ital al Flow Managem ement ent Measures es (CFMs) become popular lar toolkit its among EAEs Capital ital Flow w Manag agemen ement Measur ures (CFMs) s)-with with announcem ouncement ent date Dec 10 10 Nov 10 10 19 19 30 30 Oct 10 10 Restric rictio tion on ex external borrowin ing 1/ Liberaliz lize Sep 10 10 outflo lows 30 30 30 30 9 Jun 10 10 Liberal raliz ize e Raise reserve Restric rict t NR Feb 10 10 outflo lows Requir ireme ment t on Dec 09 09 Nov 09 09 portfolio lio inve vest t 12 12 fore reign ign deposits its 13 13 & ST bond Furthe ther 18 18 Liberaliz lize 10 10 Withholding ing 9 Limit it priva vate te outflo lows IN PH tax on bonds FX ex exposure Withholding ing Limit it NR 16 16 TH Restric rictio tion tax on bonds ID Liberal raliz ize e access ess to on on ex external (in Jan 11 11) outflo lows KR TW deposit it borrowin ing Min holding account MY periods ds & limit access ess to to CB bills Note: : 1/ Bank k levy in 2011H2 2 for Korea; Limit it bank k ST borrowin ing in Mar 11 11 and raise RR on FCD in Mar & Jun 11 11 for Indonesia ia

22 Is there re external ernality ity of the CFMs?

Previ vious us litera ratur tures s on spillover effect cts s of the CFMs • Forbes et al. (2011 2011) found that capital control in Brazil cause investors to change their portfolio allocation in such away that: 1. reduce allocation to Brazil 2. reduce allocation to countries believed to be more likely to use control 3. Increase allocation to other country in Latin America that have fewer restriction on capital flows • Effect of capital control is through signaling rather than direct cost of the controls

Previ vious us litera ratur tures s on spillover effect cts s of the CFMs • Accor ordin ing g to Edison on and Reinha hart t (2001 2001), equity ity markets ets continu ntinue to be inter erna nationa tionall lly linked ed, despite ite the introduc oductio tion n or escala lation tion of capital l contr trols ols during ing the Asia ian n financ ncia ial l crisis is. • In addition tion, follo lowin ing g the introduc oductio tion n of the capita tal l contr ntrols ols, one should ould expect t the follo lowing ing phenom omena ena in the financ ncia ial l variable les 1) A decline ne in volatili tility ty spillo lover; ; 2) A structu tural l breaks ks around nd the introdu oduction tion of controls ols 3) less contem tempor poraneo neous moveme ment nt with h inter erna nationa tional l varia iable les