Asia-Paifi Ineres Rae Movemens: A Tale Of A Two-Horse Sleig Do Quo To Nguyen, Ti Tu Ha Pi, Tuy-Duong Tô * Absra We sudy e spillover effes of e U.S. s and e euro area s maroeonomi news on e Asia-Paifi ineres raes. We find a iger an expeed news from ese wo eonomies raises e ondiional mean, and mos news, regardless of signs, eliis e assoiaed volailiy of daily reurns of e Asia-Paifi ineres raes. We also find supporing evidene for e wo possible annels of e spillover effes of maroeonomi news, namely e poliy reaion expeaion and e inegraion meanisms. More imporanly, we reveal a e U.S. s unemploymen rae news and e euro area s inflaion news ave e sronges impa on bo e firs wo momens of e Asia-Paifi ineres raes daily reurns. Tese findings are onsisen wi e relaive size of ese wo eonomies, and e moneary poliy frameworks ondued a e U.S. s Federal Reserve and e European Cenral Bank. JEL lassifiaions: E44, G14, G15 Key words: Spillover effes; Maroeonomi news; Asia-Paifi region; euro area; EGARCH * Do Quo To Nguyen (Corresponding auor), Rouen Business Sool, Mon-Sain-Aignan-Cedex, Normandy 76825, Frane, Email: ng@rouenbs.fr, Tel: + 33 2 32821776. Ti Tu Ha Pi, Erns & Young Ausralia, Sydney, NSW 2000, Ausralia, Email: a.pi@au.ey.om, Tel: +61 2 92485542. Tuy-Duong Tô, Ausralian Sool of Business, Universiy of New Sou Wales, Ausralia, Email: d.o@unsw.edu.au, Tel: +61 2 93855865. Te auors would like o ank Hali Edison and Sijn Claessens for elpful ommens and suggesions. Te remaining errors, if any, are our own.
1. Inroduion Tere is ample evidene a maroeonomi announemens ave signifian effes on finanial markes, bo on asse reurns and e assoiaed volailiy. Amongs differen finanial markes, e bond marke as araed mu aenion wi resear sudies falling in eier of wo srands. Te firs bran invesigaes e prie impas of differen maroeonomi news, see e.g. Beeey and Wrig (2009), Gürkaynak e al. (2005), Hess (2004), Balduzzi e al. (2001), Flemming and Remolona (1999b), and Edison (1997). On e oer and, wile e exising sudies fous primarily on e prie impa, ineres rae volailiy as beome an inreasing onern o poliymakers and finanial marke pariipans alike. Along is rend, researers ave sifed e fous o e impas of e U.S. maroeonomi news on e volailiy of e U.S. Treasury bonds, see e.g. Brenner e al. (2009), Arsanapalli e al. (2006), de Goeij and Marquering (2006), Flemming and Remolona (1999a), Jones e al. (1998). Te lieraure onsisenly sows a e U.S. s bond markes appear no o rea o e expeed bu o e unexpeed omponen (or news) of e releases. 1 Reenly, given srong evidene as been esablised for e informaion leadersip of e U.S. eonomy in e global eonomy, 2 ere is an emerging sool of resear invesigaing e spillover impas of e U.S. maroeonomi announemens on foreign markes. In priniple, e U.S. s news mig affe oer naional markes roug various annels of poliy aniipaion and (real and/or finanial) inegraion. Te firs annel implies a wen ere is a sok in e U.S. s poliy, marke pariipans would reasonably 1 Tere are only few sudies a empirially esed e U.S. s maroeonomi news impas on bo e firs wo momens of e U.S. s markes, see e.g. Kim e al. (2004). 2 See iner alia Arsanapalli e al. (1995) and Ng (2000). 1
expe a possible moneary poliy response and/or foreign exange inervenion by e foreign moneary auoriies, see iner alia Engel and Frankel (1984) and Hardouvelis (1988). For insane, wen e Fed deides o raise is arge ineres rae, marke pariipans mig expe a moneary auoriies of exange rae argeing eonomies, e.g. Singapore or Hong Kong, will rea aordingly o mainain eir exange raes agains e U.S. dollar. Terefore, i is likely a e Singapore and Hong Kong sor-erm raes would inrease o mainain e ineres rae pariy ondiion. Te seond annel implies a due o is relaive imporane in e world eonomy, e U.S. s maroeonomi news will send a signal abou global soks a will influene e res of e world, and us sould indue foreign ineres raes o follow a similar paern wi e U.S. s ineres raes. In fa, ere is some evidene on e spillover effes of e U.S. s maroeonomi news on foreign ineres raes. For example, Beker e al. (1995) sow a some U.S. maroeonomi news signifianly affe German, Briis and Japanese ineres raes. Goldberg and Leonard (2003) and Ermann and Frazer (2004) find a e U.S. maroeonomi news signifianly affe e euro area/german ineres rae reurns. Reenly, Andrizky e al. (2007) find a emerging markes bond spreads ave been signifianly affeed by e U.S. Fed s arge rae, wile all oer maroeonomi announemens signifianly influene volailiy. Given e inreasing imporane of e Asia-Paifi region in e world eonomy as well as e growing real and finanial inegraion wi e U.S. eonomy, aside from domesi news, ere is a srong noion a e U.S. s news as also been losely moniored by no only poliy makers bu also praiioners in e Asia-Paifi region. However, surprisingly, ere are only a andful of sudies on e poenial spillover effe of e U.S. s 2
maroeonomi news on e Asia-Paifi ineres raes. 3 Furermore, ese sudies are limied o e invesigaion of eier firs or seond momen effes and mainly fous on a erain number of individual advaned eonomies in is diversified region. Anoer sorfall in e exan lieraure is a sine e euro area blo is inreasingly playing as anoer key driver aside from e U.S. eonomy, i is likely a e euro area s maroeonomi news would also poenially spillover o e oer regions. However, ere are only sparse sudies in e urren lieraure o es for e spillover impas of e euro area s news on e Asia-Paifi region. For insane, Kim and Nguyen (2009) find a e European Cenral Bank s arge rae news signifianly affe e Asia-Paifi sok markes reurns and volailiy. Terefore, in a similar vein, i is possible a oer maroeonomi news from e euro area mig also influene e Asia-Paifi markes due o e imporane of e news onen in su announemens as well as e growing inegraion beween e wo regions. Agains is bakground, we aim o address ese sorfalls in e lieraure by providing ompreensive evidene for e spillover effes of bo e U.S. s and e euro area news on e firs wo momens of daily anges of ineres raes of six represenaive ounries in e Asia-Paifi region wi differen eonomi developmen levels, insiuional arrangemens as well as inegraion levels wi e U.S., e euro area and e world eonomy. Speifially, we aim o sed lig on e following imporan issues. Firsly, weer e U.S. s and e euro area s maroeonomi news ave any impas on e firs wo momens of e Asia-Paifi ineres rae reurns? Seondly, amongs a baske of several maroeonomi news, wi news as e mos signifian effe? In oer words, for poliy making, rading and edging sraegies, wi maroeonomi news from ese wo blos sould Asia-Paifi 3 For insane, Kim and Seen (2000) invesigae e impas of e U.S. s news on e Ausralian markes. 3
marke poliy makers and invesors wa losely? Tirdly, weer e Asia-Paifi ineres raes respond o marke expeaions of e likeliood of orresponding poliy aions and/or inervenions from e Asia-Paifi moneary auoriies or o e worldwide sok signals from e U.S. s and e euro area s eonomies? Finally, weer e Asia-Paifi ineres raes respond differenly o e U.S. s news and o e euro area s news? Using EGARCH(1,1) meodology o invesigae e impas of five U.S. and four euro area s maroeonomi news on e firs wo momens of daily anges of six Asia- Paifi eonomies sor-, medium- and long-erm ineres raes from January 1999 o Deember 2008, we firsly find a iger an expeed maroeonomi news from bo e U.S. and e euro area raises ineres raes in e Asia-Paifi region. Furermore, mos maroeonomi news from ese wo eonomies eliis e ondiional volailiy of e Asia- Paifi ineres raes. Te findings are also supporing evidene for wo possible annels of e spillover effes of maroeonomi news from e U.S. and e euro are o e Asia- Paifi ineres raes markes, namely e poliy reaion expeaion and e inegraion meanisms. More imporanly, we onsisenly sow a amongs several maroeonomi news, e real eonomi indiaor of e U.S. s unemploymen rae is e key driving fore a moves e Asia-Paifi ineres raes. Tis is onsisen wi e fa a, as Boyd e al. (2005) poin ou, e U.S. s unemploymen rae is viewed as newswory sine, iner alia, e key goal of e Fed s moneary poliy is o promoe e objeive of maximum employmen (Federal Reserve, 2005).Aordingly, e unemploymen figure as frequenly been referred o in e Federal Open Marke Commiee s (FOMC) arge rae deision making proess, and as been e arge of wide speulaion amongs praiioners. Our finding also suggess a due o is relaive size in e world eonomy, news relaed o e U.S. s real eonomi aiviy sends a signal abou e global soks on e aggregae demand su a e Asia-Paifi, as 4
e U.S. s larges rading parner, would follow aordingly. In oer words, e Asia-Paifi real ineres raes will adjus in alignmen wi e U.S. s real eonomi indiaors, and ulimaely revealed in e nominal ineres raes. On e oer and, e euro area s inflaion news is robusly sown o ave e sronges impa on e Asia-Paifi ineres raes. Tis suggess a under is offiial inflaion argeing framework, e euro area s inflaion news mig send a signal abou is enral bank ommimen o fig agains inflaion, su a e marke pariipans would expe e Asia-Paifi moneary auoriies o rea o mainain e ineres rae pariy ondiions. Terefore, su indued poliy aions would influene e marke s expeaions on e fuure pa of inflaion in e Asia-Paifi region, wi in urn leads o anges in e nominal ineres rae. Overall, our findings sugges a e Asia- Paifi believes a e U.S. arges eonomi aiviy wile e euro area arges inflaion. As a resul, for poliy making, rading and edging sraegies, e Asia-Paifi marke pariipans sould losely monior e U.S. s unemploymen rae and e euro area s inflaion news. Finally, we find a e U.S. s and e euro area s news ave similar effes on bo momens of daily anges of e Asia-Paifi ineres raes even oug e euro area s news mig be influened by e U.S. s news. Tis suggess a, aside from e U.S., anks o is growing role in e world eonomy, e euro area s news are informaive for e Asia-Paifi region. Te res of is paper proeeds as follows. Seion 2 disusses daa and empirial modeling issues. Seion 3 repors and analyzes e esimaion resuls. Seion 4 disusses several robusness eks and oer empirial ess. Seion 5 onludes e paper. 5
2. Daa 2.1. Ineres raes We olle sor, medium and long-erm ineres raes of six Asia-Paifi markes onsising of Ausralia, Hong Kong, Japan, Korea, New Zealand, and Singapore from Bloomberg for e period from January 1999 o Deember 2008. In pariular, we olle daily losing daa for e following ineres raes: 90-day Dealer Bill, 3 and 10-year Commonweal bonds for Ausralia; 3-mon Exange Fund Bill, 3 and 10-year Exange Fund Noes for Hong Kong; 3-mon Bank Bill, 3 and 10-year Governmen Bonds for Japan; 91-day Commerial Paper, 5 and 10-year Treasury Bonds for Korea; 3-mon Treasury Bill, 5 and 10-year Governmen Bonds for New Zealand; and 3-mon Treasury Bill, 5 and 10-year Treasury Bonds for Singapore. 2.2. Maroeonomi news We also olle daa on e announemen daes, release imes, and aual announed figures for five maroeonomi announemens for e U.S. and four announemens for e euro area from orresponding reporing auoriies as sown in Table 1. In pariular, we use a se of five U.S. s announemens from January 1999 o Deember 2008 inluding e FOMC s arge rae announemens (ereafer Fed ) released afer e FOMC Board meeings, quarerly Gross Domesi Produ perenage level (ereafer GDP, final release) released for e previous quarer, monly Consumer Prie Index (ereafer CPI ) perenage ange, and Unemploymen (ereafer UE ) perenage level as well as billion of dollar ange in Trade defii eadlines (ereafer TD ). For e euro area, we olle daa from January 2002 o Deember 2008 for e ECB s Governing Counil s arge raes (ereafer ECB ), GDP (inluding flas esimaes, firs esimae and seond esimae), Moneary Union 6
Index of Consumer Pries (CPI), and Unemploymen rae (UE). 4 Tese are e mos imporan eonomi announemens for bond reurns as poined ou by Balduzzi e al. (2001). As sown in Table 1, wile e Fed s funds arge is released a 14.15 pm EST afer e FOMC s meeings, all oer variables are released a 8.30 am EST (GMT-5). For e euro area, e ECB s arge ineres rae is announed a 14.30 pm CET (GMT+1) and e oer maroeonomi announemens are released a 11.00 am CET. Table 2 sows frequeny of maroeonomi announemens released on e same daes for e U.S. and e euro area. Te U.S. s TD and CPI announemens ave been released on e same daes more frequenly an e oers. Meanwile, e ECB s arge raes and e euro area s GDP ave been released onurrenly relaively frequenly. Te expeaion ypoesis of e erm sruure suggess a nominal ineres raes embody expeed fuure real ineres raes and inflaion expeaions. Terefore16, o properly model e responses of nominal ineres raes o maroeonomi news, i is neessary o invesigae e impas of maroeonomi news on ese wo aspes. Following Beeey and Wrig's (2009) and Edison s (1997) lassifiaions, we group e maroeonomi news ino ree aegories: news a are indiaive of fuure pa of inflaion, news abou moneary poliy and news assoiaed wi general level of real eonomi aiviy. Te firs group is represened by e CPI news a mig aler marke pariipans expeaion on e fuure pa of inflaion. For example, iger an expeed CPI announemens would lead o a iger expeaion of fuure inflaion, is in urn leads o a iger nominal ineres rae. In addiion, in e onex of iger an expeed inflaion, marke pariipans mig reasonably believe a e moneary auoriies will rea o akle e inflaion ike by 4 Te euro area s maroeonomi announemens are olleed from e Eurosa websie, e Saisial Offie of e European Communiies. Te earlies available dae for ese daa is January 2002. 7
raising e arge ineres rae. As a resul, iger an expeed inflaion mig lead o iger ineres raes roug is annel of expeaion. Te seond group onains e Fed s and e ECB s arge ineres raes. Tese news mainly reveal informaion abou marke pariipans beliefs of e enral bank s reaion funion a would affe e formaion of marke expeaions on e fuure real ineres raes 5. As Ermann and Frazer (2004) empasize, e effe of a moneary poliy deision on long-erm raes mig be differen from a on sor-erm raes. Canges a e longer end of e yield urve an a leas in par be aribued o e marke s views on e enral bank s redibiliy or is apabiliy o onrol inflaion. Terefore, a igening of moneary poliy an be ompaible wi a reduion in long-erm ineres raes if markes pereive e igening as a redible sep by moneary auoriies o redue inflaion in e long-run, see e.g. Tornon (1998). However, as Flemming and Remolona (1999b) find a ump-saped effe of maroeonomi announemens on e U.S. s yield urve, i suggess a e marke would expe moneary auoriy o respond in sor- o medium-erm. Te ird group is assoiaed wi e general level of real eonomi aiviy inluding announemens on GDP, UE and TD. Higer an expeed GDP grow rae and lower an expeed UE rae and TD level mig lead marke pariipans o revise upward eir beliefs of fuure eonomi grow rae. Tis in urn mig lead o a iger demand for funding. If is is e ase, iger ineres raes would also be expeed. Addiionally, iger an expeed eonomi aiviy mig indiae a e eonomy is overeaing, marke pariipans mig expe e moneary auoriies o ool i down by raising e arge ineres raes. 6 5 See e.g. Haldane and Read (2000). 6 See e.g. Edison (1997). 8
In line wi e urren lieraure, o derive e maroeonomi news omponen from e aual announemens of e U.S. marke, we use e median of e survey series obained from Money Marke Servies (MMS) Inernaional as a proxy for e marke s expeaion and e news omponen is e differene beween e aual announemen and e survey median. Te MMS surveys are ondued on e Friday of e week prior o e seduled release of ea variable under onsideraion. Te MMS surveys ave been used exensively in e lieraure su as Ederingon and Lee (1993), MQueen and Roley (1993), Edison (1997), Flemming and Remolona (1999a, b), Li and Engle (1998), Balduzzi e al. (2001), Flannery and Proopapadakis (2002), Hess (2004), and de Goeij and Marquering (2006). Furermore, e MMS surveys properies ave been empirially esed in a number of sudies, e.g. Balduzzi e al. (2001), and empirial evidene sows a e MMS foreass are unbiased, effiien and auraely represen e marke s onsensus upon e upoming announemen. However, sine aquired by Informa o form Informa Global Markes in 2003, MMS as sopped providing survey servies. In addiion, MMS does no provide survey daa for e euro area, exep from e German markes. Terefore, for e U.S. from 2003 onwards, and for e euro area, we use Reuers Daa Surveys. Tis daa is also widely used, see for example Ermann and Frazer (2004). Le M i denoe e median of e surveys and A i is e aual announemen for maroeonomi variable i, e news omponen of maroeonomi news i is measured as: i = A i - M i (1) On e oer and, in line wi e reen developmens in e arge ineres rae news lieraure, we derive e news omponen of e enral bank s arge ineres rae news using 9
Kuner (2001) s meodology. 7 Te news omponen of e Fed s and e ECB s arge rae announemens on day d of mon m an be derived from e implied ange in e fuures onra s prie. We use e urren mon Fed funds fuures onras o derive e Fed s news as Krueger and Kuner (1996) find a e Fed funds fuures rae is an effiien predior of e Fed funds arge rae. For e ECB news, we use e 3-mon EURIBOR fuures following Berno and Von Hagen (2004) s finding a e 3-mon EURIBOR fuures rae is an unbiased predior of e euro area poliy rae anges. Sine e Fed funds fuures selemen prie is based on e monly average of e spo Fed funds raes, e number of days affeed by e announemen in a pariular mon is saled as in equaion (2). On e oer and, as e 3-mon EURIBOR fuures selemen prie is alulaed based on e referene ineres rae (EURIBOR) for e 3-mon euro erm deposis on e las rading day, e news omponen of e ECB s arge rae announemen is alulaed as in equaion (2) wiou e saling faor D/(D-d). were 0 0 ( f f ) u D i = m, d m, d 1, (2) D d u i is e unexpeed arge rae anges; f 0 m, d is e urren-mon Fed funds fuures rae for e Fed and 3-mon EURIBOR fuures rae for e ECB; f 0 m, d-1 is e fuures rae as of e day prior o e announemen; D is e number of days in e mon; and D-d is e number of days in e mon affeed by e announemen. 2.3. Empirial models and ypoeses Tis paper dually examines e news impas on bo e ondiional mean and e ondiional volailiy of daily reurns. Furermore, as e volailiy effes of e news 7 As a robusness ek, we also use e MMS surveys median o derive e arge rae news for e Fed s funds arge and e ECB s arge. However, e resuls are generally onsisen wi wa repored in e paper. 10
depending on e role of e news in resolving e eerogeneiy of beliefs and expeaions of marke pariipans, e ypoeses esed in is paper inlude no only e volailiy inreasing bu also e volailiy reduing influene of e news. Terefore, e empirial model employed mus enable U.S. no only o simulaneously es e news impas on bo e firs wo momens, bu also allow U.S. o idenify e news asymmeri impas on e ondiional volailiy. In e eonomeri lieraure, e Nelson (1991) EGARCH (1,1) model allowing for bo posiive and negaive oeffiiens on e ondiional volailiy is bes suied o is paper s empirial esing purposes. Equaion (3) sows e EGARCH (1,1) model used for e empirial es of e maroeonomi news spillover effes. Following Balduzzi e al. (2001), we es e impa of ea news separaely. Te ondiional mean equaion for e Asia-Paifi ineres rae anges ( y ) is expressed as a funion of pas anges in relevan markes; e Monday dummy variable ( Mon ) is used o aoun for e weekend effe and akes e value of one for Mondays and zero oerwise; and e news variable ( ) for ea of e U.S. s and e euro area s maroeonomi variable disussed in Seion 2.2. Te ondiional variane equaion for e anges in e finanial marke series ( ) is expressed as a funion of one period lag of e variane and e residuals, e Monday effe dummy and variables. Te variable is lagged by one period o aoun for e rading ime differene beween e U.S., e euro area and e Asia-Paifi region. Te U.S. s and e euro area s announemens are released wile e Asia-Paifi markes are losed, us e laer would respond o e news from e oer wo wen ey open for rading one alendar day afer e 11
announemens. Terefore, e spillover impas, if any, would sow up in e following day. 8 ln β y p = + α Lag, i y i + α Mon Mon + α 1 i= 1 β α + β β 1 1 = + ln 1 + 1 + 2 + Mon Mon + 1 1 1 β β (3) As disussed in Seion 1, ere are a leas wo annels for e spillover effes of e U.S. s and euro area s news on e Asia-Paifi ineres raes via marke expeaion on e reaion funions of e Asia-Paifi moneary auoriies and via inegraion wi e U.S. s and e euro area s eonomies. Esimaing equaion (3), we empirially es e following ypoeses: H1 - Poliy reaion expeaion ypoesis. Sine e overnig ineres rae is used as poliy insrumen in mos of Asia-Paifi moneary auoriies, e Asia-Paifi s sor- o medium-erm ineres raes would respond o e U.S. s and e euro area s news as a resul of e marke expeaions a e Asia-Paifi moneary auoriies will rea aordingly o mainain e ineres rae pariy ondiions and onrol inflaion. H2 - Inegraion ypoesis. Te U.S. s and euro area s real eonomi news would reveal informaion abou e global soks a affe e Asia-Paifi s longer-erm ineres raes as longer-erm ineres raes are likely o be deermined by longer-erm eonomi fundamenals. 8 Te inlusion of day-of-e-week dummies as in Kim e al. (2004) does no maerially affe our resuls. Furermore, we also esimae equaion (3) wi a dummy variable equals o 1 for exange rae argeing ounries, e.g. Singapore and Hong Kong, and zero oerwise. However, e resuls are qualiaively similar o wa repored in e paper. 12
In general, we expe a iger an expeed inflaion and eonomi aiviies would ave a posiive impa on e Asia-Paifi s ineres raes. However, e news effe on e ondiional volailiies would depend on weer e news adds o or resolves unerainy in e markes. If e news leads o furer speulaions in e marke, is inreased eerogeneiy would be sown in a rise in e volailiy. Furermore, depending on e informaional role a e U.S. s and e euro area s news play in e Asia-Paifi markes, a marke alming effe ould be observed if e news resolves unerainy and onribues o e formaion of a onsensus view on e fuure pa of ese wo eonomies. Te responses of e Asia-Paifi s sor- and longer-erm ineres raes o e U.S. s and e euro area s news are inerpreed as supporing evidene for ypoeses H1 and H2. However, as we expe similar responding signs for bo ypoeses, i would be ard o idenify supporing evidene for ea ypoesis; erefore, we ondiionally assume a H1 effe dominaes H2 effe a e sor-end and H2 effe dominaes H1 effe a e long-end of e yield urve. 3. Empirial resuls 3.1. Wi U.S. s news affes e Asia-Paifi raes? Panel A of Table 3 sows e esimaions for e spillover of e U.S. s maroeonomi news on e Asia-Paifi ineres raes. To mainain e paper s fous, we only repor e oeffiiens and e assoiaed P-values. In general, all five key maroeonomi news ave signifian impas on bo e firs wo momens of e ineres rae reurns in six Asia-Paifi markes under onsideraion. Consisen wi e lieraure on e impas of e U.S. s maroeonomi news on is domesi ineres raes, all signifian ases ave expeed signs were iger an expeed Fed s arge rae and e GDP, CPI and TD news ave a posiive impa on e Asia-Paifi ineres raes anges. On e oer and, iger an expeed U.S. s UE rae would resul in reduions in e Asia-Paifi ineres rae anges. We also find evidene supporing ypoesis H1 were more sor-erm raes 13
responding o e news an medium and long-erm raes wi 15, 11 and 10 signifian responses in ea sperum. In erms of magniude, in general, longer-erm ineres raes respond more srongly o e news an sorer-erm raes. Tis is onsisen wi Gürkaynak e al. (2005) s and Kim and Seen (2000) s finding of e exess sensiiviy of long-erm ineres raes, and suggess a marke pariipans adjus eir expeaions of e long-run inflaion rae in response o maroeonomi news from e ow soures. Furermore, we observe a e U.S. s news as sronger impas on eonomies wose urrenies are pegged o e U.S. dollar, i.e. Hong Kong and Singapore. Tis furer onfirms H1 ypoesis a marke pariipans expe a moneary auoriies of ese wo eonomies will adjus eir ineres raes following e U.S. s movemens o mainain e arge exange raes. On e ondiional volailiy, wile news on e Fed s arge rae, unemploymen, and Trade defii elii volailiy in e Asia-Paifi ineres raes, e GDP and CPI news end o lower volailiy in ese markes. Te laer finding is onsisen wi Andrizky e al. (2007) s finding a e U.S. s maroeonomi news redue unerainy in emerging markes. Toug ere is a growing real inegraion wi e U.S., e GDP and CPI news, wi refle e real oupu as well as e inflaionary pressure on e eonomy, seem o provide more insigs abou e U.S. eonomy a elp o resolve eerogeneiy in beliefs among marke pariipans in e Asia-Paifi region. Unemploymen, oug also refles real aiviies of e eonomy, does no ave e same alming effe. I may be beause unemploymen is a sensiive ype of news, and marke pariipans are widely differen on eir expeaion on e reaion of e Fed. On e oer and, e Trade defii news only impas e volailiy of Japan and Korea. We onjeure a is is due o e ig elasiiy of e produs a Japan and Korea expor o e U.S. marke. In general, we find evidene supporing ypoesis H1 and is onsisen wi e lieraure were e impas are weaker a e long end of e yield urve. Among e five 14
U.S. s news, e GDP news as e sronges impa on e ondiional mean, wi is onsisen wi e relaive size of e U.S. eonomy in e global eonomy. On e oer and, e UE news as e sronges impa on e ondiional volailiy. Tis is onsisen wi Ederingon and Lee (1993) finding a during e period from lae 1980s o early 1990s, e U.S. s unemploymen news as e greaes impas on e volailiy of e U.S. s ineres raes. On bo momens, e Unemploymen news as e greaes impa wile e Trade defii s news as e leas impa on bo e ondiional mean and volailiy of e Asia- Paifi s ineres rae anges. Tis is supporing of e H2 ypoesis and onsisen wi e newsworiness of e U.S. s unemploymen news and is relaive size o e world eonomy su a e unemploymen news would reveal informaion relaed o e global soks a indue e Asia-Paifi marke o rea aordingly. 3.2. Wi euro area s news affes e Asia-Paifi raes? Panel B of Table 3 sows e esimaions for e spillover effes of e euro area s maroeonomi news on e Asia-Paifi ineres raes. We find evidene similar o e U.S. s news were in general euro area s news as expeed impas on e Asia-Paifi ineres raes. Higer an expeed ECB s arge rae, GDP grow rae and CPI inflaion rae raise ineres raes wile iger an expeed UE rae leads o lower ineres raes. We also find supporing evidene for e H1 ypoesis were more sor-erm ineres raes respond o e news wi 13 ou of 28 signifian responses are sor erm raes. Consisen wi e U.S. s news spillover effes and Gürkaynak e al. (2005) s and Kim and Seen (2000) s finding of e exess sensiiviy of long-erm ineres raes, we also find a e longer e mauriy, e sronger is e reaion. For example, a 1% unexpeed ike in e ECB s arge rae leads o in a orresponding inrease of 9.73 basis poins in e Ausralian 3-mon rae bu 33.60 basis poins for e 3-year rae. Similarly, a 1% iger an expeed CPI inflaion rae resuls in an 15
inrease of 1.5, 4.28 and 5.99 basis poins in e Singapore s 3-mon, 5 and 10-year raes, respeively. On e ondiional volailiy, wile e GDP news as unlear impas, iger an expeed ECB s arge rae, CPI and UE raes raises volailiy in mos of e Asia-Paifi ineres raes. Overall, if all news are no released a e same ime, e euro area s inflaion news ave e greaes impa wile e GDP news as e lowes impa among e four news examined. Tis an be inerpreed as supporing evidene for e H1 ypoesis and i suggess a e euro area s inflaion news would indue e Asia-Paifi moneary auoriies o ake similar aions aordingly o mainain e ineres rae pariy ondiions. 4. Robusness eks 4.1. Sandardized effes Sine measuremen unis differ aross maroeonomi news, o failiae omparabiliy and inerpreaion, following Balduzzi e al. (2001) s meodology we sandardize e news by dividing i by e sandard deviaion aross all observaions. Le S i denoes sandardized news and σ i is e sandard deviaion of maroeonomi news i, e sandardized news S i is measured as: S i = i / σ i (4) We en esimae equaion (3) wi S i insead of i as sown in equaion (5) ln y p = + α Lag, i y i + α Mon Mon + α S 1 i= 1 α + 1 1 = β ln 1 1 2 Mon Mon S 1 1 1 (5) Aordingly, e regression oeffiien is o be inerpreed as e response o a onesandard deviaion ange in e news variable. Table 4 sows evidene onsisen wi e findings repored in Seion 3. Speifially, all e U.S. s and e euro area s news ave expeed effes on e ondiional mean of e Asia-Paifi ineres rae anges were 16
iger an expeed news on e Fed and e ECB arge ineres raes, GDP, CPI and TD lead o iger ineres raes in mos of Asia-Paifi markes, wile iger an expeed UE resuls in lower ineres raes in e Asia-Paifi region. On e ondiional volailiy, wile all euro area s maroeonomi news elii volailiy, e U.S. s GDP and CPI inje some marke alming effes o e Asia-Paifi markes. We onsisenly observe a e U.S. s unemploymen rae news and e euro area s inflaion news ave e sronges effes on bo e firs wo momens of daily anges of e Asia-Paifi ineres raes. 4.2. Common-sized effes As e U.S. s rade defii news is measured in billions of dollar ange wile e oers are measured in perenage, anoer way o failiae a dire omparison is o normalize all news omponens relaive o eir aual announemens. Ta is, le C i denoes ommon-sized news and A i is e aual announemen for maroeonomi variable i, e ommon-sized news C i is measured as: C i = i / A i (6) We en esimae equaion (3) wi C i insead of i as sown in equaion (7) ln y p = + α Lag, i y i + α Mon Mon + α C 1 i= 1 α + 1 1 = β ln 1 1 2 Mon Mon C 1 1 1 (7) Te U.S. s ommon-sized news effes are sown in Table 5, Panel A. Te resuls sugges a wa we already repored in Seion 3.1 is relaively robus wi e ommonsized news measures. In erms of signifian responses and e reaion magniude, e ommon-sized news effes on e ondiional mean and ondiional volailiy of e Asia- Paifi ineres rae anges are similar o wa was repored in Seion 3.1. Te U.S. s unemploymen news is onsisenly observed as e sronges driving fore for e Asia- Paifi ineres rae anges. 17
In general, in erms of number of signifian responses, e euro area s ommon-sized effes on e ondiional mean of e Asia-Paifi ineres rae anges as sown in Panel B of Table 5 are sligly weaker an wa was repored in Seion 3.2, wile su effes on e ondiional volailiy are fairly sronger. We also onfirm e finding a e euro area s inflaion news as e sronges impa on bo momens of e Asia-Paifi ineres rae anges. Remarkably, we observe a e GDP news from bo e U.S. and e euro area elps o lower volailiy in mos of e Asia-Paifi s ineres raes. As e wo mos imporan drivers for e global eonomy grow, GDP news from ese wo blos would inje valuable informaion upon e underlying eal of e wo eonomies and fuure demand for bo maerials and apial o keep ese wo engines running. Tis en would elp o larify unerainy in e Asia-Paifi region wi ulimaely resuls in a lower volailiy level. 4.3. Conurren effes We noie a ere are irumsanes were several news ave been announed a e same ime as sown in Table 2. Terefore, i is neessary o ek if individual news effes are robus in e presene of e oers in a pariular rading day. Following Flemming and Remolona's (1999a) pooling meod, we esimae simulaneously for all U.S. s news and all euro area news, respeively. 9 Speifially, for e U.S. s news effes, we esimae equaions (8) y p TD = + α Lag, i y i + α Mon Mon + α j j, 1 i= 1 j= Fed α + 9 We also esimae equaion (3) wi all U.S. s and euro area s news ogeer, i.e. a pooling of e rig and side of equaions (8) and (9). However, e resuls are qualiaively similar o e individual esimaions of equaions (8) and (9). Tis suggess a e euro area s news as own informaive meris for e Asia-Paifi markes. 18
ln TD 1 1 = β ln 1 1 2 Mon Mon + β j j, 1 1 1 j= Fed (8) Similarly, we esimae equaion (9) for e euro area news effes ln y p UE = + α Lag, i y i + αmonmon + α j j, 1 i= 1 j= ECB α + UE 1 1 = β ln 1 1 2 Mon Mon + β j j, 1 1 1 j= ECB (9) Panel A of Table 6 sows e resuls for e U.S. s news onurren effes. Conrolling for e onurren effes, e resuls disussed in Seion 3.1 are relaively robus wi similar number of signifian responses wi expeed signs. However, we find a amongs e five news, e Fed s news as e sronges impa. Tis mig be due o e iming advanage of e Fed s news were i is released 5 ours and 45 minues afer e oer news ave been already released. By e ime e Fed s news is e markes, e oer news ave been already digesed by marke pariipans, us e Fed s news would bring in new informaion and ave sronger impas. We also observe a ere are sligly more mediumerm raes responding signifianly e U.S. s news, wi is onsisen wi Flemming and Remolona's (1999b) finding of a ump-saped effe of maroeonomi announemens on e U.S. s yield urve. On e oer and, e U.S. s news as sligly sronger impa on e ondiional volailiy an wa was repored in Seion 3.1. In general, e U.S. s news raises volailiy in mos Asia-Paifi ineres raes. However, ere are more ases wi lower volailiy an previously repored. We onjeure a inensive informaion injeed by several news onurrenly announed would elp e Asia-Paifi marke pariipans o ave a beer undersanding abou e underlying eal of e U.S. eonomy. Remarkably, e U.S. s unemploymen news onsisenly is e greaes driving fore on bo e ondiional mean and volailiy of e Asia-Paifi ineres rae anges. 19
Panel B of Table 6 sows e euro area s onurren impas. Wen e euro area s maroeonomi news are released on e same dae, e effe on e Asia-Paifi s ineres raes is marginally sronger an wa was repored in Seion 3.2. Of wi, similarly o e Fed s arge rae effes, e ECB s arge rae news as e greaes impa on e mean wile e CPI news as e mos influene on e ondiional variane. More imporanly, e euro area s inflaion news onsisenly as e sronges effe on bo e ondiional mean and volailiy of daily anges of e Asia-Paifi s ineres raes. 5. Conlusion Tis paper invesigaes e spillover impas of e U.S. s and e euro area s maroeonomi news on bo e ondiional mean and volailiy of daily anges of sor, medium and long-erm ineres raes of six eonomies in e Asia-Paifi region. We find a iger an expeed news from bo e U.S. s and euro area s eonomies are assoiaed wi iger ineres raes in e Asia-Paifi region. Furermore, maroeonomi news from ese wo eonomies ends o raise volailiy in mos of e Asia-Paifi ineres raes. More imporanly, we reveal a wile e U.S. s unemploymen rae news as e sronges impa on bo e mean and e volailiy of e Asia-Paifi ineres raes, e euro area s CPI news is onsisenly sown e sronges fore a moves e Asia-Paifi ineres raes. Tese findings are onsisen wi e noion a e U.S. eonomy is e key driving fore in e world eonomy due o is relaive size. Terefore, unemploymen soks revealing e underlying eal of e U.S. eonomy would sake no only is domesi markes bu also foreign markes. On e oer and, under e inflaion argeing framework, obviously e euro area s inflaion news would ave signifian impas as ese news no only inje more informaion direly assoiaed o e fuure pa of inflaion a elps invesors o revise eir disoun raes aordingly, bu also indirely reveal informaion relaed o e underlying eal of e euro area s eonomy and e likeliood of poliy 20
aions a e ECB mig ake o onrol inflaion. Te resuls sugges a e Asia-Paifi believes a e U.S. arges eonomi aiviy wile e euro area arges inflaion. Tese findings are onsisen wi e relaive size of e U.S. s and e euro area s eonomies as well as e moneary poliy frameworks ondued a e Fed and e ECB. We also find a e U.S. s and e euro area s news ave similar effes on bo momens of daily anges of e Asia-Paifi ineres raes even oug e euro area s news mig be influened by e U.S. s news. Tis suggess a aside from e U.S., e euro area is anoer key informaional leader a e Asia-Paifi would also monior losely. In pariular, among several maroeonomi news released every day, marke pariipans in e Asia-Paifi region would losely monior no only ose news relaed o e U.S. s unemploymen rae, bu also inflaion news from e euro area. However, furer resear is needed o furer explore e fundamenal reasons explaining e spillover effes of e U.S. s and e euro area s news. 21
Referenes Andrizky, J.R., Banniser, G.J., Tamirisa, N.T., 2007. Te impa of maroeonomi announemens on emerging marke bonds. Emerging Markes Review 8, 20-37. Arsanapalli, B., d'ouville, E., Fabozzi, F., Swizer, L., 2006. Maroeonomi news effes on ondiional volailiies in e bond and sok markes. Applied Finanial Eonomis 16, 377-384. Arsanapalli, B., Doukas, J., Lang, L.H.P., 1995. Pre and pos-oober 1987 sok marke linkages beween U.S. and Asian markes. Paifi Basin Finane Journal 3, 57-73. Balduzzi, P., Elon, E.J., Green, T.C., 2001. Eonomi and Bond pries: Evidene from e U.S. Treasury Marke. Journal of Finanial and Quaniaive Analysis 36, 523-543. Beker, K., Finnery, J., Kopeky, K., 1995. Domesi maroeonomi news and foreign ineres raes. Journal of Inernaional Money and Finane 14, 763-783. Beeey, M.J., Wrig, J.H., 2009. Te ig-frequeny impa of news on long-erm yields and forward raes: Is i real? Journal of Moneary Eonomis 56, 535-544. Berno, K., Von Hagen, J., 2004. Euribor fuures marke: effiieny and e impa of ECB poliy announemens. Inernaional Finane 7, 1-24. Boyd, J.H., Hu, J., Jagannaan, R., 2005. Te sok marke's reaion o unemploymen news: Wy bad news is usually good for soks. Journal of Finane 60, 649-672. Brenner, M., Pasquariello, P., Subramanyam, M., 2009. On e volailiy and omovemen of U.S. finanial markes around maroeonomi news announemens. Journal of Finanial and Quaniaive Analysis 44, Foroming. de Goeij, P., Marquering, W., 2006. Maroeonomi announemens and asymmeri volailiy in bond reurns. Journal of Banking and Finane 30, 2659-2680. Ederingon, L.H., Lee, J.H., 1993. How markes proess informaion: releases and volailiy. Journal of Finane 48, 1161-1191. 22
Edison, H., 1997. Te reaion of exange raes and ineres raes o news releases. Inernaional Journal of Finane and Eonomis 2, 87-100. Ermann, M., Frazer, M., 2004. Equal size, equal role? Ineres rae inerdependene beween e euro area and e Unied Saes. European Cenral Bank Working Paper series 342., Engel, C., Frankel, J., 1984. Wy ineres raes rea o moneary announemens: An explanaion from e foreign exange marke Journal of Moneary Eonomis 13, 31-39. Federal Reserve Sysem, 2005. Te Federal Reserve Sysem: Purposes and Funions. Federal Reserve Sysem Websie. Flannery, M.J., Proopapadakis, E.H., 2002. Maroeonomi faors do influene aggregae sok reurns. Review of Finanial Sudies 15, 751-782. Flemming, M.J., Remolona, E.M., 1999a. Prie Formaion and Liquidiy in e U.S. Treasury Marke: Te responses o Publi Informaion. Journal of Finane 54, 1901-1915. Flemming, M.J., Remolona, E.M., 1999b. Wa moves e Bond marke? Journal of Porfolio Managemen 25, 28-38. Goldberg, L., Leonard, D., 2003. Wa Moves Sovereign Bond Markes? Te Effes of Eonomi on U.S. and German Yields. Curren issues in Eonomis and Finane, Federal Reserve Bank of New York 9, Gürkaynak, R.S., Sak, B., Swanson, E., 2005. Te sensiiviy of Long-Term Ineres raes o eonomi news: Evidene and Impliaions for Maroeonomi Models. Amerian Eonomi Review 95, 425-436. Haldane, A., Read, M., 2000. Moneary Poliy Surprises and e Yield Curve. Bank of England Working Paper No. 106 23
Hardouvelis, G., 1988. Eonomi news, exange raes and ineres raes. Journal of Inernaional Money and Finane 7, 23-35. Hess, D., 2004. Deerminans of e relaive prie impa of unaniipaed informaion in U.S. maroeonomi releases. Journal of Fuures Markes 24, 609-629. Jones, C.M., Lamon, O., Lumsdaine, R.L., 1998. Maroeonomi news and bond marke volailiy. Journal of Finanial Eonomis 47, 315-337. Kim, S.-J., MKenzie, M.D., Faff, R.W., 2004. Maroeonomi news announemens and e role of expeaions: evidene for U.S. bond, sok and foreign exange markes. Journal of Mulinaional Finanial Managemen 14, 217-232. Kim, S.-J., Nguyen, D.Q.T., 2009. Te Spillover Effes of Targe Ineres Rae from e U.S. Fed and e European Cenral Bank on e Asia-Paifi Sok Markes. Journal of Inernaional Finanial Markes, Insiuions and Money 9, 415-431. Kim, S.-J., Seen, J., 2000. Inernaional Linkages and maroeonomi news effes on ineres rae volailiy - Ausralia and e U.S.. Paifi Basin Finane Journal 8, 85-113. Krueger, J., Kuner, K., 1996. Te Fed funds fuures rae as a predior of federal reserve poliy. Journal of Fuures Markes 16, 865-879. Kuner, K., 2001. Moneary poliy surprises and ineres raes: evidene from e Fed funds fuures marke. Journal of Moneary Eonomis 47, 523-544. Li, L., Engle, R.F., 1998. Maroeonomi Announemens and Volailiy of Treasury Fuures. Universiy of California a San Diego, Eonomis Working Paper Series 98-27. MQueen, G., Roley, V.V., 1993. Sok pries, news and business ondiions. Review of Finanial Sudies 6, 683-707. Nelson, D.B., 1991. Condiional Heerosedasiiy in Asse Reurns: A new approa. Eonomeria 62, 1-41. 24
Ng, A., 2000. Volailiy spillover effes from Japan and e U.S. o e Paifi Basin. Journal of Inernaional Money and Finane 19, 207-233. Tornon, D.L., 1998. Tess of e marke s reaion o Federal Funds rae arge anges. Federal Reserve Bank of S. Louis Review 80, 25-36. 25
Table 1 Maroeonomi announemens Tis able repors number of releases, release imes and reporing soures of maroeonomi announemens from e U.S. (from Jan 1999 o De 2008), and e euro area (from Jan 2002 o De 2008). Daa release US's Maroeonomi announemens Euro area's Maroeonomi announemens Observaions Release Time (EST) Reporing auoriy Observaions Time (CET) Reporing auoriy Targe ineres raes 85 14.15 Federal Reserve Board 156 14.30 European Cenral Bank Gross Domesi Produ (GDP) 40 8.30 Bureau of Eonomi Analysis 84 11.00 Eurosa Consumer Prie Index (CPI) 120 8.30 Bureau of Labor Saisis 84 11.00 Eurosa Unemploymen (EU) 120 8.30 Bureau of Labor Saisis 84 11.00 Eurosa Trade defii (TD) 120 8.30 Bureau of Census, Bureau of Eonomi Analysis Toal 485 408 26
Table 2 Conurren announemens Tis able repors e frequeny of onurren announemens of maroeonomi variables from e U.S. and e euro area. Fed is e FOMC s arge rae announemen, ECB is e ECB s Governing Counil s arge rae announemen, GDP is e Gross Domesi Produ, CPI is e Consumer Prie Index, UE is e Unemploymen rae, and TD is e U.S. s Trade defii. Panel A Te U.S. s maroeonomi announemen marix Fed GDP CPI UE TD Fed funds arge raes 85 2 0 0 0 Gross Domesi Produ (GDP) 2 40 0 0 0 Consumer Prie Index (CPI) 0 0 120 0 10 Unemploymen (EU) 0 0 0 120 1 Trade defii (TD) 0 0 10 1 120 Panel B Te euro area s maroeonomi announemen marix ECB GDP CPI UE ECB's arge raes 156 11 0 0 Gross Domesi Produ (GDP) 11 84 3 0 Consumer Prie Index (CPI) 0 3 84 3 Unemploymen (EU) 0 0 3 84 27
Table 3 Te news spillover effes Tis able repors e esimaion resuls of e EGARCH models desribed in equaion (3) for e spillover effe of e U.S. s and e euro area s news on daily anges of e Asia- Paifi ineres raes. Fed is e FOMC s arge rae announemen, ECB is e ECB s Governing Counil s arge rae announemen, GDP is e Gross Domesi Produ, CPI is e Consumer Prie Index, UE is e Unemploymen rae, and TD is e U.S. s Trade defii. P-values are in pareneses. *, **, *** denoe signifiane a 10%, 5%, and 1%, respeively. ln y p = + α Lag, i y i + α Mon Mon + α 1 i= 1 α + 1 1 = β ln 1 1 2 Mon Mon 1 1 Panel A Te U.S. s news spillover effes Ausralia Hongkong Japan Korea New Zealand Singapore 3M 3Y 10Y 3M 3Y 10Y 3M 3Y 10Y 3M 5Y 10Y 3M 5Y 10Y 3M 5Y 10Y Condiional mean equaion 1 α Fed 0.0015 0.0029 0.0819 *** 0.0142 *** 0.0188 0.0258 0.0021 0.0024 0.0092 *** 0.0479 *** 0.0607 *** 0.0325 0.0037 0.0237 0.0341 0.0075 * 0.0045 * 0.0098 * (0.4111) (0.7738) (0.0000) (0.0000) (0.5651) (0.4127) (0.1762) (0.5562) (0.0000) (0.0000) (0.0030) (0.2477) (0.7537) (0.1334) (0.2240) (0.0966) (0.0900) (0.0698) α GDP 0.0085 *** 0.0091 * 0.0117 0.0012 *** 0.0049 0.0013 0.0019 ** 0.0009 0.0091 ** 0.0006 0.0200 *** 0.0054 * 0.0024 0.0057 0.0043 0.0003 ** 0.0044 0.0112 *** (0.0000) (0.0837) (0.3736) (0.0000) (0.4564) (0.8598) (0.0220) (0.7086) (0.0243) (0.3894) (0.0088) (0.0890) (0.4120) (0.2241) (0.5760) (0.0390) (0.1028) (0.0077) α CPI 0.0001 0.0016 0.0076 0.0139 ** 0.0131 ** 0.0014 0.0003 0.0029 ** 0.0004 0.0030 0.0038 0.0048 0.0020 *** 0.0097 * 0.0001 0.0038 * 0.0047 0.0010 (0.3778) (0.8311) (0.5981) (0.0318) (0.0335) (0.8264) (0.1121) (0.0211) (0.9906) (0.6344) (0.3824) (0.3188) (0.0000) (0.0705) (0.2520) (0.0696) (0.2324) (0.7489) α UE -0.0230 ** -0.0122-0.0563-0.7484 *** -0.1170-0.0704 *** 0.0026-0.0110 *** -0.0272 ** -0.0035-0.0008 *** -0.0159 0.0106-0.0329-0.0483 *** -0.0357 0.0061-0.0204 (0.0112) (0.7125) (0.2515) (0.0000) (0.2786) (0.0000) (0.4789) (0.0000) (0.0113) (0.5938) (0.0000) (0.7276) (0.7039) (0.3753) (0.0000) (0.1558) (0.8638) (0.4541) α TD 0.0001 0.0002 0.0007 *** 0.0011 0.0008 *** 0.0004 0.0001 *** 0.0002 0.0002 0.0001 *** 0.0004 0.0031 0.0008 *** 0.0006 ** 0.0004 0.0001 0.0003 0.0002 (0.5536) (0.6919) (0.0000) (0.4762) (0.0086) (0.1796) (0.0000) (0.6797) (0.9718) (0.0018) (0.7056) (0.3311) (0.0064) (0.0328) (0.7841) (0.7362) (0.1794) (0.1303) Condiional variane equaion β Fed -0.3385 0.0194 0.3653 *** 0.8070 * -0.0844 0.3607-2.1400 *** 0.0397 0.2266 1.8994 *** 0.7152 *** 0.9409 *** 0.7537-0.1970-0.1452-0.5343 ** 0.1160-0.0874 (0.3962) (0.9200) (0.0012) (0.0687) (0.8139) (0.3607) (0.0003) (0.5499) (0.2781) (0.0024) (0.0001) (0.0001) (0.3548) (0.2254) (0.4492) (0.0112) (0.6022) (0.6856) β GDP 0.4064 *** -0.0120-0.0897-0.4132 *** -0.1753-0.1006 ** -0.8650 *** 0.1781 0.2732 * -1.2942 *** 0.0108-0.3888 *** -0.2703 * -0.0540-0.1512 0.1280 0.1028 0.0817 (0.0000) (0.9420) (0.7036) (0.0000) (0.2720) (0.0443) (0.0000) (0.4111) (0.0978) (0.0000) (0.9282) (0.0005) (0.0681) (0.6546) (0.1403) (0.4961) (0.2582) (0.7954) β CPI -0.3068-0.0781 0.4221 0.1137 0.0094-0.3106 ** 0.0456 *** -0.0421-0.0739 ** 0.6344 *** -0.2973-0.4724 ** -0.2716-0.0926 *** -0.1613-0.1207 *** 0.0126-0.0821 (0.1094) (0.5805) (0.1129) (0.5172) (0.9264) (0.0319) (0.0000) (0.8377) (0.0394) (0.0000) (0.1120) (0.0439) (0.2437) (0.0001) (0.2385) (0.0061) (0.9343) (0.6883) β UE 2.3721 ** 0.9859-0.4289 2.4452 ** 1.4410 ** 2.3993 ** -0.0119-1.7833 ** -1.2576 ** -3.8487 ** 1.1860 *** -1.6148 1.3770 1.5059 *** 1.2434 * -1.3325 ** -0.3258-0.1753 (0.0122) (0.2042) (0.7484) (0.0104) (0.0361) (0.0438) (0.9862) (0.0168) (0.0461) (0.0212) (0.0000) (0.1258) (0.1289) (0.0000) (0.0599) (0.0424) (0.6533) (0.8274) β TD 0.0135 0.0059 0.0690 0.0194 0.0179 0.0021 0.1751 *** 0.0378 *** 0.0037 0.0602 *** 0.0049 0.0525 0.0280 0.0008 0.0061 0.0225 0.0084 0.0019 (0.2894) (0.3693) (0.3125) (0.5596) (0.3387) (0.8915) (0.0002) (0.0000) (0.8315) (0.0002) (0.6334) (0.3722) (0.4099) (0.9346) (0.1907) (0.1531) (0.5645) (0.9084) (3) 28
Table 3 Coninued Panel B Te euro area news spillover effes Ausralia Hongkong Japan Korea New Zealand Singapore 3M 3Y 10Y 3M 3Y 10Y 3M 3Y 10Y 3M 5Y 10Y 3M 5Y 10Y 3M 5Y 10Y Condiional mean equaion α ECB 0.0973 *** 0.3360 ** 0.2220 0.0141 *** 0.3192 0.2424 0.0069 * 0.0439 ** 0.1004 0.1122 * 0.0112 0.1771 0.0077 0.2045 * 0.2660 ** 0.0544 0.0421 0.1208 (0.0000) (0.0102) (0.3531) (0.0000) (0.2502) (0.1884) (0.0668) (0.0127) (0.2182) (0.0812) (0.9335) (0.2140) (0.8719) (0.0580) (0.0373) (0.1823) (0.5620) (0.2658) α GDP 0.0059 0.0060 0.0516 *** 0.0167 0.0335 0.0300 0.0013 0.0011 0.0003 0.0056 * 0.0187 0.0013 0.0213 0.0501 ** 0.0286 0.0033 0.0116 0.0435 *** (0.1595) (0.7583) (0.0000) (0.7352) (0.6316) (0.4984) (0.1373) (0.8812) (0.9835) (0.0752) (0.3432) (0.9407) (0.1699) (0.0117) (0.2057) (0.5586) (0.5148) (0.0001) α CPI 0.0138 * 0.0076 0.0064 0.0292 *** 0.0458 0.0091 0.0015 * 0.0025 0.0089 0.0011 0.0297 0.0456 * 0.0364 *** 0.0315 0.0245 0.0150 ** 0.0428 ** 0.0599 ** (0.0865) (0.8413) (0.7674) (0.0000) (0.3920) (0.8234) (0.0504) (0.7086) (0.5433) (0.7084) (0.3271) (0.0647) (0.0000) (0.1172) (0.2741) (0.0139) (0.0258) (0.0393) α UE -0.0504 * -0.1183-0.2087-0.1051-0.2183-0.2486 *** -0.0079 ** -0.0205 *** 0.0229-0.0213-0.1751 ** -0.1708 *** -0.0472-0.0643-0.0360-0.0723 * -0.1348 * -0.1682 (0.0527) (0.3481) (0.2795) (0.3147) (0.4728) (0.0078) (0.0474) (0.0000) (0.4736) (0.2372) (0.0255) (0.0045) (0.1664) (0.5121) (0.7734) (0.0994) (0.0556) (0.8191) Condiional variane equaion β ECB 1.0789 1.1496 0.6773-4.9668-1.2283 6.1795 *** 0.1870-2.2100 *** 0.9686 4.2322 *** 0.9814 2.7670 *** -2.6559 *** 0.2024 0.7579-1.8964-3.1848-3.2015 (0.4259) (0.5307) (0.1752) (0.4382) (0.8928) (0.0000) (0.5399) (0.0038) (0.6686) (0.0000) (0.7302) (0.0003) (0.0000) (0.4168) (0.3277) (0.5252) (0.1320) (0.3690) β GDP 1.1792 *** 0.2976 1.7987 * 0.8979 0.6900 1.2608-0.3093-0.9696 *** 0.3544-0.3053 *** -0.7262-0.3749 *** 0.4643 ** 0.5183 *** -0.0561-0.3624 *** 0.2899 0.0224 (0.0000) (0.4411) (0.0742) (0.5304) (0.7443) (0.2031) (0.7306) (0.0000) (0.4868) (0.0000) (0.3207) (0.0000) (0.0243) (0.0000) (0.9175) (0.0013) (0.6723) (0.9734) β CPI 0.2858 0.5862 3.2230 *** 0.5957 1.4861 1.2096 1.5085 *** 0.4275 1.2653 * 3.9511 ** 1.5008 * 2.6053 *** 2.2711 1.5576 *** 1.3983 ** 0.6550 0.9047 0.6408 (0.1704) (0.5030) (0.0000) (0.5546) (0.4651) (0.2567) (0.0000) (0.4751) (0.0513) (0.0152) (0.0522) (0.0000) (0.1267) (0.0015) (0.0241) (0.6871) (0.2966) (0.5473) β UE 1.1913 *** 0.1780 2.2218 5.0833 0.4687 0.3473 1.1277 1.6422 *** -2.9263 3.6503 *** 1.6805 2.7236 1.0140 *** 2.1900 * 2.1466 * 0.9101 0.5993 1.1578 (0.0000) (0.8986) (0.2038) (0.3131) (0.9261) (0.9099) (0.9815) (0.0000) (0.2812) (0.0004) (0.5127) (0.3368) (0.0000) (0.0810) (0.0863) (0.7594) (0.6979) (0.2941) 29
Table 4 Sandardized news effes Tis able repors e esimaion resuls of e EGARCH models desribed in equaion (5) for e spillover effe of e U.S. s and e euro area s news on daily anges of e Asia- Paifi ineres raes using Balduzzi e al. (2001) sandardized news measures sown in equaion (4). Fed is e FOMC s arge rae announemen, ECB is e ECB s Governing Counil s arge rae announemen, GDP is e Gross Domesi Produ, CPI is e Consumer Prie Index, UE is e Unemploymen rae, and TD is e U.S. s Trade defii. P- values are in pareneses. *, **, *** denoe signifiane a 10%, 5%, and 1%, respeively. ln y p = + α Lag, i y i + α Mon Mon + α S 1 i= 1 α + 1 1 = β ln 1 1 2 Mon Mon S 1 1 Panel A Te U.S. s sandardized news effes Ausralia Hongkong Japan Korea New Zealand Singapore 3M 3Y 10Y 3M 3Y 10Y 3M 3Y 10Y 3M 5Y 10Y 3M 5Y 10Y 3M 5Y 10Y Condiional mean equaion 1 α Fed 0.0001 0.0001 0.0044 *** 0.0021 * 0.0021 * 0.0014 0.0001 *** 0.0007 0.0005 *** 0.0026 *** 0.0033 *** 0.0002 0.0002 0.0013 0.0018 *** 0.0004 0.0002 0.0005 (0.4841) (0.8281) (0.0002) (0.0908) (0.0708) (0.2579) (0.0000) (0.7083) (0.0000) (0.0000) (0.0040) (0.7566) (0.7566) (0.1307) (0.0024) (0.4478) (0.7490) (0.4723) α GDP 0.0012 ** 0.0016 ** 0.0016 0.0003 * 0.0007 0.0011 0.0005 *** 0.0001 0.0013 ** 0.0001 0.0028 *** 0.0003 0.0003 0.0007 0.0006 0.0004 ** 0.0006 0.0016 *** (0.0162) (0.0183) (0.2958) (0.0960) (0.4148) (0.2526) (0.0012) (0.6903) (0.0358) (0.4144) (0.0001) (0.3487) (0.3487) (0.3856) (0.4677) (0.0338) (0.1805) (0.0091) α CPI 0.0003 0.0001 0.0012 0.0022 ** 0.0021 ** 0.0002 0.0001 0.0005 ** 0.0001 0.0004 0.0006 0.0003 0.0003 0.0016 * 0.0009 0.0008 * 0.0008 0.0002 (0.4541) (0.9409) (0.5004) (0.0374) (0.0335) (0.8135) (0.8016) (0.0354) (0.9890) (0.6557) (0.4633) (0.2968) (0.2968) (0.0683) (0.3146) (0.0841) (0.2157) (0.8053) α UE -0.0006 * -0.0007-0.0015-0.0004 *** -0.0030-0.0018 ** -0.0001-0.0003 ** -0.0007 * -0.0001-0.0002-0.0206 *** -0.0003-0.0009-0.0013 *** -0.0009-0.0002-0.0005 (0.0940) (0.6525) (0.3003) (0.0002) (0.1747) (0.0296) (0.4092) (0.0201) (0.0675) (0.5552) (0.8156) (0.0000) (0.6190) (0.3666) (0.0001) (0.2059) (0.8277) (0.5634) α TD 0.0001 0.0003 0.0010 *** 0.0014 0.0010 *** 0.0005 0.0001 *** 0.0003 0.0005 0.0002 *** 0.0005 0.1027 0.0011 0.0008 ** 0.0008 *** 0.0002 0.0004 0.0003 (0.1365) (0.6165) (0.0000) (0.3801) (0.0061) (0.9657) (0.0000) (0.6502) (0.8069) (0.0000) (0.7535) (0.4075) (0.1245) (0.0357) (0.0000) (0.7145) (0.2423) (0.1159) Condiional variane equaion β Fed -0.0182 0.0010 0.0199 ** 0.0470 *** -0.0054 0.0194-0.1226 *** 0.0022 0.0122 0.1023 *** 0.0384 *** 0.0176 0.0176-0.0106-0.0078-0.0287 0.0062-0.0047 (0.3024) (0.9293) (0.0156) (0.0000) (0.7537) (0.2007) (0.0000) (0.5530) (0.2828) (0.0001) (0.0001) (0.4550) (0.4550) (0.1405) (0.2203) (0.2969) (0.5535) (0.6956) β GDP 0.0572 ** 0.0040-0.0126-0.0591 *** -0.0247-0.0115-0.0532 *** 0.0250 0.0383 * -0.1729 *** 0.0015-0.0380 * -0.0380 * -0.0077-0.0212 0.0177-0.0230 0.0115 (0.0126) (0.7796) (0.7415) (0.0000) (0.2061) (0.7710) (0.0000) (0.8441) (0.0921) (0.0000) (0.9239) (0.0511) (0.0511) (0.6347) (0.1917) (0.4345) (0.2349) (0.7381) β CPI -0.0485 * 0.0600 *** 0.0670 0.0179 0.0015-0.0493 ** 0.0702 *** -0.0067-0.0117 0.3133 * -0.0472 * -0.0435-0.0435-0.0146 *** -0.0257-0.0194-0.0020-0.0130 (0.0547) (0.0004) (0.1098) (0.5772) (0.9441) (0.0336) (0.0000) (0.8526) (0.5298) (0.0794) (0.0971) (0.1805) (0.1805) (0.0002) (0.2526) (0.5487) (0.9330) (0.6753) β UE 0.0616 *** 0.0257-0.0111 0.0635 * 0.0374 ** 1.9712 ** -0.0106-0.0466 *** -0.0327 ** -0.0999 ** 0.0310-1.9519 0.0358 0.0390 ** 0.0324-0.0346 ** -0.0085-0.0045 (0.0000) (0.1628) (0.7168) (0.0751) (0.0272) (0.0335) (0.5166) (0.0000) (0.0246) (0.0136) (0.1592) (0.2508) (0.1377) (0.0172) (0.3140) (0.0265) (0.6219) (0.8236) β TD -0.0185-0.0077-0.0929-0.0262-0.0240-0.0028 0.5740 *** 0.0509 ** -0.0053-0.0884 *** -0.0065 0.8524 * 0.0378 *** -0.0010-0.0087 0.0298 0.0113 0.0025 (0.1553) (0.5445) (0.3408) (0.5025) (0.3010) (0.8638) (0.0000) (0.0215) (0.8607) (0.0004) (0.6566) (0.0600) (0.0000) (0.9251) (0.3822) (0.1014) (0.5508) (0.9117) (5) 30