Macroeconomic Surprises and International Financial Market Returns

Transcription

1 Inernaional Journal of Business and Social Science Volume 8 Number 8 Augus 2017 Macroeconomic Surprises and Inernaional Financial Marke Reurns Kyung-Chun Mun School of Business Truman Sae Universiy 100 E. Normal Kirksville, MO 63501, USA Absrac This paper invesigaes he simulaneous effec of macroeconomic developmens on sock and foreign exchange (FX) marke reurns in a sysem ha is characerized by dynamic ineracion among asse reurns. Using US and UK daa, we find ha US sock marke reurns are significanly responsive o domesic macroeconomic developmens in oupu growh, ineres raes, and unemploymen rae. US sock marke reurns are also responsive o UK surprises in money growh and inflaion and he surprises in FX markes such as he ones in ineres raes and rade balance. We also find ha he dollar/pound exchange rae is asymmerically responsive o he money growh surprise in US and UK and in he FX marke as well. Keywords: Macroeconomic Surprise; Sock Marke; Foreign Exchange Marke 1. Inroducion Macroeconomic surprises ha originae from a counry impac boh sock and foreign exchange (FX) markes of ha counry and hen can be ransmied o he corresponding markes of oher counries. As documened by Anderson, Bollerslev, Diebold, and Vega (2007), sock reurn and exchange rae dynamics are joinly linked o macroeconomic fundamenals and he linkage is significanly conemporaneous beween sock and FX markes. Thus, i is imporan o recognize ha he marke response o macroeconomic developmens should be modeled wihin a simulaneous framework linking sock and FX markes. Exan research on he relaion beween macroeconomic surprises and subsequen marke responses have focused on he effec of he surprises eiher on he sock marke or on he FX marke in a separae framework wihou linking he wo markes. For example, one class of he sudies focuses on he connecion beween macroeconomic surprises and subsequen movemens in sock prices (see, for example, Jones and Kaul (1996), Errunza and Hogan (1998), Flannery and Proopapadakis (2002), Boyd e al. (2005), Basisha and Kurov (2008), and Gilber (20)), while he oher class of he sudies examines he influence of macroeconomic surprises on he exchange rae movemen (see, for example, Good har e al. (1993), Tanner (1997), Almeida e al. (1998), Anderson, Bollerslev, Diebold, and Vega (2003), Evans and Lyons (2003), Simpson e al. (2005), Bergin (2006), Chen and Gau (2010)). This isolaed analysis of only a paricular marke response ignores cross-marke informaion effecs of macroeconomic surprises and he resuls may no simulaneously hold rue. In conras o he separae approach, he simulaneous response of sock and FX markes o macroeconomic surprises has been largely ignored. Noable excepions are hepapers by Anderson, Bollerslev, Diebold, and Vega (ABD&V, 2007) and Mun (2012), for which FX markes as well as he domesic and foreign sock markes are characerized by he join response o US macroeconomic surprises. ABD&V (2007) measures he magniude of response coefficiens using wo-sep weighed leas squares (WLS) procedure wih error erms following ARCH process and find ha bad news has a negaive impac on sock markes during economic conracions, bu a posiive impac during expansions. Mun (2012) invesigaes he join response of sock and FX markes o macroeconomic surprises using US and Japanese daa. Ye, hese sudies neglec he possibiliy ha macroeconomic surprises can affec he volailiy of sock reurns and exchange raes, which in urn feeds hrough o changes in sock reurns and exchange raes via increased risk premia. The purpose of our paper is o invesigaes he simulaneous response of sock and FX marke reurns o macroeconomic surprises in a sysem ha is characerized by dynamic ineracion among asse reurns. 1

2 ISSN (Prin), (Online) Cener for Promoing Ideas, USA In paricular, we develop a unified framework by inegraing wo srands of he lieraure on marke response o macroeconomic surprises: sudies on he sock marke response o macroeconomic surprises and sudies on he FX marke response o macroeconomic surprises. Our empirical sraegy is based on he vecor-auoregressive model wih error erms following GARCH-M process ha allows for asymmeries of marke responses. We exend previous research by esimaing he model for sock marke reurns and exchange rae changes in which he endogenous variables are allowed o be simulaneously affeced by macroeconomic developmens no only a home bu also abroad. The invesigaion of he impac of foreign macroeconomic surprises on US sock and FX markes can provide imporan evidence on several hypoheses abou he cross-counry and cross-marke ransmission of macroeconomic surprises. In our model, sock marke reurns and exchange rae movemens are simulaneously influenced by surprises of macroeconomic variables common in boh US and foreign economy. We use US and UK daa for six macroeconomic variables common in boh counries: (i) money growh, (ii) oupu growh, (iii) inflaion, (iv) ineres raes, (v) unemploymen rae, and (vi) rade balance. Esimaion resuls indicae ha o he exen ha revelaion of he surprises are direcly impounded in reurns, US sock marke reurns are significanly responsive o domesic macroeconomic developmens in oupu growh, ineres raes, and unemploymen rae. Also, US sock marke reurns are responsive o UK surprises in money growh and inflaion and he surprises in FX markes such as he ones in ineres raes and rade balance. This indicaes ha some of macroeconomic developmens in UK and in FX markes feed hrough o US sock markes. We find ha he dollar/pound exchange rae is direcly and asymmerically responsive o he money growh surprise in US and UK and in he FX marke as well. The resuls indicae ha a higher-han-expeced money growh in US leads o weaker dollar (or sronger pound) han a lower-han-expeced money growh of he same magniude. The paper is organized as follows. Secion 2 provides daa and empirical models o be esimaed. Secion 3 presens empirical resuls for effecs of macroeconomic surprises. Finally, conclusions are given in secion Daa and Mehodology 2.1 Daa To invesigae he simulaneous impac of a macroeconomic surprise on he sock and FX marke, i is necessary o choose a se of facors ha are common in boh sock and FX markes. By examining facors previously idenified in he lieraure, we selec he following se of macroeconomic facors as a fair and parsimonious represenaion of he macroeconomic fundamenals ha can influence boh sock and FX markes commonly: (i) money growh; (ii) oupu growh; (iii) inflaion; (iv) ineres raes; (v) unemploymen rae; and (vi) rade balance. For our analysis, we use he seasonally-adjused M1, he Consumer Price Index (CPI), he seasonally-adjused Indusrial Producion Index (IPI), and 3-monh T-bill raes as proxies for he money supply, price level, oupu, and ineres raes, respecively. 1 The sock marke indices used are he S&P 500 Index and he FTSE 100 Index of UK. We obain hese daa from Inernaional Financial Saisics and Global Financial Daabase (GFD). The daa se we analyze is he monhly series (excep for rade balance for which i is he quarerly series) for he period from December 1988 o December This sample period is consisen wih he so-called Grea Moderaion, during which period oupu, inflaion, invesmen, and oher macroeconomic variables were sharply less volaile across G7 counries han hey were during he prior 20-year period (see, for insance, Jusiniano and Primiceri (2008)). 2 This shif in he volailiy of macroeconomic variables owards sabiliy during he Grea Moderaion reduces possible abrup breaks in he saisical esimaion of macroeconomic shocks and conribues o an increase in he precision of forecass of he macroeconomic variables (Sock and Wason (2002 and 2003) and Jusiniano and Primiceri (2008)).Sock marke reurns are compued as he difference of he logarihm of he sock marke index. Exchange raes are expressed as US dollar per uni of foreign currency. Money growh, inflaion, and oupu growh are measured as he difference of he logarihm of he M1 (M4 for UK), he consumer price index, and he oupu growh index, respecively. Oher economic series such as ineres raes and unemploymen rae are logged. The rade balance is scaled as a raio o he gross domesic produc (GDP). 1 We use he seasonally-adjused M4 running from June 1989 o December 2006, defined by he Inernaional Financial Saisics,for UK money supply. 2 Alhough here is no consensus on is exac duraion, he beginning of he Grea Moderaion is ofen suggesed o be he mid 1980s (Sock and Wason (2002)). Economiss generally agree ha he Grea Moderaion ended wih he collapse of subprime morgages in

3 Inernaional Journal of Business and Social Science Volume 8 Number 8 Augus Empirical Model Developmens in sock markes are known o influence foreign exchange markes and vice versa, and inerrelaionships exis among inernaional sock markes. This implies ha a macroeconomic surprise in one counry can simulaneously affec boh sock and FX markes a home, and can be ransmied o oher counries. To saisically capure he cross-counry ransmission and simulaneous response of sock and FX markes o a macroeconomic developmen, one should employ a sysem mehod of esimaion. Also, given ha risk-averse agens require compensaion for holding a risky asse, asse reurns should depend on he risk premium which is an increasing funcion of he condiional variance of reurns. One model ha can effecively incorporae hese phenomena is he vecor auoregressive model wih error erms following GARCH in-mean process. Following Mun (2012), he specificaion we use for he means of sock marke (boh US and UK) reurns and exchange rae changes is given by: Y = ~ N(0, p iy i h i1 H ) Z ( DZ ) (1) H = CC K 1 1K F H 1F (2) Where r ( i) ( i) ( i) r a a12 a Y = r ; = ( i) ( i) ( i) r ; i = a a22 a ; = 22 s ( i) ( i) ( i) s a a a z = 22 ; = 22 ; Z = z ; z z rr, I3 if he surprise is negaive D = = r r, 03 oherwise ss, hrr, h h h rr, rs, rr, H = h h h ; h r r, r r, r s, = h r r, ; hsr, h h sr, ss, h ss, c c12 c13 k k12 k13 f f12 f13 C= 0 c22 c ; K = k k22 k ; F = f f22 f ; 0 0 c 33 k k k 33 f f f 33 where r is he logarihmic reurn on he US sock marke. r r r 1; r 1 s s 1 r r s, r is he logarihmic reurn on he UK sock marke., he appreciaion rae of he UK pound relaive o he US dollar a ime. s is he exchange rae expressed in US dollars per uni of he UK pound a ime. z ( z ) = he US (UK) macroeconomic surprise a ime. ( z z ) = he surprise of he differenial in macroeconomic developmens beween US and UK a ime. This represens he surprise in FX markes associaed wih macroeconomic developmens. 3

4 ISSN (Prin), (Online) Cener for Promoing Ideas, USA D = he dummy variable marix ha akes on he (3 3) ideniy marix, I 3, if he surprise is negaive or he (3 3) null marix, 03, oherwise. In equaion (1), he condiional mean of US sock marke reurns is described as a funcion of he pas hisory of sock reurns for boh domesic & foreign markes and exchange rae changes, he pas hisory of error erms given by rr,,, and r r, ss,, and heir condiional variances given by h rr,, h, r r, and h ss,, respecively. Thus, his model explicily considers possible reurn and volailiy ransmissions among he hree endogenous variables. The condiional means of UK sock marke reurns and exchange rae changes are also represened by he pas hisory of he hree variables and heir error erms, and heir condiional sandard deviaions. Equaions (1) and (2) presen ha surprises emanaing from muliple counries (he US and UK in his sudy) as well as he FX marke are allowed o simulaneously influence he sock marke a home and abroad and he FX marke. Specifically, he off-diagonal parameers of he (3 3) marix in equaion (1) effecively capure possible cross-counry and cross-marke ransmission of he macroeconomic surprise. For example, he saisically significan value of implies ha a US macroeconomic surprise influences exchange rae movemens, while he significan value of 13 indicaes ha a macroeconomic surprise in he FX marke influences US sock marke reurns. The model can be esimaed using maximum likelihood mehods given ~ N(0, H ), where 4 is he informaion se available a ime. The BEKK (Engle and Kroner (1995)) H, posiive definie for all values parameerizaion of equaion (2) guaranees he condiional covariance marix, of. The off-diagonal parameers of marices of i and K in equaions (1) and (2) allow for ess for various hypoheses concerning cross-marke spillovers. For example, joinly significan values of () i a and a in marix i imply ha here are reurn spillovers from sock o FX markes. Tess for volailiy spillovers from one marke o anoher can be performed employing he off-diagonal parameers of marix K in equaion (2). For example, he join hypohesis of k = k =0 implies ha here is no volailiy spillover from sock o FX markes.the (3 1) vecor of z in equaion (1) capures surprises induced by macroeconomic developmens. Since asse reurns/exchange raes are affeced only by an unanicipaed change in he level of economic variables, he value of he surprise in equaion (1) is measured as follows: z = Acual Expeced (3) Where = he sandard deviaion of he macroeconomic surprise a ime. The values of z for macroeconomic surprises are saisically esimaed using he VAR model wih he following forma: y = a p i1 by i i e (4) Where y = he vecor for macroeconomic variables for each counry and he FX marke; a = he 3 1 vecor of consans; b = he 33 parameer marices; and e i = he 3 1 vecor of errors. The ime pah of residuals and heir sandard deviaion are obained from equaion (4).We ake he surprise o be posiive (negaive) if he acual value of a macroeconomic variable is greaer (less) han is expeced value. For he FX marke, we posulae he surprise o be posiive (negaive) if he acual value of he erms of rade raio beween he US and UK is larger (smaller) han expeced. To capure possible asymmeric responses of sock and FX markes o macroeconomic surprises, a dummy variable is added o he equaion (1) as an addiional exogenous variable. Specifically, he diagonal parameers of he (33) marix in equaion (1) capure poenial asymmeric responses of sock and FX markes o macroeconomic surprises. For example, significanly negaive values of for oupu growh imply ha lower-han-expeced US oupu growh leads o lower domesic sock marke reurns by more han equivalen posiive surprises increase he reurns. The off-diagonal parameers of he marix capure possible ransmission of negaive surprises from one marke or counry o he oher marke or counry. () i

5 Inernaional Journal of Business and Social Science Volume 8 Number 8 Augus 2017 Table1. Hypohesized Effecs of Macroeconomic Surprises on Sock and FX Markes Macroeconomic facor surprise US sock marke Sock marke of foreign counerpar FX marke Money supply b. Foreign counerpar c. Differenial beween US and foreign counerpar Indusrial Producion b. Foreign counerpar c. Differenial beween US and foreign counerpar Inflaion b. Foreign counerpar c. Differenial beween US and foreign counerpar Ineres raes b. Foreign counerpar c. Differenial beween US and foreign counerpar Unemploymen rae b. Foreign counerpar c. Differenial beween US and foreign counerpar Trade balance b. Foreign counerpar c. Differenial beween US and foreign counerpar As described in Table 1, we nex hypohesize for our own purposes ha he surprise is posiive if oupu growh is higher han expeced and negaive if i is lower han expeced. Similarly, he surprise is negaive if money growh, inflaion, ineres raes, unemploymen rae, or rade defici is higher han expeced. Earlier sudies have documened ha if US inflaion is unexpecedly higher han UK inflaion, he dollar/pound exchange rae rises or equivalenly US dollar depreciaes relaive o he UKpound. If US ineres raes are unexpecedly higher han UK ineres raes, US dollar appreciaes. This is because exising models (excep for he flexible-price moneary model) for exchange rae deerminaion predic ha higher domesic ineres raes relaive o foreign ineres raes imply an appreciaion of domesic currency. 3. Empirical Resuls 3.1 Summary Saisics Table 2 repors summary saisics for various economic variables. Panel A of Table 2 shows ha over he sample period. US sock marke reurns were higher han UK sock reurns and he dollar/pound exchange rae increased by 0.19% monhly, implying depreciaion of he dollar relaive o he pound. Panel B presens ha US had higher levels in oupu growh, inflaion, and rade defici relaive o UK. During he sample period US had an ineres rae of 4.72% while UK had 7.42%. Panel C repors he magniude of macroeconomic surprises and he direcion of surprises over he sample period. US had higher-han-expeced levels in money growh, inflaion, and unemploymen while having lower-han-expeced levels in oupu growh, ineres raes, and rade balance. On he oher hand, UK had higher-han-expeced levels in money growh, oupu growh, unemploymen, and rade balance while having lower-han-expeced levels in inflaion and ineres raes. In paricular, posiive money growh surprise was he larges in magniude in US while negaive ineres rae surprise was in UK. 5

6 ISSN (Prin), (Online) Cener for Promoing Ideas, USA Table2. Summary Saisics A. Mean of Sock Index Reurns and Exchange Rae Changes (12/ /2013) US UK Sock index reurns (3.3885) (3.8765) Exchange rae changes (2.4950) Noe: Figures in parenheses are sandard deviaions Counry Money growh US (0.7977) UK (7.8145) B. Mean of Macroeconomic Facors (12/ /2013) Oupu growh (0.09) (0.1020) Inflaion (0.2654) (0.4427) Ineres raes 4.72 (0.9672) (3.1366) Unemploy-men rae (0.9914) 7.44 (2.2007) Trade balance (0.3789) (0.4156) Noe: Money growh, oupu growh, and inflaion are monhly daa. Ineres raes and unemploymen raes are annualized. Trade balance is scaled as a raio o he gross domesic produc (GDP) and is a quarerly daa. Figures in parenheses are sandard deviaions. C. Mean of Macroeconomic Surprises (1 2/ /2013) Counry Money growh Oupu growh Inflaion Ineres raes Unemploymen rae Trade balance US (0.9998) (1.0018) 0.09 (0.9990) (1.0019) (1.0019) (1.0049) UK (1.0020) (1.0017) (1.0088) (1.0338) (1.0005) 0.08 (1.0126) Differenial surprises for US/UK US/UK (1.0025) (1.0014) (1.0447) (1.0428) (1.0020) (1.0101) Noe: Figures in parenheses are sandard deviaions. 3.2Effecs of Macroeconomic Surprises on Sock and FX Markes In his secion, we examine he effecs of US and UK macroeconomic surprises on US and UK sock markes as well as FX markes o he exen ha hese surprises are impounded in reurns. Table 3 presens he resuls of esimaing equaion (1) for various macroeconomic surprises. Because he model in equaions (1) and (2) conains so many variables, i would be counerproducive o repor all he parameer esimaes. Insead, we presen key parameer esimaes ha are of imporance for invesigaing he marke response in reurns direcly o macroeconomic surprises. Y = p iy i h i1 H = CC K 1 1K F H 1F Z ( DZ ) 6

7 Inernaional Journal of Business and Social Science Volume 8 Number 8 Augus 2017 Parameer Money Oupu growh growh (-1.308) (2.142) (2.229) (1.666) (1.643) (1.553) (-1.180) (-0.552) (-2.144) (0.537) (1.1) (0.488) (4.220) (-0.741) (-3.361) (-1.047) (5.4) (0.967) (-0.629) (-4.565) (-0.5) (0.410) (-1.745) (-1.944) (0.775) (2.272) (-1.525) (-0.0) (-1.351) (0.414) (4.689) (-0.293) (-10.84) (-0.8) (16.73) (0.514) Noe: Figures in parenheses are -saisics Response o Money Growh Surprises Table3. Esimaion Resuls of Direc Effecs Inflaion (-1.934) (2.065) (0.303) (-2.7) (-2.250) (-0.2) (-0.362) (-0.291) (0.384) (-1.771) (-0.5) (-0.209) (-2.7) 0.01 (2.139) (1.192) (0.100) (0.206) (-0.575) Ineres raes (-2.477) (-0.149) (0.471) (-0.174) (0.173) (0.865) (1.481) (1.196) (0.913) (-2.253) (0.438) (0.390) (-0.7) (-0.334) (-0.777) (-0.701) (-0.530) (-0.686) Unemploymen (-2.018) (0.668) (1.882) (-2.030) (-1.068) (1.690) (-0.818) (-1.335) (2.060) (-2.240) (-2.207) (-1.103) 0.07 (1.603) (-0.554) (-0.987) 0.00 (0.813) (-0.918) (-1.144) Trade balance (1.357) (-0.701) (-0.1) (2.227) (0.352) (-1.430) (-0.425) 0.00 (1.827) (-0.4) (-1.752) (0.402) 0.05 (0.694) (-1.254) (0.624) (1.019) (-0.670) (-1.499) (1.077) The parameer esimaes from down o 13 for money growh in Table 3 indicae ha US sock marke reurns are affeced by moneary surprises in UK bu no by domesic and FX marke developmens. This implies ha while he moneary surprise in UK feeds hrough o US sock markes, i is no direcly impounded in domesic sock reurns and has limied informaion value for sock marke paricipans abou he fuure course of he economy or economic policies. In conras, UK sock reurns appear o be direcly affeced by domesic moneary surprise. The dollar/pound exchange rae is significanly affeced by boh US and UK money surprises. For example, he parameer esimaes of and for moneary growh indicae ha he US moneary surprise has a significanly posiive direc impac on he dollar/pound exchange rae, while he UK moneary surprise is significanly negaively relaed o he exchange rae, implying a US dollar depreciaion and a pound appreciaion. This is, of course, consisen wih basic heoreical predicions. I is noeworhy ha he dollar/pound exchange rae is posiively affeced by he moneary surprise in FX markes, implying ha he US moneary surprise is dominanly affecing he dollar/pound exchange rae relaive o he UK surprise. 7

8 ISSN (Prin), (Online) Cener for Promoing Ideas, USA Also, he significance of,, and 33 for money growh indicaes ha exchange raes are asymmerically responsive o US and UK moneary surprises. For example, he significanly posiive value of indicaes ha a higher-han-expeced US money growh leads o higher dollar/pound exchange raes han a lower-han-expeced US money growh of he same magniude. Noe ha a posiive coefficien in he regressions implies ha negaive surprise is associaed wih a depreciaion of he dollar (or an appreciaion of he pound). Similarly, he values of (significanly negaive) and 33 (significanly posiive) for money growh sugges ha he dollar/pound exchange rae also displays an asymmeric response o he UK moneary surprise and he moneary surprise in he FX marke Response o Oupu Growh Surprises As suggesed by he parameer esimae of for oupu growh, US sock marke reurns are significanly posiively relaed o domesic oupu growh surprises. This esimaed response is consisen wih marke paricipans assessmens of a sronger economy wih higher-han-expeced oupu growh. All of he off-diagonal parameer esimaes of marix for inflaion are saisically insignifican, implying ha here is no direc crossmarke and cross-counry feedback relaion in response o an oupu surprise. I is also shown in Table 4 for oupu growh ha boh US and UK sock marke reurns appear o be asymmerically responsive o US oupu surprises bu in he opposie direcion. Tha is, a lower-han-expeced US oupu growh is followed by lower sock marke reurns in US bu higher reurns in UK han a higher-han-expeced US oupu growh of he same magniude. Thus, i appears ha he good (bad) surprise in US oupu growh is received in UK sock markes as a bad (good) surprise. In he FX marke, he oupu growh surprise eiher in US or in UK does no seem o direcly affec exchange raes Response o Inflaion Surprises The parameer esimaes of 12,, and for inflaion reveal ha UK sock marke reurns are affecing US sock markes posiively bu negaively responsive o domesic and US inflaion surprises and he response is asymmeric. The US inflaion surprise appears o be ransmied o he UK sock marke and a higher-hanexpeced US inflaion surprise leads o lower UK sock marke reurns han a lower-han-expeced US inflaion of he same magniude. This indicaes ha here is a direc cross-marke and cross-counry feedback relaion in response o an inflaion surprise Response o Oupu Growh Surprises The parameer esimaes of and for ineres raes sugges ha he US ineres rae surprise has a negaive and direc impac on US sock marke reurns and he marke response displays own asymmery. Tha is, a higherhan-expeced US ineres rae is followed by lower US sock marke reurns han a lower-han-expeced US ineres rae of he same magniude. I appears ha here is no direc cross-marke and cross-counry feedback relaion in response o an ineres rae surprise. In he FX marke, he inflaion surprise eiher in US or in UK does no seem o direcly affec exchange raes Response o Unemploymen and Trade Balance Surprises As evidenced by he parameer esimaes of and for unemploymen, he US unemploymen surprise has a negaive effec on US sock marke reurns and is ransmied direcly o UK sock marke reurns. In he FX marke, he unemploymen surprise eiher in US or in UK does no seem o direcly affec exchange raes, while he unemploymen surprise in he FX marke is significanly posiively relaed o he dollar/pound exchange rae. The significanly posiive value of 33 for unemploymen indicaes ha he exchange rae is dominanly affeced by he US unemploymen surprise relaive o UK surprise. Also, he significance of and 12 for unemploymen indicaes ha US sock marke reurns are negaively and direcly responsive o unemploymen surprises boh in US and UK. For example, he significanly negaive value of indicaes ha a higher-hanexpeced US unemploymen leads o lower US sock marke reurns han a lower-han-expeced unemploymen of he same magniude. 8

9 Inernaional Journal of Business and Social Science Volume 8 Number 8 Augus 2017 I appears ha UK sock marke reurns are posiively responsive o heus rade balance surprise, alhough no responsive direcly o he domesic rade balance surprise. In he FX marke, he rade balance surprise eiher in US or in UK does no seem o direcly affec exchange raes. 4. Conclusions In his sudy we invesigae he simulaneous effec of macroeconomic developmens on sock and foreign exchange (FX) marke reurns in a sysem ha is characerized by dynamic ineracion among asse reurns. Unlike exan research on he marke response o macroeconomic surprises. we esimae he model for sock marke reurns and exchange rae movemens in which he endogenous variables are allowed o be simulaneously affeced by macroeconomic developmens no only a home bu also abroad. We use US and UK daa for six series of macroeconomic variables common in boh counries: money growh, oupu growh, inflaion, ineres raes, unemploymen and rade balance. Evidence presened in his paper indicaes ha US sock marke reurns are significanly responsive o domesic macroeconomic developmens in oupu growh, ineres raes, and unemploymen raes o he exen ha revelaion of hose surprises is impounded in reurns. In paricular, as measured by he magniude of he parameer esimaes, he US ineres rae developmen appears o be he surprise causing he larges direc response o US sock marke reurns, followed by he surprise in unemploymen and oupu growh. US domesic macroeconomic developmens such as money growh, inflaion, and rade balance are no significanly relaed o US sock marke reurns. We find ha US sock reurns are responsive o UK macroeconomic developmens in money growh and inflaion surprises and UK sock reurns are significanly and direcly responsive o US macroeconomic developmens in inflaion, unemploymen, and rade balance. This indicaes ha some of US and UK macroeconomic surprises feed hrough o counerpary s sock markes. UK sock marke reurns are significanly responsive o domesic developmens in money growh and inflaion, wih inflaion surprise causing he larges response. The dollar/pound exchange rae is significanly responsive o he money growh surprise in US and UK and in he FX marke as well. The response of exchange raes o moneary surprises is significanly asymmeric. References Almeida, A., C. Goodhar, and R. Payne The effecs of macroeconomic news on high frequency exchange rae behavior, Journal of Financial and Quaniaive Analysis 33, Anderson, T.G., Bollerslev T., Diebold, F.X., &Vega C. (2003). Micro effecs of macro announcemens: Real-ime price discovery in foreign exchange, American Economic Review, 93, Anderson, T.G., Bollerslev T., Diebold, F.X., &Vega C. (2007). Real-ime discovery in global sock, bond and foreign exchange markes, Journal of Inernaional Economics, 73, Basisha, A. &Kurov, A. (2008). Macroeconomic cycles and he sockmarke s reacion omoneary policy. Journal of Banking and Finance,, Bergin, P.R. (2006). How well can be he new open economy macroeconomics explain he exchange rae and curren accoun?journal of Inernaional Money and Finance, 25, Boyd, J.H., Hu J., &Jagannahan R. (2005). The sock marke s reacion o unemploymen news: Why bad news is usually good for socks, Journal of Finance, 60, Engle, R.F. &Kroner K.(1995). Mulivariae simulaneous generalized ARCH.Economeric Theory,, Errunza, V. &Hogan K. (1998). Macroeconomic deerminaion of European sock marke volailiy, European Financial Managemen, 4, Evans, M.D. & Lyons, R.K.(2008). How is macro news ransmied o exchange raes? Journalof Financial Economics, 88, Flannery, M.J. &Proopapadakis A.A. (2002). Macroeconomic facors do influence aggregae sock reurns, Review of Financial Sudies, 15, Gilber, T.(20). Informaion aggregaion around macroeconomic announcemens: Revisions maer. Journal of Financial Economics, 101, 4-1. Goodhar, C.A.E., Hall, S.G., Henry S.G.B., &Pesaran B. (1993). News effecs in a high-frequency model of he serling-dollar exchange rae, Journal of Applied Economerics, 8, Jones, C.M.&Kaul G.(1996). Oil and he sock markes, Journal of Finance, 51, Jusiniano, A.&Primiceri, G.(2008). The ime varying volailiy of macroeconomic flucuaions. American Economic Review, 98, Mun, K.C. (2012). The join response of sock and foreign exchange markes o macroeconomic surprises: Using US and Japanese daa, Journal of Banking and Finance,36, Simpson, M., Ramchander S.,&Chaudhry M. (2005). The impac of macroeconomic surprises on spo and forward foreign exchange markes, Journal of Inernaional Money and Finance, 24, Sock, J.H. & Wason M.W.(2002). Has he business cycle changed and why? In:Gerler, M., Rogoff, K. (Eds.), NBER Macroeconomic Annual, MIT Press: Cambridge, MA. Sock, J.H. & Wason M.W.(2003). Has he business cycle changed? Evidence and expecaions. Federal Reserve Bank of Kansas Ciy Symposium, Jackson Hole, Wyoming. Tanner, G. (1997). A noe on economic news and inraday exchange raes.journal of Banking and Finance,,