7 5 Global VAR M E-mail: hime.lll.9@gmail.com
5............................ 7................................. 8 3 3............................... 3......................... 3 3....................................... 4 3.......................... 5 3.3................................... 9 3.3.................................. 3.3.................................. 3 3.3.3............................... 6 3.4............................ 3 4 34 35 35 Data Appendix 38
(, 3) (income-absorption effect demandaugmenting effect) ) ) Rey (6)
(expenditure-switching effect) (wealth effect) (search for yield, Rajan 5) ) (VAR) ) (3)
VAR Kim () Maćkowiak (6) VAR (marginal approach) Canova (5) Maćkowiak (7) (block exogeneity) marginal approach block exogeneity (unidirectional) VAR Global VAR (GVAR) (Pesaran et al. 4) GVAR VAR VARX* GVAR VARX* GVAR VARX* VARX* block exogeneity GVAR VARX* GVAR VARX* GVAR GVAR 3
3) GVAR Eickmeier and Ng (5) GVAR VARX* VARX* GVAR (VARX*) (GVAR) GVAR (, 3) Pesaran et al. (4) GVAR 3 GVAR 3 3) GVAR Canova and Ciccarelli (4, 9) (Panel VAR) Mumtaz and Surico (9) Factor Augmented VAR (FAVAR) 4
4 Pesaran et al. (4) Dees et al. (7) GVAR GVAR N + (i =,,,..., N) (VARX*) i VARX*(p i, q i ) Φ i (L, p i )x it = a i + a i t + Λ i (L, q i )x it + u it, () i =,,,..., N, t =,,..., T, u it iid(, Σ ui ). a i a i t x it i (k i ) x it i (k i ) L Φ i (L, p i ) = I ki p i l= Φ ill l Λ i (L, q i ) = q i l= Λ ill l VARX* p i q i p i q i i VAR x it i j w ij x it = N w ij x jt, j= N j= w ij = w ii = 4) () A i (L, p i, q i ) = [Φ i (L, p i ), Λ i (L, q i )] z it = (x it, x it ) Ψ it = a i + a i t () A i (L, p i, q i )z it = Ψ it + u it () 4) GAVR ( + ) w ij,t 5
K = N j= k j (K ) x t = (x t, x t,..., x Nt) x t i VARX* (k i + k i K) W i = ( ) Iki w i I k i w i I k i w in I k i w in I k i z it = W i x t (3) p = max(p, p,..., p N, q, q,..., q N ) A i (L, p i, q i ) = A i (L, p), { Φ i = (p p i > ) Λ i = (p q i > ) (4) 5) (3) (4) () A i (L, p)w i x t = Ψ it + u it (5) (5) VARX* GVAR GVAR A (L, p)w Ψ t A (L, p)w G(L, p) =., Ψ Ψ t t =., A N (L, p)w N Ψ Nt u t u t u t. u Nt G(L, p)x t = Ψ t + u t (6) 6) 5) A i (L, p) = p l= A ill l = (I ki, Λ i ) p l= (Φ il, Λ il )L l A l W A l W 6) G(L, p) = p l= G ll l G l =. A Nl W N 6
. GVAR G(L, p) () x it GVAR () a i a i Φ i (L, p i ) Λ i (L, q i ) N + G(L, p) GVAR GVAR () VARX* (VECMX*) x it = c i α i β i[z it γ i (t )] + Λ i x it + Γ i (L, p) z it + u it (7) β i [z it γ i (t )] β i z it r i (k i +k i r i) α i r i (k i r i ) r i α i β i z it r i VECMX* 7) α i β i = A i I ki Γ i (L, p) = p l= Γ ill l Γ il = p j=l+ A j VECMX* VARX* G(L, p) Ψ t i x it () (7) x it () (7) 8) x it (7) z it = ( ci c i ) ( αi α i ) β i[z it γ i (t )] + ( ) Λi x it + ( ) Γi (L, p) Γ z i (L, p) it + ( uit u it ) (8) 7) Harbo et al. (998) Pesaran et al. () 8) 7
(8) z it x it α i x it x it x it x it x it x it α i F. () (6) GVAR 9) VAR x t (Orthogonal impulse response function) VARX* GVAR VARX* Pesaran et al. (4) (Generalized impulse response function) Dees et al. (7) VARX* Eickmeier and Ng (5) Dees et al. (7) VARX* GVAR GVAR VARX* VARX* GVAR VARX* Eickmeier and Ng (5) GVAR VARX* 9) G 8
VARX* GVAR ) ξ it () i VARX* u it : ξ it = P i u it. (9) ξ i (9) P i P i P i = Σ ui () k i (k i +)/ P i k i P i Σ ui P i P i Σ ui P i P i u it VAR ) Canova (5) Chudik and Fidora () GVAR VAR GVAR GVAR GVAR 9
P C R ξ t = P u t = RP C u t () ) P () R R R ( RP C ) R R ) R R Fry and Pagan (7) Median Target Approach (MT Approach) R (Multiple models problem, Fry and Pagan ()) MT Approach R 3) VARX* P i GVAR GVAR ξ t = (ξ t, ξ t,..., ξ Nt ) GVAR u t = (u t, u t,..., u Nt ) : ξ t = P G u t. () P G N + P i P P P G =..... P N P N (3) (K K) (6) GVAR ) R = R RR = I Σ ξ = I () P C RRP C = P C P C = Σ u P C Σ u ) Rubio-Ramirez et al. () QR R 3) R ( RP C) R R
: P G G(L, p)x t = P G Ψ t + ξ t. (4) GVAR GVAR x t VAR GVAR VARX* ξ it ξ it u t VARX* () x it ξ it () ξ t = P G G ϵ t (5) x t G ξ it x it Λ i G GVAR ξ t W i i x it ξ t A i W i ξ t ξ t x it Λ i w i ξ t 4) GVAR 4) Dees et al. (7) second-round effects of the shocks return impacts
3 3. 996 3 6 ( ) IMF Direction of Trade Statistics (DOTS) 5 (y) (p) (m) (r) (ex) (eq) (exp) (po) VARX* x it x it (p i, q i ) 5) ex VARX* ex 6) VARX* ex ex ex VARX* ex m Dees et al. (7) r eq VARX* 7) 5) y ( = ) p ( = ) m ( = ) r 3 ex ( = ) eq MSCI ( = ) exp ( = ) po WTI r Data Appendix 6) VARX* ex 7) 5 34 95% VARX*
i x it x it p i q i (y, p, m, r, ex, eq, exp) (y, p, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp) (y, p, m, r, eq, exp, po) (y, p, r, ex, eq, exp, po) (y, p, m, exp) (p, ex) (y, p, m, r, eq, exp, po) i x it x it (p i, q i ) 3. VARX* VARX* 3
3.. P : u y u p u m u r u t = u ex u eq u exp = P ξ t = + + + + + + + + + + + + + ξ AD ξ AS ξ MD ξ MS ξ EX ξ eq ξ exp + (free parameter) ξ AD ξ AS y p r ( ) y p ( ) ξ MD ξ MS m r m eq r ex ( ) ξ EX r eq exp ex ξ eq ξ exp ξ AD ξ AS ξ MD ξ MS y p ξ MS y p ex eq ξ EX eq exp (ξ AD ) (ξ AS ) (ξ MS ) (ξ MD ) (ξ EX ). 4
3.. 3 (ξ AD ) IT 7 3 (ξ AS ) 997 4 4 4 4 4 997 4 3 (ξ MS ) 3 6 3 3 4 3 7 8 VARX* VARX* ( ) 3 5
( ) ( ) VARX* GVAR 6
ξ AD ξ AS.8.8.6.6.4.4.. -. -. -.4 -.6 -.8 - Jan-996 Jan-998.8.6 Jan- Jan- Jan-4 Jan-6 Jan-8 ξ MS Jan- Jan- Jan-4 Jan-6 -.4 -.6 -.8 - Jan-996 Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6.4. -. -.4 -.6 -.8 - Jan-996 Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 ( ) ξ AD ξ AS ξ MS ( ) 7
ξ AD y ξ AS y ξ MS y.5-4 36 48 ξ AD p. -. 4 36 48 ξ AD m - -4 4 36 48 ξ AD r.4. 4 36 48 ξ AD ex 4 36 48 ξ AD eq - -4 4 36 48 ξ AD exp - 4 36 48 4 36 48 ξ AS p -. -. 4 36 48 ξ AS m - - 4 36 48 5-3 ξ AS r -5 4 36 48 ξ AS ex.5 -.5 4 36 48 ξ AS eq.5 -.5 4 36 48 ξ AS exp.5 4 36 48 4 36 48 ξ MS p.5 -.5 4 36 48 ξ MS m 4 -. 4 36 48 ξ MS r -. 4 36 48 ξ MS ex - - 4 36 48 ξ MS eq 4 36 48 ξ MS exp 4 36 48 ( ) VARX* 8
3.3 3 x it 3 9
.5 -.5.6.4..5 -.5 ξ AD y 4 36 48 ξ AD p 4 36 48 ξ AD r - 4 36 48 ξ AD ex - 4 36 48 ξ AD eq - -4 4 36 48 ξ AD exp - 4 36 48.4. -. ξ AS y 4 36 48 ξ AS p.6.4. -. 4 36 48 ξ AS r.6.4. -. -.4 4 36 48 ξ AS ex.4. -. -.4 4 36 48 ξ AS eq.5 -.5-4 36 48 ξ AS exp.5 -.5 4 36 48.5 ξ MS y -.5 4 36 48 ξ MS p -.5 - -.5.5 -.5 4 36 48 ξ MS r 4 36 48 ξ MS ex - 4-4 36 48 ξ MS eq 4 36 48 ξ MS exp 4 36 48 ( ) 3 3.3. 4
5 8) 6 8) 36
.5.5.5. -. -.5.5 -.5 -.5 -.4 - - -.6 -.5 - -.8 -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.5.5.5.5.5.5.5 -.5 -.5 -.5 -.5 -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.5.5.5 -.5 -.5 - -.5 -.5 -.5 - - -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48..5..6.4 -. -.. -.4 -.5.5 -.4 -.6 -.6 -.8 -. -.8 - -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.5.5 ( ) ±SE 4 (ξ AD exp)
3.3. 5 3 6 3 () 5 6 3 3
.5.3.8.3..6....4.. -.5 -. -. -. -. 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.5..5 -. -.5 -. 4 36 48 4 36 48.5.5.6.4. 4 36 48. -. 4 36 48.5..3...5.. -.5 -. -. -. -. -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48. -. 4 36 48.5.5. -. 4 36 48. -. 4 36 48.. -. 4 36 48.5.5 ( ) ±SE 5 (ξ AS y) 4
. -. -.4.5 -.5 -... 4 36 48 4 36 48 4 36 48.5 -.5 -.5 -...5 -.5 4 36 48.4. -.. -. -.4 4 36 48.5 -.5..5 4 36 48.8.6.4. 4 36 48 -.6 -.5 -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.8.6.4. 4 36 48.. 4 36 48..5.5 -.5 4 36 48.5..5.4..5.5...5.5 -. -.5 -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48. -. -. 4 36 48 ( ) ±SE 6 (ξ AS p) 5
3.3.3 7 4 8 9 8 ( 3 ) 9 6
7 7
. -.5 -. - 4 36 48 4 36 48.. 4 36 48 -. -.4 4 36 48.8.6.4. 4 36 48....6. -..4 -..5 -.4. -. -.6 -. -.8 -. -.3 4 36 48 4 36 48.5 4 36 48 4 36 48.4.6.4.5...5.5 -. -.5 -.4 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.8.6.4. -. 4 36 48.5.5.5 -.5.5 4 36 48.8.6.4. 4 36 48.8.6.4. 4 36 48.5.5 ( ) ±SE 7 (ξ MS y) 8
.5.5.5.5 -.5.5 - -.5 -.5 4 36 48 4 36 48 4 36 48 4 36 48.5 4 36 48.5.5.5 - -.5.5 -.5 - - -.5 -.5 - -.5 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.5.5 4 36 48 4 36 48.5.5 -.5 -.5-4 36 48 4 36 48 4 36 48. -.5.5.5.5 -. - -.4 -.5 -.6 -.5 -.8 4 36 48 4 36 48 4 36 48 4 36 48 4 36 48.5.5 ( ) ±SE 8 (ξ MS exp) 9
.5.5 4 36 48 4 36 48-3 4 36 48 4 36 48 3 4 36 48 4 36 48 4 4 36 48-4 36 48 6 4 4 36 48-4 36 48-4 36 48 3 4 36 48 6 4 4 36 48 4 36 48 3 4 36 48 4 4 36 48 - - -3 4 36 48 6 4 4 36 48 4 36 48 3 4 36 48.5.5 ( ) ±SE 9 (ξ MS eq) 3
.5..5 4 36 48.5 -.5. -..4. 4 36 48 4 36 48.8.6.4.. -. -.4.. -. 4 36 48 4 36 48 4 36 48...5 -.5 -.5 4 36 48 4 36 48 4 36 48 -. -.5 4 36 48 4 36 48 4 36 48. -..5 -.5.5.5 -..5....5.. 4 36 48 4 36 48 4 36 48.....6.4..3.. 4 36 48 4 36 48 4 36 48 4 36 48.5.5 ( ) ±SE (ξ MS r) 3
3.4 GVAR VARX* (, 3) 3
.6.4 y.5 p 3.5 3 m -3 r 5..8.6..5..5.5-5.4.5 -..5 4 36 48 4 36 48 4 36 48-5 4 36 48 -.5 ex 4 3.5 3 eq 3.5 exp -.5.5 -.5.5 -.5.5 4 36 48 4 36 48 4 36 48 ( ) GVAR ±SE VARX* 33
4 GVAR GVAR VARX* ( ) GVAR GVAR 34
,, () -3 DSGE -, No.-J-7 (), 3 3 (3), 4 8 35
Canova, F. (5) The transmission of US shocks to Latin America, Journal of Applied econometrics, Vol., No., pp.9-5. Canova, F., and Ciccarelli, M. (4) Forecasting and turning point predictions in a Bayesian panel VAR model, Journal of Econometrics, Vol., No., pp.37-359. Canova, F., and Ciccarelli, M. (9) Estimating multicountry VAR models, International Economic Review, vol.5, No.3, pp.99-959. Chudik, A., Fidora, M. () Using the global dimension to identify shocks with sign restrictions, ECB Working Paper Series No.38. Dees, S., Mauro, F. D., Pesaran, M. H., and Smith, L. V. (7) Exploring the international linkages of the euro area: a global VAR analysis, Journal of Applied Econometrics, Vol., No., pp.-38. Eickmeier, S., and Ng, T. (5) How do US credit supply shocks propagate internationally? A GVAR approach, European Economic Review, Vol.74, pp.8-45. Fry, R., and Pagan, A. (7) Some issues in using sign restrictions for identifying structural VARs, National Centre for Econometric Research Working Paper, No.4. Fry, R., and Pagan, A. () Sign restrictions in structural vector autoregressions: A critical review, Journal of Economic Literature, Vol.49, No.4, pp.938-96. Harbo, I., Johansen, S., Nielsen, B., and Rahbek, A. (998) Asymptotic inference on cointegrating rank in partial systems, Journal of Business & Economic Statistics, Vol.6, No.4, pp.388-399. Kim, S. () International transmission of US monetary policy shocks: Evidence from VAR s, Journal of Monetary Economics, Vol.48, No., pp.339-37. Maćkowiak, B. (6) What does the Bank of Japan do to East Asia? Journal of International Economics, Vol.7, No., pp.53-7. Maćkowiak, B. (7) External shocks, US monetary policy and macroeconomic fluctuations in emerging markets, Journal of Monetary Economics, Vol.54, No.8, pp.5-5. Mumtaz, H., and Surico, P. (9) The transmission of international shocks: a factor augmented VAR approach, Journal of Money, Credit and Banking, Vol.4, No.s, 36
7-. Pesaran, M. H., Schuermann, T., and Weiner, S. M. (4) Modeling regional interdependencies using a global error-correcting macroeconometric model, Journal of Business & Economic Statistics, Vol., No., pp.9-6. Pesaran, M. H., Shin, Y., and Smith, R. J. () Structural analysis of vector error correction models with exogenous I () variables, Journal of Econometrics, Vol.97, No., pp.93-343. Rajan, R. G. (5) Has financial development made the world riskier? NBER Working Paper, No.w78. Rey, H. (6) International Channels of Transmission of Monetary Policy and the Mundellian Trilemma, IMF Economic Review, Vol.64, No., pp.6-35. Rubio-Ramirez, J. F., Waggoner, D. F., and Zha, T. () Structural vector autoregressions: Theory of identification and algorithms for inference, The Review of Economic Studies, Vol.77, No., pp.665-696. 37
Data Appendix y = Datastream p = (X3-ARIMA) IFS m = r Datastream ex = BIS eq MSCI JAPAN = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = Datastream r 3 Datastream ex = BIS eq MSCI AUSTRALIA = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS r Selic Datastream ex = BIS eq MSCI BRAZIL = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS 38
r Datastream ex = BIS eq MSCI CANADA = Datastream exp = Datastream y = World Bank p = (X3-ARIMA) IFS i) ex = BIS eq MSCI CHINA = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI GERMANY = Datastream exp = (X3-ARIMA) Datastream y = Datastream ii) p = (X3-ARIMA) IFS r 3 Datastream ex = BIS eq MSCI HONG KONG = Datastream exp = (X3-ARIMA) Datastream y = OECD i) IFS ii) Bloomberg 39
p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI INDIA = Datastream exp = (X3-ARIMA) Datastream p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI INDONESIA = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI KOREA = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS r 3 Datastream ex = BIS eq MSCI MALAYSIA = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS 4
r 9 Cetes Datastream ex = BIS eq MSCI MEXICO = Datastream exp = Datastream y = Datastream p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI PHILIPINNES = Datastream exp = (X3-ARIMA) Datastream y = OECD p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI RUSSIA = Datastream exp = (X3-ARIMA) Datastream y = Datastream p = (X3-ARIMA) IFS r IFS ex = BIS eq MSCI SINGAPORE = Datastream exp = (X3-ARIMA) IES y = Datastream p = (X3-ARIMA) IFS 4
r Datastream ex = BIS eq MSCI SOUTH AFRICA = Datastream exp = (X3-ARIMA) Datastream y = OECD p = (X3-ARIMA) IFS r 3 Datastream ex = BIS eq MSCI SWEDEN = Datastream exp = (X3-ARIMA) Statistics Sweden p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI THAILAND = Datastream exp = (X3-ARIMA) Datastream y = IFS p = (X3-ARIMA) IFS r Datastream ex = BIS eq MSCI TURKEY = Datastream exp = (X3-ARIMA) Datastream y = IFS p = (X3-ARIMA) IFS r Sterling Datastream 4
ex = BIS eq MSCI UK = Datastream exp = (X3-ARIMA) Datastream y = IFS p = (X3-ARIMA) IFS r FF Datastream ex = BIS eq MSCI US = Datastream exp = Datastream po WTI (X3-ARIMA) FRED p = (X3-ARIMA) IFS ex = Datastream 43
Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 5 5 y Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 5 5 p Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 3 m Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 5 5 r Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 5 5 ex Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 5 5 eq Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 5 5 exp Jan-998 Jan- Jan- Jan-4 Jan-6 Jan-8 Jan- Jan- Jan-4 Jan-6 4 6 8 po 44