## _wp1699 — GVAR analysis of Japan’s QQE spillovers (Sections II–VI and Technical Appendix)

## Source details

**Canonical URL:** [_wp1699 — GVAR analysis of Japan’s QQE spillovers (Sections II–VI and Technical Appendix)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1699.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1699.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1699.pdf.json)

---

### Literature review: channels and prior evidence
- Theoretical transmission channels of unconventional monetary policy (UMP): stock market, exchange rates, portfolio flows, and interest rates.
- Expected international spillovers of Japan’s QQE to emerging economies through: capital flows, currency appreciation, bond yields, and stock price effects.
- Mechanism described: increased liquidity → lower interest rates and yen depreciation → Japanese investors rebalance toward higher yields abroad → capital inflows to Emerging Asia; stock-price increases via improved expectations and portfolio rebalancing.
- Empirical findings summarized from the literature:
  - Chen, Filardo and Zhu (2015): Global VAR finds U.S. QE contributed to overheating, currency appreciation, and strong capital inflows to emerging countries.
  - Bauer and Neely (2014): Fed UMP announcements impacted international bond yields.
  - Gambacorta, Hofman and Peersman (2014): central bank balance sheet expansion at the ZLB caused temporary price and activity increases.
  - Glick and Leduc (2012): depreciation of the US dollar and British pound after LSAP announcements and decline in long-term interest rates.
  - Japan-focused studies:
    - Fujiwara (2014): QQE raised inflation expectations modestly short-run.
    - Lam (2011): prior BoJ QE raised stock prices strongly but not inflation or exchange rate.
    - Joo, Kee and Choi (2014): little impact of QQE on Korea trade and capital flows.
    - Ree, Hong and Choi (2015): Yen depreciation could increase Japan’s price competitiveness long-run; may harm South Korean firms if Won appreciates.
  - GVAR literature common findings: rapid transmission of financial shocks, stronger effects on equities and bond markets than GDP/inflation; many studies rely on BIS cross-border bank lending data or trade weights for Emerging Asia.

### Data and VARX setup — overview
- Sample frequency and period: monthly data from January 2000 up to December 2014.
- Country-specific endogenous variables (per country, except Japan and U.S. special cases): real GDP, equity prices, CPI inflation, bank credit, the short-term interest rate, the exchange rate, the monetary base, and capital inflows.
- Global variable: price of oil.
- GDP handling: quarterly GDP converted to monthly by the Chow-Lin interpolation method (industrial production index as reference).
- Exchange rates:
  - For all countries except Japan and the United States: bilateral rate vis-à-vis the Japanese Yen, obtained from Bloomberg.
  - For Japan: nominal effective exchange rate.
- CPI: IFS consumer price index (CPI) for all items (2010 index year).
- Bank credit proxy: IFS claims to the private non-financial sector (monthly). BIS quarterly credit to private sector interpolated for robustness; IFS monthly series used to avoid excess seasonality.
- Equity prices: MSCI index specific to each country.
- Monetary base: expressed in percentages.
- Capital inflows: net flows to Emerging Asia and Japan; positive = inflows, negative = outflows. Negative observations removed to permit natural logarithm transformation.
- Seasonal adjustment: Eviews’s Census X13.
- Sample transformations (notation preserved from source):
  - Real GDP: ݕ ൌln ܲܦܩ ௜௧ ܫܲܥ ௜௧
  - Inflation: dp = ln( (CPI_t) ) - ln( (CPI_{t-12}) )
  - Equity prices: eq = ln( MSCI_t / MSCI_{t-1} )
  - Exchange rate: ex = ln(ER_t / ER_{t-1})
  - Short-term interest rate: r = 0.25 * n_t(1 + R_{t,0}/100)
  - Bank credit: bc = ln(credit_t / credit_{t-1})
  - Price of oil: oil = ln(price_oil)
  - Capital inflows: cf = ln(flows_t / flows_{t-1})

### UMP shock identification — proxies and rationale
- Conventional proxies (interest rates or term spreads) problematic at the zero lower bound (ZLB); expectation channel matters at ZLB (Eggertsson 2008).
- Chosen alternative proxies for Japan’s QQE:
  - Increase in equity prices (stock-market channel): captures expectations and stylized facts (large equity price increases during Japan’s QEs).
  - Increase in the monetary base: consistent with QQE intermediate target; Japan almost doubled its monetary base in recent years.
- Rationale:
  - Equity price shocks capture expectation-driven transmission and wealth effects.
  - Monetary base shocks capture balance-sheet adjustments; may be less effective if banks hoard reserves.

### Weighting schemes in the GVAR
- Foreign variables for each domestic variable are weighted averages of partners’ corresponding variables; weights reflect economic relationships.
- Flow channels used to build weights: combination of foreign direct investment (FDI) flows from IMF’s CDIS and trade flows from IMF direction of trade statistics.
- Robustness checks: trade weights and FDI weights used separately; combined trade+FDI results similar to individual schemes.
- Bilateral trade weights computed from sum of imports and exports; financial weights from sum of FDI inward and outward flows.
- Fixed weights constructed for 2011-2014; varying the time frame did not significantly change results.
- Empirical examples of Japan linkages in Emerging Asia (from source Table 1):
  - Trade-FDI weight between Thailand and Japan: 0.28.
  - Trade-FDI weight between the Philippines and Japan: 0.22.
  - For China, Japan ranks third after the Euro zone and the United States.

### VARX model specification and estimation
- Non-Japan, non-U.S. VARX*: 8 domestic endogenous variables — GDP (y), CPI inflation (dp), equity prices (eq), bank credit (bc), monetary base (mb), short-term interest rates (r), exchange rate (ex), and capital inflows (cf) — augmented with foreign (weakly exogenous) variables and global oil price.
- U.S. VARX: includes all domestic variables except exchange rate; only y* and ex* enter as foreign variables in the U.S. VARX*; price of oil specified as endogenous in the U.S. VARX*.
- Foreign variables (except oil) are weighted averages of other economies’ variables using flow-based weights.
- Lag orders: domestic lag order (p) and foreign lag order (q) selected by Akaike Information Criterion (AIC), with maximum domestic lag order set to p = 2.
- Cointegration and unit-root procedures:
  - Unit root tests: Augmented Dickey-Fuller and Weight Symmetric ADF tests; majority of variables I(1).
  - VECMX (error-correction) forms estimated allowing cointegration; Johansen’s trace statistic used for cointegration ranks.
- Diagnostic checks reported:
  - Stability: eigenvalues of F matrix lie on or inside unit circle.
  - Weight smallness condition: Σ_j w_{ij}^2 → 0 as N → ∞ largely satisfied.
  - Idiosyncratic shocks weakly correlated: VECMX* residuals show low cross-sectional correlation.
- Computational tools: GVAR toolbox by Vanessa Smith and Alessandro Galesi used for estimation.

### Identification of structural shocks and impulse response estimation
- Structural impulse responses obtained by recursive Cholesky scheme to orthogonalize shocks (following Dees, diMauro, Pesaran and Smith (2007)).
- Japan bloc ordering (one specification): short-term interest rate, equity prices, the exchange rate, bank credit, capital inflows, monetary base, inflation, and output.
- Robustness: alternative orderings (equity last, equity second) and short-run restrictions tested; results largely robust.
- Japan block placed first in the GVAR (motivated by interest in Japan → Emerging Asia spillovers); placing U.S. first yields similar results.

### Dynamic analysis — main impulse response findings
- Shock proxies tested and qualitative interpretation:
  - One negative standard deviation shock to Japan’s short-term interest rate, 2-year government bond yield, 10-year government bond yield, and term spread: no statistically significant results for Japan or Emerging Asia variables; did not reflect equity price increases or yen depreciation. Interpretation: transmission mechanism via interest rates is broken at ZLB.
  - One positive standard deviation shock to Japan’s monetary base: significant increase in Japan inflation of the order of 0.05 percent, but weak / not statistically significant spillovers on financial variables elsewhere. Possible reasons: transmission primarily via expectations; injected money hoarded by banks.
  - One positive standard deviation shock to Japan’s equity prices: produced the strongest and most widespread effects.
- Quantitative magnitudes from the equity-price shock (median estimates):
  - Japan:
    - GDP increased by 0.5 percent.
    - Equity prices rose approximately by 5 percent.
    - Inflation picked up slightly by 0.05 percent (not statistically significant).
    - Exchange rate depreciated by 1 to 1.5 percent (nominal effective exchange rate).
    - Capital inflows increased by about 30 percent in the first two periods.
    - Bank credit largely unmoved.
  - Emerging Asia and other regions:
    - Equity prices increased in China, Indonesia, Malaysia, Singapore, Thailand, and Indonesia in the 2-5 percent range.
    - U.S., the Euro zone, and advanced European countries: equity prices up by about 1 percent.
    - Exchange rates: all countries in sample except Norway experienced appreciation vis-à-vis the yen.
    - GDP: generally positive impacts; statistically significant for Indonesia, Malaysia, the Philippines, and Thailand.
    - Inflation: limited spillovers; temporary first-month increases of about 0.05 percent that do not persist.
    - Capital inflows: spikes in first month(s), peaking as high as 10 percent in Indonesia and 15 percent in Thailand, then reverting close to original levels; increase statistically significant for most countries only for the first period.
    - China registered capital outflows (potential redirection of investment toward Japan).
- Interpretations emphasized:
  - Confidence/expectations channel (captured by equity price increases) dominates exchange-rate expenditure-switching negative effects in many cases.
  - Japan’s role in global value chains means yen depreciation can lower import costs of intermediate inputs for China, offsetting competitiveness channels.
  - Heterogeneous country-specific responses in bank credit; generally limited permanent spillovers on inflation.

### Counterfactual analysis — approach and results
- Purpose: assess magnitude of spillovers had Japan’s QQE not occurred.
- Method: conditional one-step-ahead GVAR forecasts assuming Japan’s equity prices (or monetary base in alternative scenarios) had remained constant from time t; compare forecasts to actuals.
- Counterfactual experiments reported for ASEAN countries and China on output, the exchange rate, and equity prices; Japan counterfactual also includes inflation.
- Key counterfactual findings (keeping equity prices constant from 2013M1 or alternative robustness scenarios):
  - For Japan:
    - Had equity prices not grown, Japan would not have experienced exchange rate depreciation, nor an increase in inflation; output would have been lower.
  - For Emerging Asia (China, Indonesia, Malaysia, Philippines, Singapore, Thailand):
    - Except for China and Thailand, countries’ output would have been lower had Japan equity prices remained constant.
    - All countries recorded actual equity prices higher than the counterfactual forecasts (i.e., QQE raised regional equity prices).
    - Counterfactual analysis reveals a currency appreciation in the region that would have been larger if Japan equity prices had remained low (i.e., actual appreciation was moderated by QQE).
  - Results robust to scenarios: keeping monetary base constant, keeping equity prices constant from 2013M2, and simultaneous constraints.
- Caveats: other contemporaneous shocks (e.g., June 2013 “taper tantrum”, decline in oil prices) also affected outcomes—distinctive dip in equity prices in 2013M06 seen across countries.

### Conclusions (primary conclusions)
- Main result: spillovers from Japan’s QQE to emerging Asian countries tended to be positive.
  - Despite appreciation of regional currencies vis-à-vis the yen, Emerging Asia GDP impacts were positive and significant in several cases.
  - Confidence/expectations channel (equity price increases) more than offset negative exchange-rate effects from expenditure switching.
- Limited spillovers in terms of higher inflation; on average increased capital inflows (short-lived).
- Proxying QQE with monetary base increases (instead of equity prices) did not yield strong or significant spillovers — suggestive that QQE affected other Asian countries primarily through expectations/confidence captured by equity prices, not via balance-sheet adjustments captured by monetary base movements.

### Technical appendix — GVAR framework and estimation highlights
- GVAR framework follows Pesaran, Schuermann and Weiner (2004) and Dees, diMauro, Pesaran and Smith (2007).
- Construction and estimation details:
  - For N countries, domestic vector x_{i,t} and foreign vector x_{i,t}* built as weighted averages: x_{i,t}* = Σ_j w_{ij} x_{j,t}, with Σ_j w_{ij} = 1.
  - Country-specific VARX*(2,2) models estimated in levels with foreign variables weakly exogenous.
  - Error-correction (VECMX) forms estimated allowing cointegration; Johansen’s trace statistic used for cointegration ranks.
  - After estimation, individual VARX* models stacked to form global GVAR and solved recursively for impulse responses and variance decompositions.
  - Unit root testing: Augmented Dickey-Fuller and Weight Symmetric ADF tests; majority of variables nonstationary (I(1)).
  - Weak exogeneity tests performed; majority of foreign variables weakly exogenous.
  - Trend coefficients restricted in cointegrating space; intercepts unrestricted (case IV in GVAR toolbox).
  - Lag order selection via AIC, maximum domestic lag p = 2.
- Diagnostic checks and computational tools:
  - Stability: eigenvalues of F matrix lie on or inside unit circle.
  - Weight smallness condition: Σ_j w_{ij}^2 → 0 as N → ∞ largely satisfied.
  - Idiosyncratic shocks weakly correlated: VECMX* residuals show low cross-sectional correlation.
  - GVAR toolbox by Vanessa Smith and Alessandro Galesi used for estimation.

*Source — IMF staff working paper content unit _wp1699 (Sections II–VI and Technical Appendix) provided in the supplied PDF excerpt.*

### Section III explains the GVAR model’s setup and the data. Section IV presents the impulse

### _wp1699 - Section III explains the GVAR model’s setup and the data. Section IV presents the impulse

### Literature review: channels and prior evidence
- Theoretical transmission channels of unconventional monetary policy (UMP): stock market, exchange rates, portfolio flows, and interest rates.
- Expected international spillovers of Japan’s QQE to emerging economies through: capital flows, currency appreciation, bond yields, and stock price effects.
- Mechanism described: increased liquidity → lower interest rates and yen depreciation → Japanese investors rebalance toward higher yields abroad → capital inflows to Emerging Asia; stock-price increases via improved expectations and portfolio rebalancing.
- Empirical literature summarized:
  - Chen, Filardo and Zhu (2015): Global VAR finds U.S. QE contributed to overheating, currency appreciation, and strong capital inflows to emerging countries.
  - Bauer and Neely (2014): Fed UMP announcements impacted international bond yields.
  - Gambacorta, Hofman and Peersman (2014): central bank balance sheet expansion at the ZLB caused temporary price and activity increases.
  - Glick and Leduc (2012): depreciation of the US dollar and British pound after LSAP announcements and decline in long-term interest rates.
  - Japan-focused studies:
    - Fujiwara (2014): QQE raised inflation expectations modestly short-run.
    - Lam (2011): prior BoJ QE raised stock prices strongly but not inflation or exchange rate.
    - Joo, Kee and Choi (2014): little impact of QQE on Korea trade and capital flows.
    - Ree, Hong and Choi (2015): Yen depreciation could increase Japan’s price competitiveness long-run; may harm South Korean firms if Won appreciates.
  - GVAR literature: Eickmeyer and Ng (2011), Dees et al (2007), Galesi and Sgherri (2009), Sun, Heinz and Ho (2013) — common findings: rapid transmission of financial shocks, stronger effects on equities and bond markets than GDP/inflation; many studies rely on BIS cross-border bank lending data or trade weights for Emerging Asia.

### Data and VARX setup — overview
- Sample frequency and period: monthly data from January 2000 up to December 2014.
- Country-specific endogenous variables (per country, except Japan and U.S. special cases): real GDP, equity prices, CPI inflation, bank credit, the short-term interest rate, the exchange rate, the monetary base, and capital inflows.
- Global variable: price of oil.
- GDP handling: quarterly GDP converted to monthly by the Chow-Lin interpolation method (industrial production index as reference).
- Exchange rates:
  - For all countries except Japan and the United States: bilateral rate vis-à-vis the Japanese Yen, obtained from Bloomberg.
  - For Japan: nominal effective exchange rate.
- CPI: IFS consumer price index (CPI) for all items (2010 index year).
- Bank credit proxy: IFS claims to the private non-financial sector (monthly). BIS quarterly credit to private sector interpolated for robustness; IFS monthly series used to avoid excess seasonality.
- Equity prices: MSCI index specific to each country.
- Monetary base: expressed in percentages.
- Capital inflows: net flows to Emerging Asia and Japan; positive = inflows, negative = outflows. Negative observations removed to permit natural logarithm transformation.
- Seasonal adjustment: Eviews’s Census X13.
- Sample transformation (variables constructed as given):
  - Real GDP: y = ln(X_t / X_{t-1})? (notation preserved as in source: Real GDP: ݕ ൌln ܲܦܩ ௜௧ ܫܲܥ ௜௧)
  - Inflation: dp = ln( (CPI_t) ) - ln( (CPI_{t-12}) ) (notation preserved in source)
  - Equity prices: eq = ln( MSCI_t / MSCI_{t-1} ) (notation preserved)
  - Exchange rate: ex = ln(ER_t / ER_{t-1}) (notation preserved)
  - Short-term interest rate: r = 0.25 * n_t(1 + R_{t,0}/100) (notation preserved)
  - Bank credit: bc = ln(credit_t / credit_{t-1}) (notation preserved)
  - Price of oil: oil = ln(price_oil)
  - Capital inflows: cf = ln(flows_t / flows_{t-1}) (notation preserved)

### UMP shock identification — proxies and rationale
- Conventional proxy (interest rates or term spreads) problematic at the zero lower bound (ZLB); expectation channel matters at ZLB (Eggertsson 2008).
- Chosen alternative proxies for Japan’s QQE:
  - Increase in equity prices (stock-market channel): captures expectations and stylized facts (large equity price increases during Japan’s QEs).
  - Increase in the monetary base: consistent with QQE intermediate target; Japan almost doubled its monetary base in recent years (statement preserved).
- Rationale:
  - Equity price shocks capture expectation-driven transmission and wealth effects.
  - Monetary base shocks capture balance-sheet adjustments; may be less effective if banks hoard reserves.

### Weighting schemes in the GVAR
- Foreign variables for each domestic variable are weighted averages of partners’ corresponding variables; weights chosen to reflect economic relationships.
- Flow channels used to build weights: combination of foreign direct investment (FDI) flows from IMF’s CDIS and trade flows from IMF direction of trade statistics.
- Robustness checks: trade weights and FDI weights used separately; combined trade+FDI results similar to individual schemes (choice of weights of secondary importance).
- Bilateral trade weights computed from sum of imports and exports; financial weights from sum of FDI inward and outward flows.
- Fixed weights constructed for 2011-2014; varying the time frame did not significantly change results.
- Empirical note: Table 1 (Trade and FDI weights) indicates notable Japan linkages in Emerging Asia (examples preserved from source):
  - Trade-FDI weight between Thailand and Japan: 0.28.
  - Trade-FDI weight between the Philippines and Japan: 0.22.
  - For China, Japan ranks third after the Euro zone and the United States.

### VARX model specification
- For each non-Japan, non-U.S. country: VARX* includes 8 domestic endogenous variables — GDP (y), CPI inflation (dp), equity prices (eq), bank credit (bc), monetary base (mb), short-term interest rates (r), exchange rate (ex), and capital inflows (cf) — augmented with foreign (weakly exogenous) variables built on flows and the global oil price.
- U.S. VARX: includes all domestic variables except exchange rate; only y* and ex* enter as foreign variables in the U.S. VARX*; price of oil specified as endogenous in the U.S. VARX*.
- Foreign variables (except oil) are weighted averages of other economies’ variables using flow-based weights.
- Lag orders: domestic lag order (p) and foreign lag order (q) selected by Akaike Information Criterion (AIC), with maximum lag order set to p = 2.

### Identification of structural shocks and impulse response estimation
- Structural impulse responses obtained by recursive Cholesky scheme to orthogonalize shocks (following Dees, diMauro, Pesaran and Smith (2007)).
- Japan bloc ordering chosen (one specification): short-term interest rate, equity prices, the exchange rate, bank credit, capital inflows, monetary base, inflation, and output.
- Robustness: experimented with alternative orderings (equity last, equity second) and short-run restrictions; results largely robust.
- Japan block placed first in the GVAR (motivated by interest in Japan → Emerging Asia spillovers); placing U.S. first yields similar results.

### Dynamic analysis — main impulse response findings
- Shock proxies tested:
  - One negative standard deviation shock to Japan’s short-term interest rate, 2-year government bond yield, 10-year government bond yield, and term spread: yielded no statistically significant results for Japan or Emerging Asia variables; did not reflect equity price increases or yen depreciation. Interpretation: transmission mechanism via interest rates is broken at ZLB.
  - One positive standard deviation shock to Japan’s monetary base: yielded a significant increase in Japan inflation of the order of 0.05 percent, but weak / not statistically significant spillovers on financial variables elsewhere. Possible reasons: transmission primarily via expectations; injected money hoarded by banks.
  - One positive standard deviation shock to Japan’s equity prices: produced the strongest and most widespread effects.
- Quantitative magnitudes from the equity-price shock (median estimates):
  - Japan:
    - GDP increased by 0.5 percent.
    - Equity prices rose approximately by 5 percent.
    - Inflation picked up slightly by 0.05 percent (not statistically significant).
    - Exchange rate depreciated by 1 to 1.5 percent (nominal effective exchange rate; depreciation expressed as decreasing impulse response).
    - Capital inflows increased by about 30 percent in the first two periods.
    - Bank credit largely unmoved.
  - Emerging Asia and other regions:
    - Equity prices increased in China, Indonesia, Malaysia, Singapore, Thailand, and Indonesia in the 2-5 percent range.
    - U.S., the Euro zone, and advanced European countries: equity prices up by about 1 percent.
    - Exchange rates: all countries in sample except Norway experienced appreciation vis-à-vis the yen.
    - GDP: generally positive impacts; statistically significant for Indonesia, Malaysia, the Philippines, and Thailand.
    - Inflation: limited spillovers; temporary first-month increases of about 0.05 percent that do not persist.
    - Capital inflows: spikes in first month(s), peaking as high as 10 percent in Indonesia and 15 percent in Thailand, then reverting close to original levels; increase statistically significant for most countries only for the first period.
    - China registered capital outflows (potential redirection of investment toward Japan).
- Interpretations emphasized:
  - Confidence/expectations channel (captured by equity price increases) dominates exchange-rate expenditure-switching negative effects in many cases.
  - Japan’s role in global value chains means yen depreciation can lower import costs of intermediate inputs for China, offsetting competitiveness channels.
  - Heterogeneous country-specific responses in bank credit; generally limited permanent spillovers on inflation.

### Counterfactual analysis — approach and results
- Purpose: assess magnitude of spillovers had Japan’s QQE not occurred.
- Method: conditional one-step-ahead GVAR forecasts assuming Japan’s equity prices (or monetary base in alternative scenarios) had remained constant from time t; compare forecasts to actuals.
- Counterfactual experiments reported for ASEAN countries and China on output, the exchange rate, and equity prices; Japan counterfactual also includes inflation.
- Key counterfactual findings (keeping equity prices constant from 2013M1 or alternative robustness scenarios):
  - For Japan:
    - Had equity prices not grown, Japan would not have experienced exchange rate depreciation, nor an increase in inflation; output would have been lower.
  - For Emerging Asia (China, Indonesia, Malaysia, Philippines, Singapore, Thailand):
    - Except for China and Thailand, countries’ output would have been lower had Japan equity prices remained constant.
    - All countries recorded actual equity prices higher than the counterfactual forecasts (i.e., QQE raised regional equity prices).
    - Counterfactual analysis reveals a currency appreciation in the region that would have been larger if Japan equity prices had remained low (i.e., actual appreciation was moderated by QQE).
  - Results robust to scenarios: keeping monetary base constant, keeping equity prices constant from 2013M2, and simultaneous constraints.
- Caveats noted: other contemporaneous shocks (e.g., June 2013 “taper tantrum”, decline in oil prices) also affected outcomes—distinctive dip in equity prices in 2013M06 seen across countries.

### Conclusions (paper’s primary conclusions)
- Main result: spillovers from Japan’s QQE to emerging Asian countries tended to be positive.
  - Despite appreciation of regional currencies vis-à-vis the yen, Emerging Asia GDP impacts were positive and significant in several cases.
  - Confidence/expectations channel (equity price increases) more than offset negative exchange-rate effects from expenditure switching.
- Limited spillovers in terms of higher inflation; on average increased capital inflows (short-lived).
- Proxying QQE with monetary base increases (instead of equity prices) did not yield strong or significant spillovers — suggestive that QQE affected other Asian countries primarily through expectations/confidence captured by equity prices, not via balance-sheet adjustments captured by monetary base movements.

### Technical appendix — GVAR framework and estimation highlights
- GVAR framework follows Pesaran, Schuermann and Weiner (2004) and Dees, diMauro, Pesaran and Smith (2007).
- Construction:
  - For N countries, domestic variable vector x_{i,t} and foreign variable vector x_{i,t}* built as weighted averages: x_{i,t}* = Σ_j w_{ij} x_{j,t}, with Σ_j w_{ij} = 1.
  - Country-specific VARX*(2,2) models estimated in levels with foreign variables weakly exogenous.
  - Error-correction (VECMX) forms estimated allowing cointegration; Johansen’s trace statistic used for cointegration ranks.
  - After estimation, individual VARX* models stacked to form global GVAR and solved recursively for impulse responses and variance decompositions.
- Estimation details:
  - Unit root testing: Augmented Dickey-Fuller and Weight Symmetric ADF tests; majority of variables nonstationary (I(1)).
  - Weak exogeneity tests performed; majority of foreign variables weakly exogenous.
  - Trend coefficients restricted in cointegrating space; intercepts unrestricted (case IV in GVAR toolbox).
  - Lag order selection via AIC, maximum domestic lag p = 2.
  - GVAR diagnostic checks satisfied:
    - Stability: eigenvalues of F matrix lie on or inside unit circle.
    - Weight smallness condition: Σ_j w_{ij}^2 → 0 as N → ∞ largely satisfied.
    - Idiosyncratic shocks weakly correlated: VECMX* residuals show low cross-sectional correlation.
- Computational tools: GVAR toolbox by Vanessa Smith and Alessandro Galesi used for estimation.

*Italic: Source — IMF staff working paper content unit _wp1699 (Sections II–VI and Technical Appendix) provided in the supplied PDF excerpt.*

### References:

### _wp1699 - References

### References
- Bauer M., Neely C. 2014. International channels of the Fed’s unconventional monetary policy. Journal of International Money and Finance 44, 24–46.
- Blinder, A., 2010. Quantitative easing: entrance and exit strategies. Federal Reserve Bank of St. Louis Review, November/December.
- Botman D., 2015. Abenomics: lessons learned from two decades of conventional and unconventional monetary policies. In “Can Abenomics succeed? Overcoming the legacy of Japan’s lost decades.”, IMF.
- Chen Q., Filardo A., Dong H., Feng Z. 2015, Financial crisis, US unconventional monetary policy and international spillovers, IMF working paper WP/15/85
- Chen Q., Lombardi M., Ross A., Feng Z., 2015. Global impact of US and Euro area unconventional monetary policies: a comparison
- Chudik, A., Fratzscher, M. 2011. Identifying the global transmission of the 2007-2009 financial crises in a GVAR model. European Economic Review 55, 325-339
- Cooley, T. F and V. Quadrini (1999a), Monetary Policy and the Financial Decisions of Firms, mimeo.
- Dees, S., di Mauro, F., Smith, V., Pesaran, H., 2007. Exploring the international linkages of the euro area: a global VAR analysis. J. Appl. Econ. 22(1), 1–38.
- Eggertsson, G. , 2008. Great Expectations and the End of the Depression. American Economic Review, 98 (4), 1476–1516.
- Eickmeier, S., Ng T., 2011. How do credit supply shocks propagate internationally? A GVAR approach. CEPR Discussion Paper, 8720.
- Fujiawara I., Nakazono Y., Ueda K., 2014. Policy Regime Change against Chronic Deflation? Policy option under a long-term liquidity trap. RIETI Discussion paper series 14-E-019
- IMF Regional Economic Outlook April 2015. Asia and Pacific, Stabilizing and Outperforming Other Regions.
- Galesi A., Sgherri S. 2009. Regional financial spillovers across Europe: A Global VAR analysis. IMF working paper WP/09/03
- Gambacorta L., Hofman B., Peersman G., 2014. The effectiveness of unconventional monetary policy at the zero lower bound: A cross-country analysis. Journal of Money, Credit and Banking, Vol. 46, No. 4
- Glick R., Leduc S., 2012. Central bank announcements of asset purchases and the impact on global financial and commodity markets. Journal of International Money and Finance 31, 2078–2101
- Hosono K., Isobe S., 2014. The financial market impact of unconventional monetary policies in the U.S., the U.K., the Eurozone, and Japan. PRI discussion paper series No.14A-05
- Kang, J. S., 2015, “What explains Japan’s export sluggishness?” Box 1 in “Japan: Staff Report for the 2015 Article IV Consultation, forthcoming IMF Country Report.
- Lam R.W., 2011. Bank of Japan’s monetary easing measures: Are they powerful and comprehensive? IMF working paper WP11.264
- Lastrapes, W., D. 1998. International Evidence on Equity Prices, Interest Rates and Money, Journal of International Money and Finance, Vol. 17, No. 3, pp. 377-406.
- Neri S. , 2004, Monetary policy and stock prices : theory and evidence. Banca D’Italia, Working paper number 513
- Pesaran, M.H., Schuermann, T., Weiner, S., 2004. Modeling regional interdependencies using a global error-correcting macroeconometric model. J. Bus. Econ. Stat. April.
- Rapach, D., E. 2001. Macro Shocks and Real Stock Prices, Journal of Economics and Business, Vol. 53, pp. 5-26.
- Ree, Joo K., Choi S. 2014, Safe-Haven Korea? - Spillover Effects from UMPs, IMF working paper WP/14/53
- Ree, Joo K., Hong G., Choi S. 2015, Should Korea worry about a permanently weak Yen? , IMF working paper WP/15/158
- Ueda, K., 2012. The Effectiveness of Non-Traditional Monetary Policy Measures: The Case of the Bank of Japan. Japanese Economic Review 63(1), 1-22.
- Yan S., Frigyes F., Giang H., 2013. Cross-Country Linkages in Europe: A Global VAR Analysis. IMF working paper, WP/13/194

### Data Appendix — Variables, Descriptions, Sources, and Notes
- Real GDP (y)
  - Description: ݕൌln ܲܦܩ ್ ್ ௜௧ ܫܲܥ ௜௧
  - Source: IMF IFS
  - Notes: The monthly time series are obtained using the Chow-Lin interpolation method, with the Industrial production index as a reference. (seasonally adjusted using Eviews X13)
- CPI inflation (dp)
  - Description: ൌln݌݀ ( ܫܲܥ ௜௧ ) ሺ ܫܲܥെlnሺ ௜,௧ିଵ ሻ
  - Source: IMF IFS
- Equity prices (eq)
  - Description: ൌln ݍ݁ ܳܧ ௜௧ ܫܲܥ ௜௧
  - Source: Bloomberg
  - Notes: The MSCI index for each country is used.
- Exchange rate (ex)
  - Description: ൌln ݔ݁ ܧ ௜௧ ܫܲܥ ௜௧
  - Source: Bloomberg and the BIS for Japan
  - Notes: The exchange rate vis-à-vis the Japanese yen (cross-rate) is used for all Asian countries. For Japan, we use the nominal effective exchange rate
- Monetary base (mb)
  - Description: Monetary base, percent.
  - Source: IMF IFS
- Short-term interest rate (r)
  - Description/formula: r= 0.25*ln(1+ ܴ ௜௧ ௦ /100)
  - Source: IMF IFS
  - Notes:
    - For China, the deposit rate is used.
    - For Malaysia, the Philippines, the treasury bill rate is used.
    - For Japan, Indonesia, Thailand, the money markets rate is used.
- Bank credit (bc)
  - Description: ൌln ܾܿ ܥܤ ௜௧ ܫܲܥ ௜௧
  - Source: IMF IFS
  - Notes: Claims on the private sector (seasonally adjusted using Eviews X13)
- Price of Oil
  - Description: US dollars per barrel, in natural logarithm
  - Source: IMF IFS
- Imports and Exports
  - Description: In millions of U.S. Dollars
  - Source: IMF CDIS Direction of Trade statistics
- Portfolio flows
  - Description: In millions of U.S. Dollars
  - Source: IMF CPIS Coordinated Portfolio Investment Survey
- FDI flows
  - Description: In millions of U.S. Dollars
  - Source: IMF Coordinated Direct Investment Survey
- Capital inflows
  - Description: In millions of U.S. Dollars; ൌln ݂ܿ ܨܥ ௜௧ ܫܲܥ ௜௧
  - Source: Haver Analytics (EPFR Asia)
  - Notes:
    - Capital flows are the sum of Equity flows and bond flows. For the years that data for one series is unavailable, we just use the data from the other series.

*Source: _wp1699 - References (PDF chapter/section).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1699.pdf_