## _wp1663 ## Source details **Canonical URL:** [_wp1663](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1663.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1663.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1663.pdf.json) --- ### Key results and context - The analysis employs a Global VAR (GVAR) framework comprising 26 region-specific VARX* models within a sample of 33 economies, which together cover more than90%of world GDP. - Country-specific VARX*(p_i,q_i) lag orders are selected by the Akaike Information Criterion (AIC) with maximum lag orders set top_max= 2 andq_max= 1. - The univariate financial stress index (FSI_t) model lag order is selected as(2;1)based on the AIC. - Cointegrating relations (r_i) are selected using trace test statistics with 95% critical values from MacKinnon (1991); in some countries r_i was reduced below the trace-statistic suggestion to ensure well-behaved persistence profiles and global-model stability. ### Lag orders and cointegrating relations (country-specific) - Argentina: p_i 2 q_i 1 relations (r_i) 1 - Australia: p_i 2 q_i 1 relations (r_i) 4 - Brazil: p_i 1 q_i 1 relations (r_i) 1 - Canada: p_i 1 q_i 1 relations (r_i) 3 - China: p_i 1 q_i 2 relations (r_i) 2 - Chile: p_i 2 q_i 1 relations (r_i) 3 - Euro Area: p_i 2 q_i 1 relations (r_i) 2 - India: p_i 2 q_i 1 relations (r_i) 2 - Indonesia: p_i 2 q_i 1 relations (r_i) 3 - Japan: p_i 1 q_i 1 relations (r_i) 4 - Korea: p_i 1 q_i 3 relations (r_i) 3 - Malaysia: p_i 1 q_i 2 relations (r_i) 2 - Mexico: p_i 1 q_i 3 relations (r_i) 3 - Norway: p_i 2 q_i 1 relations (r_i) 3 - New Zealand: p_i 2 q_i 1 relations (r_i) 3 - Peru: p_i 2 q_i 1 relations (r_i) 2 - Philippines: p_i 2 q_i 1 relations (r_i) 2 - South Africa: p_i 2 q_i 1 relations (r_i) 1 - Saudi Arabia: p_i 2 q_i 1 relations (r_i) 1 - Singapore: p_i 2 q_i 1 relations (r_i) 2 - Sweden: p_i 2 q_i 1 relations (r_i) 3 - Switzerland: p_i 2 q_i 1 relations (r_i) 2 - Thailand: p_i 1 q_i 1 relations (r_i) 2 - Turkey: p_i 1 q_i 3 relations (r_i) 3 - UK: p_i 1 q_i 3 relations (r_i) 3 - USA: p_i 2 q_i 1 relations (r_i) 2 ### Model specification and variables - Country domestic variable vector x_it includes up to six variables: y_it (log real GDP), π_it (inflation), eq_it (log real equity prices), r_Sit (short-term interest rate), r_Lit (long-term interest rate), ep_it (real exchange rate). - Foreign variables x*_it are cross-sectional averages constructed with bilateral weights w_ij; alternative weights explored include trade, trade in value added (TiVa), and financial weights. - The United States model excludes foreign financial variables eq*_US,t, r*_SUS,t, and r*_LUS,t; the U.S. model includes e*_US,t - p*_US,t as a weakly exogenous foreign variable instead of the real exchange rate. - A common financial stress index (FSI_t) is included as a weakly exogenous common variable in all country models, with allowance for feedback effects from macro variables; oil price (log nominal oil price in U.S. dollars) is also included as a common variable. ### Empirical findings — effects of a 1 percent permanent negative GDP shock in China - Global aggregate effects (GVAR estimated over 1981Q1–2013Q1): - Global growth reduces by0:23percentage points in the short-run following a one percent permanent negative Chinese GDP shock. - Oil prices fall by around2:8%in the long run. - Evidence of falls in global inflation and short-term interest rates; the median effect on global equity prices is negative but not statistically significant. - Country/region-specific average annual output elasticities over the first year (following a 1% permanent negative Chinese GDP shock): - ASEAN-5 countries (Indonesia, Malaysia, Philippines, Singapore, Thailand): large and statistically significant impacts for all except the Philippines; output elasticities range between0:23and0:35percent. - Other Asia-Pacific countries (except India): statistically significant elasticities ranging between0:06and0:17percent. - India: negligible spillover (output reduction less than0:1 percent after one year), reflecting weak trade links with China. - Systemic economies (average elasticities): euro area 0:12percent, U.K. 0:04percent, U.S. 0:07percent. - Robustness to alternative weight constructions: - Elasticity magnitudes are very similar when foreign variables x*_it are constructed using trade weights (Trade), trade in value added weights (TiVa), or mixed (trade for y*_it and π*_it; financial weights for eq*_it, r*_Sit, r*_Lit). Choice of weights is of second-order importance; trade weights retained for main results. ### Effects of a surge in global financial market volatility (FSI_t shock) - Shock definition: - A one standard deviation increase in the financial stress index (FSI_t) is used to represent a surge in global financial market volatility. - Global aggregate impacts (short-run, first four quarters unless noted): - World output falling by around0:29%below the pre-shock level on average over the first year. - Negative peak effects in the first quarter after the shock: global equity prices 3:7%, oil prices 6:5%, long-term interest rates 0:03%. - The FSI_t index measures price movements relative to trend with a historical average value of zero; the magnitude of the shock is comparable to the 2002 episode of market volatility in advanced economies and is much smaller than the Global Financial Crisis shock. - Country-level responses (average output responses over the first year): - Asia: output falls between0:17and0:47percent below the pre-shock level; effects statistically significant for all countries except Korea. - China: real output falls by around0:17percent on average over the first year. - Japan: real output falls by around0:37percent on average over the first year. - Australia and New Zealand (commodity exporters): output falls by0:17and0:21percent respectively. - India (large commodity importer): real output falls by about0:19percent on average over the first year. - Euro area, U.K., U.S.: output falls by0:32,0:24and0:31percent respectively. - Robustness and channel identification: - Heterogeneities reflect trade and financial linkages. - Alternative configurations of cross-country linkages (bilateral trade weights, trade in value weights, mixed trade-financial weights) yield generally similar magnitudes and directions of impulse responses. ### Time variation and comparative magnitudes - Median generalized impulse responses to a one standard deviation fall in China’s GDP are reported quarterly; figures show larger median effects for 2002/07/12 weights versus 1980s/1990s weights. - Estimates based on trade weights from the 1980s and 1990s are not statistically significant, whereas estimates based on the 2002/07/12 weights are significant. - Median effects of a Chinese GDP shock are generally much larger than those of three decades ago, reflecting China’s growing role in the global economy and evolving trade linkages. ### Policy-relevant implications - China’s slowdown generates negative spillovers to the rest of the world, with considerable cross-country heterogeneity and larger effects for countries with large trade exposures to China. - Growth-spillover elasticities have become larger over time due to evolving trade patterns and China’s expanding role in world commodity markets. - A sharp increase in global financial market volatility can lead to a short-run global growth decline of around0:29percentage points and significant negative spillovers to emerging market economies through trade and financial channels. - Policy considerations implied by results: - Countries with strong trade exposures to China and those sensitive to commodity-price fluctuations should account for larger spillovers from Chinese growth shocks. - Sound fundamentals and policy frameworks are important but may not fully insulate countries from sudden global financial market volatility, especially amid monetary policy normalization uncertainty in advanced economies. - Monitoring and preparing for commodity-price and financial-stress channels can help mitigate short-run adverse effects on output and inflation. *Source: IMF Working Paper content unit _wp1663 - References and related figures/results as reported in the source.* ### References ............................................................................................................. ### _wp1663 - References ............................................................................................................... ### Figures - 1. Average Output Elasticities Over the First Year Following a Negative GDP Shock in China (Using Various Weights) ........................................................................................................ 12 - 2. Average Output Elasticities Over the First Year Following a Negative GDP Shock in China (Using Time-Varying Bilateral Trade Weights) .................................................................... 13 - 3. Implications of a Negative China GDP Shock for the Global Economy ............................ 14 - 4. Implications of an Increase in Financial Market Volatility for the Global Economy ......... 16 - 5. Average Output Responses to an Increase in Global Financial Market Volatility over the First Year (Using Various Weights)........................................................................................ 16 ### Tables - 1. Countries and Regions in the GVAR Model ......................................................................... 8 *Source: _wp1663 - References.* ### 2. Lag Orders of the Country-Specific VARX*(p,q) Models Together with the Number of ### 2. Lag Orders of the Country-Specific VARX*(p,q) Models Together with the Number of ### Key results and context - The analysis employs a Global VAR (GVAR) framework comprising 26 region-specific VARX* models within a sample of 33 economies, which together cover more than90%of world GDP. - Country-specific VARX*(p_i,q_i) lag orders are selected by the Akaike Information Criterion (AIC) with maximum lag orders set top_max= 2 andq_max= 1. - The univariate financial stress index (FSI_t) model lag order is selected as(2;1)based on the AIC. - Cointegrating relations (r_i) are selected using trace test statistics with 95% critical values from MacKinnon (1991); in some countries r_i was reduced below the trace-statistic suggestion to ensure well-behaved persistence profiles and global-model stability. ### Lag orders and number of cointegrating relations (Table 2 items) - Argentina: p_i 2 q_i 1 relations (r_i) 1 - Australia: p_i 2 q_i 1 relations (r_i) 4 - Brazil: p_i 1 q_i 1 relations (r_i) 1 - Canada: p_i 1 q_i 1 relations (r_i) 3 - China: p_i 1 q_i 2 relations (r_i) 2 - Chile: p_i 2 q_i 1 relations (r_i) 3 - Euro Area: p_i 2 q_i 1 relations (r_i) 2 - India: p_i 2 q_i 1 relations (r_i) 2 - Indonesia: p_i 2 q_i 1 relations (r_i) 3 - Japan: p_i 1 q_i 1 relations (r_i) 4 - Korea: p_i 1 q_i 3 relations (r_i) 3 - Malaysia: p_i 1 q_i 2 relations (r_i) 2 - Mexico: p_i 1 q_i 3 relations (r_i) 3 - Norway: p_i 2 q_i 1 relations (r_i) 3 - New Zealand: p_i 2 q_i 1 relations (r_i) 3 - Peru: p_i 2 q_i 1 relations (r_i) 2 - Philippines: p_i 2 q_i 1 relations (r_i) 2 - South Africa: p_i 2 q_i 1 relations (r_i) 1 - Saudi Arabia: p_i 2 q_i 1 relations (r_i) 1 - Singapore: p_i 2 q_i 1 relations (r_i) 2 - Sweden: p_i 2 q_i 1 relations (r_i) 3 - Switzerland: p_i 2 q_i 1 relations (r_i) 2 - Thailand: p_i 1 q_i 1 relations (r_i) 2 - Turkey: p_i 1 q_i 3 relations (r_i) 3 - UK: p_i 1 q_i 3 relations (r_i) 3 - USA: p_i 2 q_i 1 relations (r_i) 2 ### Model specification and variables - Country domestic variable vector x_it includes up to six variables: y_it (log real GDP), π_it (inflation), eq_it (log real equity prices), r_Sit (short-term interest rate), r_Lit (long-term interest rate), ep_it (real exchange rate). - Foreign variables x*_it are cross-sectional averages constructed with bilateral weights w_ij; alternative weights explored include trade, trade in value added (TiVa), and financial weights. - The United States model excludes foreign financial variables eq*_US,t, r*_SUS,t, and r*_LUS,t; the U.S. model includes e*_US,t - p*_US,t as a weakly exogenous foreign variable instead of the real exchange rate. - A common financial stress index (FSI_t) is included as a weakly exogenous common variable in all country models, with allowance for feedback effects from macro variables; oil price (log nominal oil price in U.S. dollars) is also included as a common variable. ### Empirical findings: effects of a one percent permanent negative GDP shock in China - Global aggregate effects (GVAR estimated over 1981Q1–2013Q1): - Global growth reduces by0:23percentage points in the short-run following a one percent permanent negative Chinese GDP shock. - Oil prices fall by around2:8%in the long run. - Evidence of falls in global inflation and short-term interest rates; the median effect on global equity prices is negative but not statistically significant. - Country/region-specific average annual output elasticities over the first year (following a 1% permanent negative Chinese GDP shock): - ASEAN-5 countries (Indonesia, Malaysia, Philippines, Singapore, Thailand): large and statistically significant impacts for all except the Philippines; output elasticities range between0:23and0:35percent. - Other Asia-Pacific countries (except India): statistically significant elasticities ranging between0:06and0:17percent. - India: negligible spillover (output reduction less than0:1 percent after one year), reflecting weak trade links with China. - Systemic economies (average elasticities): euro area 0:12percent, U.K. 0:04percent, U.S. 0:07percent. - Robustness to alternative weight constructions: - Elasticity magnitudes are very similar when foreign variables x*_it are constructed using trade weights (Trade), trade in value added weights (TiVa), or mixed (trade for y*_it and π*_it; financial weights for eq*_it, r*_Sit, r*_Lit). Choice of weights is of second-order importance; trade weights retained for main results. ### Effects of a surge in global financial market volatility (FSI_t shock) - A one standard deviation shock to the financial stress index translates into: - World output falling by around0:29%below the pre-shock level on average over the first year. - Negative peak effects in the first quarter after the shock: global equity prices 3:7%, oil prices 6:5%, long-term interest rates 0:03%. - The FSI_t index measures price movements relative to trend with a historical average value of zero; the magnitude of the shock is comparable to the 2002 episode of market volatility in advanced economies and is much smaller than the Global Financial Crisis shock. *Source: Author’s estimations as reported in the specified content unit.* ### 2012. Figure 2 shows that based on trade weights more than 30 years ago, a negative Chinese ### _wp1663 - 2012. Figure 2 shows that based on trade weights more than 30 years ago, a negative Chinese ### A. Global effects of a negative Chinese GDP shock - Median generalized impulse responses are computed to a one standard deviation fall in China’s GDP; impact reported in percentage and horizon is quarterly. - Global growth impact: - Following a permanent1%negative Chinese GDP shock, global growth reduces by0:23%in the short run. - Commodity prices and channels: - Oil prices fall by2:8percent below their pre-shock levels in response to a negative China output shock. - China’s rebalancing affects commodity-exporting economies mainly through global demand for commodities and associated prices, translating into lower overall economic growth for those countries. - Other macro variables: - There is evidence for a short-run fall in both global inflation and short-term interest rates. - The median effect on global equity prices is negative but not statistically significant. - Time-varying spillovers: - Estimates based on trade weights from the 1980s and 1990s are not statistically significant, whereas estimates based on the 2002/07/12 weights are significant. - Median effects of a Chinese GDP shock are generally much larger than those of three decades ago, reflecting China’s growing role in the global economy and evolving trade linkages. - Cross-country heterogeneity (short-run elasticities / percentage points): - ASEAN-5 countries (except the Philippines): growth elasticities range between0:23and0:35percentage points (largest impact). - Other Asia-Pacific countries (except India): effects statistically significant and fall between0:06and0:17percentage points. - Systemic countries (average elasticities): euro area0:12, U.K.0:04, U.S.0:07percentage points. ### B. Effects of a surge in global financial market volatility - Shock definition: - A one standard deviation increase in the financial stress index (FSI_t) is used to represent a surge in global financial market volatility. - Global aggregate impacts (short-run, first four quarters unless noted): - World output falls on average by0:29%below the pre-shock level in the first four quarters after the surge; this effect is statistically significant. - Oil prices fall by6:5% (peak effect in the first quarter). - Global equity prices fall by3:7% (peak effect in the first quarter). - Long-term interest rates fall by0:03% (peak effect in the first quarter). - Widening of the output gap and lower commodity prices likely moderate global inflation slightly. - Country-level responses (average output responses over the first year; 16th and 84th percentile error bands reported in figures): - Asia: output falls between0:17and0:47percent below the pre-shock level; effects statistically significant for all countries except Korea. - China: real output falls by around0:17percent on average over the first year. - Japan: real output falls by around0:37percent on average over the first year. - Australia and New Zealand (commodity exporters): output falls by0:17and0:21percent respectively, reflecting the commodity-price channel. - India (large commodity importer): real output falls by about0:19percent on average over the first year. - Euro area, U.K., U.S.: output falls by0:32,0:24and0:31percent respectively. - Robustness and channel identification: - Heterogeneities reflect trade and financial linkages. - Alternative configurations of cross-country linkages (bilateral trade weights, trade in value weights, mixed trade-financial weights) yield generally similar magnitudes and directions of impulse responses. ### C. Methodology and scope - Model and sample: - A GVAR model is estimated for 26 countries/regions over the period 1981Q1–2013Q1. - Impulse response reporting: - Figures report median generalized impulse responses together with the 16th and 84th percentile error bands; impact measured in percentage and horizon is quarterly. - Auxiliary notes: - Figures showing statistical significance for different weights are not reported in the text but are available from the authors upon request. ### D. Concluding remarks and policy-relevant implications - Overall findings: - China’s slowdown generates negative spillovers to the rest of the world, with considerable cross-country heterogeneity and larger effects for countries with large trade exposures to China. - Growth-spillover elasticities have become larger over time due to evolving trade patterns and China’s expanding role in world commodity markets. - A sharp increase in global financial market volatility can lead to a short-run global growth decline of around0:29percentage points and significant negative spillovers to emerging market economies through trade and financial channels. - Policy implications (implicit from results): - Countries with strong trade exposures to China and those sensitive to commodity-price fluctuations should account for larger spillovers from Chinese growth shocks. - Sound fundamentals and policy frameworks are important but may not fully insulate countries from sudden global financial market volatility, especially amid monetary policy normalization uncertainty in advanced economies. - Monitoring and preparing for commodity-price and financial-stress channels can help mitigate short-run adverse effects on output and inflation. *Italic: IMF Working Paper content unit _wp1663 - 2012 (figures and numerical results as reported in the source).* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1663.pdf_