## wpiea2023206-print-pdf — 1. Introduction

## Source details

**Canonical URL:** [wpiea2023206-print-pdf — 1. Introduction](https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023206-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2023/english/wpiea2023206-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2023/english/wpiea2023206-print-pdf.pdf.json)

---

### Background: China’s growing global role
- China’s share of global production rose from 2 percent in 1995 to 16 percent in 2018.
- Over 2000-14 China accounted for one-third of global growth.
- China’s role as a source of final demand increased from 0.3 percent of global production in 1995 to 2.2 percent by 2018.
- For ASEAN countries the share of output absorbed by Chinese domestic demand rose from 1 percent to 6 percent.
- China’s medium-term growth is projected to slow from a 7.7 percent average during 2010-2019 to a medium-term average of approximately 5 percent (IMF October 2022 WEO).
- Recent slowdowns reflect slowing productivity growth, a rapidly ageing population, and a policy shift from investment towards domestic consumption.
- COVID-19 introduced both supply-side (lockdown-induced capacity constraints) and demand-side (confidence declines, real estate weakness) elements.

### Research gap and contribution
- Existing studies largely estimate cross-country spillovers from China’s GDP growth without distinguishing demand versus supply shocks.
- This paper estimates distinct spillovers from supply- versus demand-driven slowdowns in China for a large sample of countries and firms.
- Novel elements:
  - Uses China Cyclical Activity Tracker (CCAT) as a broad proxy of Chinese activity.
  - Distinguishes demand and supply shocks using domestic data and sign restrictions in an SVAR.
  - Provides firm-level spillover estimates using quarterly balance sheet data for about 20,000 firms across 61 countries and 20 sectors.

### Methodology overview
- Activity measure:
  - CCAT is the first principal component of a panel including exports, imports, air passengers, electricity consumption, credit extension, rail use, retail sales, industrial production, government revenue, and highway usage.
  - All series are detrended year-over-year growth rates, normalized (mean zero, unit standard deviation).
- Identification of shocks:
  - SVAR with four variables: CCAT, China CPI, global GDP growth excluding China, global inflation excluding China.
  - Sample period: 2001Q1-2022Q1.
  - Sign restrictions: demand shocks → prices and quantities move in the same direction; supply shocks → prices and quantities move in opposite directions.
  - Domestic shocks imposed to have no impact on global variables on impact.
  - Lags: p = 2 (robust to 1 to 8 lags).
- Macro spillovers:
  - Panel local projection framework (Jordà 2005) for 50 advanced and emerging countries.
  - Shocks scaled to be equivalent to a 1 percent decline in Chinese GDP (CCAT standardized to unit standard deviation; scaled by 1/2.35 to match 1 percent of headline GDP).
  - Controls include financial conditions indices (excluding China), VIX, and three lags of world export-weighted GDP growth (excluding China).
- Firm-level analysis:
  - Quarterly firm balance sheet and revenue data from Capital IQ for about 20,000 firms across 61 countries and 20 sectors.
  - Flexible specification distinguishing demand-output, demand-input, supply-output, and supply-input channels.
  - Interaction with export share to China and measures of sector dependence on Chinese inputs versus final demand.

### Key SVAR and shock characteristics
- Domestic China impulse responses:
  - Demand shocks: on-impact effect on activity about 0.6 percent; persistence about 4 quarters.
  - Supply shocks: on-impact effect on activity about 0.45 percent; persistence about 6 quarters.
  - Negative demand shocks → short-term decline in inflation; negative supply shocks → short-term increase in inflation.
- SVAR global response:
  - Peak effect on global growth to Chinese supply shocks: 0.45 percent.
  - Peak effect on global growth to Chinese demand shocks: 0.34 percent.
- Historical decomposition (CCAT):
  - Demand shocks dominant until recently (e.g., GFC decline and rebound mainly driven by domestic demand).
  - From 2015 and especially during COVID-19, supply shocks played a more prominent role, explaining the majority of lockdown and reopening movements.
  - Weak demand weighed on activity from 2021Q2-2022Q1.

### Aggregate spillover findings (local projections)
- Shock scaling: shocks equivalent to a 1 percent decline in Chinese GDP.
- CCAT (headline activity) spillovers:
  - On-impact: neither real GDP nor investment in other countries respond significantly on impact.
  - After 2 years: cumulative impact is negative but small.
  - After 4 years: output falls by 0.5 percent and investment falls by 1 percent.
- Headline activity slowdown summary:
  - Peak impact of 0.6 percent after 2 years and moderates to 0.3 percent after 4 years.
- Structural demand shock spillovers:
  - Very similar to CCAT-based spillovers: delayed impact on GDP and investment with a similar size after 4 years.
- Structural supply shock spillovers:
  - Larger and more front-loaded than demand shocks.
  - Supply shocks lead to larger and quicker declines in GDP and investment (decline in GDP significantly negative after the first year).
  - Real investment shows a peak fall of around 1.5 percent after 2 years, moderating to 1 percent after 4 years.
- Labor market and inflation:
  - The unemployment rate increases broadly similarly across shocks and cumulates to approximately 0.15 percentage point over four years.
  - No significant evidence of disinflation following a demand shock; (negative) supply shocks produce a moderate increase in inflation in the first six quarters.

### Conditional results (export exposure to China)
- Interaction with export exposure (equation (4)):
  - For the headline activity measure, response of real GDP in partner countries to a slowdown in China is around 0.15 percent larger when accounting for export exposure.
  - For the supply shock, the additional peak impact is 0.2 percent.
  - For the demand shock, little additional impact on baseline and some evidence the medium-term effect may be slightly weaker for countries with higher export exposure.
- Aggregate export exposure conflates final-consumption suppliers and intermediate-input suppliers; firm-level analysis is used to disentangle these channels.

### Robustness of aggregate results
- Alternative local-projections specifications:
  1. No control variables except country fixed effects.
  2. Full model excluding the correction for future shocks.
  3. Full model excluding lags of past shocks.
- Supply shock impact broadly similar across specifications: impact felt within the first year and peak effect in region of 0.5 percent.
- Demand shock results less stable across specifications; baseline that controls for past and future shocks is preferred.
- A larger SVAR including investment and consumption preserves result that supply shocks have dominated Chinese activity since around 2014, especially around lockdowns.

### Firm-level data and measures
- Main dataset: S&P Capital IQ (CIQ), quarterly frequency.
- Sample period and coverage:
  - 2001Q3 to 2021Q4.
  - 63 countries (29 AEs and 34 EMDEs).
  - More than 20,000 firms after filtering.
  - 20 CIQ-defined industries (excluding financial, insurance and utilities).
- Firm-level variables:
  - Main variable: total revenue (IQ_TOTAL_REV).
  - Investment measure: capital expenditure (IQ_CAPEX).
- Data cleaning and summary statistics:
  - Revenue (yoy, %): No. of Obs. 1,528,030; Mean 6.84; Std. Dev. 41.61; 25th Pctile -7.69; Median 6.06; 75th Pctile 21.74.
  - Capital Expenditure (yoy, %): No. of Obs. 742,350; Mean 4.95; Std. Dev. 119.52; 25th Pctile -57.98; Median 4.44; 75th Pctile 67.11.
  - Investment Ratio (yoy, %): No. of Obs. 725,623; Mean -0.49; Std. Dev. 7.86; 25th Pctile -2.06; Median -0.08; 75th Pctile 1.498.
- Trade exposure measures derived from MRIO tables (ADB) and WIOD augmentation:
  - Country-industry export share to China, input dependence coefficients (input share supplied by China), output dependence coefficients (share of global demand for country-industry products coming from China).
  - Linkage measures are de-trended and standardized so results reflect one standard-deviation differences.

### Firm-level empirical specifications
- Local projection method (Jordà 2005) used to estimate revenue and investment effects and heterogeneity by trade linkages.
- Controls include firm-quarter dummies, shocks in preceding year and future year, four lags of dependent variable, financial conditions indices (IMF GFSR, excluding China), VIX, and three lags of world export-weighted GDP growth.
- Conditional specifications interact shocks with export ties to China and with input/output linkage measures to estimate four channels: demand-output, demand-input, supply-output, supply-input.

### Firm-level spillover findings
- Average firm revenue responses to a negative China shock equivalent to 1 percent of GDP:
  - Peak impact after 8 quarters:
    - Supply shock: -1.4 percent.
    - Demand shock: -1.6 percent.
  - Direct impacts of supply and demand shocks on revenue are not significantly different in short and medium terms and are similarly persistent.
- Firm investment responses:
  - Impact of both shocks on investment is less persistent than for revenue.
  - Supply shock impact on investment is substantially larger and twice as persistent, with a peak decline of almost 3 percent after 7 quarters.
- Heterogeneity by export exposure:
  - Following a negative demand shock, firms in sectors with higher exports to China experienced larger revenue declines.
  - The decline in revenues is 3.4 percent larger for firms at the 90th percentile of export exposure than for firms at the 10th percentile.
  - Export exposure plays a smaller and shorter-lived role for supply shocks (significant only through the 5th quarter).
- GVC linkage channels:
  - Demand shocks: larger negative impacts on firms with relatively strong output linkages to China (final-demand channel).
  - Supply shocks: larger negative impacts on firms with relatively strong input linkages to China (intermediate-input channel).
  - Demand-input and supply-output channels are not significant.
  - Magnitudes for a negative supply shock equivalent to 1 percent of Chinese GDP over four years:
    - A firm one standard-deviation more dependent on Chinese demand than average sees revenue fall by 0.2 percentage point more.
    - A firm one standard-deviation more dependent on Chinese inputs than average sees revenue fall by 0.6 percentage point more.

### Firm-level robustness checks
- Robustness variants include dropping controls, excluding future shock corrections, alternative lag specifications (2 and 6 lags), dropping countries with less than 250 firms, dropping the two biggest sectors (Materials and Capital Goods), using median de-trending, and including raw supply and demand shocks.
- Results across robustness checks are qualitatively similar.

### Synthesis, conclusions, and policy implications
- Substantial aggregate and firm-level spillovers from a slowdown in China, concentrated in countries and firms with stronger trade links to China.
- Supply-driven slowdowns in China are particularly significant:
  - Investment responds more and more quickly at both the aggregate and firm levels to supply shocks.
  - Firms in industries closely linked to Chinese suppliers are more strongly affected.
- Conceptual framing:
  - A negative demand shock in the world’s largest source of demand propagates through consumption and financial linkages, whereas a supply shock in the world’s largest producer propagates through production linkages.
- Key conclusions:
  - Chinese supply shocks have faster, larger aggregate and firm-level spillovers than demand shocks, particularly through input linkages in GVCs.
  - Demand spillovers remain substantial for firms with strong sales linkages to China.
  - Over the medium term, stronger supply-side spillovers—together with projections of lower Chinese TFP and labor force growth—are likely to be especially important.
- Suggested avenues for future work:
  - Study spillovers via financial linkages using a similar two-stage approach.
  - Identify supply shocks using sectoral Chinese data to test whether shocks in specific Chinese sectors generate disproportionately large international spillovers.
  - Explore heterogeneous firm responses by characteristics such as debt levels and firm size.

*Source: wpiea2023206-print-pdf — 1. Introduction*

### 1. Introduction

### 1. Introduction

### Background: China’s growing global role
- China’s share of global production rose from 2 percent in 1995 to 16 percent in 2018.
- Over 2000-14 China accounted for one-third of global growth.
- China’s role as a source of final demand increased from 0.3 percent of global production in 1995 to 2.2 percent by 2018.
- For ASEAN countries the share of output absorbed by Chinese domestic demand rose from 1 percent to 6 percent.
- China’s medium-term growth is projected to slow from a 7.7 percent average during 2010-2019 to a medium-term average of approximately 5 percent (IMF October 2022 WEO).
- Recent slowdowns reflect slowing productivity growth, a rapidly ageing population, and a policy shift from investment towards domestic consumption.
- COVID-19 introduced both supply-side (lockdown-induced capacity constraints) and demand-side (confidence declines, real estate weakness) elements.

### Research gap and contribution
- Existing studies largely estimate cross-country spillovers from China’s GDP growth without distinguishing demand versus supply shocks.
- This paper estimates distinct spillovers from supply- versus demand-driven slowdowns in China for a large sample of countries and firms.
- Novel elements:
  - Uses a broad proxy of Chinese activity (CCAT) rather than official GDP alone.
  - Distinguishes demand and supply shocks using domestic data and sign restrictions in an SVAR.
  - Provides firm-level spillover estimates using quarterly balance sheet data for about 20,000 firms across 61 countries and 20 sectors.

### Methodology overview
- Activity measure:
  - China Cyclical Activity Tracker (CCAT) is the first principal component of a panel including exports, imports, air passengers, electricity consumption, credit extension, rail use, retail sales, industrial production, government revenue, and highway usage.
  - All series are detrended year-over-year growth rates, normalized (mean zero, unit standard deviation).
- Identification of shocks:
  - Structural Vector Autoregression (SVAR) with four variables: CCAT, China inflation (CPI), global GDP growth excluding China, and global inflation excluding China.
  - Sample period: 2001Q1-2022Q1.
  - Sign restrictions: demand shocks → prices and quantities move in the same direction; supply shocks → prices and quantities move in opposite directions.
  - Domestic shocks are imposed to have no impact on global variables on impact.
  - Lags: p = 2 (robust to 1 to 8 lags).
- Macro spillovers:
  - Panel local projection framework (Jordà 2005) for 50 advanced and emerging countries.
  - Shocks scaled to be equivalent to a 1 percent decline in Chinese GDP (CCAT standardized to unit standard deviation; scaled by 1/2.35 to match 1 percent of headline GDP).
  - Controls include financial conditions indices (excluding China), VIX, and three lags of world export-weighted GDP growth (excluding China).
- Firm-level analysis:
  - Quarterly firm balance sheet and revenue data from Capital IQ for about 20,000 firms across 61 countries and 20 sectors.
  - Estimation of firm-level analogues of aggregate regressions and a flexible specification distinguishing demand-output, demand-input, supply-output, and supply-input channels.
  - Interaction with export share to China and measures of sector dependence on Chinese inputs versus final demand.

### Key SVAR and shock characteristics
- SVAR impulse responses (domestic China):
  - Demand shocks: on-impact effect on activity about 0.6 percent; persistence about 4 quarters.
  - Supply shocks: on-impact effect on activity about 0.45 percent; persistence about 6 quarters.
  - Negative demand shocks → short-term decline in inflation; negative supply shocks → short-term increase in inflation.
- SVAR global response:
  - Peak effect on global growth to Chinese supply shocks: 0.45 percent.
  - Peak effect on global growth to Chinese demand shocks: 0.34 percent.
- Historical decomposition (CCAT):
  - Demand shocks dominant until recently (e.g., GFC decline and rebound mainly driven by domestic demand).
  - From 2015 and especially during COVID-19, supply shocks played a more prominent role, explaining the majority of lockdown and reopening movements.
  - Weak demand weighed on activity from 2021Q2-2022Q1.

### Aggregate spillover findings (local projections)
- Shock scaling: shocks are scaled to be equivalent to a 1 percent decline in Chinese GDP.
- CCAT (headline activity) spillovers:
  - On-impact: neither real GDP nor investment in other countries respond significantly on impact.
  - After 2 years: cumulative impact is negative but small.
  - After 4 years: output falls by 0.5 percent and investment falls by 1 percent.
- Structural demand shock spillovers:
  - Very similar to CCAT-based spillovers: delayed impact on GDP and investment with a similar size after 4 years.
- Structural supply shock spillovers:
  - Larger and more front-loaded than demand shocks.
  - Supply shocks lead to larger and quicker declines in GDP and investment (decline in GDP significantly negative after the first year).

### Firm-level spillover findings
- Aggregate firm response:
  - Two years after a demand shock equivalent to a 1 percent decline in Chinese GDP, average firm revenues across countries decline by 1.6 percent.
  - Two years after a supply shock equivalent to a 1 percent decline in Chinese GDP, average firm revenues across countries decline by 1.4 percent.
  - The decline is slightly faster for supply shocks than demand shocks.
- Trade-link amplification:
  - Interacting shocks with the export share to China of each firm’s industry shows that stronger trade linkages amplify negative spillovers.
- Supply vs. demand through production networks:
  - Flexible specification estimating marginal impacts for demand-output, demand-input, supply-output, supply-input:
    - Demand shocks have negative impacts on firms with strong output linkages to China (final-demand channel).
    - Supply shocks have negative impacts on firms with strong input linkages to China (intermediate-input channel).
    - The other two channels are statistically insignificant.
  - Magnitudes:
    - The total impact after four years of Chinese supply shocks through input linkages is roughly three times the size of the impact of Chinese demand shocks through output linkages.
    - A firm in a sector that is one standard-deviation more dependent on Chinese inputs than average sees revenue fall by 0.6 percentage points more in response to a negative supply shock equivalent to a 1 percent decline in Chinese GDP.

### Synthesis and interpretation
- Substantial aggregate and firm-level spillovers from a slowdown in China, concentrated in countries and firms with stronger trade links to China.
- Supply-driven slowdowns in China are particularly significant:
  - Investment responds more and more quickly at both the aggregate and firm levels to supply shocks.
  - Firms in industries closely linked to Chinese suppliers are more strongly affected.
- Conceptual framing:
  - Analogous but inverted counterpart to the ‘global financial cycle’ (Rey, 2013): a negative demand shock in the world’s largest source of demand propagates through consumption and financial linkages, whereas a supply shock in the world’s largest producer propagates through production linkages.

### Relation to existing literature
- Aggregate spillovers in this paper are in the upper range of existing estimates (literature estimates spillovers from a 1 percentage point decline in Chinese growth in the range 0.15 to 0.8 percentage points).
- Examples from literature (ranges and point estimates reported in source):
  - GVAR and related estimates: 0.2 percentage point global decline (Cashin and others 2016; Dizioli and others 2016).
  - Duval and others (2014): spillovers to Asia of 0.3 percent vs. 0.15 percent for weaker value-added trade link economies.
  - Furceri, Jalles, and Zdzienicka (2017): average decline of 0.4 percent in GDP after 3 years.
  - Barcelona and others (2022): 0.3 percent increase in global GDP from a 1 percent of GDP expansion of credit in China.
  - Huidrom and others (2020): global spillover of 0.8 percent for China.
  - Sznajderska and Kapuscinski (2020): Bayesian VAR estimates range 0 to 1.4 percent; GVAR estimates range 0 to 0.5 percent.
- Firm-level literature is sparse; comparable findings exist in Ahuja and Nabar (2012) for industrial production and stock prices. The paper’s finding that Chinese supply shocks impact downstream customer firms aligns with production network literature (e.g., Acemoglu and others, 2012).

*Source: wpiea2023206-print-pdf - 1. Introduction*

### 0.3 percent and peaks at a 0.6 percent after 2 years. After 4 years the impact is more moderate at

### wpiea2023206-print-pdf - 0.3 percent and peaks at a 0.6 percent after 2 years. After 4 years the impact is more moderate at

### Key findings on aggregate spillovers
- A headline activity slowdown in China produces a peak impact of 0.6 percent after 2 years and moderates to 0.3 percent after 4 years.
- Real investment shows a more persistent response: a peak fall in investment of around 1.5 percent after 2 years, moderating to 1 percent after 4 years.
- The unemployment rate increases broadly similarly across shocks and cumulates to approximately 0.15 percentage point over four years.
- No significant evidence of disinflation following a demand shock; (negative) supply shocks produce a moderate increase in inflation in the first six quarters.

### Conditional results (export exposure to China)
- Interacting shocks with export exposure to China (equation (4)) yields additional impacts relative to baseline:
  - For the headline activity measure, the response of real GDP in partner countries to a slowdown in China is around 0.15 percent larger when accounting for export exposure.
  - For the supply shock, the additional peak impact is 0.2 percent.
  - For the demand shock, little additional impact on baseline and some evidence the medium-term effect may be slightly weaker for countries with higher export exposure.
- Aggregate export exposure conflates final-consumption suppliers and intermediate-input suppliers; firm-level analysis is used to disentangle these channels.

### Robustness of aggregate results
- Three alternative local-projections specifications were estimated:
  1. Model with no control variables except country fixed effects.
  2. Full model excluding the correction for future shocks.
  3. Full model excluding lags of past shocks.
- Supply shock impact is broadly similar across specifications: impact felt within the first year and peak effect in region of 0.5 percent.
- Demand shock results are less stable:
  - Simple model without controls indicates short-term spillovers similar to long-run baseline (0.3 percent v. 0.4 percent in baseline).
  - Controlling for past but not future shocks indicates a positive response to a negative demand shock in first two years before baseline long-run impact is recovered.
  - Specification (3) that controls for future shocks is broadly similar to baseline and avoids the implausible positive response.
- The reduced form CCAT shock may produce faster impacts, but potential endogeneity suggests baseline (which controls for past and future CCAT) is more reliable.
- A larger SVAR including investment and consumption preserves the result that supply shocks have dominated Chinese activity since around 2014, especially around lockdowns; addition of consumption and investment reduces the contribution of generic demand shocks.

### Firm-level analysis: scope and data
- Main firm dataset: S&P Capital IQ (CIQ), quarterly frequency.
- Sample period and coverage:
  - 2001Q3 to 2021Q4.
  - 63 countries (29 AEs and 34 EMDEs).
  - More than 20,000 firms after filtering.
  - 20 CIQ-defined industries (after excluding financial, insurance and utilities sectors).
- Firm-level variables:
  - Main variable: total revenue (IQ_TOTAL_REV).
  - Investment measure: capital expenditure (IQ_CAPEX).
- Data cleaning:
  - Remove firms with negative assets or debt in any year.
  - Observations with incorrect sign for revenue or capital expenditure set to missing.
  - Winsorize firm-level variables at the 1st and 99th percentiles.
- Trade exposure measures derived from MRIO tables (ADB) and WIOD augmentation:
  - Country-industry export share to China: 푋...
  - Input dependence coefficients (input share supplied by China): 푖푛푝푢푡...
  - Output dependence coefficients (share of global demand for country-industry products coming from China): 표푢푡푝푢푡...
- Linkage measures are de-trended by replacing each variable with fixed-effect averages (훼 푖,푗) to purge time-varying integration trends; standardized across sample so results reflect one standard-deviation differences.

### Empirical specifications (firm-level)
- Local projection method (Jordà 2005) used to estimate revenue effects and heterogeneity by trade linkages.
- Unconditional firm-level revenue response (equation (8)):
  - Dependent variable: log revenue 푦 of firm f in country c and sector i.
  - Shocks: 푒 supply and 푒 demand as identified in Section 2.
  - Controls: firm-quarter dummies, shocks in preceding year and future year, four lags of dependent variable, financial conditions indices (IMF GFSR, excluding China), VIX, three lags of world export-weighted GDP growth.
- Conditional specification by export exposure (equation (9)) includes interaction of shocks with export ties to China and country-time fixed effects.
- Full GVC linkage specification (equation (10)) interacts shocks with both input and output linkage measures to estimate four channels: demand-output, demand-input, supply-output, supply-input.
  - Both country-time and industry-time fixed effects included, along with firm-quarter fixed effects.
  - To reduce collinearity, shocks are orthogonalized each quarter so only two of the four coefficients are estimated at a time (footnote methodology).

### Firm-level results
- Average firm revenue responses to a negative China shock equivalent to 1 percent of GDP:
  - Peak impact after 8 quarters:
    - Supply shock: -1.4 percent.
    - Demand shock: -1.6 percent.
  - Direct impacts of supply and demand shocks on revenue are not significantly different in short and medium terms and are similarly persistent.
- Firm investment responses differ:
  - Impact of both shocks on investment is less persistent than for revenue.
  - Supply shock impact on investment is substantially larger and twice as persistent, with a peak decline of almost 3 percent after 7 quarters.
- Heterogeneity by export exposure (specification (9)):
  - Following a negative demand shock, firms in sectors with higher exports to China experienced larger revenue declines.
  - The decline in revenues is 3.4 percent larger for firms at the 90th percentile of export exposure than for firms at the 10th percentile.
  - Export exposure plays a smaller and shorter-lived role for supply shocks (significant only through the 5th quarter and smaller magnitude).
- GVC linkage channels (equation (10), Figure 12):
  - Demand shocks: persistently larger negative impacts on firms with relatively strong output linkages to China.
  - Supply shocks: persistently larger negative impacts on firms with relatively strong input linkages to China.
  - Demand-input and supply-output channels are not significant.
  - Magnitudes for a negative supply shock equivalent to 1 percent of Chinese GDP over four years:
    - A firm one standard-deviation more dependent on Chinese demand than average sees revenue fall by 0.2 percentage point more.
    - A firm one standard-deviation more dependent on Chinese inputs than average sees revenue fall by 0.6 percentage point more.

### Firm-level robustness checks
- Variants shown in appendices:
  - Drop all controls from equation (8) (Figure A4.1).
  - Exclude Teulings and Zubanov (2014) future correction shocks (Figure A4.2).
  - Omit four-quarter lags of shocks (Figure A4.3).
  - Drop countries with less than 250 firms (Figure A4.5).
  - Drop the two biggest sectors (Figure A4.6).
  - Use 2 and 6 lags of dependent variable and shocks (Figures A4.7 and A4.8).
  - Take the simple median of time-varying input/output coefficients instead of regression-based de-trending (Figure A4.9).
  - Include raw (non-orthogonalized) supply and demand shocks in every period of specification (10) (Figure A4.10).
- Results across robustness checks are qualitatively similar.

### Conclusion and policy implications
- Main contributions:
  - Use of a broad measure of domestic activity in China.
  - Characterization of spillovers by shock type (supply vs demand).
  - Analysis at both aggregate and firm levels across advanced and emerging economies.
- Key conclusions:
  - Chinese supply shocks have faster, larger aggregate and firm-level spillovers than demand shocks, particularly through input linkages in GVCs.
  - Demand spillovers remain substantial for firms with strong sales linkages to China, relevant given Chinese property sector weakness and investment slowdown.
  - Over the medium term, stronger supply-side spillovers—together with projections of lower Chinese TFP and labor force growth—are likely to be especially important.
- Suggested avenues for future work:
  - Study spillovers via financial linkages using a similar two-stage approach.
  - Identify supply shocks using sectoral Chinese data to test whether shocks in specific Chinese sectors generate disproportionately large international spillovers.
  - Explore heterogeneous firm responses by characteristics such as debt levels and firm size.

*Source: wpiea2023206-print-pdf (extracted content).*

### References

### wpiea2023206-print-pdf - References

### References (selected entries as in source)
- Acemoglu, D., Carvalho, V.M., Ozdaglar, A., Tahbaz-Salehi, A. (2012). “The network origins of aggregate fluctuations”. Econometrica.
- Acemoglu, D., Akcigit, U., Kerr., W. (2015). “Networks and the Macroeconomy: An Empirical Exploration”. NBER Macroeconomics Annual.
- Acemoglu, D., Ozdaglar, A., Tahbaz-Salehi, A. (2016). “Networks, shocks, and systemic risk”. The Oxford Handbook of the Economics of Networks, Oxford University Press, pp. 568-608.
- Ahuja, A., and Nabar, M., (2012). “Investment-Led Growth in China: Global Spillovers”, IMF Working Papers 2012/267, International Monetary Fund.
- Alesina, A., Favaro, C. and Giavazzi, F. (2014). “The output effect of fiscal consolidation plans”. Journal of International Economics, 96, pp.S19-S42
- Arbatli Saxegaard, E. C., Firat, M., Furceri, D., Verrier, J. (2022). U.S. Monetary Policy Shock Spillovers: Evidence from Firm-Level Data. International Monetary Fund Working Paper
- Autor, Dorn and Hanson, (2013). “The China syndrome: Local labor market effects of import competition in the United States”. American Economic Review.
- Barcelona, William L., Danilo Cascaldi-Garcia, Jasper J. Hoek and Eva Van Leemput (2022). “What Happens in China Does Not Stay in China,” International Finance Discussion Papers 1360. Washington: Board of Governors of the Federal Reserve System, https://doi.org/10.17016/IFDP.2022.1360.
- Bloom, Draca and Van Reenen, (2016). “Trade induced technical change? The impact of Chinese imports on innovation, IT and productivity”. The Review of Economic Studies.
- Cashin, Paul, Mohaddes, Kamiar and Raissi, Mehdi (2016) “"China’s slowdown and global financial market volatility: is world growth losing out?”" Globalization Institute Working Papers 270, Federal Reserve Bank of Dallas.
- Dauth, Findeisen and Suedekum (2017). “Trade and manufacturing jobs in Germany”. American Economic Review.
- Dieppe, Alistair and Legrand, Romain and van Roye, Björn (2016), The BEAR Toolbox. ECB Working Paper No. 1934, Available at SSRN: https://ssrn.com/abstract=2811020 or http://dx.doi.org/10.2139/ssrn.2811020
- Duval, R., K. Cheng, K. H. Oh, R. Saraf, and D. Seneviratne (2014). Trade Integration and Business Cycle Synchronization: A Reappraisal with Focus on Asia. IMF Working Paper WP/14/52.
- Fernald, John, Eric Hsu, and Mark Spiegel. (2021) “Is China Fudging its GDP Figures? Evidence from Trading Partner Data.” Journal of International Money and Finance, 114 (June): 102406.
- Forbes, Kristin, Hjortsoe, Ida, and Nenova, Tsvetelina, (2018) “The shocks matter: Improving our estimates of exchange rate pass-through”, Journal of International Economics, Volume 114, Pages 255-275, ISSN 0022-1996, https://doi.org/10.1016/j.jinteco.2018.07.005.
- Furceri, Davide, João Tovar Jalles, and Aleksandra Zdzienicka. (2017) “China Spillovers: New Evidence from Time-Varying Estimates.” Open Economies Review, 28: 413–29.
- Huidrom, R, Ayhan Kose, M, Matsuoka, H, Ohnsorge, FL. ( 2020) How important are spillovers from major emerging markets? International Finance, 23: 47– 63.
- Iacovone, Rauch and Winters, (2013). “Trade as an engine of creative destruction: Mexican experience with Chinese competition”. Journal of International Economics.
- International Monetary Fund (IMF). (2022a). “People’s Republic of China: 2021 Article IV Consultation-Press Release; Staff Report; and Statement by the Executive Director for the People’s Republic of China”, International Monetary Fund, Washington, DC.
- International Monetary Fund (IMF). (2022b). “Regional Economic Outlook for Asia and the Pacific”, October 2022, International Monetary Fund, Washington, DC.
- International Monetary Fund (IMF). (2022c). “World Economic Outlook”, October 2022, International Monetary Fund, Washington, DC
- International Monetary Fund (IMF). (2023). “People’s Republic of China: Selected Issues”, February 2023, International Monetary Fund, Washington, DC.
- Lane, N. (forthcoming). “Manufacturing Revolutions: Industrial Policy and Industrialization in South Korea”. Quarterly Journal of Economics.
- Maliszewski , Wojciech & Dizioli , Allan and Hunt, Benjamin L. (2016).“"Spillovers from the Maturing of China’s Economy”" IMF Working Papers 2016/212, International Monetary Fund.
- Miranda-Agrippino, Silvia, and Helene Rey (2020). “U.S. Monetary Policy and the Global Financial Cycle”, Review of Economic Studies, Volume 87, Issue 6, November 2020, Pages 2754–2776.
- Rey, Helene (2013). “Dilemma not Trilemma: The Global Financial Cycle and Monetary Policy Independence”, Jackson Hole Conference Proceedings.
- Sinclair, Tara (2012). “Characteristics and Implications of Chinese Macroeconomic Data Revisions.” Manuscript, George Washington University. http://www.gwu.edu/~iiep/assets/docs/papers/Sinclair_IIEPWP2012-09.pdf.
- Sznajderska, A, Kapuściński, M. (2020), “Macroeconomic spillover effects of the Chinese economy.” Rev Int Econ., 28: 992– 1019. https://doi.org/10.1111/roie.12479
- Teulings, C.N. and Zubanov, N. (2014), Is Economic Recovery a Myth? Robust Estimation of Impulse Responses. J. Appl. Econ., 29: 497-514. https://doi.org/10.1002/jae.2333

### Figures, Tables, and Annex — Key data and methodological details
- Figure 1: Share of global demand met by production in USA and China — time series plotted from 1995 through 2018 (visual; sources: OECD TiVA, IMF staff calculations).
- Figure 2: Share of output absorbed by Chinese domestic demand — ASEAN defined as average across Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam (sources: OECD TiVA, IMF staff calculations).
- Figure 3: China and global growth 2001-2027 — sources: IMF WEO October 2022, IMF staff calculations (chart includes Global, Europe, North America, ASEAN on right axis; labels include "China growth" and "World growth", and "Forecast").
- Figures 4–6: SVAR impulse responses, historical decomposition, and demand/supply shocks — notes cite IMF WEO October 2022; Fernald, Hsu, and Spiegel (2021); IMF staff calculations.
- Figures 7–12: Firm- and country-level responses to a 1 percent of GDP shock in China:
  - Repeated note: y-axis in percent.
  - Figure 7: Impact of negative 1 percent of GDP shock in China on level of real GDP and investment in other countries — estimation follows equation (3); standard errors clustered by country; shaded areas display the 68% and 90% confidence intervals.
  - Figure 8: Differential by export exposure for impact of a 1 percent of GDP shock in China on real GDP — estimation follows equation (4); standard errors clustered by country; shaded areas display the 68% and 90% confidence intervals.
  - Figures 9–12: Average effects and differentials on firm revenue and firm investment — estimation follows equations (8), (9), and (10); standard errors two-way clustered on firms and country-time or clustered by firm as noted; gray areas display the 90% confidence intervals.
- Annex 1: SVAR Identification — Table A1 sign restrictions (as presented):
  - Variables vs Shocks:
    - Global GDP: Global Supply (+), Global Demand (+), Domestic Supply (0*), Domestic Demand (0)
    - Global CPI: Global Supply (-), Global Demand (+), Domestic Supply (0), Domestic Demand (0)
    - China Activity: Domestic Supply (+), Domestic Demand (+)
    - China CPI: Domestic Supply (-), Domestic Demand (+)
  - Note: * Short-run zero restriction that applies to the first period of the shock.
- Annex 2: Data — Selected tables and figures
  - Table A2.1. Country Sample for Macro Analysis lists categories "Advanced Economies" and "Emerging Market Economies" with country names as in source (including China listed under Emerging Market Economies).
  - Table A2.2. Macro Data Sources — data definitions and sources include:
    - China Cyclical Activity Tracker: Federal Reserve Bank of San Francisco (2021); Developed by Fernald, Hsu, and Spiegel
    - Real GDP: Haver Analytics
    - Consumer price index: Haver Analytics
    - Volatility index: The Chicago Board Options Exchange
    - World export-weighted GDP growth: Haver Analytics; Author’s own calculations
    - Financial conditions indices: IMF GFSR
    - Investment: Haver Analytics
  - Table A2.3. Firm-Level Data Summary Statistics (exact values preserved):
    - Revenue (yoy, %): No. of Obs. 1,528,030; Mean 6.84; Std. Dev. 41.61; 25th Pctile -7.69; Median 6.06; 75th Pctile 21.74
    - Capital Expenditure (yoy, %): No. of Obs. 742,350; Mean 4.95; Std. Dev. 119.52; 25th Pctile -57.98; Median 4.44; 75th Pctile 67.11
    - Investment Ratio (yoy, %): No. of Obs. 725,623; Mean -0.49; Std. Dev. 7.86; 25th Pctile -2.06; Median -0.08; 75th Pctile 1.498
  - Table A2.4. Number of Firms and Observations by Country — selected counts (as listed):
    - United States 5,076
    - India 3,516
    - Japan 2,860
    - Korea 2,043
    - Taiwan 1,822
    - Canada 1,522
    - Hong Kong 1,380
    - Australia 1,339
    - Malaysia 883
    - United Kingdom 873
    - Sweden 709
    - Thailand 628
    - Indonesia 604
    - Poland 541
    - Singapore 499
    - Vietnam 495
    - Germany 488
    - France 461
    - Israel 395
    - Pakistan 313
    - Turkey 294
    - Italy 274
    - Brazil 242
    - Bangladesh 212
    - Philippines 199
    - Switzerland 198
    - Sri Lanka 195
    - Norway 187
    - South Africa 164
    - Egypt 157
    - Greece 147
    - Saudi Arabia 142
    - Spain 140
    - Denmark 137
    - Finland 130
    - Russia 121
    - Jordan 116
    - Chile 109
    - Mexico 101
    - Netherlands 98
    - New Zealand 98
    - Nigeria 95
    - Kuwait 93
    - Belgium 86
    - Cayman Islands 76
    - Peru 75
    - Bulgaria 66
    - Ireland 62
    - Romania 62
    - Oman 61
    - Cyprus 60
    - Mauritius 54
    - Croatia 53
    - United Arab Emirates 52
    - Argentina 50
    - Austria 49
    - Luxembourg 48
    - Tunisia 45
    - Brunei 41
    - Colombia 39
    - Portugal 35
    - Malta 31
    - Hungary 25
  - Table A2.5. Number of Firms and Observations by Sector — counts (as listed):
    - Capital Goods 5,234
    - Materials 5,027
    - Software and Services 2,469
    - Technology Hardware and Equipment 2,363
    - Pharmaceuticals and Biotechnology 2,247
    - Consumer Durables and Apparel 2,185
    - Food, Beverage and Tobacco 2,027
    - Media and Entertainment 1,708
    - Health Care Equipment and Services 1,558
    - Energy 1,531
    - Consumer Services 1,460
    - Retailing 1,423
    - Professional Services 1,307
    - Transportation 1,010
    - Automobiles and Components 932
    - Semiconductors 836
    - Household and Personal Products 468
    - Food and Staples Retailing 412
    - Telecommunication Services 382
  - Figures A2.1 and A2.2: China’s share in total demand and China’s share in total input usage — notes indicate horizontal bars represent share by country-sector pair.
- Annex 3: Robustness — Macro Analysis
  - Figure A3.1: Impact of negative 1 percent of GDP shock in China on the price level — y-axis in percent; results follow estimation of equation (3); standard errors clustered by country; shaded areas display the 68% and 90% confidence intervals.
  - Figure A3.2: Sensitivity of spillovers to GDP to model specification — comparison of specifications labeled "(1) No controls, only fixed effect", "(2) No future shocks correction", "(3) No lagged shocks correction"; y-axis in percent; results follow estimation of equation (3); standard errors clustered by country; shaded areas display the 68% and 90% confidence intervals.
- Annex 4: Robustness — Firm-Level Analysis
  - Figures A4.1–A4.10: Multiple robustness checks for firm revenue and investment responses to a 1 percent of GDP shock in China:
    - Variants include "No Controls", "With Future Shocks Correction", "No Lagged Shocks", "Remove countries for which we have limited firms: dropping countries with less than 250 firms", "Dropping the Two Biggest Sectors: Materials and Capital Goods", different lag specifications (2 lags, 6 lags), using simple median of time-varying linkage coefficients, and including raw supply and demand shocks in every period.
    - Common notes: y-axis in percent; estimations follow equations (8) or (10) as applicable; standard errors two-way clustered on firms and country-time or clustered by firm; gray areas display the 90% confidence intervals.

*China Spillovers: Aggregate and Firm-Level Evidence — Working Paper No. WP/2023/206*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023206-print-pdf.pdf_