## wp18281

## Source details

**Canonical URL:** [wp18281](https://www.imf.org/-/media/files/publications/wp/2018/wp18281.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18281.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18281.pdf.json)

---

### Contributions and key findings
- Paper quantifies the dynamic impact of global trade uncertainty on investment in the euro area, tracing impulse responses over several quarters after an uncertainty shock.
- Focuses on a narrower definition of uncertainty specific to international trade rather than a synthetic economic uncertainty index.
- Identification exploits variation in countries’ exposure to global trade via interaction of a global trade uncertainty index and country trade openness.
- Main empirical findings:
  - Trade uncertainty has a negative impact on investment that stays significant for about a year.
  - The investment-to-GDP ratio is 0.8 percentage points lower for four to five quarters following a standard deviation increase in the level of trade uncertainty in a median trade openness regime (examples: Germany).
  - When trade openness is very high (examples: Ireland, Netherlands), the decline in the investment-to-GDP ratio can reach 1.6 percentage points in the fifth year following the uncertainty shock, before becoming statistically indistinguishable from zero.
- Results are robust to: presence of various control variables, other global shocks, various uncertainty measures, and alternative measures of trade openness that focus on supply chain links.

### Econometric approach and identification strategy
- Objective: quantify how much the investment ratio changes following a sudden increase in trade-related uncertainty and trace dynamic effects.
- Methodology: Local Projections (LP) method (Jorda, 2005) to identify impulse response functions of the investment ratio following a trade uncertainty shock.
- Baseline LP specification (informal description):
  - Dependent variable: change in non-residential investment-to-GDP ratio, y_{i,t+h} = I_{i,t+h} - I_{i,t-1}, with I as non-residential gross fixed capital formation divided by GDP.
  - U: logarithmic measure of global trade uncertainty common to all countries.
  - O: standard trade openness (exports + imports of goods and services normalized by GDP), averaged over several quarters.
  - Coefficients of interest: θ_h O_{i,t-q}, representing conditional impacts of trade uncertainty for a given degree of trade openness at each horizon h.
  - Sample: euro area countries observed over 1998Q1 to 2018Q2.
- Fixed effects and controls:
  - Country-fixed effects (u_i) to control for time-invariant unobservables.
  - Quarter (time) fixed effects (λ_t) to absorb time-varying common factors across countries (e.g., rising protectionism worldwide, geopolitical shocks, oil prices, global financial crisis, global market risk aversion).
  - Interaction term U_t × O_{i,t-q} is the identification crux: uncertainty shock common to all countries interacted with lagged trade openness specific to each country.
  - Lagged trade openness is included additively in X to ensure identification.
  - Trade openness measured as the average of exports and imports in percent of GDP averaged over the past 16 quarters (four years).
  - Other control variables included: quarterly real GDP growth (including five lags), corporate bond spreads, domestic lending rates, 10-year sovereign bond yields, and five lags of the change in the investment ratio.
- Interpretation: treatment effects measured at various levels of trade openness; assumes zero statistical effect of a spike in global trade uncertainty in autarky and a marginal negative effect rising with trade openness.

### Measurement of trade uncertainty
- Trade uncertainty variable created by Hlatshwayo (2018), aggregating references to trade protectionism in newspapers using a four-part article metric:
  1. country name or demonym within eight words of a trade-policy related term;
  2. within eight words of an uncertainty-related term;
  3. excludes mentions of declines in uncertainty, references to alternative policies, or equity-market volatility;
  4. article length must be over 99 words to avoid ticker articles.
- Trade-policy terms include trade war, trade barrier, and protectionism.
- Index normalized by broader country news coverage over time and ranges between 0 and 100.
- Search algorithm rejects articles that mention other types of uncertainty (e.g., fiscal or monetary) to avoid conflation.
- Global measure construction:
  - Aggregate trade uncertainty of the EU, the U.S., China, and the U.K. as a proxy for generalized trade uncertainty.
  - Aggregation performed using principal component analysis (PCA); the first principal component explains 65 percent of the overall variance and is retained as the global measure of generalized trade protectionism uncertainty.
- All uncertainty measures in the paper are rescaled to be comprised between 0 and 100 before applying the natural log transformation ln(1+u).

### Baseline results
- Trade uncertainty increased since the first quarter of 2017 by around 1.4 log points, corresponding to an increase of about 300 percent = 100×(e^{1.4}−1).
- For the median country in terms of trade openness:
  - Estimated reduction in the investment-to-GDP ratio by up to 0.8 percentage points of GDP in the fifth quarter following a typical trade uncertainty spike (middle chart).
  - This corresponds to about one-fifth of the typical standard deviation of the investment-to-GDP ratio in the sample.
  - A typical trade uncertainty spike (of one standard deviation) translates into an increase in trade uncertainty of 0.9 log points, i.e., roughly 146 percent.
  - After five quarters, point estimates are no longer statistically different from zero.
- For countries with highest trade openness regimes (where exports and imports sum to 200 percent of GDP on average):
  - Investment decline can reach 1.6 percentage points of GDP after four quarters.

### Robustness checks
- A. Alternative Measure of Trade Uncertainty: BlackRock Index of Trade Tensions
  - BlackRock index aims to capture market attention to trade risk (heavy weight to brokerage reports).
  - A one standard deviation increase in the BlackRock index corresponds to an increase in trade uncertainty of 0.7 log points, i.e., roughly 100 percent.
  - Using BlackRock index (log-transformed), impact on investment remains statistically significant, negative, and robust up to the fourth quarter following the shock.
  - Estimated decline in investment-to-GDP ratio reaches 0.7 (1.3) percentage points of GDP in the second quarter following the increase in trade uncertainty in the median (high) trade openness regime.
- B. Controlling for Other Non-Linear Forms of Global Uncertainty
  - Augmented specification includes GU (global economic policy uncertainty, in log form) interacted with trade openness.
  - Controlling for GU × trade openness does not materially affect baseline estimates.
  - For the median country in trade openness:
    - Reduction in investment-to-GDP ratio up to 0.7 percentage points of GDP in the fifth quarter following a typical trade uncertainty spike (about one-fifth of typical standard deviation).
  - For the highest trade openness regime:
    - Investment decline can reach 1.3 percentage points of GDP (about 30 percent of the standard deviation of the investment rate in the sample).
- C. Excluding Intra-Euro Area Trade Flows
  - Replacing trade openness with trade exposure outside the euro area yields broadly similar estimates to baseline.
  - Excluding intra-euro area trade flows did not materially change sensitivity of investment-to-GDP to global trade uncertainty.
- D. Alternative Measure of Trade Openness: Participation in Supply Chains
  - Supply chain indicator (Koopman et al., 2010) = share of foreign inputs (backward participation) and domestically produced inputs used in third countries' exports (forward participation), normalized by GDP; data from OECD TiVA.
  - Model substitutes trade openness with SC (averaged over past 16 quarters) and controls for GU × SC.
  - Results echo baseline: sudden increase in trade uncertainty reduces investment ratio for about five quarters.
  - Magnitudes: for median and high levels of supply chain participation, reduction in investment-to-GDP ratio is near 0.5 and 1 percentage points of GDP in the fourth and fifth quarter following the shock, respectively.

### What the estimates imply for recent bouts of uncertainty
- Benchmark measure of trade uncertainty increased by 1.4 log points between Q1 2017 and Q2 2018 (about 300 percent; double the typical standard deviation).
- Using preferred estimates (Section B: controlling for other global uncertainty interacted with trade openness):
  - Investment-to-GDP could be lowered by 1 and 2 percentage points one year from now for countries at the medium and high level of trade openness, respectively.
  - This corresponds to a quarter (half) of the typical standard deviation of the investment-to-GDP ratio in the sample for countries at median (high) trade openness.
- Results conditional on trade talks resuming: trade uncertainty can dissipate quickly and rebound in investment could be very large.

### Conclusion
- Trade uncertainty reduces investment more in countries more dependent on global trade.
- Using text-mined measures and alternative openness metrics (including supply chain participation), the reduction in investment ratio following a surge in trade uncertainty is visible up to four to five quarters after the shock.
- Peak decline in investment:
  - 0.8 percentage points of GDP for median trade openness (roughly a year after the shock).
  - 1.6 percentage points of GDP for high trade openness.
- Policy implication: avoiding protectionism and finding cooperative solutions to promote continued trade growth are essential to preserve near-term growth and long-term production capacity.

### Key statistics (Appendix: Descriptive Statistics)
- Investment-to-GDP: Obs 1729; Mean 11.6; Std. Dev. 3.9; Min -5.2; Max 41.3
- Trade uncertainty (log): Obs 1692; Mean 2.0; Std. Dev. 0.9; Min 0; Max 4.6
- Trade uncertainty (BlackRock) (log): Obs 990; Mean 3.1; Std. Dev. 0.7; Min 0; Max 4.6
- Economic Policy Uncertainty (log): Obs 1566; Mean 111.2; Std. Dev. 41.3; Min 55.6; Max 244.0
- Trade openness: Obs 1476; Mean 104.0; Std. Dev. 52.8; Min 39.0; Max 314.9
- Participation in supply chains: Obs 1728; Mean 21.3; Std. Dev. 11.3; Min 4.7; Max 56.0
- GDP growth: Obs 1674; Mean 5.3; Std. Dev. 6.0; Min -26.8; Max 39.3
- Corporate bond spreads: Obs 966; Mean 0.1; Std. Dev. 2.3; Min -20.3; Max 3.2
- Bank lending rate: Obs 1086; Mean 3.9; Std. Dev. 1.6; Min 1.3; Max 12.4
- 10-year sovereign spread: Obs 1921; Mean 6.0; Std. Dev. 4.0; Min -0.1; Max 25.4

*Source: Unpacking Policy Uncertainty: Evidence from European Firms. IMF Working Paper (forthcoming).*

### 0.2 percentage points of GDP). Using the estimated impact of uncertainty on investment in

### wp18281 - 0.2 percentage points of GDP). Using the estimated impact of uncertainty on investment in

### Contributions and key findings
- Paper quantifies the dynamic impact of global trade uncertainty on investment in the euro area, tracing impulse responses over several quarters after an uncertainty shock.
- Focuses on a narrower definition of uncertainty specific to international trade rather than a synthetic economic uncertainty index.
- Identification exploits variation in countries’ exposure to global trade via interaction of a global trade uncertainty index and country trade openness.
- Main empirical findings:
  - Trade uncertainty has a negative impact on investment that stays significant for about a year.
  - The investment-to-GDP ratio is 0.8 percentage points lower for four to five quarters following a standard deviation increase in the level of trade uncertainty in a median trade openness regime (examples: Germany).
  - When trade openness is very high (examples: Ireland, Netherlands), the decline in the investment-to-GDP ratio can reach 1.6 percentage points in the fifth year following the uncertainty shock, before becoming statistically indistinguishable from zero.
- Results are robust to: presence of various control variables, other global shocks, various uncertainty measures, and alternative measures of trade openness that focus on supply chain links.

### Econometric approach and identification strategy
- Objective: quantify how much the investment ratio changes following a sudden increase in trade-related uncertainty and trace dynamic effects.
- Methodology: Local Projections (LP) method (Jorda, 2005) to identify impulse response functions of the investment ratio following a trade uncertainty shock.
- Baseline LP specification (informal description):
  - Dependent variable: change in non-residential investment-to-GDP ratio, y_{i,t+h} = I_{i,t+h} - I_{i,t-1}, with I as non-residential gross fixed capital formation divided by GDP.
  - U: logarithmic measure of global trade uncertainty common to all countries.
  - O: standard trade openness (exports + imports of goods and services normalized by GDP), averaged over several quarters.
  - Coefficients of interest: θ_h O_{i,t-q}, representing conditional impacts of trade uncertainty for a given degree of trade openness at each horizon h.
  - Sample: euro area countries observed over 1998Q1 to 2018Q2.
- Fixed effects and controls:
  - Country-fixed effects (u_i) to control for time-invariant unobservables.
  - Quarter (time) fixed effects (λ_t) to absorb time-varying common factors across countries (e.g., rising protectionism worldwide, geopolitical shocks, oil prices, global financial crisis, global market risk aversion).
  - Interaction term U_t × O_{i,t-q} is the identification crux: uncertainty shock common to all countries interacted with lagged trade openness specific to each country.
  - Lagged trade openness is included additively in X to ensure identification.
  - Trade openness measured as the average of exports and imports in percent of GDP averaged over the past 16 quarters (four years).
  - Other control variables included: quarterly real GDP growth (including five lags), corporate bond spreads, domestic lending rates, 10-year sovereign bond yields, and five lags of the change in the investment ratio.
- Interpretation: treatment effects measured at various levels of trade openness; assumes zero statistical effect of a spike in global trade uncertainty in autarky and a marginal negative effect rising with trade openness.

### Measurement of trade uncertainty
- Trade uncertainty variable created by Hlatshwayo (2018), aggregating references to trade protectionism in newspapers using a four-part article metric:
  1. country name or demonym within eight words of a trade-policy related term;
  2. within eight words of an uncertainty-related term;
  3. excludes mentions of declines in uncertainty, references to alternative policies, or equity-market volatility;
  4. article length must be over 99 words to avoid ticker articles.
- Trade-policy terms include trade war, trade barrier, and protectionism.
- Index normalized by broader country news coverage over time and ranges between 0 and 100.
- Search algorithm rejects articles that mention other types of uncertainty (e.g., fiscal or monetary) to avoid conflation.
- Global measure construction:
  - Aggregate trade uncertainty of the EU, the U.S., China, and the U.K. as a proxy for generalized trade uncertainty.
  - Aggregation performed using principal component analysis (PCA); the first principal component explains 65 percent of the overall variance and is retained as the global measure of generalized trade protectionism uncertainty.
- All uncertainty measures in the paper are rescaled to be comprised between 0 and 100 before applying the natural log transformation ln(1+u).

### Limitations and interpretation caveats
- Average effects of trade uncertainty on investment cannot be directly traced because the key parameter is associated with an interaction term (U × O) after controlling for other global shocks.
- Identification yields treatment effects conditional on levels of trade openness rather than a single average effect.
- Assumption that spike in global trade uncertainty has no statistical effect on domestic investment under autarky; marginal negative effect increases with trade openness.
- Measurement challenges inherent to text-based uncertainty measures persist despite the specificity to trade-related uncertainty.

*Source: wp18281 - 0.2 percentage points of GDP). Using the estimated impact of uncertainty on investment in*

### 100. We finally converted the index into natural log (ln (1 + u)) to smooth out the effects of

### wp18281 - 100. We finally converted the index into natural log (ln (1 + u)) to smooth out the effects of

### Baseline Results
- Trade uncertainty increased since the first quarter of 2017 by around 1.4 log points, corresponding to an increase of about 300 percent = 100×(e^{1.4}−1).
- For the median country in terms of trade openness:
  - Estimated reduction in the investment-to-GDP ratio by up to 0.8 percentage points of GDP in the fifth quarter following a typical trade uncertainty spike (middle chart).
  - This corresponds to about one-fifth of the typical standard deviation of the investment-to-GDP ratio in the sample.
  - A typical trade uncertainty spike (of one standard deviation) translates into an increase in trade uncertainty of 0.9 log points, i.e., roughly 146 percent.
  - After five quarters, point estimates are no longer statistically different from zero.
- For countries with highest trade openness regimes (where exports and imports sum to 200 percent of GDP on average):
  - Investment decline can reach 1.6 percentage points of GDP after four quarters.

### Robustness Checks
A. Alternative Measure of Trade Uncertainty: BlackRock Index of Trade Tensions
- BlackRock index aims to capture market attention to trade risk (heavy weight to brokerage reports).
- A one standard deviation increase in the BlackRock index corresponds to an increase in trade uncertainty of 0.7 log points, i.e., roughly 100 percent.
- Using BlackRock index (log-transformed), impact on investment remains statistically significant, negative, and robust up to the fourth quarter following the shock.
- Estimated decline in investment-to-GDP ratio reaches 0.7 (1.3) percentage points of GDP in the second quarter following the increase in trade uncertainty in the median (high) trade openness regime.

B. Controlling for Other Non-Linear Forms of Global Uncertainty
- Augmented specification includes GU (global economic policy uncertainty, in log form) interacted with trade openness.
- Controlling for GU × trade openness does not materially affect baseline estimates.
- For the median country in trade openness:
  - Reduction in investment-to-GDP ratio up to 0.7 percentage points of GDP in the fifth quarter following a typical trade uncertainty spike (about one-fifth of typical standard deviation).
- For the highest trade openness regime:
  - Investment decline can reach 1.3 percentage points of GDP (about 30 percent of the standard deviation of the investment rate in the sample).

C. Excluding Intra-Euro Area Trade Flows
- Replacing trade openness with trade exposure outside the euro area yields broadly similar estimates to baseline.
- Excluding intra-euro area trade flows did not materially change sensitivity of investment-to-GDP to global trade uncertainty.

D. Alternative Measure of Trade Openness: Participation in Supply Chains
- Supply chain indicator (Koopman et al., 2010) = share of foreign inputs (backward participation) and domestically produced inputs used in third countries' exports (forward participation), normalized by GDP; data from OECD TiVA.
- Model substitutes trade openness with SC (averaged over past 16 quarters) and controls for GU × SC.
- Results echo baseline: sudden increase in trade uncertainty reduces investment ratio for about five quarters.
- Magnitudes: for median and high levels of supply chain participation, reduction in investment-to-GDP ratio is near 0.5 and 1 percentage points of GDP in the fourth and fifth quarter following the shock, respectively.

### What the Estimates Imply for Recent Bouts of Uncertainty
- Benchmark measure of trade uncertainty increased by 1.4 log points between Q1 2017 and Q2 2018 (about 300 percent; double the typical standard deviation).
- Using preferred estimates (Section B: controlling for other global uncertainty interacted with trade openness):
  - Investment-to-GDP could be lowered by 1 and 2 percentage points one year from now for countries at the medium and high level of trade openness, respectively.
  - This corresponds to a quarter (half) of the typical standard deviation of the investment-to-GDP ratio in the sample for countries at median (high) trade openness.
- Results conditional on trade talks resuming: trade uncertainty can dissipate quickly and rebound in investment could be very large.

### Conclusion
- Trade uncertainty reduces investment more in countries more dependent on global trade.
- Using text-mined measures and alternative openness metrics (including supply chain participation), the reduction in investment ratio following a surge in trade uncertainty is visible up to four to five quarters after the shock.
- Peak decline in investment:
  - 0.8 percentage points of GDP for median trade openness (roughly a year after the shock).
  - 1.6 percentage points of GDP for high trade openness.
- Policy implication: avoiding protectionism and finding cooperative solutions to promote continued trade growth are essential to preserve near-term growth and long-term production capacity.

### Key Statistics (from Appendix: Descriptive Statistics)
- Investment-to-GDP: Obs 1729; Mean 11.6; Std. Dev. 3.9; Min -5.2; Max 41.3
- Trade uncertainty (log): Obs 1692; Mean 2.0; Std. Dev. 0.9; Min 0; Max 4.6
- Trade uncertainty (BlackRock) (log): Obs 990; Mean 3.1; Std. Dev. 0.7; Min 0; Max 4.6
- Economic Policy Uncertainty (log): Obs 1566; Mean 111.2; Std. Dev. 41.3; Min 55.6; Max 244.0
- Trade openness: Obs 1476; Mean 104.0; Std. Dev. 52.8; Min 39.0; Max 314.9
- Participation in supply chains: Obs 1728; Mean 21.3; Std. Dev. 11.3; Min 4.7; Max 56.0
- GDP growth: Obs 1674; Mean 5.3; Std. Dev. 6.0; Min -26.8; Max 39.3
- Corporate bond spreads: Obs 966; Mean 0.1; Std. Dev. 2.3; Min -20.3; Max 3.2
- Bank lending rate: Obs 1086; Mean 3.9; Std. Dev. 1.6; Min 1.3; Max 12.4
- 10-year sovereign spread: Obs 1921; Mean 6.0; Std. Dev. 4.0; Min -0.1; Max 25.4

*Source: Unpacking Policy Uncertainty: Evidence from European Firms. IMF Working Paper (forthcoming).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18281.pdf_