## Good News Travels Fast: Global Demand Shocks, Oil Futures, and Emerging Markets Dynamics — Working Paper No. WP/2025/253

## Source details

**Canonical URL:** [Good News Travels Fast: Global Demand Shocks, Oil Futures, and Emerging Markets Dynamics — Working Paper No. WP/2025/253](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025253-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025253-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025253-source-pdf.pdf.json)

---

### Econometric framework and identification
- Model setup:
  - Structural VAR (SVAR) in structural form: PX_t = sum_{i=1}^p Q_i X_{t−i} + ε_t.
  - Reduced form: X_t = sum_{i=1}^p R_i X_{t−i} + u_t, with Ω = E[u_t u′_t].
  - Shock of interest: the global demand shock is the disturbance in the first equation (ε_{r=1,t}).
  - Objective: estimate the first column of P^{-1} to trace effects of the identified structural shock.
- Proxy-SVAR / external instrument requirements:
  - Relevance: E[W_t ε_{r=1,t}] = φ.
  - Exclusion restriction: E[W_t ε_{r≠1,t}] = 0.
- Methodological precedent: builds on Stock and Watson (2012), Mertens and Ravn (2013), and Gertler and Karadi (2015).

### External instrument for global demand shocks
- Instrument construction:
  - W_t = O_{t,d} − O_{t,d−1}, where O is the closing price of the 3-month oil futures contract on the announcement day.
  - Instrument = sum of movements in oil futures around US and Euro Area employment announcements within the same month.
  - Daily change assigned uniformly across the month (VAR estimated at monthly frequency).
- Identification assumptions and rationale:
  - Within immediate window after labor market releases, no other structural shocks systematically affect market expectations about future oil prices (exclusion restriction).
  - Measurement error allowed but assumed not systematically correlated with the instrument.
  - Labor market surprises in US and EA provide timely signals of global economic activity; oil futures are forward-looking.
- Empirical support:
  - Distributional comparison (2000–2019): three-month futures changes around labor announcements display greater dispersion and fatter tails than unconditional daily changes.
  - Placebo robustness: shifting event window to +4, +5, +6, and +7 days yields placebo results not statistically significant.
- Notable illustrative episodes:
  - May 2008: oil futures rose by more than 4 dollars, over four times the average absolute change (0.91) around US announcements.
  - June 2012: oil futures fell by more than 3 dollars, over three times the average response.
  - September 2008 (EA): oil futures fell by more than 5 dollars, nearly six times the average absolute response (0.88).
  - September 2019: oil futures declined by about 3.5 dollars, roughly four times the average response.
  - Measurement error note: June 6th, 2008 treated as non-systematic measurement error.

### Data and estimation details
- Sample: monthly data spanning February 2000 to December 2019.
- Baseline global VAR variables: real price of oil, world industrial production, world financial conditions index (FCI), VIX, world oil inventories.
- EMEs additional variables: industrial production (IP), consumer prices (CPI), real effective exchange rates (REER), sovereign credit spreads (EMBI Global).
- Coverage: emerging markets sample covers about 88 percent of the total EMEs GDP, and about 51 percent of global GDP.
- Variable treatment:
  - All variables except the world financial conditions index and the VIX included in log levels.
  - An increase in the REER interpreted as a depreciation of EMEs’ currencies.
- Lag order: two months (selected by the Hannan–Quinn criterion).
- Descriptive statistics and key moments (selected):
  - Oil prices: Average absolute monthly change: 6.6%; Maximum monthly swing: 28.6%.
  - World industrial production: Average growth: 0.2% per month (approximately 2.4% annualized).
  - VIX: Average absolute monthly change: 12.5%.
  - EMEs aggregates: IP 0.4% monthly growth (roughly 5.2% annualized); CPI 0.3% per month (3.7% annualized); REER average monthly appreciation of 0.11% (std dev > 1%); sovereign spreads ranged from 160 to nearly 980 basis points, stabilized around 400 basis points in second half of sample.
- Event-comparison (sharp real oil price increases defined as monthly changes of 10±1%; 16 episodes, 2000–2019):
  - After 12 months (median):
    - EMEs annual industrial production growth declines by 0.8 percentage points.
    - Inflation rises by approximately 0.2 percentage points.
    - Sovereign spreads show a mild decline of 22 basis points.
    - REER appreciates by 2.2%.

### Sign-restriction VAR alternative (summary of identification)
- Sign restrictions over six months: world industrial production (+), VIX (-), world oil inventories (-).
- Results for a 10 percent increase in the real oil price under sign-restriction Global Demand shock (six-month horizon):
  - World industrial production: roughly 2% over six months.
  - VIX: -5 points over six months.
  - World oil inventories: -1.5% over six months.
  - World financial conditions: improvement of about -0.6 percentage points by month sixth.
- Non-restricted variables follow patterns similar to the external-instrument identification; effects can be even more pronounced.

### Instrument strength diagnostics (selected first-stage F-statistics)
- Global VAR first-stage F-statistic = 22.34 (joint US and EA instrument, 3-month futures).
- EMEs VAR first-stage F-statistic = 27.35.
- Disaggregated F-statistics (Exporters / Importers, joint 3-month instrument): Exporter 14.64; Importer 22.22.
- Note: following Stock, Wright, and Yogo (2002), a value less than 10 represents a weak instrument.

### Results — Global VAR (normalized 10% oil price increase, 40-month horizon)
- Normalization: impulse responses produce a 10% increase in the real price of oil on impact; all variables (except FCI and VIX) in logs interpreted as elasticities.
- Key impulse responses:
  - World industrial production: rises sharply, peaking at around 0.6% in the ninth month, then gradually returning to baseline.
  - VIX: falls significantly, indicating reduced global uncertainty and increased investor risk appetite.
  - World financial conditions: responses consistent with more relaxed global financial conditions.
  - World oil inventories: initially dip slightly, then gradually return to baseline.
- Overall inference: pattern supports interpretation that the identified shock is demand-driven, associated with improved growth expectations and a positive global outlook.

### Macroeconomic effects in Emerging Markets (EMEs specification)
- Aggregate responses (40-month horizon):
  - EMEs industrial production: rises significantly, peaking around 0.4% in month four; activity boost lasts between 18 and 24 months.
  - EMEs consumer prices: increase moderately and significantly on impact, then gradually return to pre-shock levels.
  - VIX: declines markedly, supporting a risk-taking / portfolio rebalancing channel that triggers capital inflows.
  - Real exchange rates: appreciate persistently after the shock.
  - Sovereign spreads (EMBI): fall strongly and persistently; response statistically and economically significant.
- Specification details:
  - First-stage F-statistic = 27.35.
  - Variables included: real oil price, world industrial production, VIX, FCI, EMEs aggregates for IP, CPI, EMBI Global spread, and REER.
  - All variables in logarithms, except VIX and FCI in levels.

### Heterogeneity: Net oil exporters vs importers (40-month horizon)
- Industrial production:
  - Oil exporters: peak around 1.3%; expansion more persistent.
  - Oil importers: peak around 0.5%; response reverses quickly after initial surge.
- Consumer prices (CPI):
  - Oil importers: CPI increases modestly.
  - Oil exporters: CPI falls mildly.
- Real exchange rate:
  - Exporters: significant and persistent appreciation.
  - Importers: weaker, less persistent appreciation with larger confidence intervals.
- Sovereign spreads (EMBI):
  - Exporters: sovereign spreads fall by nearly 14%.
  - Importers: sovereign spreads fall by about 7%.
- Mechanisms:
  - Exporters benefit from terms-of-trade improvements, fiscal revenues, and stronger industry linkages.
  - Importers face higher input costs but may benefit from higher world demand for other exports.
  - Identified shock appears to raise prices of other non-oil commodities as well.

### Robustness checks (model and instrument variations)
- Variations tested:
  - Lag structures: 1, 2 (baseline), 3, and 12 lags.
  - Futures maturities: 1, 2, 3 (baseline), 6, and 12 months.
  - Alternative instruments: US-only, EA-only, and joint US+EA employment announcement instruments.
- Robustness findings — Global VAR:
  - Direction of dynamics unchanged across lag choices: world industrial production rises, VIX declines, oil inventories fall initially then revert.
  - Longer futures maturities tend to reinforce persistence.
  - Euro Area-based instruments generate somewhat larger effects than US-based instruments.
- Robustness findings — EMEs and exporter/importer splits:
  - EMEs: results consistent across alternative lags, futures maturities, and separate US/EA instruments.
  - Exporter-importer split: direction of responses unchanged; longer futures maturities strengthen qualitative patterns (exporters: persistent production gains, CPI falls; importers: modest production gains, CPI rises).
  - Euro Area instrument produces systematically more pronounced effects than US instrument.
- Selected first-stage F-statistics across robustness checks (examples):
  - Global, joint 3-month instrument: 22.34 (baseline); alternative lags: 33.81 (1 lag), 21.00 (3 lag), 15.32 (12 lag).
  - EMEs, joint 3-month instrument: 27.35 (baseline); alternative maturities: 29.31 (1M), 27.77 (2M), 27.35 (3M), 23.95 (6M), 21.93 (12M).
  - Exporters (joint 3-month): 14.64; Importers (joint 3-month): 22.22.

### Data sources and classification (selected)
- Oil futures: Daily Light Sweet Crude Oil Futures Price for the 3-month contract settlement (CME via Haver Analytics, PZTEXF3@DAILY).
- World oil inventories: monthly proxy based on OECD petroleum stocks (EIA).
- World industrial production: Baumeister and Hamilton (2019) global IP index.
- VIX: CBOE Volatility Index (VIXCLS from FRED).
- World FCI: average of Bloomberg Financial Conditions Indices for US, EA, UK, Asia excluding Japan, multiplied by -1; each index is a z-score.
- EMEs aggregates: GDP-weighted averages across EMEs (Table A.1); REER increase interpreted as depreciation.
- Net oil exporter/importer classification: based on International Energy Agency data.
- Coverage and data availability (Table A.1 note): IP 73%, CPI 95%, REER 99%, EMBI 88% of total EMEs GDP.

### Summary statistics (selected from Table A.2, sample 2000–2019)
- WTI Price (∆%): Mean 0.67 / Abs. Mean 6.56 / Std 8.44 / Min -28.59 / Max 23.85.
- WTI Price (∆Price): Mean 0.01 / Abs. Mean 0.90 / Std 1.29 / Min -9.79 / Max 10.28.
- World IP (∆%): Mean 0.19 / Abs. Mean 0.48 / Std 0.62 / Min -3.21 / Max 1.86.
- VIX (∆%): Mean 1.27 / Abs. Mean 12.54 / Std 19.15 / Min -31.13 / Max 103.07.
- EMEs IP (∆%): Mean 0.42 / Abs. Mean 0.62 / Std 0.68 / Min -2.57 / Max 4.06.
- EMEs CPI (∆%): Mean 0.31 / Abs. Mean 0.32 / Std 0.19 / Min -0.29 / Max 1.21.
- EMEs REER (∆%): Mean -0.11 / Abs. Mean 0.78 / Std 1.00 / Min -2.31 / Max 3.28.
- EMEs EMBI (bps): Mean 411.81 / Abs. Mean 411.81 / Std 173.49 / Min 159.95 / Max 979.21.
- Combined Surprise (∆Price) (instrument): Mean 0.03 / Abs. Mean 1.34 / Std 1.88 / Min -7.40 / Max 9.42.

### Main conclusions and suggested extensions
- Main empirical findings:
  - Global demand shocks identified via the oil-futures employment-announcement instrument lead to a significant and broad expansion in real activity, improved international financial conditions, reduced global uncertainty, and reductions in oil inventories.
  - In EMEs: industrial production and consumer prices increase, sovereign spreads compress, and real exchange rates appreciate, with marked asymmetries between oil-exporting and oil-importing economies.
- Methodological contribution:
  - High-frequency financial data around well-defined macroeconomic announcements provide a credible external instrument for identifying global demand shocks.
- Suggested extensions:
  - Apply the approach to other forms of macroeconomic news, incorporate cross-country interdependencies more explicitly, or investigate other transmission mechanisms of global demand shocks.

*Source: wpiea2025253-source-pdf*

### 3.1    Econometric Framework

### 3.1    Econometric Framework

### Identification and VAR setup
- Structural VAR (SVAR) in structural form:
  - PX_t = sum_{i=1}^p Q_i X_{t−i} + ε_t
  - Reduced form (pre-multiplying by P^{-1}): X_t = sum_{i=1}^p R_i X_{t−i} + u_t, with Ω = E[u_t u′_t].
- Shock of interest: the global demand shock is taken as the disturbance in the first equation (ε_{r=1,t}).
- Objective: estimate the first column of P^{-1} to trace effects of the identified structural shock across variables:
  - X^{(1)}_t = sum_{i=1}^p R^{(1)}_i X^{(1)}_{t−i} + p^{-1}_{(1)} ε^{(1)}_t
- Proxy-SVAR / external instrument requirements:
  - Relevance: E[W_t ε_{r=1,t}] = φ (instrument strongly correlated with the structural shock).
  - Exclusion restriction: E[W_t ε_{r≠1,t}] = 0 (instrument orthogonal to other structural innovations).
- Methodological precedent: approach builds on Stock and Watson (2012), Mertens and Ravn (2013), and Gertler and Karadi (2015).

### External instrument designed for global demand shocks
- Instrument construction:
  - W_t = O_{t,d} − O_{t,d−1}, where O denotes the closing price of the 3-month oil futures contract on the announcement day.
  - The global demand instrument is constructed as the sum of the movements in oil futures around employment announcements in the United States (US) and the Euro Area (EA) within the same month.
  - The observed daily change is assigned uniformly across the corresponding month because the VAR is estimated at monthly frequency.
- Identification assumption:
  - Within the immediate time window after labor market data releases, no other structural shocks systematically affect market expectations about future oil prices (exclusion restriction).
  - Measurement error is allowed (other events could influence oil futures in the same narrow window), but such events are assumed not to be systematically correlated with the instrument.
- Rationale:
  - Labor market surprises in US and EA provide timely signals of global economic activity given their central role in global output, trade, and financial markets.
  - Oil futures are forward-looking; residual adjustments immediately after announcements are interpreted as revisions in beliefs about future demand for oil.
- Empirical evidence supporting the instrument:
  - Distributional comparison (2000–2019): three-month futures changes around labor announcements display greater dispersion and fatter tails than unconditional daily changes, indicating larger adjustments in oil price expectations tied to employment news.
  - Placebo robustness: shifting event window to +4, +5, +6, and +7 days after releases yields placebo results that are not statistically significant, indicating prompt market reactions to labor data releases.

- Notable episodes (illustrative responsiveness of oil futures to employment surprises):
  - May 2008: oil futures rose by more than 4 dollars, over four times the average absolute change (0.91) observed around US announcements.
  - June 2012: oil futures fell by more than 3 dollars, over three times the average response.
  - September 2008 (EA): oil futures fell by more than 5 dollars, nearly six times the average absolute response (0.88).
  - September 2019: oil futures declined by about 3.5 dollars, roughly four times the average response.
  - Measurement error note: June 6th, 2008 is an outlier where oil futures increased despite a large unemployment increase; treated as non-systematic measurement error.

### Data
- Sample: monthly data spanning February 2000 to December 2019.
- Baseline VAR variables (global model): real price of oil, world industrial production, world financial conditions index (FCI), VIX, world oil inventories.
- EMEs additional variables: industrial production (IP), consumer prices (CPI), real effective exchange rates (REER), sovereign credit spreads (EMBI Global).
- Coverage: emerging markets sample covers about 88 percent of the total EMEs GDP, and about 51 percent of global GDP.
- Variable treatment:
  - All variables except the world financial conditions index and the VIX are included in log levels.
  - An increase in the REER is interpreted as a depreciation of EMEs’ currencies.
- Lag order: two months (selected by the Hannan–Quinn criterion).
- Descriptive statistics and key data moments:
  - Oil prices:
    - Average absolute monthly change: 6.6%
    - Maximum monthly swing: 28.6% (during the global financial crisis)
  - World industrial production:
    - Average growth: 0.2% per month (approximately 2.4% annualized)
  - VIX:
    - Average absolute monthly change: 12.5%
  - EMEs aggregates:
    - Industrial production: 0.4% monthly growth (roughly 5.2% annualized)
    - Consumer price inflation: 0.3% per month (3.7% annualized)
    - REER: average monthly appreciation of 0.11%, with standard deviation exceeding 1%
    - Sovereign spreads: ranged from 160 to nearly 980 basis points; stabilized around 400 basis points in the second half of the sample
- Event-comparison summary (EMEs around large oil price increases):
  - Definition: sharp increases in the real price of oil defined as monthly changes of 10±1%.
  - Based on 16 episodes (2000–2019), median dynamics over 12 months:
    - EMEs annual industrial production growth declines by 0.8 percentage points twelve months after the shock.
    - Inflation rises by approximately 0.2 percentage points after one year.
    - Sovereign spreads show a mild decline of 22 basis points after one year.
    - REER appreciates by 2.2% over the same horizon.
- Interpretation:
  - Net oil importers: higher oil prices increase input costs, depress output, and raise inflation via energy pass-through.
  - Oil exporters and financially open EMEs: higher oil prices can improve terms of trade and attract capital inflows, strengthening currencies and reducing risk premia.
  - When oil price increases are driven by stronger global demand: associated with rising world output, falling uncertainty, and improved macroeconomic conditions in EMEs.
  - Supply-driven oil shocks: associated with lower output, inflation pressures, and financial tightness in oil-importing economies.

### Results: Global VAR (effects of a demand-driven 10% oil price shock)
- Experimental design:
  - Impulse responses are normalized to generate a 10% increase in the real price of oil on impact.
  - All variables (except world financial conditions index and VIX) are in logs and interpreted as elasticities.
  - Solid lines in figures denote point estimates; shaded areas represent 68% confidence intervals based on robust standard errors.
  - First-stage F-statistics reported to assess instrument strength.
- Global VAR specification: real oil price, world industrial production, VIX, world oil inventories, world financial conditions (FCI).
- Instrument strength:
  - First-stage F-statistic = 22.34
  - Note on weak instruments: following Stock, Wright, and Yogo (2002), a value of less than 10 represents a weak instrument.
- Key impulse response findings (40-month horizon):
  - World industrial production:
    - Rises sharply, peaking at around 0.6% in the ninth month, then gradually returning to baseline.
  - VIX:
    - Falls significantly, indicating a reduction in global uncertainty and increased investor risk appetite.
  - World financial conditions:
    - Responses consistent with more relaxed global financial conditions following the demand shock.
  - World oil inventories:
    - Initially dip slightly during the first months, then gradually return to baseline, consistent with transitory imbalances between supply expectations and realized demand.
- Overall inference:
  - The pattern of responses supports the interpretation that the identified shock is demand-driven, associated with improved growth expectations and a positive shift in the global outlook.

*Source: wpiea2025253-source-pdf - 3.1    Econometric Framework*

### 4.2    Macroeconomic Effects in Emerging Markets

### 4.2    Macroeconomic Effects in Emerging Markets

### Aggregate macroeconomic and financial responses in EMEs
- Industrial production in EMEs rises significantly, with a peak around 0.4% in month four.
- The activity boost in EMEs appears to last between 18 and 24 months, longer than the global industrial production boost.
- Consumer prices in EMEs increase moderately and significantly on impact, then gradually return to pre-shock levels.
- The VIX declines markedly following the shock, indicating lower uncertainty and supporting a risk-taking or portfolio rebalancing channel that triggers capital inflows into EMEs.
- Real exchange rates appreciate persistently after the shock (supported by improved terms of trade, capital inflows, and stronger fundamentals).
- Sovereign spreads in EMEs (EMBI) fall strongly and persistently; the response is both statistically and economically significant.
- Specification and measurement details:
  - First-stage F-statistic = 27.35.
  - Instrument: 3-month futures price variation around US and Euro Area employment announcements.
  - Variables included: real oil price, world industrial production, CBOE Volatility Index (VIX), world financial conditions (FCI), EMEs aggregates for industrial production, consumer price index, EMBI Global spread, and real exchange rate (increase means EMEs’ currency depreciation).
  - All variables are in logarithms, except the VIX and FCI, which are in levels.
  - All figures show percent responses to the initial shock over a 40-month horizon.

### Interpretation
- The combination of stronger external demand, booming global economy, falling VIX, and improved world financial conditions points to an overall improvement in macroeconomic and financial conditions across EMEs following a global demand shock.
- Heterogeneity is expected in inflation and exchange rate responses across countries, depending on external trade position (net oil trade).

---

### 4.3    Dynamic Effects Across Net Oil Exporting and Importing Economies

### Key asymmetric responses (exporters vs importers)
- Industrial production:
  - Oil exporters: peak around 1.3%.
  - Oil importers: peak around 0.5%.
  - Expansion is more persistent in exporters; response in importers reverses quickly after the initial surge.
- Consumer prices (CPI):
  - Oil importers: CPI increases modestly.
  - Oil exporters: CPI falls mildly.
- Real exchange rate:
  - Exporters: significant and persistent appreciation.
  - Importers: weaker, less persistent appreciation with larger confidence intervals.
- Sovereign spreads (EMBI):
  - Exporters: sovereign spreads fall by nearly 14%.
  - Importers: sovereign spreads fall by about 7%.

### Mechanisms and interpretation
- Oil exporters benefit directly from higher oil prices through terms-of-trade improvements, fiscal revenues, and stronger industry linkages—leading to output gains, currency appreciation, lower inflation, and declining sovereign spreads.
- Oil importers face upward cost pressures from higher oil prices, raising inflation and import costs, but stronger global demand can also raise prices for other exported commodities, partially offsetting negative effects.
- The identified shock appears to reflect global demand forces broadly: in response to the same global demand shock, prices of other non-oil commodities (agricultural goods, livestock, industrial metals, and precious metals) also rise.

### Specification and measurement details
- First-stage F-statistic = (a) 14.64 and (b) 22.22.
- Classification between oil exporters and importers based on EMEs net oil trade position, using data from the International Energy Agency.
- All figures show percent responses to the initial shock over a 40-month horizon.

---

### 4.4    Sign-Restriction VAR Identification of Global Demand Shocks

### Alternative identification and key quantitative findings
- Sign-restriction VAR imposes sign restrictions over a six-month horizon: world industrial production (+), VIX (-), world oil inventories (-).
- Results for a 10 percent increase in the real oil price under the sign-restriction Global Demand shock:
  - World industrial production: roughly 2% over six months.
  - VIX: -5 points over six months.
  - World oil inventories: -1.5% over six months.
  - World financial conditions: improvement of about -0.6 percentage points by month sixth.
- Non-restricted variables follow a pattern similar to the external-instrument identification; effects can be even more pronounced under the sign-restriction VAR.

### Specification details
- Instrument: Global Demand Instrument as described in Section 3.2.
- Variables included: real oil price, world industrial production, CBOE Volatility Index (VIX), world oil inventories, and world financial conditions (FCI).
- All variables are in logarithms, except the VIX and FCI, which are in levels.
- All figures display percent responses to the initial shock over a 40-month horizon.

---

### 4.5    Robustness Checks

### Model and instrument variations tested
- Lag structures in the VAR: baseline two lags versus one, three, and twelve lags.
- Futures contract maturities for the instrument: 1, 2, 3, 6, and 12 months (baseline 3-month).
- Alternative instruments: employment announcements from the US and the Euro Area separately.

### Robustness findings — Global VAR
- Results are highly robust to changing the number of lags; direction of dynamics remains unchanged (world industrial production rises, VIX declines, oil inventories fall initially then revert).
- Varying futures maturities preserves dynamics; longer maturities tend to reinforce persistence of effects.
- Instruments based on Euro Area surprises generate somewhat larger effects than those based on US announcements.

### Robustness findings — EMEs specification and exporters/importers split
- EMEs: industrial production, consumer prices, exchange rates, and sovereign spreads move consistently with the baseline across alternative lags, futures maturities, and separate US/Euro Area instruments.
- Exporter-importer split:
  - Direction of responses unchanged when changing lags.
  - Longer futures maturities strengthen the qualitative patterns (exporters: persistent production gains, CPI falls; importers: modest production gains, CPI rises).
  - Euro Area instrument produces systematically more pronounced effects than the US instrument.

---

### 5    Conclusion

- A new high-frequency external instrument is proposed to isolate global demand shocks using oil futures price reactions to labor market announcements from the United States and the Euro Area.
- Main empirical findings:
  - Global demand shocks lead to a significant and broad expansion in real activity, improved international financial conditions, reduced global uncertainty, and oil inventories reductions.
  - In EMEs: industrial production and consumer prices increase, sovereign spreads compress, and real exchange rates appreciate, with notable asymmetries between oil-exporting and oil-importing economies.
- Methodological contribution:
  - High-frequency financial data around well-defined macroeconomic announcements provide a credible external instrument for identifying global demand shocks.
- Suggested extensions:
  - Apply the approach to other forms of macroeconomic news, incorporate cross-country interdependencies more explicitly, or investigate other transmission mechanisms of global demand shocks.

*Source: wpiea2025253-source-pdf - 4.2    Macroeconomic Effects in Emerging Markets*

### References

### References

### Key bibliographic sources cited
- Ahmed, S., Coulibaly, B., & Zlate, A. (2017). International financial spillovers to emerging market economies: How important are economic fundamentals? Journal of International Money and Finance, 76, 133-152.
- Alquist, R., & Kilian, L. (2010). What do we learn from the price of crude oil futures? Journal of Applied Econometrics, 25(4), 539-573.
- Baumeister, C., & Hamilton, J. D. (2019). Structural Interpretation of Vector Autoregressions with Incomplete Identification: Revisiting the Role of Oil Supply and Demand Shocks. American Economic Review, 109(5), 1873–1910.
- Blanchard, O., & Quah, D. (1989). The dynamic effects of aggregate demand and aggregate supply. The American Economic Review(79), 655-673.
- Caldara, D., & Herbst, E. (2019). Monetary Policy, Real Activity, and Credit Spreads: Evidence from Bayesian Proxy SVARs. American Economic Journal: Macroeconomics, 11(1), 157–192.
- Christiano, L. J., Eichenbaum, M., & Evans, C. L. (1999). Chapter 2 monetary policy shocks: What have we learned and to what end? In (Vol. 1, p. 65-148). Elsevier.
- Gertler, M., & Karadi, P. (2015). Monetary policy surprises, credit costs, and economic activity. American Economic Journal: Macroeconomics, 7(1), 44–76.
- Kilian, L. (2009). Not All Oil Price Shocks Are Alike: Disentangling Demand and Supply Shocks in the Crude Oil Market. American Economic Review, 99(3), 1053–69.
- Miranda-Agrippino, S., & Ricco, G. (2021). The Transmission of Monetary Policy Shocks. American Economic Journal: Macroeconomics, 13(3), 74–107.
- Nakamura, E., & Steinsson, J. (2018). High-frequency identification of monetary non-neutrality: the information effect. The Quarterly Journal of Economics, 133(3), 1283–1330.
- Uhlig, H. (2005). What are the effects of monetary policy on output? results from an agnostic identification procedure. Journal of Monetary Economics, 52(2), 381-419.
- (Additional references listed in source omitted here for brevity; full list appears in the source document.)

### Data sources and description (Appendix A)
- Oil Price:
  - Spot Crude Oil Price: West Texas Intermediate (WTI), measured in dollars per barrel, monthly frequency, not seasonally adjusted (WTISPLC from FRED).
- Oil Futures:
  - Daily Light Sweet Crude Oil Futures Price for the 3-month contract settlement (end-of-period), measured in dollars per barrel. Data obtained from CME via Haver Analytics (PZTEXF3@DAILY).
- World Oil Inventories:
  - Monthly proxy for global crude oil inventories based on OECD petroleum stocks, following Kilian and Murphy (2014). Series obtained from US Energy Information Administration (EIA).
- World Industrial Production:
  - Global industrial production index constructed by Baumeister and Hamilton (2019).
- VIX:
  - CBOE Volatility Index, reported monthly and not seasonally adjusted (VIXCLS from FRED).
- World Financial Conditions Index:
  - Average of Bloomberg Financial Conditions Indices for the United States (BFCIUS Index), Euro Area (BFCIEU Index), United Kingdom (BFCIGB Index), and Asia excluding Japan (BFCIAXJ Index). The average is multiplied by -1 so that a positive (negative) value indicates tighter (accommodative) financial conditions relative to pre-crisis levels. Each index is a z-score.
- Emerging Markets Industrial Production Index:
  - Monthly IP index, excluding construction, seasonally and working day adjusted, obtained via Haver Analytics. Constructed as a GDP-weighted average across EMEs (see Table A.1).
- Emerging Markets Consumer Price Index:
  - Monthly CPI, seasonally adjusted, obtained via Haver Analytics. Constructed as a GDP-weighted average across EMEs (see Table A.1).
- Emerging Markets Real Exchange Rate Index:
  - Monthly Bruegel’s Real Effective Exchange Rate index, not seasonally adjusted, obtained via Haver Analytics. Constructed as a GDP-weighted average across EMEs. Throughout the study, an increase in real exchange rate is an EMEs currency depreciation.
- Emerging Markets J.P. Morgan EMBI Spread:
  - EMBI Global Spread, sourced from Bloomberg using JPEGSOSD ticker; constructed as a GDP-weighted average across EMEs.
- Employment Announcements:
  - Data obtained from Bloomberg using USURTOT Index and UMRTEMU Index tickers; capture release dates of US Bureau of Labor Statistics’ Employment Situation Report and Euro Area Eurostat Unemployment Statistics.
- Net Oil Exporter and Importers Classification:
  - Countries classified as net oil importers or exporters based on most recent annual crude oil trade data from EIA.

### Emerging Markets classification (Table A.1)
- The 30 largest EMEs by GDP (PPP) and their weights (shares of total EMEs GDP, normalized to sum to 100%):
  - China 36.5% — Importer
  - India 15.5% — Importer
  - Russia 6.6% — Exporter
  - Brazil 4.5% — Exporter
  - Indonesia 4.5% — Importer
  - Türkiye 3.3% — Importer
  - Mexico 3.2% — Exporter
  - Egypt 2.1% — Exporter
  - Saudi Arabia 2.0% — Exporter
  - Poland 1.8% — Importer
  - Thailand 1.7% — Importer
  - Islamic Republic of Iran 1.6% — Exporter
  - Bangladesh 1.6% — Importer
  - Vietnam 1.6% — Importer
  - Pakistan 1.5% — Importer
  - Nigeria 1.4% — Exporter
  - Malaysia 1.3% — Exporter
  - Philippines 1.3% — Importer
  - Colombia 1.1% — Exporter
  - South Africa 0.9% — Importer
  - Romania 0.9% — Importer
  - United Arab Emirates 0.8% — Exporter
  - Kazakhstan 0.8% — Exporter
  - Algeria 0.8% — Exporter
  - Iraq 0.7% — Exporter
  - Chile 0.6% — Importer
  - Ukraine 0.6% — Importer
  - Peru 0.6% — Importer
  - Argentina* 0.0% — Exporter
  - Venezuela* 0.0% — Exporter
- Note: Argentina and Venezuela are excluded due to persistent inflation and data inconsistencies, assigned GDP share of zero. Data availability coverage for macro variables: IP 73%, CPI 95%, REER 99%, EMBI 88% of total EMEs GDP.

### Summary statistics of key variables and instruments (Table A.2)
- Sample: monthly data from 2000 to 2019.
- Variables and exact summary statistics (Unit / Mean / Abs. Mean / Std / Min / P25 / Median / P75 / Max):
  - WTI Price (∆%) — 0.67 / 6.56 / 8.44 / -28.59 / -4.74 / 1.72 / 5.82 / 23.85
  - WTI Price (∆Price) — 0.01 / 0.90 / 1.29 / -9.79 / -0.61 / 0.04 / 0.67 / 10.28
  - World IP (∆%) — 0.19 / 0.48 / 0.62 / -3.21 / -0.13 / 0.23 / 0.55 / 1.86
  - VIX (∆%) — 1.27 / 12.54 / 19.15 / -31.13 / -9.55 / -2.22 / 7.21 / 103.07
  - World FCI (Z-score) — 0.37 / 0.80 / 1.31 / -1.17 / -0.39 / -0.01 / 0.75 / 8.17
  - EMEs IP (∆%) — 0.42 / 0.62 / 0.68 / -2.57 / 0.14 / 0.43 / 0.78 / 4.06
  - EMEs CPI (∆%) — 0.31 / 0.32 / 0.19 / -0.29 / 0.20 / 0.31 / 0.42 / 1.21
  - EMEs REER (∆%) — -0.11 / 0.78 / 1.00 / -2.31 / -0.79 / -0.19 / 0.39 / 3.28
  - EMEs EMBI (bps) — 411.81 / 411.81 / 173.49 / 159.95 / 303.51 / 360.69 / 443.04 / 979.21
  - US Announcement Surprise (∆Price) — 0.09 / 0.91 / 1.33 / -4.04 / -0.57 / 0.11 / 0.75 / 10.28
  - EA Announcement Surprise (∆Price) — -0.07 / 0.88 / 1.25 / -5.58 / -0.58 / 0.10 / 0.65 / 3.48
  - Combined Surprise (∆Price) — 0.03 / 1.34 / 1.88 / -7.40 / -0.91 / 0.06 / 0.96 / 9.42
- Note: The first set of variables (WTI Price, World IP, VIX, EMEs IP, EMEs CPI, EMEs REER) are month-over-month percentage changes. EMBI in basis points. REER increase interpreted as depreciation. Instruments are surprise movements in oil futures prices around employment announcements (daily changes).

### Figures and robustness analyses (Appendix B)
- Figure B.1:
  - Shows empirical dynamics over 12 months around a 10±1% month-to-month real oil price increase.
  - Analysis includes 16 episodes in 2000–2019: Jun 2000, Jun 2003, May 2004, Aug 2004, Aug 2005, Jan 2006, Apr 2006, Jul 2007, Sep 2007, Mar 2008, May 2008, Aug 2009, Oct 2016, Nov 2017, Jan 2018, Apr 2019.
  - An increase in the REER is interpreted as a depreciation of EMEs’ currencies.

- Instrument relevance and first-stage F-statistics (Table B.1; exact values shown):
  - US Employment Announcement — Chng. in 1-month WTI future: Global L=1 25.30, L=2 22.57, L=3 18.32; EMEs 28.48 / 25.03 / 20.98; Exporter 20.44 / 16.13 / 15.24; Importer 28.70 / 22.69 / 18.91.
  - US Employment Announcement — Chng. in 2-month WTI future: Global 25.45 / 22.09 / 18.49; EMEs 27.98 / 23.61 / 19.82; Exporter 19.79 / 14.56 / 14.07; Importer 27.25 / 21.30 / 18.10.
  - US Employment Announcement — Chng. in 3-month WTI future: Global 25.25 / 21.79 / 18.43; EMEs 27.51 / 22.78 / 19.24; Exporter 19.65 / 14.09 / 13.70; Importer 26.60 / 20.64 / 17.48.
  - EA Employment Announcement — Chng. in 1-month WTI future: Global 10.89 / 4.25 / 5.15; EMEs 11.07 / 6.68 / 6.79; Exporter 10.17 / 2.78 / 3.13; Importer 5.12 / 4.34 / 4.59.
  - EA Employment Announcement — Chng. in 2-month WTI future: Global 10.17 / 4.15 / 4.86; EMEs 10.66 / 6.74 / 7.04; Exporter 9.37 / 2.76 / 3.16; Importer 5.10 / 4.39 / 4.63.
  - EA Employment Announcement — Chng. in 3-month WTI future: Global 10.08 / 4.28 / 5.00; EMEs 10.61 / 6.99 / 7.38; Exporter 9.20 / 2.95 / 3.32; Importer 5.17 / 4.63 / 4.87.
  - US and EA Joint instruments — Chng. in 1-month WTI future: Global 35.67 / 23.42 / 21.61; EMEs 38.77 / 29.31 / 26.18; Exporter 30.91 / 16.26 / 16.24; Importer 29.89 / 23.88 / 21.44.
  - US and EA Joint instruments — Chng. in 2-month WTI future: Global 34.27 / 22.46 / 20.95; EMEs 37.14 / 27.77 / 25.16; Exporter 28.70 / 14.76 / 15.09; Importer 28.12 / 22.44 / 20.45.
  - US and EA Joint instruments — Chng. in 3-month WTI future: Global 33.81 / 22.34 / 21.00; EMEs 36.49 / 27.35 / 25.03; Exporter 28.23 / 14.64 / 14.98; Importer 27.61 / 22.22 / 20.23.
- Lag length selection criteria for baseline VAR (Table B.2):
  - Reported for Global, EMEs, Exporter, Importer specifications across lags 1–12 using Schwarz (SIC) and Hannan-Quinn (HQ) criteria. (Exact criterion values reported in table; bold entries indicate selected lag order under each criterion.)

- Selected events and Bloomberg Relevance Values (Table B.3 highlights top indicators):
  - United States top relevance values: Change in Nonfarm Payrolls NFP TCH Index 99.3; Initial Jobless Claims INJCJC Index 98.6; FOMC Rate Decision FDTR Index 98.0; CPI MoM CPI CHNG Index 97.3; GDP Annualized QoQ GDP CQOQ Index 96.6; ISM Manufacturing NAPMPMI Index 95.2; (additional indicators and exact Relevance Values listed in table).
  - Euro Area top relevance values: ECB MRO Announcement Rate EURR002W Index 97.6; EA Real GDP QoQ EUGNEMUQ Index 90.5; HCOB EA Manufacturing PMI SAMPMIEZMA Index 90.0; EA MUICP All Items MoM NSAECCPEMUM Index 85.7; (additional indicators and exact Relevance Values listed in table).
  - Note: Bloomberg Relevance Values retrieved on August 20, 2025.

### Robustness: Global VAR (Appendix B.1)
- Impulse response robustness checks to a 10% increase in real oil price, instrumented using futures variation around US and EA employment announcements:
  - Alternative lag specifications considered: 1, 2, 3, and 12 lags.
  - First-stage F-statistics reported: 33.81 (1 lag), 22.34 (2 lag), 21.00 (3 lag), 15.32 (12 lag).
  - Variables: real oil price, world industrial production, VIX, world financial conditions (FCI), world oil inventories. All variables in logs except VIX and FCI (levels). Responses over 40-month horizon.

- Alternative futures maturities considered: 1M, 2M, 3M, 6M, 12M.
  - First-stage F-statistics: 23.42 (1M), 22.46 (2M), 22.34 (3M), 19.96 (6M), 21.31 (12M).

- Separate vs joint instruments (US and EA):
  - First-stage F-statistics: Joint Instrument 22.34, US 21.79, EA 4.28.

- Alternative specification including three real commodity indexes compared to baseline (black line) that includes EMEs aggregates for IP, CPI, EMBI Global, and REER.

### Robustness: EMEs VAR (Appendix B.2)
- Impulse response robustness checks for EMEs specification to a 10% real oil price increase:
  - Lag alternatives: 1, 2, 3 lags. First-stage F-statistics: 36.49 (1 lag), 27.35 (2 lag), 25.03 (3 lag).
  - Futures maturities: First-stage F-statistics: 29.31 (1M), 27.77 (2M), 27.35 (3M), 23.95 (6M), 21.93 (12M).
  - Separate vs joint instruments: First-stage F-statistics: Joint Instrument 27.35, US 22.78, EA 6.99.
  - Specification includes: real oil price, world industrial production, VIX, FCI, and emerging market aggregates for IP, CPI, EMBI Global spread, and REER. All variables in logs except VIX and FCI (levels). REER increase interpreted as EMEs currency depreciation. Responses over 40-month horizon.

### Robustness: EMEs disaggregation VAR (Appendix B.3)
- Disaggregated analyses for Net Oil-Exporters and Net Oil-Importers:
  - Lag alternatives (Net Oil-Exporters): First-stage F-statistic = 28.23 (1 lag), 14.64 (2 lag), 14.98 (3 lag).
  - Lag alternatives (Net Oil-Importers): First-stage F-statistic = 27.61 (1 lag), 22.22 (2 lag), 20.23 (3 lag).
  - Futures maturities (Exporters): First-stage F-statistic = 16.26 (1M), 14.76 (2M), 14.64 (3M), 12.80 (6M), 11.61 (12M).
  - Futures maturities (Importers): First-stage F-statistic = 23.88 (1M), 22.44 (2M), 22.22 (3M), 19.40 (6M), 17.00 (12M).
  - Separate vs joint instruments:
    - Exporters: First-stage F-statistic = 14.64 (Joint), 14.09 (US), 2.95 (EA).
    - Importers: First-stage F-statistic = 22.22 (Joint), 20.64 (US), 4.63 (EA).
  - Specification variables and interpretation same as EMEs VAR; classification based on EMEs net oil trade position (Table A.1). Responses over 40-month horizon.

*Good News Travels Fast: Global Demand Shocks, Oil Futures, and Emerging Markets Dynamics — Working Paper No. WP/2025/253*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025253-source-pdf.pdf_