## _wp15179 - 18. Net Commodity Exporters and Real Domestic Demand Growth

## Source details

**Canonical URL:** [_wp15179 - 18. Net Commodity Exporters and Real Domestic Demand Growth](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15179.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15179.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15179.pdf.json)

---

### I. Introduction: question and main transmission channel
- Data: 1970–2014 across emerging market and developing economies (63 countries).
- Main empirical observation:
  - During periods of U.S. dollar appreciation, real GDP growth in emerging markets slows; U.S. dollar depreciations are associated with stronger emerging market growth.
- Core transmission channel:
  - Income effect via the U.S. multilateral real exchange rate’s impact on global commodity prices:
    - As the dollar appreciates, dollar commodity prices tend to fall.
    - Lower dollar commodity prices reduce real (dollar) income of commodity exporters, depressing domestic demand and real GDP.
- Additional empirical findings:
  - Expenditure-switching effects from relative currency depreciation of emerging markets do not offset the income effect found.
  - An increase in the U.S. interest rate reduces emerging market growth, conditional on U.S. real exchange rate appreciation and U.S. real GDP growth.
  - Effects are stronger in countries with more rigid exchange rate regimes and strongest for net commodity exporters; countries reliant on imported capital or inputs are also affected, though marginally less.
- Policy implication highlighted:
  - With a persistent strong dollar and anticipated increases in U.S. interest rates, emerging markets’ growth is likely to remain subdued.

### II. Historical context and U.S. REER cycle identification
- Decade dynamics (qualitative):
  - 1970s: Dollar depreciation, U.S. real interest rates ~2 percent, South America real GDP growth averaging over 6 percent.
  - 1980s: Volcker disinflation, U.S. real interest rates reaching 8 percent, dollar appreciation, commodity prices dropped; South America growth ~2½ percent.
  - 1990s: Sustained U.S. expansion, commodity prices mostly weak, South America growth ~3 percent.
  - 2000s: Low real interest rates, depreciating dollar, strong commodity prices; South America growth ~4¼ percent until 2008–9 crisis.
  - 2010s: Dollar appreciation since mid-2014, weakening commodity prices; prospects subdued into the medium term.
- Markov-switching identification (1970–2014):
  - Regime 1 (appreciation): average annual real appreciation Coeff. 3.179 (3.2 percent per year); t-stat 3.07; p-value 0.00; average duration 6.33 years; total 19 years (42.22%).
  - Regime 2 (depreciation): average annual real depreciation Coeff. -3.769 (-3.8 percent per year); t-stat -4.06; p-value 0.00; average duration 8.67 years; total 26 years (57.78%).
  - Transition probabilities:
    - Regime 1 → Regime 1: 0.833
    - Regime 2 → Regime 2: 0.876
  - Identified cycles (Table 2):
    - Depreciation Cycles: 1970–1978; 1986–1992; 2002–2011.
    - Appreciation Cycles: 1979–1985; 1996–2001; 2012–2014.
  - Note: 1992–1995 not assigned to either cycle.

### III. Stylized empirical facts
- Stylized Fact I: Stronger U.S. dollar associated with lower emerging markets’ growth.
  - Effect strongest for net commodity exporting regions (South America); strong in emerging Asia; weaker in emerging Europe; intermediate for MENA.
- Stylized Fact II: Stronger U.S. dollar associated with softer real domestic demand growth.
  - Domestic demand is the proximate driver of GDP responses; impact weaker for MENA, stronger for Latin America; Emerging Asia and Emerging Europe intermediate.
- Stylized Fact III: Higher U.S. interest rates associate with a more appreciated U.S. dollar.
- Stylized Fact IV: Higher U.S. interest rates sometimes associated with stronger U.S. growth, but not systematically; net effect: faster U.S. growth benefits EMs, but higher U.S. real interest rates and a stronger dollar can mitigate benefits.
- Stylized Fact V: Stronger dollar, slower EM growth driven by weaker commodity prices.
  - During U.S. appreciation episodes, average U.S. real GDP growth about the same as during depreciation episodes; real interest rates slightly higher during appreciation but not substantially different.
  - Latin America real GDP growth much stronger when the U.S. dollar is more depreciated; commodity prices weaker during U.S. appreciation episodes.

### IV. Commodity REER and purchasing-power channel
- Construction:
  - Commodity purchasing power measured by replacing trading-partner CPI weights in the standard REER with trade-weighted commodity terms of trade (commodity REER).
- Observation and interpretation:
  - When commodity prices increase, the commodity REER rises relative to the standard REER (higher purchasing power); when commodity terms of trade weaken, commodity REER falls below the standard REER.
  - A stronger U.S. dollar reduces the dollar purchasing power of commodity exports (dollars received per unit exported fall), lowering dollar real income and domestic demand in commodity-exporting EMs.

### V. Event analysis: dynamics of GDP and domestic demand through U.S. REER cycles
- Event setup:
  - Identified appreciation/depreciation cycles standardized: first observation t0, last year t4; indices computed at t0, t1 (¼), t2 (½), t3 (¾), t4 (t).
- Key event-based patterns:
  - Except for Central America and Mexico, every other emerging market region shows lower real GDP during U.S. dollar appreciation.
  - Latin America (aggregate) and South America: real GDP about 25 percent lower toward the end of the cycle during appreciation cycles than during depreciation cycles.
  - MENA: differences about 50 percentage points.
  - Emerging Asia: differences about 20 percentage points.
  - Real domestic demand growth is much stronger when the U.S. dollar is more depreciated (except CAM and Mexico); domestic demand decreases or remains flat during U.S. dollar appreciation in many regions, supporting an income-effect-dominant interpretation.

### VI. Theoretical framework: supply, demand, and response to a U.S. appreciation shock
- Small open emerging market model elements:
  - Produces nontradable goods (N) and home tradable goods (H); imports M enter production of H.
  - Export demand X increases with purchasing power of the U.S., pM/pH (X' > 0).
- Responses to U.S. real appreciation (higher pM/pH):
  - Terms-of-trade deterioration for the emerging market; income effect reduces wages and domestic demand.
  - Supply-side elasticity matters: Leontieff, Cobb-Douglas, and linear specifications imply different quantitative outcomes.
  - Two counteracting effects:
    - Income effect (negative): lower wages and domestic demand.
    - Expenditure-switching effect (potential offset): higher external demand for home goods via relative price changes.
  - Financial channel:
    - Higher U.S. real interest rates and appreciating dollar can trigger capital flows to the U.S., depreciating EME real exchange rates, amplifying effects through credit squeezes and more expensive imports.
- Empirical implication: net effect on domestic activity is ambiguous in theory but resolved empirically toward a net negative impact.

### VII. Econometric specification and main regression results
- Baseline pooled panel regression (country fixed effects):
  - y_it = α_i + β X_US,t + δ Z_t + ε_it
  - X_US,t = [Δ log U.S. REER (t-1), RIR (t-1), RGDPgr_US (t-1)]'
  - Controls Z_t: lagged real GDP per capita (1 RGDPpc_it-1), China real GDP growth (CH_t RGDPgr), net capital inflows (1 KI_it-1), U.S. stock market volatility (US_t Vol).
- Main empirical findings:
  - U.S. real exchange rate appreciation (stronger dollar) is associated with lower real GDP growth in most regions, conditional on controls.
  - Higher U.S. real interest rates reduce EME economic activity.
  - Stronger U.S. real GDP growth increases EME real GDP growth.
- Selected coefficient magnitudes (preserved exactly as reported):
  - Latin America (Table 3):
    - U.S. REER coefficient ~ -0.206*** to -0.209*** (standard errors ~ 0.025–0.026).
    - U.S. real interest rate (t-1) ~ -0.270*** to -0.430***.
    - U.S. real GDP growth (t-1) ~ 0.154** to 0.222***.
    - Observations: 946 to 858; Adjusted R2 range 0.0699 to 0.1056.
    - Interpretation: conditional on other factors, a one percent real appreciation of the U.S. dollar reduces average real GDP growth in emerging markets by 0.2 percentage points.
  - South America (Table 4):
    - U.S. REER coefficient ~ -0.247*** to -0.248*** (std. errors ~ 0.027–0.031).
    - U.S. real interest rate (t-1) ~ -0.417*** to -0.621***.
    - Observations: 516; Adjusted R2 up to 0.1452.
  - Central America and Mexico (Table 5):
    - U.S. REER coefficient ~ -0.131** to -0.121** (std. errors ~ 0.041–0.044).
  - Emerging Asia (Table 6):
    - U.S. REER coefficient generally negative but mostly not statistically significant; marginal significance in some specifications (-0.087+ to -0.088+ in column 6).
  - Emerging Europe (Table 7):
    - U.S. REER coefficient ~ -0.167** to -0.146** (std. errors ~ 0.052–0.059).
  - MENA (Table 8):
    - U.S. REER coefficient generally negative but not statistically significant in real GDP regressions.
- Real domestic demand regressions (Tables 9–14): larger coefficients on U.S. REER than for real GDP in many regions:
  - Latin America (Table 9): U.S. REER ~ -0.587*** to -0.596*** (std. errors ~ 0.131–0.146).
  - South America (Table 10): U.S. REER ~ -0.813*** to -0.837***.
  - Central America and Mexico (Table 11): U.S. REER ~ -0.290** to -0.227*.
  - Emerging Asia (Table 12): U.S. REER ~ -0.849*** to -0.848*** in some specifications.
  - Emerging Europe (Table 13): U.S. REER ~ -0.755** to -0.886***.
  - MENA (Table 14): U.S. REER ~ -0.632** to -0.639**.
  - U.S. real interest rate (t-1) negatively impacts domestic demand strongly in many regions.

### VIII. Heterogeneity and subgroup findings
- Exchange rate rigidity (Table 15):
  - U.S. REER (Fixed reg.) coefficients ~ -0.171*** to -0.183***.
  - U.S. REER (Floating reg.) coefficients ~ -0.151*** to -0.141**.
  - Implication: stronger dollar impacts economies with less exchange rate flexibility more on GDP.
- Exchange rate regimes and real domestic demand (Table 16):
  - U.S. REER (Fixed reg.) on domestic demand ~ -0.531*** to -0.505***.
  - U.S. REER (Floating reg.) on domestic demand ~ -0.733*** to -0.803***.
  - Interpretation: negative effect of stronger dollar on domestic demand larger in more flexible regimes (trade-off with synchronization noted).
- Net commodity exporters vs net commodity importers (Tables 17–18):
  - Real GDP growth (Table 17):
    - U.S. REER (net comm. exp.) coefficients ~ -0.135*** to -0.139***.
    - U.S. REER (net comm. imp.) coefficients ~ -0.104*** to -0.123***.
  - Real domestic demand growth (Table 18):
    - U.S. REER (net comm. exp.) coefficients ~ -0.683*** to -0.744***.
    - U.S. REER (net comm. imp.) coefficients ~ -0.690*** to -0.682***.
  - Interpretation: negative impact exists for both groups; economically larger for net commodity exporters; commodity income channel important but other channels (imports of inputs, financial) also relevant.

### IX. Robustness and dynamics
- Robustness checks summary:
  - Including lagged EME real exchange rate: stronger effects on domestic demand; baseline results hold.
  - 5-year averages: baseline results persist.
  - Alternative volatility measure (VIX): similar results.
  - Including commodity terms of trade index (Gruss 2014): not always significant when U.S. REER included; U.S. REER remains significant.
  - Rolling-expanding regressions:
    - Negative coefficient on U.S. REER on real GDP growth stable over time and almost flat since the late 1990s.
    - Negative effect of U.S. real interest rates systematic and increasing in absolute value.
    - Positive association between U.S. real GDP growth and EME activity stable.
    - China real GDP growth becomes positively significant only in the second half of the 1990s.
  - Results robust to crisis dummies, time effects, clustering by time, and restricting sample start to 1999 (though significance may diminish with smaller sample).

### X. Conclusions, implications, and policy-relevant points
- Empirical conclusion:
  - Conditional on the positive effect of stronger U.S. growth, periods of a stronger U.S. dollar result in subdued growth in emerging markets—and vice versa—via an income channel that reduces purchasing power for commodity exporters and dampens domestic demand.
- Outlook and probabilistic statements:
  - The U.S. dollar appears to be on an appreciating cycle since mid-2014.
  - Based on historical estimations, the probability of the dollar remaining appreciated in the short- and medium-term is high (above 80 percent).
  - Appreciating cycles in the U.S. dollar are reported as about 68 years (cycle length mentioned).
  - Implication: commodity prices are expected to remain weak; domestic demand and real GDP growth in emerging markets are likely to be slower than otherwise across regions.
- Policy-relevant implications and recommendations:
  - Countries with stronger fiscal frameworks, credible monetary policy, and flexible exchange rates should be better prepared to navigate less benign external conditions.
  - Exchange rate flexibility mitigates synchronization with U.S. dollar cycles by enabling expenditure-switching; however, larger immediate income effects on domestic demand may occur through nominal adjustments—policy design should consider trade-offs.
  - The negative income effect is particularly strong for commodity exporters and countries with more rigid exchange rate regimes; these countries may need targeted policies to buffer income shocks (e.g., fiscal buffers, stabilization funds, diversified exports).
  - Higher U.S. real interest rates amplify financial channels (capital flow reversals, credit squeezes); macrofinancial policies should account for potential tightening of external financing conditions.
- Final empirical takeaway:
  - The empirical evidence resolves the tension between the income effect and the expenditure-switching effect in favor of a net negative impact of a stronger U.S. dollar on EME domestic demand and real GDP in many regions, especially commodity exporters and less flexible exchange rate regimes.

*Source: _wp15179 - 18. Net Commodity Exporters and Real Domestic Demand Growth (IMF working paper, using data 1970–2014).*

### References _______________________________________________________________39

### _wp15179 - References _______________________________________________________________39

### Figures
- 1. A Historical Perspective of U.S. Real Exchange Rate Cycles _______________________7
- 2. U.S. Effective Exchange Rates Appreciation/Depreciation Cycles __________________10
- 3. U.S. Dollar Strength and Real GDP Growth in Emerging Markets __________________12
- 4. U.S. Dollar Strength and Real Domestic Demand Growth in Emerging  
 Markets ____________________________________________________________13
- 5. U.S. Dollar Strength, Real GDP Growth, and Real Interest Rates ___________________14
- 6. Appreciation and Depreciation Cycles ________________________________________15
- 7. Purchasing Power of Commodity Real Exchange Rates __________________________16
- 8. Real GDP During U.S. Dollar Appreciation and Depreciation Cycles _______________18
- 9. Real Domestic Demand During U.S. Dollar Appreciation and  
 Depreciation Cycles __________________________________________________19

### Tables
- 1. U.S. Dollar Appreciation Cycles _____________________________________________9
- 2. U.S. Appreciation and Depreciation Cycles ____________________________________10
- 3. Baseline Regression: Real GDP Growth in Latin America ________________________26
- 4. Baseline Regression: Real GDP Growth in South America ________________________28
- 5. Baseline Regression: Real GDP Growth in Central America and Mexico _____________29
- 6. Baseline Regression: Real GDP Growth in Emerging Asia ________________________29
- 7. Baseline Regression: Real GDP Growth in Emerging Europe ______________________30
- 8. Baseline Regression: Real GDP Growth in MENA ______________________________30
- 9. Baseline Regression: Real Domestic Demand Growth in Latin America _____________31
- 10. Baseline Regression: Real Domestic Demand Growth in South America ____________31
- 11. Baseline Regression: Real Domestic Demand Growth in Central America and Mexico _32
- 12. Baseline Regression: Real Domestic Demand Growth in Emerging Asia ____________32
- 13. Baseline Regression: Real Domestic Demand Growth in Emerging Europe __________33
- 14. Baseline Regression: Real Domestic Demand Growth in MENA __________________33
- 15. Exchange Rate Rigidity __________________________________________________34
- 16. Exchange Rate Regimes and Real Domestic Demand Growth ____________________35
- 17. Net Commodity Exporters and Real GDP Growth ______________________________36

*Source: _wp15179 - References _______________________________________________________________39*

### 18. Net Commodity Exporters and Real Domestic Demand Growth ___________________36

### 18. Net Commodity Exporters and Real Domestic Demand Growth ___________________36

### I. Introduction: question and main transmission channel
- Data: 1970–2014 across emerging market and developing economies (63 countries).
- Main empirical observation: during periods of U.S. dollar appreciation, real GDP growth in emerging markets slows; U.S. dollar depreciations are associated with stronger emerging market growth.
- Core transmission channel argued: an income effect via the U.S. multilateral real exchange rate’s impact on global commodity prices:
  - As the dollar appreciates, dollar commodity prices tend to fall.
  - Lower dollar commodity prices reduce real (dollar) income of commodity exporters, depressing domestic demand and real GDP.
- Additional factors:
  - Expenditure-switching effects from relative currency depreciation of emerging markets do not offset the income effect found.
  - An increase in the U.S. interest rate reduces emerging market growth, conditional on U.S. real exchange rate appreciation and U.S. real GDP growth.
  - Effects are stronger in countries with more rigid exchange rate regimes and strongest for net commodity exporters; countries reliant on imported capital or inputs are also affected, though marginally less.
- Policy implication raised: with a persistent strong dollar and anticipated increases in U.S. interest rates, emerging markets’ growth is likely to remain subdued.

### II. Historical context and “decade-dynamics”
- Decade summaries (qualitative):
  - 1970s: Dollar depreciation, U.S. real interest rates ~2 percent, South America real GDP growth averaging over 6 percent (oil shocks).
  - 1980s: Volcker disinflation, U.S. real interest rates reaching 8 percent, dollar appreciation, commodity prices dropped; South America growth ~2½ percent.
  - 1990s: Sustained U.S. expansion, commodity prices mostly weak, South America growth ~3 percent.
  - 2000s: Low real interest rates (Greenspan’s put), depreciating dollar, strong commodity prices (China-driven); South America growth ~4¼ percent until 2008–9 crisis.
  - 2010s: Dollar appreciation since mid-2014, weakening commodity prices; prospects subdued into the medium term.
- Illustrative recent fact: the U.S. dollar appreciated about 13 percent between April 2014 and April 2015, while average emerging market and developing economies growth for 2015 was revised down by over 1 percent (IMF World Economic Outlook).

### III. Identifying U.S. dollar appreciation/depreciation cycles: Markov-switching framework
- Model setup:
  - Two regimes labeled “appreciation” (Regime 1) and “depreciation” (Regime 2).
  - Annual U.S. REER growth (Δreert) is modeled with regime-dependent intercepts μ1, μ2 plus an impulse dummy for 2009 (I2009).
  - Sample period: 1970–2014 (IMF International Financial Statistics).
- Key estimated results (Table 1 summaries preserved):
  - Average annual real appreciation (Regime 1): 3.2 percent per year (Coeff. 3.179; t-stat 3.07; p-value 0.00).
  - Average duration of appreciation regimes: 6.33 years (total 19 years, 42.22% of sample).
  - Average annual real depreciation (Regime 2): -3.8 percent per year (Coeff. -3.769; t-stat -4.06; p-value 0.00).
  - Average duration of depreciation regimes: 8.67 years (total 26 years, 57.78% of sample).
  - Transition probabilities:
    - Regime 1 -> Regime 1: 0.833 (period of real appreciation is 83.3 percent likely to remain appreciating in the following period).
    - Regime 2 -> Regime 2: 0.876 (about 87.6 percent probability of continuation).
  - Diagnostic tests for scaled residuals: Normality Chi^2(2) = 5.5771 [0.0615]; ARCH 1-1 F(1,37) = 0.41741 [0.5222]; Portmanteau(6) Chi^2(6) = 10.635 [0.1003].
- Identified U.S. cycles (Table 2):
  - Depreciation Cycles: 1970–1978; 1986–1992; 2002–2011.
  - Appreciation Cycles: 1979–1985; 1996–2001; 2012–2014.
- Note: 1992–1995 was not assigned to either cycle to avoid arbitrary classification.

### IV. Stylized facts (empirical co-movements and interpretation)
- Stylized Fact I: Stronger U.S. dollar, lower emerging markets’ growth.
  - Periods of U.S. dollar appreciation coincide with softer real GDP growth across EM regions; depreciated dollar coincides with stronger EM activity.
  - Effect strongest for net commodity exporting regions (South America), strong in emerging Asia, weaker in emerging Europe, intermediate for MENA.
- Stylized Fact II: Stronger U.S. dollar, softer real domestic demand growth.
  - Real domestic demand growth falls in periods of dollar appreciation, suggesting domestic demand is the proximate driver of GDP responses.
  - Impact on domestic demand appears weaker for MENA, stronger for Latin America; Emerging Asia and Emerging Europe are intermediate.
- Stylized Fact III: Higher U.S. interest rates associate with a more appreciated U.S. dollar.
  - Periods of higher U.S. interest rates tend to coincide with a stronger dollar (capital inflows seeking higher yields).
- Stylized Fact IV: Higher U.S. interest rates are sometimes associated with stronger U.S. growth, but not systematically.
  - When U.S. growth is strong, interest rates may rise to mitigate inflationary pressure; however this relationship varies over time.
  - Net effect: faster U.S. growth benefits EMs, but higher U.S. real interest rates and a stronger dollar can mitigate those benefits.
- Stylized Fact V: Stronger dollar, slower EM growth driven by weaker commodity prices.
  - During U.S. appreciation episodes, average U.S. real GDP growth is about the same as during depreciation episodes; real interest rates are slightly higher during appreciation but not substantially different.
  - Latin America example: Latin America real GDP growth is much stronger when the U.S. dollar is more depreciated.
  - Commodity prices are weaker during U.S. appreciation episodes; lower commodity prices accompany weaker growth in commodity exporters.

### V. Commodity purchasing power and the REER channel
- Construction: commodity purchasing power measured by replacing trading-partner CPI weights in the standard REER with trade-weighted commodity terms of trade (commodity REER).
- Observation: when commodity prices increase, the commodity REER rises relative to the standard REER (higher purchasing power); when commodity terms of trade weaken, commodity REER falls below the standard REER.
- Interpretation: a stronger U.S. dollar reduces the dollar purchasing power of commodity exports (dollars received per unit exported fall), lowering dollar real income and domestic demand in commodity-exporting EMs.

### VI. Event analysis and dynamics (setup)
- Approach: use identified appreciation/depreciation cycles to construct events; standardize event timing so the first observation of a cycle is t0 and the last year is t4; real GDP in t-1 is standardized (details of event discretization follow in the source).
- Purpose: to add dynamic analysis to the average cycle comparisons and to trace responses of GDP and domestic demand through the cycle phases.

### VII. Theoretical and empirical context
- Supporting literature and theory cited in the source:
  - Frenkel (1986): U.S. monetary easing (dollar depreciation) results in higher commodity prices and vice versa.
  - Dornbusch (1986), Borensztein and Reinhart (1994), Akram (2009): nominal and real commodity prices depend negatively on the U.S. real exchange rate.
  - Zhang and others (2008): a stronger U.S. dollar lowers the real price of oil.
  - Engel and Hamilton (1990): documented long swings in dollar values.
- Novelty claim: limited systematic evidence exists linking the strength of the U.S. dollar over its cycles to EM economic activity and documenting the commodity-price income transmission channel; this paper aims to bridge that gap.

*Source: _wp15179 - 18. Net Commodity Exporters and Real Domestic Demand Growth (IMF working paper, using data 1970–2014).*

### 100. Given the data for real GDP growth rates for each (PPP-weighted) real GDP, we

### _wp15179 - 100. Given the data for real GDP growth rates for each (PPP-weighted) real GDP, we

### Event analysis: real GDP and domestic demand during U.S. dollar appreciation/depreciation cycles
- Method:
  - Reconstruct indexed real GDP for each region using PPP-weighted real GDP growth rates.
  - Discretize time-space to compute indices at t0, t1 (¼), t2 (½), t3 (¾), and t4 (t).
  - Repeat for an index of real domestic demand.
- Key empirical patterns:
  - Except for Central America and Mexico, every other emerging market region shows that real GDP is lower during periods of U.S. dollar appreciation.
  - This pattern holds for Latin America as an aggregate and especially for South America (a strong commodity exporter).
  - The pattern also holds for emerging countries in the Middle East and North Africa Region (MENA) and emerging Europe; it is present, though to a lesser extent, in emerging Asia.
- Magnitudes reported from event analysis:
  - Latin America: real GDP about 25 percent lower toward the end of the cycle during appreciation cycles than during depreciation cycles.
  - South America: differences of similar order of magnitude (around 25 percent).
  - MENA: differences about 50 percentage points.
  - Emerging Asia: differences about 20 percentage points.
- Interpretation of channels:
  - Trade link: external demand for goods.
  - Tourism link: external demand for services.
  - Remittances: transfer resources from the U.S. to Mexico, Central America, and the Caribbean, potentially offsetting negative income effects of a stronger dollar.
  - Countries with hard pegs or dollarization (e.g., Ecuador, El Salvador, Panama) are more synchronized with the U.S. business cycle.
- Domestic demand:
  - Figure 9 (event analysis) shows real domestic demand growth is much stronger when the U.S. dollar is more depreciated (except CAM and Mexico).
  - In many regions domestic demand decreases or remains flat during U.S. dollar appreciation, indicating a negative impact of a stronger dollar on purchasing power of domestic demand and supporting an income-effect-dominant interpretation in some cases.

### Theoretical framework (sketch): supply, demand, and response to U.S. appreciation shock
- Model setup:
  - Small open emerging market economy producing nontradable goods (N) and home tradable goods (H).
  - Nontradable production: gN(yt; L) increasing, concave in labor L.
  - Home tradable production: fH(yt; L, M) increasing in labor L and imports M, with decreasing marginal products.
  - Prices pN and pH taken as given; optimality conditions imply factor-price relationships: fL/fM = pM/pH and gL = w.
  - Consumer utility u(cN, cH, cM) with budget constraint wL = pN cN + pH cH + pM cM.
  - Market clearing: yN = cN; yH = cH + X where X is exports; trade balance pH X = pM cM + ...
  - Export demand X increases with purchasing power of the U.S., pM/pH (X' > 0).
- Response to a U.S. real appreciation (represented by higher pM/pH, equivalently an increase in pM with pH held constant):
  - Terms-of-trade deterioration for the emerging market: 0_M pΔ >, 0_H pΔ =.
  - Supply-side elasticity matters:
    - Leontieff (fixed proportions): higher import prices require lower wages; reduced labor demand → negative income effect → lower aggregate consumption and nontradable production; implies Δc < 0 with relative consumption changes Δc_N > Δc_H = > Δc_M per equation (11).
    - Cobb-Douglas: ambiguous impact on home goods production; partial derivative calculations show ambiguous sign.
    - Linear/perfect substitutability: corner solutions possible; overall production may remain unaltered; exports could raise output.
  - Two counteracting effects:
    - Income effect: lower wages, reduced income, lower domestic demand and output.
    - Expenditure-switching effect: higher external demand for home goods (exports) due to relative price change, improving trade balance and potentially offsetting income effect.
  - Financial channel:
    - Higher U.S. real interest rates (and appreciating dollar) trigger capital flows to the U.S., depreciating EME real exchange rates, amplifying effects by making nontradables cheaper and imports more expensive.
    - Credit squeeze due to capital outflows can dampen investment and activity.
- Empirical implication: net effect on domestic activity ambiguous and empirical.

### Econometric model: baseline specification and identification
- Baseline pooled panel regression with country fixed effects:
  - Equation (12): y_it = α_i + β X_US,t + δ Z_t + ε_it
  - X_US,t includes the log change in U.S. real effective exchange rate (REER), lagged U.S. real interest rate (RIR; proxied by 10-year U.S. treasury real return), and lagged U.S. real GDP growth:
    - X_US,t = [Δ log U.S. REER (t-1), RIR (t-1), RGDPgr_US (t-1)]'
  - Controls Z_t include lagged real GDP per capita (1 RGDPpc_it-1), China real GDP growth (CH_t RGDPgr), net capital inflows (1 KI_it-1, financial account balance percent of GDP), and U.S. stock market volatility (US_t Vol, S&P std. dev.).
- Estimation details:
  - Panel regressions with country fixed effects and robust standard errors clustered by country.
  - Models run for each region separately.
  - Data sources: IMF World Economic Outlook, International Financial Statistics, Information Notice System; interest rates from FRED; volatility from Bloomberg.

### Econometric results: real GDP growth and domestic demand growth
- Main finding:
  - U.S. real exchange rate appreciation (stronger dollar) is associated with lower real GDP growth in most emerging market regions, conditional on controls.
  - Higher U.S. real interest rates reduce EME economic activity.
  - Stronger U.S. real GDP growth increases EME real GDP growth.
- Economic magnitudes (selected estimates, preserving reported coefficients and significance):
  - Latin America (Table 3):
    - U.S. real effective exchange rate coefficient ~ -0.206*** to -0.209*** (standard errors ~ 0.025–0.026).
    - U.S. real interest rate (t-1) ~ -0.270*** to -0.430***.
    - U.S. real GDP growth (t-1) ~ 0.154** to 0.222***.
    - Observations: 946 to 858 across specifications; Adjusted R2 range 0.0699 to 0.1056.
    - Interpretation: conditional on other factors, a one percent real appreciation of the U.S. dollar reduces average real GDP growth in emerging markets by 0.2 percentage points.
  - South America (Table 4):
    - U.S. REER coefficient ~ -0.247*** to -0.248*** (std. errors ~ 0.027–0.031).
    - U.S. real interest rate (t-1) ~ -0.417*** to -0.621***.
    - Observations: 516; Adjusted R2 up to 0.1452.
  - Central America and Mexico (Table 5):
    - U.S. REER coefficient ~ -0.131** to -0.121** (std. errors ~ 0.041–0.044).
    - Coefficient smaller in absolute value and less statistically significant than SLAC.
  - Emerging Asia (Table 6):
    - U.S. REER coefficient generally negative but mostly not statistically significant; marginal significance in some specifications (-0.087+ to -0.088+ in column 6).
    - China real GDP growth often not statistically significant in many specifications.
  - Emerging Europe (Table 7):
    - U.S. REER coefficient ~ -0.167** to -0.146** (std. errors ~ 0.052–0.059).
  - MENA (Table 8):
    - U.S. REER coefficient generally negative but not statistically significant in real GDP regressions; financial effects appear important.
- Real domestic demand regressions (Tables 9–14):
  - Coefficients on U.S. REER are larger in absolute magnitude compared to real GDP regressions in many regions:
    - Latin America (Table 9): U.S. REER ~ -0.587*** to -0.596*** (std. errors ~ 0.131–0.146).
    - South America (Table 10): U.S. REER ~ -0.813*** to -0.837***.
    - Central America and Mexico (Table 11): U.S. REER ~ -0.290** to -0.227*.
    - Emerging Asia (Table 12): U.S. REER ~ -0.849*** to -0.848*** in some specifications (note: Table 12 labels vary by panel).
    - Emerging Europe (Table 13): U.S. REER ~ -0.755** to -0.886***.
    - MENA (Table 14): U.S. REER ~ -0.632** to -0.639**.
  - U.S. real interest rate (t-1) negatively impacts domestic demand strongly in many regions.
- Heterogeneity analyses:
  - Exchange rate rigidity (Table 15):
    - U.S. REER (Fixed reg.) coefficients ~ -0.171*** to -0.183***.
    - U.S. REER (Floating reg.) coefficients ~ -0.151*** to -0.141**.
    - Absolute value larger for fixed regimes; implication: stronger dollar impacts more on economies with less exchange rate flexibility.
  - Exchange rate regimes and real domestic demand growth (Table 16):
    - U.S. REER (Fixed reg.) on domestic demand ~ -0.531*** to -0.505***.
    - U.S. REER (Floating reg.) on domestic demand ~ -0.733*** to -0.803***.
    - Interpretation: negative effect of a stronger dollar on domestic demand is larger in more flexible regimes.
  - Net commodity exporters vs net commodity importers (Tables 17–18):
    - Real GDP growth (Table 17):
      - U.S. REER (net comm. exp.) coefficients ~ -0.135*** to -0.139***.
      - U.S. REER (net comm. imp.) coefficients ~ -0.104*** to -0.123***.
    - Real domestic demand growth (Table 18):
      - U.S. REER (net comm. exp.) coefficients ~ -0.683*** to -0.744***.
      - U.S. REER (net comm. imp.) coefficients ~ -0.690*** to -0.682***.
    - Interpretation: negative impact exists for both groups, larger economically for net commodity exporters; commodity income channel important but other channels (imports of inputs, financial) also relevant.

### Robustness checks and dynamics
- Robustness exercises performed (summary):
  - Included lagged real exchange rate of each EME; stronger effects on domestic demand but not always significant on GDP; baseline results hold.
  - 5-year averages used; baseline results persist.
  - Alternative volatility measure (VIX) yields similar results.
  - Included commodity terms of trade index (Gruss 2014): not always significant when U.S. REER included; U.S. REER remains significant.
  - Rolling-expanding regressions:
    - Negative coefficient on U.S. REER on real GDP growth is stable over time and almost flat since the late 1990s.
    - Negative effect of U.S. real interest rates is systematic and increasing in absolute value over the sample.
    - Positive association between U.S. real GDP growth and EME activity stable.
    - China real GDP growth becomes positively significant only in the second half of the 1990s.
  - Results robust to crisis dummies, time effects, clustering by time, and restricting sample start to 1999 (though significance may diminish with smaller sample).

### Conclusions, implications, and policy-relevant points
- Summary conclusion:
  - Conditional on the positive effect of stronger U.S. growth, periods of a stronger U.S. dollar result in subdued growth in emerging markets—and vice versa—via an income channel that reduces purchasing power for commodity exporters and dampens domestic demand.
- Probabilistic and outlook statements from the paper:
  - The U.S. dollar appears to be on an appreciating cycle since mid-2014.
  - Based on historical estimations, the probability of the dollar remaining appreciated in the short- and medium-term is high (above 80 percent).
  - Appreciating cycles in the U.S. dollar are reported as about 68 years (cycle length mentioned).
  - Implication: commodity prices are expected to remain weak; domestic demand and real GDP growth in emerging markets are likely to be slower than otherwise across regions.
- Policy-relevant implications and recommendations:
  - Countries with stronger fiscal frameworks, credible monetary policy, and flexible exchange rates should be better prepared to navigate less benign external conditions.
  - Exchange rate flexibility mitigates the synchronization with U.S. dollar cycles by enabling expenditure-switching; however, larger immediate income effects on domestic demand may occur through nominal adjustments—policy design should consider trade-offs.
  - The negative income effect is particularly strong for commodity exporters and countries with more rigid exchange rate regimes; these countries may need targeted policies to buffer income shocks (e.g., fiscal buffers, stabilization funds, diversified exports).
  - Higher U.S. real interest rates amplify financial channels (capital flow reversals, credit squeezes); macrofinancial policies should account for potential tightening of external financing conditions.
- Final empirical takeaway:
  - The tension between the income effect (negative for EMEs when the dollar appreciates) and the expenditure-switching effect (potentially positive) is resolved empirically in favor of a net negative impact of a stronger U.S. dollar on EME domestic demand and real GDP in many regions, especially commodity exporters and less flexible exchange rate regimes.

*Source: IMF staff calculations and analysis as presented in the provided content unit.*

### REFERENCES

### _wp15179 - REFERENCES

### References cited
- Akram, Q. F. (2009), “Commodity Prices, Interest Rates, and the Dollar,” Energy Economics, Vol. 31, pp. 838–851.  
- Borensztein, E. and C. Reinhart (1994), “The Macroeconomic Determinants of Commodity Prices,” IMF Staff Papers. Vol. 41, No. 2 (June), pp. 236–261.  
- Calvo, G., L. Lederman, and C. Reinhart (1996), “Inflows of Capital to Developing Countries in the 1990s,” The Journal of Economic Perspectives, Vol. 10, No. 2. (Spring), pp. 123-139.  
- Calvo, G., L. Lederman, and C. Reinhart (1993), “Capital Inflows and Real Exchange Rate Appreciation in Latin America: The Role of External Factors,” IMF Staff Papers, Vol. 40, No. 1 (March), pp. 108151.  
- Chinn, M. (2015), “How Much More Dollar Appreciation?,” Econonobrowser (http://econbrowser.com/archives/2015/04/how-much-more-dollar-appreciation), April.  
- Dornbusch, R. (1986), “Inflation, Exchange Rates and Stabilization." Essays in International Finance, No. 165, (October), pp. 1-24. International Finance Section, Princeton, NJ: Princeton University Press.  
- Engel, C. and J. Hamilton (1990), “Long Swings in the Dollar: Are they in the Data and Do Markets Know it?,” American Economic Review, September, pp. 689—713.  
- Frenkel, J. (1986), “Expectations and Commodity Price Dynamics: the Overshooting Model,” Journal of Agricultural Economics, Vol. 68, No. 2 (May), pp. 344-348.  
- Husain, A. R. Arezki, P. Breuer, V. Haksar, T. Helbling, P. Medas, and M. Sommer (2015), “Global Implications of Lower Oil Prices,” IMF Staff Discussion Note, forthcoming.  
- Ilzeztky, I., C. Reinhart, and K. Rogoff (2012), “Exchange Rate Arrangements into the 21st Century: Will the Anchor Currency Hold?,” mimeo (2012).  
- International Monetary Fund (2015), World Economic Outlook (Washington, April).  
- Magud, N. C. Reinhart, and E. Vesperoni (2014), “Capital Inflows, Exchange Rate Flexibility, and Domestic Credit,” Review of Development Economics, Vol. 18, No. 3 (August), pp. 415-430. (Based on NBER WP 17670 and IMF WP 12/41.)  
- Magud, N. and S. Sosa (2015), Investment in Emerging Markets: We are not in Kansas anymore... Or are we?” IMF Working Paper 15/77 (Washington, April).  
- Magud, N. And E. Vesperoni (2015), “Exchange Rate Flexibility and Credit during Capital Inflows Reversals: Purgatory, ... not Paradise,” Journal of International Money and Finance, pp. 88-110, doi:10.1016/j.jimonfin.2015.02.010.  
- Rogoff, K. (1996), “The Purchasing Power Parity Puzzle,” Journal of Economic Literature, Vol .34, No. 2(June), pp. 647–668.  
- Taylor, M., D. Peel, and L. Sarno. (2011), “Nonlinear Mean-Reversion in Real Exchange Rates: Toward a Solution to the Purchasing Power Parity Puzzles.” International Economic Review 42(4), pp. 1015-42.  
- Zhang, Y., Y. Fang, H. Tsai, and Y. Wei (2008), “Spillover Effect of U.S. Dollar Exchange Rate on Oil Policies,” Journal of Policy Modeling, Vol. 30, pp. 973–991.

### Annex A.1 — List of Countries (by region)
- LAC:
  - Argentina
  - Belize
  - Bolivia
  - Brazil
  - Chile
  - Colombia
  - Costa Rica
  - Dominican Republic
  - Ecuador
  - Guatemala
  - Guyana
  - Honduras
  - Jamaica
  - Mexico
  - Nicaragua
  - Panama
  - Peru
  - Paraguay
  - El Salvador
  - Suriname
  - Uruguay
  - Venezuela
- DevAsia:
  - Bangladesh
  - Bhutan
  - Fiji
  - Indonesia
  - India
  - Lao P.D.R.
  - Sri Lanka
  - Maldives
  - Mongolia
  - Malaysia
  - Nepal
  - Philippines
  - Papua New Guinea
  - Solomon Islands
  - Thailand
  - Tonga
  - Samoa
  - Vanuatu
- MENA:
  - United Arab Emirates
  - Bahrain
  - Algeria
  - Egypt
  - Iran
  - Jordan
  - Kuwait
  - Lebanon
  - Libya
  - Morocco
  - Oman
  - Qatar
  - Saudi Arabia
  - Sudan
  - Tunisia
- DevEuro:
  - Albania
  - Bulgaria
  - Hungary
  - Montenegro, Rep. of
  - Poland
  - Romania
  - Turkey

### Annex A.2 — U.S. Dollar Appreciation Cycles — Without Current Appreciation Cycle
- Model and regime coefficients:
  - MuI (2009)
  - Regime 1 Δ REER Coeff. 3.439.49 3.60 (Appreciation)
    - t-stat 2.98 2.55 8.24
    - p-value 0.01 0.02 0.00
  - Regime 2 Δ REER Coeff. -3.74 9.49 3.60 (Depreciation)
    - t-stat -4.34 2.55 8.24
    - p-value 0.00 0.02 0.00
- Descriptive statistics for scaled residuals:
  - Normality test: Chi^2(2)  =   4.3971 [0.1110]
  - ARCH 1-1 test: F(1,34)   =  0.22521 [0.6381]
  - Portmanteau( 6): Chi^2(6)  =   10.302 [0.1125]
- Regime durations and probabilities:
  - Regime 1
    - 1979 - 1985: years 7, avg.prob. 0.898
    - 1993 - 2001: years 9, avg.prob. 0.875
    - Total: 16 years (38.10%) with average duration of 8.00 years.
  - Regime 2
    - 1970 - 1978: years 9, avg.prob. 0.977
    - 1986 - 1992: years 7, avg.prob. 0.940
    - 2002 - 2011: years 10, avg.prob. 0.953
    - Total: 26 years (61.90%) with average duration of 8.67 years.
- Transition probabilities:
  - Regime 1, (t) → Regime 1, (t+1): 0.827
  - Regime 1, (t) → Regime 2, (t+1): 0.173
  - Regime 2, (t) → Regime 1, (t+1): 0.083
  - Regime 2, (t) → Regime 2, (t+1): 0.917

*Content derived from _wp15179 - REFERENCES (IMF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15179.pdf_