## ch2

## Source details

**Canonical URL:** [ch2](https://www.imf.org/-/media/files/publications/esr/2019/english/ch2.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/esr/2019/english/ch2.pdf.md)
- [Structured JSON version](/-/media/files/publications/esr/2019/english/ch2.pdf.json)

---

### Introduction and research focus
- Examines how exchange rate movements affect external adjustment, revisiting the Mundell-Fleming framework in light of:
  - Dominant role of certain currencies in trade invoicing, especially the US dollar.
  - Growing importance of global value chains (GVCs), where imports of intermediate goods and foreign-value-added content can alter sensitivity of gross trade flows to exchange rate movements.
- Empirical scope and data:
  - Panel of bilateral manufacturing trade among 37 advanced and emerging market economies during 1990–14.
  - Uses newly constructed bilateral prices and quantities and novel measures of value-chain-related exchange rate shocks.
  - Country sample accounts for about 85 percent of world GDP.
- Caveats:
  - Focuses on manufacturing trade elasticities; services trade and balance-sheet vulnerabilities are not considered.
  - Takes invoicing and GVC integration as given; these are interdependent and influenced by country-specific factors.
- Main authors and contributors are listed in the chapter.

### Currency of invoicing, pricing scenarios, and empirical setup
- The currency of invoicing determines exchange rate pass-through when nominal prices are sticky.
- Pricing scenarios and short-term implications:
  - Producer-currency pricing: depreciation raises import prices in domestic currency, lowering import demand; trading partners face lower prices in their currencies, raising exports for the depreciating country—balanced import and export response.
  - Dominant (third-country) currency pricing: depreciation raises import prices in domestic currency and lowers import demand, but trading partners’ local-currency prices remain unchanged if their rates vis‑à‑vis the dominant currency do not change—exports of the depreciating country may not respond, producing an unbalanced response.
- Empirical framework:
  - Panel specification models prices and quantities of bilateral manufacturing trade among 37 economies for 1990–14.
  - Disentangles price and quantity responses to bilateral and US dollar exchange rates from exporter and importer perspectives; computes trade balance responses.
  - Explores contemporaneous and lagged effects up to three years to assess short- and medium-term dynamics.

### Empirical findings on exchange rate pass-through
- Short-term (same year) estimates (exporter-currency prices):
  - A 1 percent change in the bilateral exchange rate → 0.2 percent change in trade prices in the exporter’s currency (average).
  - A 1 percent change in the exchange rate vis-à-vis the US dollar → 0.45 percent change in trade prices in the exporter’s currency (average).
  - Results are consistent from the importer perspective; starker in unweighted regressions.
- Medium-term (three-year horizon) dynamics:
  - US dollar pass-through weakens as US dollar prices become more flexible:
    - US dollar pass-through to export prices falls from 0.45 (short term) to 0.25 (medium term).
    - Bilateral exchange rate pass-through rises slightly from 0.2 (short term) to 0.25 (medium term).
  - Reduced importance of the US dollar in the medium term visible from importer perspective.
- Heterogeneity by US dollar invoicing:
  - High US dollar invoicing:
    - Bilateral pass-through to export-currency prices averages 0.1.
    - US dollar pass-through averages 0.7.
  - Low US dollar invoicing:
    - Bilateral: 0.3; US dollar: 0.2.
  - Over the medium term, US dollar invoicing effects persist but are less pronounced.

### Short-term effects of a 10 percent depreciation vis-à-vis all other currencies
- Indirect (average) estimation (prices and volumes combined):
  - Exports prices: 6.31 percent (*** p < 0.01)
  - Imports prices: 7.95 percent (*** p < 0.01)
  - Export volumes: 0.516 percent
  - Import volumes: –2.88 percent (*** p < 0.01)
  - Trade Balance: 0.322 percent of GDP
- Direct estimation by US dollar invoicing:
  - Low US dollar invoicing:
    - Exports prices: 4.81 percent (***)
    - Imports prices: 6.84 percent (***)
    - Export volumes: 1.26 percent (***)
    - Import volumes: –2.16 percent (***)
    - Trade Balance: 0.256 percent of GDP
  - High US dollar invoicing:
    - Exports prices: 8.28 percent (***)
    - Imports prices: 8.96 percent (***)
    - Export volumes: –0.59 percent
    - Import volumes: –2.77 percent (***)
    - Trade Balance: 0.276* percent of GDP
- Key short-term mechanics:
  - US dollar dominance makes bilateral depreciations have limited effects on export volumes, while depreciation vis‑à‑vis the US dollar raises import prices in the importer’s currency and reduces import demand.
  - Higher US dollar invoicing → unbalanced volume responses and a shift toward price/markup adjustments rather than export-quantity adjustments.

### Medium-term effects of a 10 percent depreciation vis-à-vis all other currencies
- Indirect (average) estimation (prices and volumes combined):
  - Exports prices: 5.07 percent (***)
  - Imports prices: 7.50 percent (***)
  - Export volumes: 4.32 percent (***)
  - Import volumes: –4.50 percent (***)
  - Trade Balance: 1.177*** percent of GDP
- Direct estimation by US dollar invoicing:
  - Low US dollar invoicing:
    - Exports prices: 3.81 percent (***)
    - Imports prices: 8.09 percent (***)
    - Export volumes: 4.56 percent (***)
    - Import volumes: –4.97 percent (***)
    - Trade Balance: 0.963*** percent of GDP
  - High US dollar invoicing:
    - Exports prices: 6.95 percent (***)
    - Imports prices: 8.62 percent (***)
    - Export volumes: 3.38 percent (***)
    - Import volumes: –4.96 percent (***)
    - Trade Balance: 1.228*** percent of GDP
- Key medium-term mechanics:
  - As prices in the invoicing currency adjust, export and import volumes become more sensitive to bilateral exchange rate movements and the influence of the US dollar exchange rate becomes economically and statistically insignificant.
  - Conventional expenditure-switching through both exports and imports reemerges in the medium term.
  - Higher US dollar invoicing has a more muted influence over the medium term; export and import volume responses become more symmetric while export/import price asymmetry increases.

### Global Value Chains (GVCs): effects on elasticities and trade balance
- Integration patterns:
  - Backward integration (import content of exports) can reach one-third to one-half in some small, highly integrated economies.
  - Large systemic economies remain dominated by traditional trade.
- Quantitative effects on medium-term volume elasticities:
  - Medium-term export volume elasticity:
    - 25th percentile of GVC integration: 0.45
    - 75th percentile of GVC integration: 0.3
  - Medium-term import volume elasticity:
    - 25th percentile of GVC integration: –0.5
    - 75th percentile of GVC integration: –0.25
- Additional GVC-related findings:
  - Greater GVC integration dampens gross trade volume elasticities in both short and medium term through backward and forward linkages.
  - GVC integration leads to somewhat higher exchange rate pass-through to both export and import prices (effects generally small).
  - USD dominance is partly related to exporters’ use of imported intermediate goods but remains significant after controlling for GVC measures.
- Trade balance implications:
  - Sensitivity of the trade balance to exchange rates falls with greater GVC integration, holding trade openness constant.
  - For the average country, a 10 percent depreciation is estimated to increase the trade balance by about 1 percentage point of GDP.
  - Over time, increases in GVC integration have been largely offset by increases in trade openness, leaving median trade-balance elasticities broadly stable since early 2001.

### Distributional and time-variation findings (2000–14)
- Trade balance response metric: "Response to a 10 percent depreciation vis-à-vis all currencies, percent of GDP."
- Time windows: 2000, 2001–04, 2005–09, 2010–14.
- Decomposition components:
  - "At constant GVC integration (Effect of changing trade openness)"
  - "At constant trade openness (Effect of changing GVC integration)"
  - "Net effect"
- Text highlights:
  - Exchange rate effects depend on GVC integration, invoicing currency, and country-specific features (including expectations about exchange rate policies).
  - Further empirical integration needed for services trade and balance-sheet vulnerabilities.

### Boxes and illustrative mechanisms
- Box 2.1 — US Dollar Shifts and Global Trade:
  - Widespread US dollar invoicing implies global US dollar movements may have short-term implications for global trade.
  - Short-term effects of a 10 percent appreciation of the US dollar:
    - United States: import prices faced by US consumers largely unchanged in short term; US exports tend to contract when US dollar appreciates.
    - Rest of world: depreciation of other currencies vis‑à‑vis the US dollar raises local-currency prices of traded goods among non‑US country pairs → import demand contracts and trade volumes among the rest of the world contract.
  - Over time, US internal adjustment becomes more balanced and effects on the rest of the world fade.
- Box 2.2 — GVC simple example:
  - Backward integration: exports from a to b that contain intermediates from c are affected by e_ab and e_ac because e_ac affects marginal costs.
  - Forward integration: intermediate exports reexported to d are affected by e_bd because e_bd determines demand.
  - Dominant currency pricing extends the dependence of exports and imports on bilateral rates and vis‑à‑vis USD rates.
- Box 2.3 — Measuring GVC-related exchange rate shocks:
  - Constructs backward (supply) and forward (demand) shifters as weighted sums of exchange rate movements of upstream suppliers and downstream buyers using weights from World Input-Output Database.
  - Δln MC_a→b = Σ_c ω_a→b,c^B ∙ Δln e_ac; Δln D_a→b = Σ_d ω_a→b,d^F ∙ Δln e_bd.
- Supply-chain flexibility illustration:
  - Inflexible (Leontief) vs flexible (trade in tasks) supply chains affect how exchange rate changes translate into demand for intermediate goods and output.

### Empirical evidence on supply-chain flexibility (Box 2.4)
- Methodology:
  - Error-correction models with short-term heterogeneous coefficients (Pesaran, Shin, and Smith 1999).
  - Test values of beta and gamma coefficients to distinguish flexible vs inflexible supply chains.
  - Data: OECD Inter-Country Input-Output Tables; 1,116 observations.
- Selected empirical estimates (short- and long-term coefficients reported with t statistics and significance notation):
  - Long term highlights (selected):
    - Importing Partners’ EER – A – B – 2.252 (–5.45)***
    - Own EER × DVA Share 0 – B – 0.607 (–4.60)***
    - Importing Partners’ EER × DVA 0 + B 1.295 (5.07)***
    - Own EER – A – B – 0.750 (–6.34)***
    - Own EER × FVA Share 0 + B 1.381 (2.31)**
  - Short term highlights (selected):
    - Error Correction Term –0.202 (–7.10)***
    - Error Correction Term –0.155 (–6.49)***
    - Importing Partners’ EER – a – b – 0.640 (–2.94)***
    - Own EER × DVA Share 0 – b – 0.477 (–4.43)***
    - Importing Partners’ EER × DVA 0 + b 0.677 (5.56)***
- Main findings:
  - Overwhelmingly rejects hypothesis that global supply chains are flexible in the short term.
  - Estimated beta and gamma coefficients are significantly different from zero and imply sizable inflexibility.
  - Reestimating for 2000–15 suggests production linkages might be fully inflexible in the short term.
  - Transition half-life from short- to long-term relationships is about three to five years; closing three-quarters of short-term deviation requires six to nine years.
  - Supply chains remain somewhat inflexible in the long term.
- Implications:
  - Supply chains’ inflexibility implies larger disruptions from trade barriers and higher costs to recreate supply chains once lost.
  - Competitiveness metrics should give greater weight to final destinations.

### Methodological and data considerations
- Econometric approach identifies average effects of exchange rate variations on prices and quantities without attributing shocks to specific sources.
- For prices, identification strengthened by assuming prices are sticky in US dollars; for quantities, omitted-variable bias is a greater concern.
- Robustness checks and rich controls (including import demand and unit labor costs) support baseline results.
- Further empirical methods and country-sample details in Online Annex 2.1.

### Conclusions and policy implications
- Measurement and analytics:
  - Dominant currency invoicing and GVC integration broaden the set of exchange rates relevant for a country’s external position; conventional effective exchange rates focusing on trading-partner currencies may be less informative for short-term dynamics.
  - Competitiveness metrics that incorporate invoicing currencies and input linkages would complement traditional measures, especially for small economies highly integrated into GVCs.
  - Improved data collection on invoicing currencies and GVC linkages is essential.
- Policy guidance:
  - Exchange rate flexibility remains key for durable external adjustment: conventional exchange rate mechanisms reappear in the medium term despite short-term modifications from invoicing and GVCs.
  - Because exchange rate changes have muted short-term effects on trade balances (notably limited export-volume responses), achieving meaningful near-term external adjustment where deficits are excessive may require:
    - Larger exchange rate movements (with attendant balance-sheet and inflationary risks), and/or
    - Tighter macroeconomic policies.
  - Exchange rate flexibility may need support from structural policies that reduce price stickiness in dominant currencies and address capacity constraints, including:
    - Improved access to credit
    - Better transportation infrastructure
  - Other temporary policies can support exchange rate flexibility in the near term but are not substitutes for exchange rate flexibility as a mechanism for durable external adjustment.

*Source: ch2 - Introduction and related sections, IMF 2019 External Sector Report (chapter content).*

### Introduction

### ch2 - Introduction

### Introduction and research focus
- The chapter examines how exchange rate movements affect external adjustment, revisiting the Mundell-Fleming framework in light of two international-trade features:
  - The dominant role of certain currencies in the invoicing of trade, especially the US dollar.
  - The growing importance of global value chains (GVCs), where imports of intermediate goods and foreign-value-added content can alter the sensitivity of gross trade flows to exchange rate movements.
- The analysis studies the response of trade prices and quantities to exchange rate movements in a panel of bilateral manufacturing trade among 37 advanced and emerging market economies during 1990–14.
- The analysis uses newly constructed data on bilateral prices and quantities and novel measures of value-chain-related exchange rate shocks.
- Caveats noted:
  - The chapter focuses on manufacturing trade elasticities; services trade and balance-sheet vulnerabilities are not considered.
  - The analysis takes invoicing and GVC integration as given and recognizes that they are interdependent and influenced by country-specific factors.

*Main authors: Gustavo Adler, Sergii Meleshchuk, and Carolina Osorio-Buitron; support from Jair Rodriguez, Kyun Suk Chang, and Zijiao Wang; contributions from Tam Bayoumi, Diego Cerdeiro, and Jelle Barkema.*

### Currency of trade invoicing: theory and empirical setup
- The currency of invoicing determines exchange rate pass-through (how exchange rate changes translate into domestic-currency prices) when nominal prices are sticky.
- Pricing scenarios and short-term implications:
  - Producer-currency pricing (as in Mundell-Fleming): depreciation raises import prices in domestic currency, lowering import demand; trading partners face lower prices in their currencies, raising exports for the depreciating country—balanced import and export response.
  - Dominant (third-country) currency pricing: depreciation raises import prices in domestic currency and lowers import demand, but trading partners’ local-currency prices remain unchanged if their exchange rates vis-à-vis the dominant currency do not change—exports of the depreciating country may not respond, producing an unbalanced response.
- Empirical framework:
  - Panel specification models prices and quantities of bilateral manufacturing trade among 37 economies for 1990–14.
  - The framework disentangles price and quantity responses to bilateral and US dollar exchange rates from exporter and importer perspectives and computes trade balance responses.
  - Contemporaneous and lagged effects up to three years are explored to assess short- and medium-term dynamics.
- Coverage note: the country sample accounts for about 85 percent of world GDP.

### Empirical findings on exchange rate pass-through
- Short-term (same year as shock) evidence:
  - The exchange rate vis-à-vis the US dollar is a statistically and economically important driver of trade prices in domestic currency even after controlling for the bilateral exchange rate.
  - Average estimated effects (exporter-currency prices):
    - A 1 percent change in the bilateral exchange rate → 0.2 percent change in trade prices in the exporter’s currency (average).
    - A 1 percent change in the exchange rate vis-à-vis the US dollar → 0.45 percent change in trade prices in the exporter’s currency (average).
  - Results are consistent from the importer perspective and are starker in unweighted regressions where smaller economies receive equal weight.
- Medium-term (three-year horizon) evidence:
  - US dollar pass-through weakens as US dollar prices become more flexible:
    - US dollar pass-through to export prices falls from 0.45 (short term) to 0.25 (medium term).
    - Bilateral exchange rate pass-through rises slightly from 0.2 (short term) to 0.25 (medium term).
  - The reduced importance of the US dollar in the medium term is also visible from the importer perspective.
- Heterogeneity by degree of US dollar invoicing:
  - For countries with high US dollar invoicing:
    - Pass-through from bilateral exchange rates to export-currency prices averages 0.1.
    - Pass-through from the US dollar exchange rate averages 0.7.
  - For countries with low US dollar invoicing:
    - Corresponding magnitudes are 0.3 (bilateral) and 0.2 (US dollar).
  - Over the medium term, US dollar invoicing effects persist but are less pronounced.

### Implications for trade volumes and external adjustment
- The dominance of the US dollar affects export and import volume responses:
  - For countries other than the United States, in the short term:
    - Bilateral export volumes respond positively to a bilateral exchange rate depreciation (i.e., appreciation of trading partner’s currency alone).
    - Bilateral exports respond negatively to a depreciation vis-à-vis the US dollar (because trading partners’ local-currency prices do not change when trade is invoiced in US dollars).
  - For the United States:
    - A depreciation of the US dollar entails limited effects through imports (prices in US dollars remain largely unchanged).
    - US exports increase on account of higher demand from the rest of the world as their prices in local currency of trading partners fall.
- Short- versus medium-term dynamics:
  - Short-term nominal rigidities make invoicing-currency effects (especially USD dominance) particularly consequential for adjustment via prices and volumes.
  - Over the medium term, as prices adjust, bilateral exchange rates play a relatively larger role in transmitting exchange rate changes into domestic-currency trade prices.

### Methodological and data considerations
- The econometric approach identifies average effects of exchange rate variations on prices and quantities without attributing shocks to specific sources.
- For prices, identification is strengthened by assuming prices are sticky in US dollars; for quantities, omitted-variable bias is a greater concern, but robustness checks and rich controls (including import demand and unit labor costs) support baseline results.
- Further details on empirical methods and country samples are provided in Online Annex 2.1.

*Source: ch2 - Introduction (PDF chapter), International Monetary Fund, July 2019.*

### 1. Average Effects2. Effects for Countries with Low/High US Dollar Invoicing

### 1. Average Effects2. Effects for Countries with Low/High US Dollar Invoicing

### Short-term effects of a 10 percent depreciation vis-à-vis all other currencies
- Indirect (average) estimation (prices and volumes combined):
  - Exports prices: 6.31 percent (*** p < 0.01)
  - Imports prices: 7.95 percent (*** p < 0.01)
  - Export volumes: 0.516 percent
  - Import volumes: –2.88 percent (*** p < 0.01)
  - Trade Balance: 0.322 percent of GDP
- Direct estimation by US dollar invoicing:
  - Low US dollar invoicing:
    - Exports prices: 4.81 percent (***)
    - Imports prices: 6.84 percent (***)
    - Export volumes: 1.26 percent (***)
    - Import volumes: –2.16 percent (***)
    - Trade Balance: 0.256 percent of GDP
  - High US dollar invoicing:
    - Exports prices: 8.28 percent (***)
    - Imports prices: 8.96 percent (***)
    - Export volumes: –0.59 percent
    - Import volumes: –2.77 percent (***)
    - Trade Balance: 0.276* percent of GDP
- Key short-term mechanics and interpretation:
  - The US dollar plays a dominant role in the near term: bilateral depreciations (in the importer’s currency alone) have limited effects on export volumes, while depreciation vis-à-vis the US dollar raises import prices in the importer’s currency and reduces import demand.
  - US dollar invoicing is associated with:
    - Unbalanced volume responses: import volumes fall with depreciation irrespective of invoicing share; export volumes react less when US dollar invoicing is greater.
    - Greater (and more symmetric) price responses across exporter and importer currencies under high US dollar invoicing.
    - A shift in the short-term composition of external adjustment toward price/markup adjustments rather than export-quantity adjustments as US dollar invoicing rises.

### Medium-term effects of a 10 percent depreciation vis-à-vis all other currencies
- Indirect (average) estimation (prices and volumes combined):
  - Exports prices: 5.07 percent (***)
  - Imports prices: 7.50 percent (***)
  - Export volumes: 4.32 percent (***)
  - Import volumes: –4.50 percent (***)
  - Trade Balance: 1.177*** percent of GDP
- Direct estimation by US dollar invoicing:
  - Low US dollar invoicing:
    - Exports prices: 3.81 percent (***)
    - Imports prices: 8.09 percent (***)
    - Export volumes: 4.56 percent (***)
    - Import volumes: –4.97 percent (***)
    - Trade Balance: 0.963*** percent of GDP
  - High US dollar invoicing:
    - Exports prices: 6.95 percent (***)
    - Imports prices: 8.62 percent (***)
    - Export volumes: 3.38 percent (***)
    - Import volumes: –4.96 percent (***)
    - Trade Balance: 1.228*** percent of GDP
- Key medium-term mechanics and interpretation:
  - As prices in the currency of invoicing adjust, export and import volumes become more sensitive to bilateral exchange rate movements and the influence of the US dollar exchange rate becomes economically and statistically insignificant.
  - The conventional expenditure-switching mechanism through both exports and imports reemerges in the medium term.
  - Higher US dollar invoicing has a more muted influence over the medium term; export and import volume responses become more symmetric while export/import price asymmetry increases.

### Global Value Chains (GVC) — effects on exchange rate elasticities and trade balance
- Integration patterns:
  - Backward integration (import content of exports) can reach one-third to one-half in some small, highly integrated economies.
  - Large systemic economies (for example, China, Japan, United States) remain dominated by traditional trade.
- Quantitative effects of greater GVC integration on volume elasticities:
  - Medium-term export volume elasticity:
    - 25th percentile of GVC integration: 0.45
    - 75th percentile of GVC integration: 0.3
  - Medium-term import volume elasticity:
    - 25th percentile of GVC integration: –0.5
    - 75th percentile of GVC integration: –0.25
- Additional GVC-related findings:
  - Greater GVC integration dampens gross trade volume elasticities in both the short term and medium term through backward and forward linkages.
  - GVC integration leads to somewhat higher exchange rate pass-through to both export and import prices, reflecting greater sensitivity of marginal costs and input demand, though effects are generally small.
  - The dominant role of the US dollar is partly related to exporters’ use of imported intermediate goods but remains significant even after controlling for GVC measures.
- Trade balance implications:
  - The sensitivity of the trade balance to exchange rates falls with greater GVC integration, holding trade openness constant.
  - For the average country (in terms of GVC integration and trade openness), a 10 percent depreciation is estimated to increase the trade balance by about 1 percentage point of GDP.
  - Over time, increases in GVC integration have been largely offset by accompanying increases in trade openness, leaving median trade-balance elasticities broadly stable since early 2001.

### Conclusions and policy implications
- Measurement and analytics:
  - Dominant currency invoicing and GVC integration broaden the set of exchange rates relevant for a country’s external position; conventional effective exchange rates that focus on trading-partner currencies may be less informative for short-term dynamics.
  - Competitiveness metrics that take invoicing currencies and input linkages into account would complement traditional measures, especially for small economies highly integrated into GVCs.
  - Improved data collection on invoicing currencies and GVC linkages is essential.
- Policy guidance:
  - Exchange rate flexibility remains key for durable external adjustment: conventional exchange rate mechanisms are present in the medium term despite short-term modifications due to invoicing and GVCs.
  - Because exchange rate changes have muted short-term effects on trade balances (notably limited export-volume responses), achieving meaningful near-term external adjustment where deficits are excessive may require:
    - Larger exchange rate movements (with attendant balance-sheet and inflationary risks), and/or
    - Tighter macroeconomic policies.
  - Exchange rate flexibility may need support from structural policies that reduce price stickiness in dominant currencies and address capacity constraints, including:
    - Improved access to credit
    - Better transportation infrastructure
  - Other temporary policies can support exchange rate flexibility in the near term but are not substitutes for exchange rate flexibility as a mechanism for durable external adjustment.

*Source: Chapter section "1. Average Effects2. Effects for Countries with Low/High US Dollar Invoicing" from the IMF 2019 External Sector Report chapter content provided.*

### 1. Distribution, 2001 and 2014

### 1. Distribution, 2001 and 2014

### Trade Balance Response — Distribution and Variation over Time, 2000–14
- Response metric: "Response to a 10 percent depreciation vis-à-vis all currencies, percent of GDP."
- Time windows reported: 2000, 2001–04, 2005–09, 2010–14.
- Components decomposed:
  - "At constant GVC integration (Effect of changing trade openness)"
  - "At constant trade openness (Effect of changing GVC integration)"
  - "Net effect"
- Figure referenced: "Figure 2.8. Trade Balance Response—Distribution and Variation over Time, 2000–14" (visual distribution and time variation across the sample period).

### Key contextual findings in chapter text
- Exchange rate effects on external adjustment depend on:
  - Integration into global value chains (GVCs).
  - Pricing strategies and invoicing currency.
  - Country features, including expectations about exchange rate policies.
- Important areas for further empirical integration:
  - Whether manufacturing trade results apply to services trade (such as tourism), which relies more on nontradable inputs.
  - Role of external balance sheet vulnerabilities in shaping exchange rate adjustment.
- Conclusion: "Further efforts are necessary to integrate empirically these additional trade and financial features."

*Authors/contributors noted in boxes: Gustavo Adler, Carolina Osorio Buitron, Sergii Meleshchuk.*

---

### Box 2.1 — US Dollar Shifts and Global Trade
- Core mechanism:
  - Widespread US dollar invoicing implies global movements in the US dollar (vis‑à‑vis all other currencies) may have short-term implications for global trade.
- Estimated short-term effects of a 10 percent appreciation of the US dollar (weighted regression; Figure 2.1.1):
  - United States:
    - Imports: "the price of imports US consumers face is largely unchanged, so will be import demand" (short term).
    - Exports: "Export volumes... tend to contract in response to the appreciation of the US dollar" because the rest of world faces higher domestic prices of tradable goods.
  - Rest of the world:
    - Depreciation of other currencies vis‑à‑vis the US dollar increases local currency prices of goods traded between country pairs excluding the United States → "import demand contracts and, thus, trade volumes among countries in the rest of the world contract."
- Time dynamics:
  - "Over time, the adjustment in the United States becomes more balanced (with both export and import volumes reacting to exchange rate movements) and the effects on the rest of the world fade away."
- Note: "This exercise sheds light on, among other things, the spill-overs of US monetary policy through trade."
- Sources for data and estimates: "Data sets from Gopinath and others (2018) and Boz and others (forthcoming); and IMF staff estimates." Point estimates and 95 percent confidence bands reported in original figure.

---

### Box 2.2 — The Economics of Global Value Chains: A Simple Example
- Traditional trade framing:
  - Bilateral exchange rate e_ab is the relevant rate if priced in either country's currency: T_a→b = f[e_ab] and T_b→a = f[e_ab].
- GVC complexity: third-country exchange rates matter via:
  - Backward integration (BWD — upstream suppliers):
    - If exports from a to b contain intermediate goods imported from c, then exports T_a→b^a are affected by e_ab and e_ac because e_ac affects a's marginal costs, MC_a ≡ MC_a(e_ac).
    - Formally: T_a→b^a = T_a→b^a(e_ab; e_ac). Greater substitutability between domestic and foreign intermediate inputs reduces the impact of e_ac on marginal costs.
  - Forward integration (FWD — downstream buyers):
    - If intermediate exports from a to b are reexported to d, then T_a→b^a is affected by e_bd (exchange rate of b vis‑à‑d) because e_bd determines demand D(e_bd).
    - Formally: T_a→b^a = T_a→b^a(e_ab; e_ac; e_bd).
- Generic characterization combining stand-alone, BWD, and FWD:
  - T_a→b^a ≡ f_a→b^a[ e_ab (stand-alone), MC_a→b^a(e_ac) (BWD), D(e_bd) (FWD) ].
- Dominant currency (for example, US dollar) pricing extension:
  - Exports and imports depend on bilateral rates and vis‑à‑vis USD rates (e_a$, e_b$), and through MC and D functions that include USD terms.
  - Export equation example: T_a→b^a = f_a→b^a[ e_ab, e_a$, MC_a→b^a(e_ac, e_a$), D_a→b(e_bd, e_b$) ].
  - Resulting decomposition of dT/d e includes stand-alone bilateral, stand-alone vis‑à‑vis USD, BWD bilateral and vis‑à‑vis USD, and FWD bilateral and vis‑à‑vis USD components.
- Table 2.2.1 (Effects of a Depreciation vis‑à‑vis All Other Currencies under GVC integration):
  - Prices (in country a’s currency) / Quantities
  - Exports (a → b): Prices: ++ (Stand-alone) ; BWD/FWD Linkages: +− (BWD) ; Quantities: (Stand-alone) +? ; BWD/FWD Linkages: (BWD) — table shows mixed signs for linkages.
  - Imports (b → a): Prices: ++ (Stand-alone) ; BWD/FWD Linkages: − + (FWD) ; Quantities: Stand-alone negative; FWD positive.
  - Note: "BWD = backward integration; FWD = forward integration. Stand-alone denotes effects on prices for a combination of producer and consumer currency pricing."
- Interpretation: Depreciation vis‑à‑vis all other currencies operates on exports and imports both directly and through BWD and FWD linkages, with potentially offsetting effects on quantities and prices depending on GVC structure and invoicing currency.

---

### Box 2.3 — Measuring Global-Value-Chain-Related Exchange Rate Shocks at the Bilateral Level
- Scope: Analysis focuses on the period 2001–14 and 37 countries for which both required data sources are available.
- Data inputs:
  - Domestic and imported intermediate inputs: 2016 World Input-Output Database.
  - Bilateral price and quantity indices: Boz and Cerutti (forthcoming).
- Construction of shifters:
  - Backward (supply) shifter:
    - "A backward (supply) shifter can be constructed as the weighted sum of exchange rate movements of exporter a vis‑à‑vis its upstream suppliers."
    - Weights ω_a→b,c^B correspond to the import content from c in exports from a to b.
    - Δln MC_a→b = Σ_c ω_a→b,c^B ∙ Δln e_ac.
  - Forward (demand) shifter:
    - "A forward (demand) shifter is the weighted sum of exchange rate movements of importer b vis‑à‑vis its downstream buyers."
    - Weights ω_a→b,d^F correspond to the exports from a to b that are reexported to d.
    - Δln D_a→b = Σ_d ω_a→b,d^F ∙ Δln e_bd.
- Weight sums:
  - Σ_c ω_a→b,c^B = import content of exports from a to b.
  - Σ_d ω_a→b,d^F = reexported content of exports from a to b.
- Each measure includes:
  - Direct component: production inputs directly imported.
  - Indirect component: import content of intermediate inputs supplied by the domestic economy.

---

### Supply‑chain flexibility — empirical investigation and illustrative cases
- Motivation: Degree of flexibility of supply chains (how easily they can reconfigure) affects how changes in competitiveness (real exchange rates) translate into changes in demand for domestic goods and output.
- Illustration (Korean flat screens example):
  - Inflexible supply chains (Leontief case): Chinese producer uses similar amounts of Korean flat screens irrespective of price → demand for Korean flat screens determined by demand for final Chinese computer in the US; depreciation of won matters only in proportion to flat screen's contribution to final good.
  - Flexible supply chains (trade in tasks): Chinese producer responds to flat screen price as much as US consumers respond to computer price → Korean flat screens effectively directly exported to the United States; value of won entirely matters for demand for flat screens; renminbi value inconsequential.
  - Implication: GVCs mute impact on gross trade but impact on output rose through the 2008 financial crisis then fell modestly afterward.
- Empirical specification to assess flexibility (annual data, 59 countries, 21 years):
  - Let FVA_it and DVA_it denote foreign and domestic components embedded in country i’s exports to final demand at time t.
  - Regression forms:
    - FVA_it = η + α REER_it* + β dva_it × REER_it + γ dva_it × REER_it* + δ X_it + ε_it.
    - DVA_it = η + α REER_it + β fva_it × REER_it* + γ fva_it × REER_it + δ X_it + ε_it.
  - Definitions:
    - REER = country i’s real effective exchange rate.
    - REER* = real effective exchange rate of country i’s intermediate-import partners.
    - dva (fva) = share of domestic (foreign) value added in country i’s gross exports to final demand.
    - X = vector of controls.
- Data note: "See Bayoumi and others (forthcoming) for details on the construction of the data set."

*Source: IMF staff, "CHAPTER 2 EXChANgE RATES ANd EXTERNAL AdjuSTMENT," 2019 EXTERNAL SECTOR REPORT, International Monetary Fund | July 2019.*

### Box 2.4. How Inflexible Are Global Supply Chains?

### Box 2.4. How Inflexible Are Global Supply Chains?

### Methodology
- Estimation approach: error-correction models with short-term heterogeneous coefficients (Pesaran, Shin, and Smith 1999).
- Key test: values of the beta and gamma coefficients. If beta = gamma = 0, supply chains are flexible (trade in tasks). If beta and gamma matter alongside alpha, supply chains are inflexible (trade in goods). Fully inflexible implies beta = −gamma = alpha.
- Controls included: foreign demand, oil price, non-oil commodity prices.
- Data source: Organisation for Economic Co-operation and Development, Inter-Country Input-Output Tables.
- Number of observations: 1,116.

### Empirical Estimates (selected coefficients from Table 2.4.1)
- Long Term (Foreign Value Added and Domestic Value Added equations):
  - Importing Partners’ EER – A – B – 2.252 (–5.45)***
  - Own EER × DVA Share 0 – B – 0.607 (–4.60)***
  - Importing Partners’ EER × DVA 0 + B 1.295 (5.07)***
  - Own EER – A – B – 0.750 (–6.34)***
  - Importing Partners’ EER × FVA 0 – B – 0.435 (–0.75)
  - Own EER × FVA Share 0 + B 1.381 (2.31)**
- Short Term:
  - Error Correction Term –0.202 (–7.10)***
  - Error Correction Term –0.155 (–6.49)***
  - Importing Partners’ EER – a – b – 0.640 (–2.94)***
  - Own EER × DVA Share 0 – b – 0.477 (–4.43)***
  - Importing Partners’ EER × DVA 0 + b 0.677 (5.56)***
  - Own EER – a – b – 0.297 (–1.54)
  - Importing Partners’ EER × FVA 0 + b – 0.719 (–1.01)
  - Own EER × FVA Share 0 + b 0.757 (1.05)
- Note on statistical notation: t statistics in parentheses; * p < 0.1 ** p < 0.05 *** p < 0.01.

### Main Findings
- The evidence "overwhelmingly rejects the hypothesis that global supply chains are flexible in the short term."
- Estimated beta (gamma) coefficients are significantly negative (positive) in both foreign- and domestic-value-added equations.
- For foreign value added, beta and gamma are approximately equal and opposite, and sizable relative to the absolute value of alpha.
  - The point estimate implies this ratio is about two-thirds over the entire 1995–2015 sample.
- Domestic value added coefficients point to a similar qualitative result but are less precisely estimated.
- Reestimating the model for 2000–15 (removing the first five years) indicates production linkages might be fully inflexible in the short term; hypotheses that alpha, beta, and gamma are all equal in absolute terms cannot be rejected in either equation.
- Short-term effects persist:
  - Estimated half-life for transition from short- to long-term relationships is about three to five years.
  - Closing three-quarters of any short-term deviation requires six to nine years.
  - Short-term coefficients remain relevant for horizons of five years.
- Supply chains remain somewhat inflexible even in the long term: some beta and gamma terms are significant in both equations, indicating persistent complementarities in production.

### Implications and Interpretation
- Supply chains are "pretty inflexible," implying:
  - Larger disruptions from trade barriers.
  - Higher costs to recreate supply chains once lost.
- Competitiveness calculations should give greater weight to final destinations (countries that consume final goods) compared with existing practice.
- The observed rising share of foreign inputs in international trade (Figure 2.4.1) is attributed to increasingly complex production chains involving increasingly specialized inputs.

*Source: IMF staff calculations; Organisation for Economic Co-operation and Development, Inter-Country Input-Output Tables.*

---


_Source: https://www.imf.org/-/media/files/publications/esr/2019/english/ch2.pdf_