## _wp07166

## Source details

**Canonical URL:** [_wp07166](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07166.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07166.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07166.pdf.json)

---

### Overview and motivation
- Frames policy debate on exchange rate regimes in East Asia following the East Asian currency crisis of 1997.
- Notes prior characterization of the region as a "soft dollar zone" (Frankel (1993); Frankel and Wei (1994)) and cites policy proposals:
  - Common basket peg (Williamson (1999); Williamson (2000)).
  - Basket peg to stabilize the real effective exchange rate (Ogawa and Ito (2002)).
  - Formation of a yen bloc (Kwan (1998)).
  - Restoring dollar-based regime with stronger bank regulation (McKinnon (2000a)).
- Institutional signals: ASEAN Hanoi Plan of Action (1998) studies on common currency; ADB seminar on a single currency; Haruhiko Kuroda (Kuroda (2004)) arguing for “intra-region exchange rate stabilization and ultimately a single currency in East Asia”; finance ministers’ tentative measures toward coordination (New York Times, May 5, 2006).
- Core research question: Does a single currency in East Asia facilitate international trade among member countries?

### Theoretical framework and empirical approach
- Theoretical basis:
  - Micro-founded gravity equation from Anderson and van Wincoop (2003) general equilibrium trade model.
  - CES utility with elasticity of substitution σ.
  - Bilateral trade barrier t_ij and multilateral trade resistance P_j.
  - Indirect utility (welfare) metric: U_j = y_j / P_j.
  - Comparative-statics: solve the general equilibrium model before and after changes in trade barriers (e.g., forming a currency union) to obtain changes in bilateral trade and welfare.
  - Model implications highlighted:
    - Increase in intra-union trade is smaller the larger the size of the union; welfare effect is larger the bigger the size of the union.
    - Increase in trade among union members is smaller the higher the pre-union level of trade; welfare effect is larger the higher pre-union trade.
- Empirical specification:
  - Logged gravity equation controlling for multilateral resistance via fixed effects; currency union dummy δ_ij enters trade-barrier specification.
  - Currency-union coefficient of interest: β2.
  - Assumptions for counterfactuals: σ = 5; border effect from Anderson and van Wincoop (2003) included.
- Data:
  - Bilateral trade among 137 countries for years 1980 and 1990.
  - Bilateral trade measured as sum of nominal exports and imports deflated by the US price index and divided by two.
  - Controls: log distance, common language, adjacency, political union, colony/colonizer, common colonizer, FTA; country/time fixed effects for multilateral resistance.

### Main estimation results and robustness
- Main estimate used for counterfactuals (Table 1, column (4)):
  - Currency Union coefficient β2 = 0.91 (standard error 0.18).
  - Note: β2 = 0.91 often interpreted as increase in trade of almost 250 percent (e^{0.91} ≈ 2.48) in partial-equilibrium; general equilibrium solutions are required for correct trade and welfare effects.
- Robustness (Table 2; currency-union estimates; standard errors in parenthesis):
  - 1980 only: 1.075 (0.22)***
  - 1990 only: 0.735 (0.27)***
  - No North-North trade: 0.741 (0.17)***
  - Only developing countries: 0.565 (0.21)**
  - Similar-size countries (by less than factor of 5): 0.928 (0.19)***
  - No Australia, France, New Zealand, United Kingdom, and United States: 0.909 (0.20)***
- Smallest reported sensitivity estimate: 0.57 (interpreted as 176 percent increase in bilateral trade, e^{0.57} ≈ 1.76).
- External studies report ranges like (.59, .80).

### General-equilibrium counterfactuals: East Asian groupings (Table 3)
- Country groupings:
  - NIES3: Hong Kong, Singapore, Korea
  - ASEAN4: Thailand, Malaysia, Indonesia, Philippines
  - ASEAN5: ASEAN4 plus Singapore
  - East Asia (EA): NIES3 and ASEAN4
- Counterfactual average percentage changes (percentage increases; standard errors in parenthesis reported in the source):
  - EA + China — Trade: 114 percent; Welfare: 3.7 percent
  - EA (without China) — Trade: 125 percent; Welfare: 2.5 percent
  - NIES3 — Trade: 1500.1 percent; Welfare: 0.1 percent
  - ASEAN4 — Trade: 1183.2 percent; Welfare: 1.0 percent
  - ASEAN5 — Trade: 1203.0 percent; Welfare: 0.9 percent
  - China + Hong Kong — Trade: 1410.9 percent; Welfare: 0.3 percent
- Interpretation:
  - Common currency among all East Asian countries raises bilateral trade substantially (e.g., 114 percent) while welfare effects are moderate (e.g., 3.7 percent when China included).
  - Trade-creating effects for East Asian unions exceed the EMU’s estimated 58 percent trade increase (Rose and van Wincoop (2001)), but welfare gains are smaller than the 11 percent average welfare increase reported for EMU members in that study.
  - Results are upper bounds if β2 is overestimated.

### Extended membership: adding Australia, New Zealand, and Japan (Table 4)
- Counterfactual percentage increases (percentage increases; standard errors in parenthesis reported in the source):
  - EA + China + Australia + New Zealand — Trade: 102 percent; Welfare: 5.0 percent
  - EA + Australia + New Zealand — Trade: 110 percent; Welfare: 4.1 percent
  - NIES3 + Australia + New Zealand — Trade: 129 percent; Welfare: 2.1 percent
  - ASEAN4 + Australia + New Zealand — Trade: 100 percent; Welfare: 5.2 percent
  - ASEAN5 + Australia + New Zealand — Trade: 104 percent; Welfare: 4.8 percent
  - EA + China + Japan + Australia + New Zealand — Trade: 93 percent; Welfare: 6.2 percent
- Key findings:
  - Including Australia and New Zealand retains large trade-creating effects (100–129 percent) and raises welfare materially (e.g., ASEAN4 + Australia + New Zealand welfare +5.2 percent).
  - Adding Japan further raises welfare (e.g., EA + China + Japan + Australia + New Zealand welfare +6.2 percent).

### Unilateral (bilateral) vs regional arrangements: yen zone vs dollar zone (Table 5)
- Unilateral scenarios: each East Asian economy forms a currency union with Japan, or with the United States (treated as exogenous).
- Average percentage increases by country (Trade: CU with Japan; CU with US — Welfare: CU with Japan; CU with US). Standard errors in parenthesis in the source:
  - Hong Kong — Trade: 149; 147 — Welfare: 0.2; 0.4  (standard errors: (43)(42)(0.05)(0.12))
  - Singapore — Trade: 150; 147 — Welfare: 0.1; 0.3  (standard errors: (43)(42)(0.03)(0.11))
  - Korea — Trade: 110; 148 — Welfare: 4.1; 0.3  (standard errors: (27)(42)(1.21)(0.08))
  - Indonesia — Trade: 142; 143 — Welfare: 0.8; 0.7  (standard errors: (40)(40)(0.24)(0.20))
  - Malaysia — Trade: 143; 128 — Welfare: 0.7; 2.1  (standard errors: (40)(34)(0.22)(0.68))
  - Thailand — Trade: 143; 144 — Welfare: 0.7; 0.6  (standard errors: (40)(41)(0.21)(0.20))
  - Philippines — Trade: 137; 134 — Welfare: 1.3; 1.6  (standard errors: (37)(36)(0.40)(0.49))
  - China — Trade: 133; 140 — Welfare: 1.7; 1.0  (standard errors: (36)(39)(0.50)(0.29))
  - EA (East Asia aggregate) — Trade: 111; 115 — Welfare: 4.0; 3.6  (standard errors: (27)(29)(1.15)(1.03))
  - EA + China — Trade: 101; 105 — Welfare: 5.2; 4.7  (standard errors: (24)(25)(1.45)(1.32))
- Comparative conclusions:
  - Trade-creating effects are large and similar for unilateral and regional arrangements (around 100–150 percent for most cases).
  - Welfare effects differ: regional currency unions generate much larger welfare gains than unilateral bilateral unions for most countries.
  - Exception: Korea — bilateral union with Japan yields welfare +4.1 percent, comparable to regional gains.
  - Regional union including Japan yields larger welfare increases for East Asian members than regional union including the United States (e.g., EA welfare ≈ 4.0 percent with Japan; with US lower in many cases).
  - Caveat: East Asia was de facto dollar-pegged in the sample period; this may bias welfare effects estimated for dollar-union scenarios upward.

### Synthesis: substantive findings and policy-relevant takeaways
- Trade effects:
  - Currency union involving only East Asian countries roughly doubles bilateral trade among members in general-equilibrium estimates (examples: EA + China trade +114 percent; EA trade +125 percent).
  - Trade-creating effects of East Asian currency unions generally larger than estimated EMU effect (+58 percent cited).
- Welfare effects:
  - Welfare gains from East Asian-only currency unions are moderate (examples: ASEAN5 +3.0 percent; EA + China +3.7 percent).
  - Including Australia, New Zealand, and especially Japan raises welfare gains materially (examples: ASEAN4 + Australia + New Zealand welfare +5.2 percent; EA + China + Japan + Australia + New Zealand welfare +6.2 percent).
  - Regional (multilateral) currency unions yield substantially larger welfare gains than unilateral bilateral currency unions for most East Asian economies.
  - Japanese participation contributes more to welfare gains than US participation in the modeled scenarios.
- Policy implication emphasized:
  - Certain regional currency arrangements are beneficial for East Asia; Japanese participation appears especially important to generate larger welfare gains.
  - As regional integration and intra-regional trade increase over time, potential gains from a currency union could rise.

### Quantitative highlights reiterated
- Welfare increase reported: "3.7 percent on average (Table 3)" for EA + China scenario.
- Adding Japan to the union increases welfare by 5.2 percent (reported in extended scenarios).
- Adding Australia and New Zealand to the union increases welfare by 5.0 percent (reported in extended scenarios).
- Marginal contribution by Japan to the welfare increase is larger than that by Australia and New Zealand combined.

### Caveats, limitations, and recommended further work
- Analysis focuses only on trade aspect and benefits associated with changes in trade costs; other potential benefits of a currency union are not taken into account.
- Costs of forming a currency union are not analyzed.
- The paper does not determine which currency (regional common currency, Japanese yen, or U.S. dollar) should be used for different arrangements.
- Results may be upper-bound if β2 is overestimated; pre-sample de facto dollar pegs may bias dollar-union welfare estimates upward.
- Recommended further work:
  - Quantify benefits beyond reductions in trade costs.
  - Estimate the costs of forming a currency union to fully evaluate feasibility.
  - Analyze currency choice issues for different currency arrangements in East Asia.

*Source: _wp07166 - IMF Working Paper excerpt.*

### References..............................................................................................................

### _wp07166 - References................................................................................................................................21

### Overview
- The document frames the policy debate on exchange rate regimes in East Asia following the East Asian currency crisis of 1997.
- It highlights interest among policymakers and regional institutions in alternative arrangements, including a common currency.

### Context and motivation
- Before the crisis, East Asia was characterized as a "soft dollar zone" with countries loosely pegging their currencies to the US dollar (Frankel (1993) and Frankel and Wei (1994)).
- Various studies cited in the text argue that this de facto dollar peg contributed to the 1997 crisis.
- Policy proposals referenced include:
  - A common basket peg among East Asian economies (Williamson (1999) and Williamson (2000)).
  - A basket peg to stabilize the real effective exchange rate (Ogawa and Ito (2002)).
  - Formation of a yen bloc in East Asia (Kwan (1998)).
  - Restoring a dollar-based regime while strengthening bank regulations (McKinnon (2000a)).
- Institutional and political signals:
  - ASEAN issued the Hanoi Plan of Action in 1998 and indicated studies on the feasibility of establishing a common currency in the region.
  - The Asian Development Bank organized a seminar on a single currency for East Asia.
  - Haruhiko Kuroda (Special Adviser to the Cabinet of Japan) argued that high regional integration calls for “intra-region exchange rate stabilization and ultimately a single currency in East Asia” (Kuroda (2004)).
  - Finance ministers from China, Japan and Korea announced tentative measures to coordinate their currencies in manners that can ultimately lead to the creation of a common regional currency (“Asian Finance Ministers Seek Common Currency,” New York Times, May 5, 2006).

### Previous literature and contrasting conclusions
- Optimum Currency Area (OCA) literature yields mixed evidence for East Asia:
  - Eichengreen and Bayoumi (1999) find East Asian countries are as plausible candidates for a currency union as EU countries based on OCA grounds.
  - Chow and Kim (2003) report Asian countries face asymmetric shocks and thus a common currency would be costly.
  - Lee and Barro (2006) conclude it is unclear whether East Asia is an optimum currency area, but their calibration shows a currency union of a broad group of East Asian economies will generate a net welfare gain for most member countries.

### Scope, approach, and core research question
- This paper focuses exclusively on the international trade aspect of currency unions, building on recent developments in the empirical trade literature.
- The paper addresses the following question:
  - 1. Does a single currency in East Asia facilitate international trade among member countries?

*Source: _wp07166 - References................................................................................................................................21*

### 2. How does a currency union in East Asia affect the welfare of member countries?

### 2. How does a currency union in East Asia affect the welfare of member countries?

### Theoretical framework (Gravity model and welfare metric)
- Model: micro-founded gravity equation derived from Anderson and van Wincoop (2003) general equilibrium trade model.
- Key primitives and definitions:
  - CES utility with elasticity of substitution σ.
  - Bilateral trade barrier t_ij; importer price p_ij = p_i t_ij.
  - Multilateral trade resistance P_j summarizes trade costs between j and all partners.
  - Indirect utility (welfare) metric: U_j = y_j / P_j — welfare is inversely related to the multilateral resistance index.
- Comparative-statics capability: once parameters are estimated, solve the general equilibrium model before and after changes in trade barriers (e.g., forming a currency union) to obtain changes in bilateral trade and welfare.
- Model implications stressed in the paper:
  - Increase in intra-union trade is smaller the larger the size of the union; welfare effect is larger the bigger the size of the union.
  - Increase in trade among union members is smaller the higher the pre-union level of trade; welfare effect is larger the higher pre-union trade.

### Empirical specification and estimation
- Estimation equation (logged and controlling for multilateral resistance via fixed effects) corresponds to the theoretical gravity equation; currency union dummy δ_ij enters the bilateral trade-barrier specification.
- Currency-union coefficient of interest: β2.
- Assumptions used in counterfactuals:
  - σ = 5 (elasticity of substitution).
  - Border effect from Anderson and van Wincoop (2003) is incorporated in total trade costs.
- Data:
  - Bilateral trade among 137 countries for years 1980 and 1990.
  - Bilateral trade measured as sum of nominal exports and imports deflated by the US price index and divided by two.
  - Controls include log distance, common language, adjacency, political union, colony/colonizer, common colonizer, FTA; country/time fixed effects used to control for multilateral resistance.

### Estimation results (key parameters and robustness)
- Main estimate (Table 1, column (4) used for counterfactuals):
  - Currency Union coefficient β2 = 0.91 (standard error 0.18) — statistically significant.
  - Interpretation note: β2 = 0.91 is often interpreted as an increase in trade of almost 250 percent (e^{0.91} ≈ 2.48) in partial-equilibrium, but general equilibrium solutions are required to obtain correct trade and welfare effects.
- Robustness checks (Table 2: only currency-union estimates shown; standard errors in parenthesis):
  - 1980 only: 1.075 (0.22)***
  - 1990 only: 0.735 (0.27)***
  - No North-North trade: 0.741 (0.17)***
  - Only developing countries: 0.565 (0.21)** 
  - Similar-size countries (by less than factor of 5): 0.928 (0.19)***
  - No Australia, France, New Zealand, United Kingdom, and United States: 0.909 (0.20)***
- Note: Several alternative estimates are smaller than 0.91; the smallest reported in this sensitivity analysis is 0.57 (interpreted as a 176 percent increase in bilateral trade, e^{0.57} ≈ 1.76). External studies report ranges like (.59, .80).

### Trade-creating and welfare effects for East Asian groupings (general-equilibrium counterfactuals)
- Country groupings used:
  - NIES3: Hong Kong, Singapore, and Korea
  - ASEAN4: Thailand, Malaysia, Indonesia, and Philippines
  - ASEAN5: ASEAN4 plus Singapore
  - East Asia (EA): NIES3 and ASEAN4
- Counterfactual average percentage changes (Table 3; percentage increases; standard errors in parenthesis reported in the source):
  - EA + China — Trade: 114 percent; Welfare: 3.7 percent
  - EA (without China) — Trade: 125 percent; Welfare: 2.5 percent
  - NIES3 — Trade: 1500.1 percent; Welfare: 0.1 percent
  - ASEAN4 — Trade: 1183.2 percent; Welfare: 1.0 percent
  - ASEAN5 — Trade: 1203.0 percent; Welfare: 0.9 percent
  - China + Hong Kong — Trade: 1410.9 percent; Welfare: 0.3 percent
- Summary interpretation from the paper:
  - A common currency among all East Asian countries raises bilateral trade substantially (e.g., 114 percent) but welfare effects are moderate (e.g., 3.7 percent when China included).
  - Trade-creating effects in East Asia appear larger than EMU’s estimated 58 percent trade increase, but welfare gains in East Asia are smaller than the 11 percent average welfare increase reported for EMU members in Rose and van Wincoop (2001).
  - Results are upper-bound if β2 is overestimated; actual effects could be smaller.

### Extended membership: adding Australia, New Zealand, and Japan
- Common-currency scenarios including Australia and New Zealand (Table 4; percentage increases; standard errors in parenthesis reported in the source):
  - EA + China + Australia + New Zealand — Trade: 102 percent; Welfare: 5.0 percent
  - EA + Australia + New Zealand — Trade: 110 percent; Welfare: 4.1 percent
  - NIES3 + Australia + New Zealand — Trade: 129 percent; Welfare: 2.1 percent
  - ASEAN4 + Australia + New Zealand — Trade: 100 percent; Welfare: 5.2 percent
  - ASEAN5 + Australia + New Zealand — Trade: 104 percent; Welfare: 4.8 percent
  - EA + China + Japan + Australia + New Zealand — Trade: 93 percent; Welfare: 6.2 percent
- Key findings:
  - Including Australia and New Zealand keeps trade-creating effects large (100–129 percent) and raises welfare effects materially (e.g., ASEAN4 + Australia + New Zealand welfare +5.2 percent).
  - Adding Japan to the extended union raises welfare further (e.g., EA + China + Japan + Australia + New Zealand welfare +6.2 percent in the reported scenario).

### Yen zone vs dollar zone — unilateral vs regional arrangements (trade and welfare comparisons)
- Unilateral (bilateral) currency-union scenarios considered: each East Asian economy forming a currency union with Japan, or with the United States (treated as exogenous).
- Table 5 — Average percentage increases by country (Trade: CU with Japan; CU with US — Welfare: CU with Japan; CU with US). Standard errors in parenthesis in the source.
  - Hong Kong — Trade: 149; 147 — Welfare: 0.2; 0.4  (standard errors: (43)(42)(0.05)(0.12))
  - Singapore — Trade: 150; 147 — Welfare: 0.1; 0.3  (standard errors: (43)(42)(0.03)(0.11))
  - Korea — Trade: 110; 148 — Welfare: 4.1; 0.3  (standard errors: (27)(42)(1.21)(0.08))
  - Indonesia — Trade: 142; 143 — Welfare: 0.8; 0.7  (standard errors: (40)(40)(0.24)(0.20))
  - Malaysia — Trade: 143; 128 — Welfare: 0.7; 2.1  (standard errors: (40)(34)(0.22)(0.68))
  - Thailand — Trade: 143; 144 — Welfare: 0.7; 0.6  (standard errors: (40)(41)(0.21)(0.20))
  - Philippines — Trade: 137; 134 — Welfare: 1.3; 1.6  (standard errors: (37)(36)(0.40)(0.49))
  - China — Trade: 133; 140 — Welfare: 1.7; 1.0  (standard errors: (36)(39)(0.50)(0.29))
  - EA (East Asia aggregate) — Trade: 111; 115 — Welfare: 4.0; 3.6  (standard errors: (27)(29)(1.15)(1.03))
  - EA + China — Trade: 101; 105 — Welfare: 5.2; 4.7  (standard errors: (24)(25)(1.45)(1.32))
- Interpretation and comparative conclusions:
  - Trade-creating effects are large and similar for unilateral and regional arrangements (around 100–150 percent for most cases).
  - Welfare effects differ sharply: regional currency unions generate much larger welfare gains than unilateral bilateral unions for most countries (exception: Korea, where a bilateral union with Japan yields welfare +4.1 percent comparable to regional gains).
  - Regional currency union with Japan yields larger welfare increases for East Asian members than a regional union with the United States. Example aggregates:
    - Regional union with Japan: welfare increase around 4 percent for EA; with China included, welfare rises by 5.2 percent.
  - The paper suggests Japanese membership in a regional currency arrangement is more important for generating larger welfare gains for East Asia than US membership.
  - Caveat: East Asia was de facto dollar-pegged in the sample period; this may bias welfare effects estimated for dollar-union scenarios upward because higher pre-union trade increases estimated welfare gains.

### Synthesis of substantive findings and policy-relevant takeaways
- Trade effects:
  - A currency union involving only East Asian countries will, in the paper’s general-equilibrium estimates, roughly double bilateral trade among members (examples: EA + China trade +114 percent; EA trade +125 percent).
  - Trade-creating effects of East Asian currency unions are generally larger than estimated EMU effects (EMU estimated +58 percent in Rose and van Wincoop (2001) as cited).
- Welfare effects:
  - Welfare gains from East Asian-only currency unions are moderate (examples: ASEAN5 +3.0 percent; EA + China +3.7 percent).
  - Including Australia, New Zealand, and especially Japan raises welfare gains materially (examples: ASEAN4 + Australia + New Zealand welfare +5.2 percent; EA + China + Japan + Australia + New Zealand welfare +6.2 percent).
  - Regional cooperation (multilateral currency union) yields substantially larger welfare gains than unilateral bilateral currency unions for most East Asian economies.
  - If East Asia forms a regional currency union, Japanese membership contributes more to welfare gains than US membership in the modeled scenarios.
- Policy implication emphasized by the paper:
  - Certain regional currency arrangements are beneficial for East Asia; Japanese participation appears especially important to generate larger welfare gains.
  - As regional integration and intra-regional trade increase over time, potential gains from a currency union could rise.

*Source: IMF Working Paper — 2. How does a currency union in East Asia affect the welfare of member countries?*

### 3.7 percent on average (Table 3), adding Japan to the union increases welfare by 5.2 percent

### 3.7 percent on average (Table 3), adding Japan to the union increases welfare by 5.2 percent

### Key quantitative findings
- Common currency in East Asia increases bilateral trade flows in the region substantially.
- Welfare increase from forming a currency union:
  - "3.7 percent on average (Table 3)"
  - "adding Japan to the union increases welfare by 5.2 percent (Table 5)"
  - "adding Australia and New Zealand to the union increases welfare by 5 percent (Table 4)"
- The marginal contribution by Japan to the welfare increase is larger than that by Australia and New Zealand combined.

### Caveats and limitations
- The analysis focuses only on the trade aspect of currency unions and benefits associated with changes in trade costs.
- Other potential benefits of a currency union are not taken into account.
- Costs of forming a currency union are not analyzed.
- The paper does not address which currency (a regional common currency, Japanese yen, or U.S. dollar) should be used for different currency arrangements in East Asia.
- More comprehensive analysis is necessary to incorporate these omitted issues and to fully assess the possibility of a common currency in East Asia.

### Conclusions and policy implications
- A region-wide currency union that includes Japan will generate more substantial welfare gains than sub-regional arrangements.
- Welfare gains are rather moderate for sub-regional currency arrangements that involve only a subset of East Asian economies.
- A regional currency union will stimulate regional trade rigorously, and increased regional trade will facilitate further regional integration in East Asia.
- Consistent with Frankel and Rose (1998): "the more integrated East Asian economies become, the higher the benefits of a common currency may be."
- Recent movements toward regional cooperation and integration are seen as a right direction if East Asian economies are to form a currency union in the future.

### Recommended further work
- Quantify benefits beyond reductions in trade costs.
- Estimate the costs of forming a currency union to fully evaluate feasibility.
- Analyze currency choice issues for different currency arrangements in East Asia.

*Source: IMF working paper excerpt.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07166.pdf_