Collateral Damage: Exchange Restrictions and Trade Flows
Shang-Jin WeiZhiwei Zhang
IntroductionAre exchange restrictions (capital controls) good for developing countries?
Adverse effects of liberalization Rodrik, Stiglitz, survey by Kose, Prasad, Rogoff, and Wei
Costs of not liberalizing Forbes study on cost of borrowing in ChileThis paper: Effects on international trade
Can exchange restrictions damage trade?
ExamplesRepatriation requirements for exporters.Anecdote with the chief of a foreign exchange control administration
Empirical research is scarceTamirisa (1999)Data limitationSpecification issue
Notable features of this paper:
Three unique databasesAREAER database for exchange controls.Country-pair specific tariff data from WITS allowing for calculation of tariff equivalent.Non-tariff barriers index from IMF/PDR/TRI.Theory-consistent gravity model that incorporates recent theoretical advancesAnderson Van Wincoop (2003)Helpman Melitz Rubinstein (2006)
Exchange Restrictions192 indicators in AREAER
Three groups of restrictions on:Trade payments.Capital transactions.FX transactions (exchange taxes and subsidies ) and others.
Main findings:Exchange restrictions have large negative effect on trade
Increasing restrictions on trade payments by 1 S.D. is equivalent to increasing tariff rate by 9 to14 percentage points.
Increasing restrictions on FX transactions and others by 1 S.D. is equivalent to raising tariff rate by 11 to 15 percentage points.
Controls on Trade Payments or Proceeds
Imports and Import Payments:1. Foreign exchange budget2. Financing requirements for imports3. Documentation requirements for release of foreign exchange for imports
Exports and export proceeds:1. Repatriation requirement2. Financing requirements3. Documentation requirements4. Export licenses5. Export taxes
Restrictions on Following Capital Transactions:
1. Capital and money market instruments2. Derivatives and other instruments3. Credit operations4. Direct investment5. Liquidation of direct investment6. Real estate transactions7. Personal capital transactions8. Provisions specific to commercial banks and other credit institutions9. Provisions specific to institutional investors10. Other controls imposed by securities laws
Restrictions on FX Transactions and others 1. Exchange tax2. Exchange subsidy3. Absence of forward exchange market4. Currency requirements for pricing & settlement5. Payments arrears6. Controls on trade in gold (coins and/or bullions)7. Controls on exports and imports of banknotes8. Controls on transfers 9. Proceeds from invisible transactions10. Resident Accounts11. Nonresident Accounts
Almost all exchange restrictions can be used as capital controls.Malaysia. (Johnson, Kochhar, Mitton & Tamirisa 2006)Currency requirements for settlement.Export proceeds.
ChinaResidents holding FX bank accounts.
Restriction IndicesFor each group of the restrictions, an index is constructed as the weighted sum of all restrictions in place.
Weights are chosen to ensure each category within the group receives equal weight.
Augmented gravity model:Y: Bilateral trade flows.X: Importers and exporters GDP, distance, Mills ratio and predicted probability of trade (Helpman-Melitz-Rubinstein), colonial ties, common languageRI: Vector of restriction indices.Tariff: Bilateral tariff rates.NTB: Non-trade barrier index from TRI.IMP, EXP, Year: fixed effects for importer, exporter, year.
Tariff Equivalent Calculations
A 10 percent increase in tariff rate is associated with a 7 percent reduction in trade volume
Increasing restrictions on trade payments by 1 standard deviation (0.18) is associated with: 0.18*(-0.537)/(-0.717)*100 = 13.9 percentage points increase in tariff.
Alternative SpecificationsCountry pair fixed effects to incorporate trade costs more generally.
Separate import price indices for two trading partners to proxy for time varying trade costs.
Model with country-pair fixed effects:
Model with time-varying import price indices
Refined categories of restrictions:Both imports and exports restrictions seem to matter.Capital transactions show different signs.Negative estimates for restrictions on cap & money market instruments, derivatives.Positive estimates for FDI, personal capital transactions, and provisions to institutional investors.Restrictions on FX transactions and others are predominantly trade-reducing.Currency requirements, exchange taxes and subsidies, and arrears.
Case Study:The financial crises in emerging markets during 1996 1999 period led governments to set more controls.
Sample includes11 emerging markets in the MSCI index that strengthened their controls on either FX transactions or capital transactions during 1996 1999 period.
Did increase in exchange controls led to less trade, after controlling for changes in tariff, NTB index, GDP, and exchange rates?
Dependent variable: change in bilateral trade, 1996 to 1999
Tentative Conclusion:Some exchange restrictions have large adverse effects on trade
Increasing restrictions on trade payments by 1 S.D. is equivalent to raising tariff by 9 to 14 percentage points.
Increasing restrictions on FX transactions and others by 1 S.D. is equivalent to raising tariff by 11 to 15 percentage points.
Further Research Needed:Non-linear effects.
Interactive effects (e.g. with governance)