Collateral Damage: Exchange Restrictions and Trade Flows

1
Collateral Damage Exchange Restrictions and
Trade Flows

• Shang-Jin Wei
• Zhiwei Zhang

2
Introduction

• Are 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 Chile
• This paper Effects on international trade

3

• Can exchange restrictions damage trade?
• Examples
• Repatriation requirements for exporters.
• Anecdote with the chief of a foreign exchange
  control administration
• Empirical research is scarce
• Tamirisa (1999)
• Data limitation
• Specification issue

4

• Notable features of this paper
• Three unique databases
• AREAER 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 advances
• Anderson Van Wincoop (2003)
• Helpman Melitz Rubinstein (2006)

5
Exchange Restrictions

• 192 indicators in AREAER
• Three groups of restrictions on
• Trade payments.
• Capital transactions.
• FX transactions (exchange taxes and subsidies )
  and others.

6
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.

7

• Controls on Trade Payments or Proceeds
• Imports and Import Payments
• 1. Foreign exchange budget
• 2. Financing requirements for imports
• 3. Documentation requirements for release of
  foreign exchange for imports
• Exports and export proceeds
• 1. Repatriation requirement
• 2. Financing requirements
• 3. Documentation requirements
• 4. Export licenses
• 5. Export taxes

8

• Restrictions on Following Capital Transactions
• 1. Capital and money market instruments
• 2. Derivatives and other instruments
• 3. Credit operations
• 4. Direct investment
• 5. Liquidation of direct investment
• 6. Real estate transactions
• 7. Personal capital transactions
• 8. Provisions specific to commercial banks and
  other credit institutions
• 9. Provisions specific to institutional investors
• 10. Other controls imposed by securities laws

9

• Restrictions on FX Transactions and others
• 1. Exchange tax
• 2. Exchange subsidy
• 3. Absence of forward exchange market
• 4. Currency requirements for pricing settlement
• 5. Payments arrears
• 6. Controls on trade in gold (coins and/or
  bullions)
• 7. Controls on exports and imports of banknotes
• 8. Controls on transfers
• 9. Proceeds from invisible transactions
• 10. Resident Accounts
• 11. Nonresident Accounts

10
Almost all exchange restrictions can be used as
capital controls.

• Malaysia. (Johnson, Kochhar, Mitton Tamirisa
  2006)
• Currency requirements for settlement.
• Export proceeds.
• China
• Residents holding FX bank accounts.

11
Restriction Indices

• For 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.

12
(No Transcript)
13
(No Transcript)
14
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 language
• RI Vector of restriction indices.
• Tariff Bilateral tariff rates.
• NTB Non-trade barrier index from TRI.
• IMP, EXP, Year fixed effects for importer,
  exporter, year.

15
Restrictions on trade payments -0.567 -0.537
(0.153) (0.159)
Restrictions on capital -0.068 0.032
transactions (0.107) (0.110)
Restrictions on FX transactions -0.314 -0.152
and others (0.171) (0.180)
Tariff -0.713 -0.724 -0.725 -0.717
(0.182) (0.182) (0.182) (0.182)
Non-tariff barrier index -0.213 -0.226 -0.221 -0.212
(0.049) (0.048) (0.048) (0.048)
16
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
  with0.18(-0.537)/(-0.717)100 13.9
  percentage points increase in tariff.

17
Alternative Specifications

• Country pair fixed effects
• to incorporate trade costs more generally.
• Separate import price indices for two trading
  partners
• to proxy for time varying trade costs.

18
Model with country-pair fixed effects
Restrictions on trade payments -0.434 -0.367
(0.103) (0.109)
Restrictions on capital -0.018 0.087
transactions (0.069) (0.072)
Restrictions on FX transactions -0.438 -0.353
and others (0.116) (0.124)
Tariff -0.774 -0.784 -0.801 -0.784
(0.107) (0.107) (0.107) (0.107)
Non-tariff barrier index -0.181 -0.189 -0.186 -0.178
(0.034) (0.034) (0.034) (0.034)
19
Model with time-varying import price indices
Restrictions on trade payments -0.633 -0.494
(0.154) (0.161)
Restrictions on capital -0.187 -0.033
transactions (0.102) (0.108)
Restrictions on FX transactions -0.608 -0.438
and others (0.162) (0.174)
Tariff -0.708 -0.726 -0.728 -0.721
(0.181) (0.182) (0.182) (0.182)
Non-tariff barrier index -0.177 -0.197 -0.178 -0.17
(0.047) (0.047) (0.047) (0.047)
20
(No Transcript)
21
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.

22
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?

23
Dependent variable change in bilateral trade,
1996 to 1999
24
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.

25
Further Research Needed

• Non-linear effects.
• Interactive effects (e.g. with governance)
• Suggestions?