Paper

Currency Wars
By Joseph E. Gagnon, Peterson Institute for International Economics

Originally published in the Milken Institute Review. Reposted with permission.

January 2013

© Milken Institute Review

The rules of global trade forbid countries from artificially boosting exports and curbing imports by
manipulating the exchange rates of their currencies. But for many reasons, policymakers have been
wary (more wary than presidential candidates, anyway) of pressing cases against abuses. That
reluctance may be coming to an end, however, as the global recession slouches on and the shadow of
chronic unemployment looms over industrialized economies.

What explains that traditional reluctance to pursue currency abuses? For one thing, more is typically at
stake in bilateral relations than commerce. For another, some groups in the “losing” countries benefit
from currency manipulation—Country A’s exporters’ unfair advantage translates into lower prices for
consumers and higher profits for retailers in Country B.

Besides, economists argue, currency manipulation may determine which industries flourish, but it
should not have much effect on the total number of unemployed in the long run. Last but not least,
policymakers sympathize with some motives for currency manipulation—in particular, for building nest
eggs of foreign currency to protect against economic (and military) shocks and for providing income for
future generations when nonrenewable resources are depleted.

However, the politics and economics of what is appropriately dubbed “currency aggression” have been
changing. Jobs do not seem to come back as quickly in the wake of economic downturns—and when
they do come back, they leave a trail of economic and social dislocation. As the biggest nonaggressor,
the United States is plainly suffering. I estimate that in 2011, currency aggression cost Americans more
than two million jobs.

The Mechanics

I know what you’re thinking, but you’re wrong. This isn’t just a problem with China. The policy took off
in a big way early in the past decade, with roughly two dozen countries getting in on the act. They
included advanced economies (Japan and Switzerland); newly industrialized economies (Israel,
Singapore, and Taiwan); developing Asian countries (Malaysia, Thailand and, of course, China); and oil
exporters (Norway, Russia, and Saudi Arabia).

To push down the exchange value of its currency, a government purchases foreign currency with its own
(which any country can do because governments create their own money). The government must hold
the foreign currency in the form of a financial asset, typically a bond or a bank deposit. Most, but not all,
of these assets are recorded in official statistics as foreign-exchange reserves. A government purchase of
foreign assets is called an “official financial outflow,” while a sale is an official financial inflow. The

1
magnitude of currency manipulation can thus be quantified as the difference between official outflows
and official inflows, or net official financial flows.

Sometimes, governments amass foreign currency to achieve quite different objectives. For example, a
government may borrow on international markets to finance a development project or to finance a
budget deficit.
But regardless of the government’s intent, cross-border official flows also have effects on the exchange
rate. (Of course, other factors, like the trade balance and expectations about the exchange rate in the
future, also affect those rates.)

The impact of official flows on the exchange rate depends on the ability and willingness of private
investors to adjust the mix of currencies in their own their portfolios. In order for a government to sell
domestic currency for foreign currency, it must induce private investors to sell the foreign currency. As
in any market, an increase in supply pushes down the price—here the exchange rate. If financial markets
are deep and liquid (as is the case with US dollars, euros, British pounds, and Japanese yen), the
resulting depreciation will be small relative to the size of the transaction. However, if legal barriers
prevent some investors from buying the domestic currency—for example, foreigners need permission to
buy Chinese renminbi—depreciation will be greater. But even when markets are highly liquid, the
impact of large net official flows on the exchange rate and trade balance can be quite large.

The Fruits of Aggression

So why do governments manipulate currencies? To improve their balance of trade. When a currency
falls in exchange value, exports become cheaper to foreigners (who then buy more), while imports
become more expensive to a nation’s own residents (who then buy less). The net result is an
improvement in the balance of trade. The higher trade balance supports more employment and adds to
economic growth through expansion of industries that export as well as industries that face less
competition from imports. This expansion is matched by economic contraction in its trade partners, who
now export less and import more. In other words, currency aggression is roughly a zero-sum game,
shifting production (and jobs) from one country to another without changing total employment or
output, at least in the short run. With time, the effects on employment in each country usually wear off,
as monetary policy in the aggressor countries must tighten to prevent inflation and monetary policy in
the nonaggressor countries must loosen to encourage employment in other industries. But these
adjustments are moving very slowly right now.

There is another, very important, aspect to this game. A country with a trade surplus is lending to the
rest of the world—a reality reflected in increased ownership (by the government or private investors) of
foreign financial assets (bonds, bank loans, and stocks). And, of course, creditors’ increased foreign
assets in surplus economies are matched by added liabilities in economies with trade deficits.

In some cases, temporary trade surpluses (matched, of course, by deficits elsewhere) may serve the
interests of all parties, creating demand in economies operating at less than full capacity and bleeding
off excess demand in overheated ones. But in order to achieve global economic balance in the long run,
the pattern of surpluses and deficits should follow from differences in the returns on investments.
Industrialized countries (where returns on new productive facilities are relatively low) should thus be
lending to developing countries (where expected returns are high). Accordingly, industrialized countries
should run trade surpluses, while developing ones should run deficits.

2
But over the past decade or so, this pattern has been turned on its head by currency aggression on the
part of poor countries fueling growth through exports. The developing world has been lending to the
industrialized world to sustain more competitive exchange rates (and export-led growth), thereby
running a huge collective surplus. Figure 1 shows the current account balance (a broader measure of the
trade balance that includes income on foreign investments) for developing countries as a group. It also
shows the magnitude of currency manipulation in these countries, as measured by their net official
financial flows. In both cases, they are shown as a percentage of GDP.

Figure 1 Currency manipulation and current account

in developing countries (percent of GDP)
10

Net official flows

8

-2
Current account

-4
1981 1986 1991 1996 2001 2006 2011

Source: IMF World Economic Outlook database.

Note that after two decades of a small but generally negative trade balance—about what one would
expect if exchange rates were largely driven by market forces—developing countries began to run large
current-account surpluses after the year 2000. This change coincided with a marked upturn in currency
manipulation.

Developing countries turned to currency manipulation as a means of resuming growth after the 1990s
financial crises in Latin America and Asia. They were also motivated by the wish to build larger stockpiles
of foreign exchange reserves to protect themselves from the rising volatility of global financial markets.
In practice, however, these reserves have not been very helpful in managing risk.

In many cases, currency aggression was not associated with a depreciation of the domestic currency but
the prevention of appreciation. Normally, the currency of a country with rapidly growing productivity

3
would appreciate, reflecting falling production costs in export markets; this enables the country to enjoy
more imported goods at lower prices. China is the most dramatic example of a country that has resisted
normal currency appreciation.

Currency Aggression Is Big

In my own research I found that, over the past 30 years, currency manipulation has been the main driver
of changes in trade surpluses and deficits. Fiscal policy (budget deficits) is a distant second. My analysis
suggests that the current-account balance increases by about 65 cents for every dollar spent on
currency manipulation; by no coincidence, between 1982 and 2011, the current accounts of developing
countries increased about two-thirds as much as the increase in net official flows.

Figure 2 illustrates that relationship: Countries that have larger net official outflows have higher current-
account balances. In 2011, net official flows of all countries that were net purchasers of official assets
totaled about $1.1 trillion. Adding in purchases by sovereign wealth funds (mostly by oil exporters)
brings the total to $1.4 trillion.

Figure 2. Current Accounts and Currency Manipulation

percent of GDP, 2002-2011 average
20

SGP
SAU
15

MAS NOR
SWI
Current Account

VEN
10

HKG
TWN
SWE RUS
CHN
5

DNK
JPN
ARG THA
KOR
EUR IDN
CHL
CAN
0

BRA
MEX IND
GBR COL
SAF
USA AUS POL
TUR
-5

-5 0 5 10 15
Net Official Flows

Note: Countries are Argentina (ARG), Australia (AUS), Brazil (BRA), Canada (CAN), Chile (CHL), China (CHN),
Colombia (COL), Denmark (DNK), euro area (EUR), Hong Kong (HKG), India (IND), Indonesia (IDN), Japan (JPN),
Korea (KOR), Malaysia (MAS), Mexico (MEX), Norway (NOR), Poland (POL), Russia (RUS), Saudi Arabia (SAU),
Singapore (SGP), South Africa (SAF), Sweden (SWE), Switzerland (SWI), Taiwan (TWN), Thailand (THA), Turkey
(TUR), United Kingdom (GBR), United States (USA), and Venezuela (VEN).

Sources: IMF International Financial Statistics; IMF Currency Composition of Official Foreign Exchange Reserves;
Finance Ministries of Norway and Singapore; and Central Bank of the Republic of China (Taiwan).

4
Global currency manipulation has been running at roughly this rate since 2006. According to the latest
data and International Monetary Fund (IMF) projections, it declined by about 20 percent in 2012, but
remains near historically high levels.

About $100 billion of these net official flows reflected repayments of loans, which I do not consider to
be aggressive behavior. Also, it is widely accepted that developing countries, particularly those
vulnerable to natural disasters or that face military threats, need to hold foreign assets as a form of
insurance. And it is only prudent for countries that are dependent on exports of non-renewable
resources, like oil, to invest some of their export revenue abroad, rather than consuming or investing it
all at home.

I made the educated guess that justifiable accumulation of reserves and other foreign assets in 2011
was roughly $300 billion, most of which reflects savings of oil exporters beyond what could be prudently
invested in their own economies. Altogether then, excessive global currency manipulation (currency
aggression) was reflected in roughly $1 trillion in net flows in 2011.

Accordingly, the trade balances of currency aggressors in 2011 were very likely elevated by $650 billion
as a result of their purchases of official assets (and the trade balances of the remaining countries were
correspondingly depressed). We do not know with any precision how much the US trade balance was
affected by currency manipulation. However, since the American economy constitutes about one-third
of the global economy (excluding currency aggressors), a first guess is that the US current-account
balance in 2011 was some $200 billion lower than it would have otherwise been.

Actually, it might have been considerably more: The United States is more strongly affected by currency
manipulation than other countries, because about 60 percent of foreign- exchange reserves held by
governments worldwide are held in dollars. If the net effect of currency aggression in 2011 was to lower
the US current-account balance by $300 billion rather than $200 billion, the resulting reduction in
domestic demand for goods and services made in the United States would imply a loss of two to three
million jobs.

To be sure, in normal times one cannot draw a direct link between the trade balance and employment,
because monetary policy in the United States and most other industrialized economies can be (and is)
used to offset trade effects on employment. However, in 2011, the Federal Reserve was still struggling
with the consequences of the Great Recession, and had stretched its capacity to expand financial
liquidity to the political limits. Hence, it simply didn’t have the firepower to offset the jobs lost through
trade.

In any event, currency aggression has serious costs even when countries are at full employment. The
resulting unsustainable trade imbalances distort the pattern of economic activity in ways that reduce
long-run economic welfare. For example, during the peak years of currency aggression in the mid-2000s,
the Federal Reserve maintained full employment by keeping interest rates low. But low rates
encouraged vast overinvestment in the US housing market, from which we are still suffering the
consequences.

5
Identifying the Scofflaws

Table 1 lists countries that were currency Table 1 Foreign exchange reserves of currency
aggressors over the 10-year-period through 2011, aggressors, 2011
along with their accumulated holdings of foreign-
exchange reserves. To be included on my list, Country Billions of Percent of GDP
countries must meet all of these criteria: US dollars
Advanced countries
 They must have foreign-exchange Denmark 78 24
reserves greater than the value of six Hong Kong 286 118
months of goods and services imports. Israel 73 30
 They must have an average current Japan 1225 21
account balance from 2002 through 2011 Korea* 335 30
that is greater than zero. Norway* 547 113
 They must have increased their official Singapore** 560 216
foreign assets relative to their GDP over Switzerland 272 43
the past decade. Taiwan 386 83
Latin America
Low-income countries, defined by the World Bank Bolivia 10 39
as those with per capita incomes of less than Developing Asia
$1,025 in terms of purchasing power, are China* 3262 45
excluded on the principle that they should be Malaysia 129 46
allowed greater latitude than other countries in Philippines 66 31
their economic policies. Thailand 166 48
Sub-Saharan Africa
As a share of GDP, foreign-exchange reserves are Angola 28 28
largest among the industrialized economy and Middle East and North Africa
Middle Eastern aggressors. (The number for Libya Algeria 181 95
is distorted by the drop in GDP in 2011 associated Kuwait* 235 133
with the regime change. In 2010, Libya’s reserves Libya 97 264
equaled 119 percent of GDP.) Qatar* 101 58
Saudi Arabia 527 91
The most striking revelation is the large number U. A. E.* 779 216
of Asian countries—both industrialized and Eastern Europe and former Soviet Union
developing—on the list. Some economists have Azerbaijan* 33 53
argued that the Asian currency aggressors are Kazakhstan* 64 36
motivated by a desire to build large war chests in Russia 443 24
order to avoid a repetition of their experience *Includes 2010 estimated foreign assets of
during the Asian financial crisis of 1997–98, when sovereign wealth funds.
they lacked sufficient foreign exchange to counter **Gross financial assets of Government of
private outflows. That may have been appropriate Singapore (may include some domestic assets).
in the early part of the last decade, but these Sources: IMF International Financial Statistics;
countries long ago passed the level of reserves Edwin Truman, Sovereign Wealth Funds: Is Asia
that might be needed for this purpose. Different? (2011); Singapore Ministry of Finance;
and Central Bank of the Republic of China
Moreover, Asia’s relatively good economic (Taiwan).
performance during and after the recent global
financial crisis of 2008–09 is mainly the result of a reduction in foreign-currency borrowing and an

6
improvement in macroeconomic- policy frameworks. It is not clear that the reserve buildup helped
much; central banks in Asia spent very little of their reserves in 2008 or 2009. A more plausible
explanation for why Asian governments continue to add to their stocks of foreign reserves is that
exporters are politically powerful and that maintaining a current-account surplus is viewed as a time-
tested way to maintain steady growth.

The table does not account for net foreign saving by oil-exporter nations and other countries facing
serious security threats, like Israel and Taiwan. It is possible that these considerations would also lead
one to exclude Angola, Kazakhstan, and Russia from the list. But the other countries affected by security
issues or depleting oil reserves all have foreign assets in excess of 50 percent of GDP—and in many
cases, in excess of 100 percent. That is more than can be justified by these concerns.

Fighting Currency Aggression

Under the IMF Articles of Agreement, Fund members agree to “avoid manipulating exchange rates or
the international monetary system in order to prevent effective balance of payments adjustment or to
gain an unfair competitive advantage over other members.” But since the adoption of the language on
currency manipulation in 1978, the IMF has never publicly declared a member to be in breach. In any
event, the IMF has no sanctions with which to enforce this prohibition, short of expelling the offending
country.

Unlike the IMF, the World Trade Organization (WTO) does provide a framework for countries to lodge
complaints and to impose sanctions on others that violate its rules. Article XV of the agreements that
underlie the
WTO states that “contracting parties shall not, by exchange action, frustrate the intent of the provisions
of this Agreement.”

However, this provision was likely intended to bar countries from imposing exchange controls to offset
the trade impact of negotiated cuts in tariffs. It has not been tested as a tool against currency
aggression. Moreover, the WTO is required to defer to the IMF on matters related to currencies and
foreign-exchange reserves, and—as discussed earlier—the IMF has been unwilling to designate
countries as aggressors. The best possible approach would be for the WTO to work with the IMF to
restrict currency aggression. Countries that believe they are harmed by such aggression should be able
to file formal complaints with the WTO. It would ask the IMF to rule on the existence and magnitude of
currency manipulation. If the IMF agreed that the target country was manipulating its currency to gain
an unfair competitive advantage, the WTO would authorize the complainants to impose countervailing
tariffs against imports from the offending country in proportion to the implied currency undervaluation.

The question then arises as to how to get members of the IMF and the WTO to agree to this new
procedure. Changes in the WTO agreements require unanimity, while changes in the IMF articles require
85 percent of voting shares. And since currency aggressors would have no incentive to cooperate,
victims of currency aggression would need to find carrots and sticks to bring them to the bargaining
table.

There are some proverbial carrots: The United States and the European Union could offer to increase
the representation and voting shares of developing countries in the IMF and to increase the size of
emergency credit lines provided by the IMF as an alternative to amassing larger foreign-exchange
reserves. One stick: Victims could fight fire with fire—that is, buy up quantities of the aggressor’s

7
currency to prevent depreciation. But this would expose their own taxpayers to the risks of holding
foreign assets. Besides, it would not work against countries that have small or closed financial markets.

Alternatively, victims could unilaterally prohibit purchases of their financial assets by currency
aggressors—or, perhaps, tax them. Such selective regulation is permissible under international law, and
the threat of imposition would give currency aggressors a powerful incentive to bargain. A
comprehensive deal might create a procedure within the IMF and WTO to impose sanctions against
aggressors while limiting the ability of all countries to impose discriminatory taxes or controls on their
own financial assets.

Of course, any approach to containing currency aggression needs to be implemented in the broader
context of managing political and economic relationships. Far more is at stake in these relationships—
everything from human rights to nuclear proliferation to climate change—than jobs or profits. This
points to the desirability, where possible, of discouraging currency aggression by nonconfrontational
means.

First, it is generally expedient to use multilateral forums in which limits on currency manipulation can be
negotiated without risking the breakdown of bilateral agreements on unrelated issues. Second, every
effort should be made to put the issue in the context of general economic welfare rather than specific
exporters’ welfare. Thus, it may not be in the interest of Chinese exporters of machinery or solar panels
or automobiles to allow the appreciation of the renminbi, but it is very much in the interest of the
Chinese people to redirect productive capacity from exports to the myriad products and services—like
health care, higher education, and transportation infrastructure—that are in short supply at home.

Sooner or later (preferably sooner) the global financial imbalances created by currency aggression must
be corrected. The trick will be to get from here to there with minimal disruption of the global order.