Is the U.S. Current Account Deficit Sustainable?

Prepared for Members and Committees of Congress

America’s current account (CA) deficit (the trade deficit plus net income payments and net
unilateral transfers) rose as a share of gross domestic product (GDP) from 1991 to a record high
of 6.1% of GDP in 2006. In 2007, it fell to 5.3% of GDP, and is projected to continue to fall in
2008. The CA deficit is financed by foreign capital inflows. Many observers have questioned
whether such large inflows are sustainable and expressed concern about the economic impact
should foreign capital inflows decline rapidly. Further, a large share of the capital inflows have
come from foreign central banks in recent years, and some are concerned about the economic and
political implications of this reliance. Some fear that a rapid decline in capital inflows would
trigger a sharp drop in the value of the dollar and an increase in interest rates that could lower
asset values and disrupt economic activity. However, economic theory and empirical evidence
suggest that the most plausible scenario is a slow decline in the CA deficit, which would not
greatly disrupt economic activity because production in the traded goods sector would be
While comprehensive data will not be available for some time, it appears that gross capital
inflows have recently fallen as a result of financial market turmoil. While financial market
turmoil has led to an economic slowdown, it has not been via a sudden decline in the dollar or a
sudden broad spike in U.S. interest rates. On the contrary, the dollar has appreciated in value and
demand for U.S. Treasury bonds has risen since the turmoil worsened in September 2008.
One long-term consequence of a large CA deficit has been the growing foreign ownership of U.S.
capital stock. A large CA deficit is not sustainable in the long run because it increases U.S. net
debt owed to foreigners, which cannot rise without limit. A larger debt can be serviced only
through more borrowing or higher net exports. For net exports to rise, all else equal, the value of
the dollar must fall. This explains why many economists believe that both the dollar and the CA
deficit will fall at some point in the future. To date, debt service has not been burdensome.
Because U.S. holdings of foreign assets have earned a higher rate of return than U.S. debt owed
to foreigners, U.S. net investment income has remained positive, even though the United States is
a net debtor nation.
Since 1980, most episodes of a declining CA deficit in industrialized countries have been
associated with slow economic growth. Only two episodes were associated with a severe
disruption in economic activity. Because most of the episodes involved small countries, these
cases may differ in important ways from any corresponding episode in the United States.
Historically, a few other countries have had a higher net foreign debt-to-GDP ratio than the
United States has at present; however, if CA deficits continue at current levels, the U.S. net
foreign debt could eventually be the highest ever recorded.
This report also reviews studies on the CA deficit’s sustainability. Some of the studies suggest
that a large dollar depreciation could eventually be required to restore sustainability. After an
inflation-adjusted 21% depreciation of the dollar from 2002-2007, the CA deficit was still higher
in 2007 than in 2002.

Introduc tion ..................................................................................................................................... 1
Economic Impact of a Declining Current Account Deficit............................................................4
Historical Parallels..........................................................................................................................6
A Review of Five Estimates............................................................................................................8
Conclusion ..................................................................................................................................... 10
Figure 1. Current Account Deficit as a Share of GDP, 1987-2007..................................................1
Author Contact Information...........................................................................................................11

America’s current account (CA) deficit (the trade deficit plus net income payments and net
unilateral transfers) rose as a share of gross domestic product (GDP) from 1991 to a record high
of 6.1% of GDP in 2005 and 2006. In 2007, it fell to 5.3% of GDP, but remained well above the
historical average (see Figure 1). It is projected to continue to fall modestly in 2008. It is too
early to tell if the decline in the past two years is caused by temporary cyclical factors or is part of
a longer-term trend. The CA has been in deficit every year but one since 1982. By accounting
identity, the CA deficit is equal to net inflows of foreign capital to the United States and reflects
the imbalance between domestic saving and investment.
Figure 1. Current Account Deficit as a Share of GDP, 1987-2007
-3f G
-1% o
1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Source: Bureau of Economic Analysis.
Some observers have questioned whether the CA deficit is sustainable. It has not prevented the
economy from generally attaining full employment—the United States has run large CA deficits
for several years, yet unemployment has remained low in most of those years. The CA deficit has
both positive and negative effects on the economy. Production of exports and import-competing
goods is arguably lower than it would be in the absence of a CA deficit, but interest rates are also
lower than they would be in the absence of foreign capital inflows. As a result, interest-sensitive
spending on capital investment, residential investment, and consumer durables (e.g., automobiles 1
and appliances) is higher.
Those expressing concern about the CA deficit typically define unsustainability to mean that the
United States would have difficulty financing the CA deficit at some point in the near future, and
the resulting adjustment process would harm the U.S. economy. Basically, the CA deficit is
sustainable as long as foreigners are willing to continue buying American assets. It is not enough
for foreigners to reduce their demand for U.S. assets, since this would cause yields on U.S. assets
to automatically rise until the market once again cleared. But if the desirability of U.S. assets
were to change rapidly (due to a loss in confidence in the U.S. economy, for example), foreign

1 For an overview, see RL31032, The U.S. Trade Deficit: Causes, Consequences, and Cures, by Craig K. Elwell.

capital inflows and the value of the dollar could decline quickly; at a minimum, foreigners would
require significantly higher interest rates than they do at present for inflows to continue. For the
U.S. to literally be unable to continue financing its current account deficit, foreign demand would
have to fall so much that asset yields could not rise enough for foreigners to be willing to hold 2
U.S. assets again.
But what would make foreign investors change their minds about U.S. assets? Federal Reserve
Chairman Ben Bernanke has argued that foreigners will continue to increase their holdings of
U.S. assets in the near term because of a global savings glut that leaves them with few other 3
desirable investment alternatives. If both lender and borrower are rational, many economists
believe that the CA deficit can be mutually beneficial—it allows the lender to enjoy a higher rate
of return than could be enjoyed at home and allows the borrower to operate with a larger capital
stock than could be financed from domestic saving. As long as those investments yield a high
enough rate of return to service the debt, borrowing should not reduce future domestic income in
absolute terms.
Some economists, however, doubt this interpretation and are concerned that the large CA deficit4 5
is symptomatic of wider economic imbalance. They argue that a country cannot persistently rely
on foreign borrowing to finance its investment needs, so the United States must eventually raise
its low saving rate. They maintain that by financing a large budget deficit and housing boom
(until 2006), much of the foreign borrowing is being used in ways that do not expand the
economy’s productive capacity, and therefore such borrowing does not enhance our ability to
service foreign debt. Because foreign borrowing is not sustainable, they argue, Americans will
eventually be forced to significantly increase their saving (equivalently, to reduce their
consumption) and reduce their investment rates, and the U.S. economy will slow. As a long-term
solution, these economists prescribe policy measures to raise saving and reduce overall spending
as the appropriate response to an excessively large CA deficit. This could be done through
microeconomic policies, such as policies to encourage higher private saving, and macroeconomic
policies, such as a tightening of monetary and fiscal policy, although this response would risk
inducing the same recession that they fear the CA deficit may eventually cause (particularly if the
economy were weak already).
The steady financing of the CA deficit has depended heavily on official capital inflows—6
purchases of U.S. assets by foreign central banks—since 2002. Official capital inflows have
averaged more than $300 billion per year from 2002 to 2007. Had these central banks decided not
to increase their dollar-denominated foreign reserves, the CA deficit and the value of the dollar
might have fallen precipitously. Although the dollar did not experience a sizable overall decline
during this period, it did fall significantly against certain currencies without disruption to the U.S.

2 It is widely assumed that a rapid change in the current account would be caused by changes in financial markets, not
goods markets. Although theoretically a rapid decline in imports could also cause the CA deficit to shrink, little
empirical evidence exists that broad trade patterns change that quickly.
3 Ben Bernanke, “The Global Saving Glut and the U.S. Current Account Deficit,” the Sandridge Lecture, Virginia
Association of Economics, March 10, 2005, available on the Federal Reserve Board of Governors website. For an
analysis, see CRS Report RL33140, Is the U.S. Trade Deficit Caused by a Global Saving Glut?, by Marc Labonte.
4 For the purposes of this report, a CA deficit is considered large if it exceeds the trend growth rate of the economy.
5 See, for example, Edwin Truman, “Postponing Global Adjustment,” Institute of International Economics, Working
Paper no. 05-6, July 2005.
6 For more information, see CRS Report RS21951, Financing the U.S. Trade Deficit: Role of Foreign Governments, by
Marc Labonte and Gail E. Makinen.

economy. Going forward, some analysts fear that official capital inflows could prove unreliable or
provide foreign governments with leverage that would undermine U.S. policy aims. Because
official inflows are likely financed by considerations other than rate of return, it is difficult to
predict how they will evolve in the future. But given the importance of the United States as a
foreign export market, deliberately taking a step that could potentially precipitate a U.S.
economic crisis could be against a foreign country’s economic self-interest.
As a consequence of large CA deficits, a growing share of the U.S. capital stock is owned by
foreigners and a rising fraction of U.S. income will need to be diverted overseas in the form of
interest and dividends to foreigners. The process cannot continue indefinitely, or else foreigners
would eventually hold a larger share of American assets in their portfolios than they desire. But
though economists may feel confident that the CA deficit will decline in the long run, long run
predictions do not offer much help in predicting short-term trends. Foreigners may wish to
increase their holdings of American assets (in which case the CA deficit would persist) in the near
term. One common assumption is that the CA deficit would, at most, continue until the share of
American assets held in foreign portfolios equaled America’s share of world output; by this
measure, foreigners still hold too few American assets. For example, citizens of the Euro Area
hold an estimated 53% of their wealth in euro-denominated assets and 19% in U.S. dollar-
denominated assets, whereas the Japanese hold an estimated 63% of their wealth in yen-7
denominated assets and 4% in dollar-denominated assets. This is referred to as a “home bias” in
saving because all countries hold more of their own assets, and fewer foreign assets, than optimal
portfolio diversification would suggest. This bias is considered unlikely to disappear entirely, in
which case CA deficits will ease before this benchmark is met. In any case, the reason why home
bias would decline for foreigners but not Americans, as continuing CA deficits would imply,
remains unclear. On the other hand, if the U.S. economy grows faster than the rest of the world in
the future, then (small) CA deficits would be needed for foreigners to maintain U.S. asset 8
holdings equal to the U.S. share of world GDP.
One reason that U.S. imports cannot exceed exports indefinitely (and the dollar could eventually
fall) is that today’s CA deficits have a consequence for future trade balances. The accumulation of
net debt that Americans owe to foreigners will need to be serviced in the future, and debt service
will take the form of a capital outflow from the United States. To offset the outflow, the United
States must export more or borrow more. But, all else equal, foreigners will only be induced to 9
buy more exports if the dollar depreciates. Net investment income payments make up a small
fraction of the CA deficit today, but economist Edwin Truman estimates that if CA deficits
continued to equal 6% of GDP, net income payments would eventually reach 4.5% of GDP, 10
leaving a trade deficit of only 1.5% of GDP. In other words, a constant trade deficit would imply
a growing CA deficit because of growing net investment income payments.

7 Olivier Blanchard, Francesco Giavazzi, Filipa Sa,The U.S. Current Account and the Dollar,” Massachusetts Institute
of Technology, Working Paper no. 05-02, May 2005.
8 Engel and Rogers argue that the CA deficit can be fully explained by investors’ anticipation that the U.S. share of
industrialized countries’ output will rise in the future. Their model omits developing countries, however, which have
played a large role in world CA deficits and growth recently. Charles Engel and John Rogers, “The U.S. Current
Account Deficit and the Expected Share of World Output,” National Bureau of Economic Research, Working Paper no.
11921, January 2006.
9 Another possibility for how the CA deficit could improve would be if the U.S. terms of trade improved in the future.
There is no consensus on any policy tools that can directly improve the terms of trade.
10 Edwin Truman, “Postponing Global Adjustment,” Institute for International Economics, Working Paper no. 05-6,
July 2005.

In 2007, the United States had a net foreign debt of $2.4 trillion, but received net investment
income of $89 billion from the rest of the world. What is surprising about these data is that the
United States still consistently has positive net investment income despite its large net debt, and
shows no long-term downward trend as the foreign debt grows. That is because U.S. holdings of
foreign assets have earned a higher rate of return than U.S. debt owed to foreigners. Between

2002 and 2004, the United States earned an estimated rate of return of 9.6% on its foreign assets 11

and paid a rate of return of 0.9% on its foreign liabilities. These estimated rates of return have
prevented foreign borrowing from becoming burdensome and suggest that the U.S. net foreign
debt could become significantly larger before debt payments would become burdensome.
Kouparitsas estimates that if these rate-of-return differentials were to continue, the United States
could maintain a current account deficit of 0.9-1.3% of GDP indefinitely without any increase in 12
its net foreign debt. However, this favorable rate-of-return differential may not continue in the
future if foreigners believe that the dollar will depreciate and demand a higher rate of return on
U.S. borrowing as compensation. If U.S. interest rates rose, the debt could become burdensome

Although some reduction in the CA deficit is inevitable (although not necessarily in the near
future), it need not be sudden. It should be emphasized that economic theory suggests that a slow
decline in the CA deficit and dollar would not be troublesome for the overall economy. In fact, a
slow decline could have an expansionary effect on the economy, because the increase in net
exports could have a more stimulative effect on aggregate demand in the short run than the
decrease in investment and other interest-sensitive spending caused by lower capital inflows.
More realistically, the trade deficit would not decline exogenously, but in response to a slowing of
overall domestic demand. Therefore, a falling trade deficit may tend to coincide with slower
economic growth in practice, but that does not indicate that a falling trade deficit has caused
growth to slow.
Historical experience seems to support the idea that a slow decline in the trade deficit need not be
harmful to the economy—the dollar declined by about 40% in real terms and the CA deficit
declined continually in the late 1980s, from 2.8% of GDP in 1986 to nearly zero during the early
1990s. Yet economic growth was strong throughout the late 1980s. Of course, the adjustment
required to balance the current account today is about twice as large, so historical experience may
not be a good guide. Beginning in 2007, the CA deficit began a gradual and modest decline, after
a long and gradual decline in the dollar.
A potentially serious short-term problem would emerge if foreigners suddenly decided to reduce
the fraction of their saving that goes to the United States in the form of a capital inflow, or if they

11 Philip Lane and Gian Milesi-Feretti, “A Global Perspective on External Positions,” National Bureau of Economic
Research, Working Paper no. 11589, August 2005. This also suggests that the attractiveness of U.S. assets may be more
perceived than actual.
12 Michael Kouparitsas, “Is the U.S. Current Account Sustainable?”, Federal Reserve Bank of Chicago, Chicago Fed
Letter, no. 215, June 2005. Curcuru et al cast doubt on whether the favorable rate of return differential can be expected
to continue in the future. Stephanie Curcuru et al, “The Stability of Large External Imbalances, National Bureau of
Economic Research, working paper 13074, May 2007.

suddenly decided to repatriate part of their liquid financial assets. The initial effect could be a
sudden and large depreciation in the value of the dollar, as the supply of dollars on the foreign
exchange market increased, and a sudden and large increase in U.S. interest rates, as an important
funding source for investment and the budget deficit was withdrawn from the financial markets.
Most likely, the direct trade effects of these shifts in lending patterns by foreigners would not
cause a recession because the dollar depreciation would lead to a trade surplus (or smaller 13
deficit), which expands aggregate demand. (Empirical evidence suggests that the full effects of
a change in the exchange rate on traded goods takes time, so the dollar may have to “overshoot”
its eventual depreciation level in order to achieve a significant adjustment in trade flows in the
short run.) However, the indirect interest rate effects, which typically only partially offset the
direct effects, could cause a recession if the change is sudden. Large increases in interest rates
could cause problems for the U.S. economy, as these increases reduce the market value of debt
securities, cause prices on the stock market to fall, and jeopardize the solvency of various debtors
and creditors. Resources may not be able to shift quickly enough from interest-sensitive sectors to 14
export sectors to make this transition fluid. The Federal Reserve could mitigate the interest rate
spike by reducing short-term interest rates, although this reduction would influence long-term
rates only indirectly and could increase inflation.
Is a scenario where the dollar crashes a likely one? Economic theory typically assumes that
financial market participants act rationally. If the rationality assumption is a good one, then the
potential for a sudden decline is slim. After all, foreigners would be demanding high rates of
return to buy U.S. assets today if they could foresee that the foreign currency value of these assets
is likely to fall sharply in the near future. However, a sudden decline in capital inflows would be
unlikely to occur in a purely rational context, in which case theory would have little predictive
power. Given the traditional role the United States has played as an investment safe haven,
sudden capital outflows seem unlikely. Investment might be attracted to U.S. assets in a liquidity
crisis because the U.S. offers more liquid financial markets (e.g., U.S. Treasury bond markets)
than do most foreign counterparts.
U.S. financial markets experienced exceptional turmoil beginning in August 2007. Over the
following year, the dollar declined by almost 8% in inflation-adjusted terms—a decline that was
not, in itself, disruptive. But as the turmoil deepened and spread to the rest of the world, the value
of the dollar began rising. Interest rates on U.S. Treasuries fell close to zero, implying excessive
investor demand. Other interest rates also remained low, although access to credit was limited for
some. Although comprehensive data will not be available for some time, a “sudden stop” in
capital inflows does not appear to have been a feature of the downturn. Problems recently
experienced in U.S. financial markets have been widely viewed as “once in a lifetime” events. If
these events failed to cause a sudden flight from U.S. assets and an unwinding of the CA deficit,
it is hard to imagine what would.

13 A sharp decline in the value of the dollar would also reduce living standards, all else equal, because it would raise the
price of imports to households. This effect, which is referred to as a decline in the terms of trade, would not be recorded
directly in GDP, however.
14 For a more formal explanation, see J. Bradford DeLong, “Some Simple Analytics for a ‘Hard Landing’,” working
paper, University of California at Berkeley, April 2005. Interestingly, one study found that countries with larger
declines in the CA deficit experienced, on average, higher GDP growth. See Hilary Croke, Steven Kamin, and Sylvain
Leduc, “Financial Market Developments and Economic Activity During Current Account Adjustments, International
Finance Discussion Papers, Federal Reserve, no. 827, February 2005.

Most comparisons to historical experience abroad are limited by the fact that the United States
economy is so much larger than those of other countries. As a result, an economic occurrence in
the United States has ramifications for the world economy that could have feedback effects for
the U.S. economy, whereas changes in the CA balance of most small countries will most likely
not affect the world economy. It also means that the amount of borrowing required to finance the
U.S. CA deficit is a much larger share of world saving. Another difference is the role that the
dollar plays as the world’s “reserve currency.” Because the dollar is the world’s preferred
currency for a store of value, medium of exchange, and unit of account, holders may be less
willing to abandon it than they would any other currency. If so, the United States may be able to
run higher sustainable CA deficits than other countries.
In the developing world, a large CA deficit has often been a leading indicator of financial and
currency crisis. This was the case in many recent crises, including Mexico, East Asia, Turkey,
Brazil, and Argentina. The applicability of these experiences to the United States may be limited,
however, because of three key differences between the United States and these countries. First,
the United States has a flexible exchange rate regime. Countries with fixed exchange rates can be
forced into crisis when their foreign exchange reserves are exhausted by “speculators” betting
against the currency; similar bets are harder to make against flexible exchange rates. Second, the
United States has traditionally been seen as a “safe haven” for investment. Third, unlike
developing countries, the United States is able to borrow in its own currency, so that depreciation 15
reduces rather than increases the burden of servicing its debt. Therefore, historical comparisons
have tended to focus on the experience of other industrialized countries.
Economist Sebastian Edwards found that since 1970, only two other industrialized countries,
Ireland and New Zealand, had high CA deficits that were long-lasting (seven and five years,
respectively). He found that large countries that experienced sharp declines in their CA also saw 16
per capita GDP growth decline by 3.6% to 5.0%.
Economists at Goldman Sachs, a financial firm, analyzed all episodes in industrialized countries
since 1980 where the CA improved by more than 2% of GDP. It found 31 cases where the
adjustment occurred under circumstances of economic disruption and 13 where there was no
comparable disruption. In the disruption episodes, the economy typically started from a position
of overheating and the output gap (the difference between actual and potential output) worsened
by an average of 3.6% of GDP, whereas in the benign episodes, the economy started from a
position of excess capacity and the output gap improved by 1.9%. The fact that the economy was
initially overheating in the disruption episodes suggests that causality may run in the opposite
direction—the CA shift may be a symptom rather than a cause of economic slowdown. In the
disruption cases, there was little real exchange rate depreciation; in the benign cases, it averaged
5.1%. In most cases, the adjustment took several years. In all cases, consumption growth was
negative on average and (surprisingly) interest rates on average fell. In only two cases (Portugal
in the early 1980s and Finland in the early 1990s) was the CA decline associated with a severe

15 In dollar terms, depreciation would not affect the value of dollar-denominated liabilities. However, in foreign
currency terms or in terms of the purchasing power that it conveys to foreign lenders, depreciation reduces the value of
U.S. liabilities.
16 Sebastian Edwards,Is the Current Account Deficit Sustainable?” National Bureau of Economic Research, Working
Paper no. 11541, August 2005. Since his study, both Ireland and New Zealand have again run large CA deficits.

recession. (The recession and CA decline in Finland were widely attributed to the collapse of the
Soviet Union, among other factors.) Some of these cases may not be applicable to the U.S.
experience, however, because the sample includes countries that had a small CA deficit or CA
surplus. Only eight of these episodes involved a larger CA deficit as a share of GDP than the U.S. 17
deficit today, and all eight episodes involved small countries.
The International Monetary Fund conducted a similar study.18 It found 42 cases where an
advanced economy had an initial CA deficit of at least 2.5% of GDP, and the CA deficit
subsequently declined by at least 50%. It found, on average, that the real exchange rate declined
by a cumulative 12.2%, causing a shift in the CA of 5.7% of GDP over the next 4.6 years, moving
the CA from deficit to surplus (on average). GDP growth fell by an annual average of 1.4%
during the reversal, causing the output gap to deteriorate by an average of 3% of GDP, from peak
to trough of the business cycle. In 11 of these cases, the slowdown in economic growth was large,
but, unlike the United States, 9 of these 11 did not have a floating exchange rate. In 10 other
cases, there was no slowdown in growth following the CA reversal.
However, not all cases of large CA deficits end in a reversal. Besides the United States, the IMF
identified five other advanced economies that have had large CA deficits that have persisted to
date: Australia (which has run a CA deficit of more than 2% since 1980), Greece (since 1996),
New Zealand (1989), Portugal (1996), and Spain (1999). All except Australia have had CA
deficits that were considerably larger than the U.S. in recent years.
In a similar study, economists Debelle and Galati found little evidence that CA adjustments
historically lead to significant disruption in financial markets. They found little change in the
composition of capital flows before adjustment, which they argue is evidence that current account 19
adjustment is caused by, rather than the cause of, broader macroeconomic imbalances.
Economists Hung and Kim use statistical regressions to estimate the probability of an exchange
rate crash based on several macroeconomic variables across many industrialized countries over
time, and estimate that the United States had a 9% probability of a dollar crash in 2006. They find
that the current account balance (and the regression overall) has relatively little predictive power 20
for a currency crash.
The size of the CA deficit in any given year may be less important in determining sustainability
than how persistent CA deficits increase a country’s net foreign debt over time. Economists
Maurice Obstfeld and Kenneth Rogoff found that in 2003 the net debt owed to foreigners was
about 23% of GDP for the United States, near an all-time high. Were CA deficits to continue at
more than 5% of GDP per year, U.S. net debt to foreigners would reach 70% of GDP within 30

17 Dominic Wilson and Roopa Purushothaman, “Do Current Account Adjustments Have to Be Painful? Global
Economics Weekly, Goldman Sachs, no. 05/04, February 2005. Similar results are found in Hilary Croke, Steven
Kamin, and Sylvain Leduc, “Financial Market Developments and Economic Activity During Current Account
Adjustments,International Finance Discussion Papers, Federal Reserve, no. 827, February 2005.
18 International Monetary Fund, “Exchange Rates and the Adjustment of External Imbalances,” World Economic
Outlook, April 2007, ch. 3.
19 Guy Debelle and Gabriele Galati, “Current Account Adjustments and Capital Flows, Bank of International
Settlements, Working Paper no. 169, February 2005.
20 Juann Hung and Young Kim,Implications of Past Currency Crises for the U.S. Current Account Adjustment,”
Congressional Budget Office, Working Paper no. 2006-7, June 2006. The authors define a currency crash as a decline
in value of at least 25% in one year that is accelerating.

years. Although this implies a relatively small yearly debt burden, many countries that have
experienced CA reversals in the postwar period had smaller net foreign debt-to-GDP ratios,
between 20% and 80% of GDP. Obstfeld and Rogoff identify only one country (Ireland) with a
net foreign debt-to-GDP ratio that has exceeded 80%. Thus, the authors conclude that large U.S. 21
CA deficits cannot be sustained indefinitely.

Five recent academic papers address the sustainability issue. In the papers, the models used to
estimate changes in the dollar and CA are not empirically derived; they are simulations based on 22
theoretical assumptions meant to be consistent with reality.
Obstfeld and Rogoff have estimated how much the dollar would need to depreciate in order to 23
make the CA deficit disappear. In their model, shocks to aggregate demand or shifts in the
demand or supply of tradeable goods could cause the CA deficit to decline; their model does not
allow for exogenous changes in the demand for U.S. assets to affect the CA deficit. They estimate
that the real exchange rate would depreciate between 14.7% and 33.6% if a CA deficit equal to
5% of GDP were eliminated by a change in aggregate demand, and between 9.8% and 25.5% if
eliminated by a change in the supply of tradeable goods. They estimate that depreciation would
be accompanied by a 3.9% to 7.1% decline in the terms of trade. The predicted dollar
depreciation is so large because about three-quarters of U.S. output is nontradeable, production
cannot be quickly shifted into tradeable goods to take advantage of the depreciation, and import
and export prices change much more slowly than the exchange rate. This model does not predict 24
how much larger the CA deficit could get or how quickly it will eventually fall.
Economists Blanchard, Giavazzi, and Sa explicitly allow asset demand to influence the exchange
rate, and they assume that assets from different countries are not perfect substitutes. In their
model, a CA deficit would eventually decline because demand for U.S. assets is finite. Although
an increase in the demand for U.S. assets would initially cause the dollar to appreciate, they
argue, it would later depreciate to finance debt service (though it would remain above its pre-
appreciation value). They estimate that a 15% decline in the dollar would be associated with a

21 Maurice Obstfeld and Kenneth Rogoff, “The Unsustainable U.S. Current Account Position Revisited,” National
Bureau of Economic Research, Working Paper no. 10869, October 2004.
22 In addition to the papers reviewed, similar research includes, Hamid Faruqee, et al.,Smooth Landing or Crash?
Model Based Scenarios of Global Current Account Rebalancing, National Bureau of Economic Research, Working
Paper no. 11583, August 2005; Richard Clarida, G-7 Current Account Imbalances: Sustainability and Adjustment
(Chicago, IL: University of Chicago Press, 2006); Robert Dekle et al, “Global Rebalancing With Gravity, National
Bureau of Economic Research, Working Paper no. 13846, March 2008.
23 Maurice Obstfeld and Kenneth Rogoff, “The Unsustainable U.S. Current Account Position Revisited,” National
Bureau of Economic Research, Working Paper no. 10869, October 2004.
24 Using a similar model and parameters, Cavallo and Tille suggest that the reduction in the current account balance and
depreciation of the dollar could be smoother and more gradual (although ultimately by a similar magnitude) if one
assumes that the net foreign debt is held constant and changes in asset valuations due to exchange rate effects are used
to temporarily finance current account deficits. They estimate that the dollar would ultimately depreciate by 31%.
Michele Cavallo and Cedric Tille, “Current Account Adjustment With High Financial Integration, Federal Reserve
Bank of San Francisco, Economic Review, 2006, p. 31.Using a similar model and parameters, Engel and Rogers argue
that if the U.S. grows faster than other industrialized countries in the future, then little dollar depreciation would be
required to reduce the CA deficit. Charles Engel and John Rogers, “The U.S. Current Account Deficit and the Expected
Share of World Output,” National Bureau of Economic Research, Working Paper no. 11921, January 2006.

decline in the CA deficit equal to 1.4% of GDP. About one-third of the decline in the CA deficit
results from U.S. debt being denominated in U.S. dollars, because a depreciation reduces its
value. Blanchard et al. estimate that stabilizing the net-debt-to-GDP ratio at 2003 levels would
require the dollar to immediately depreciate by 56% and the CA deficit to decline to 0.75% of
GDP. However, assuming foreigners desire to maintain holdings of U.S. assets at their current
share, their model predicts that the depreciation would be stretched over a few decades,
depreciating by 2.7% a year, at most. If foreigners decided to reduce their holding of U.S. assets, 25
the model predicts a larger, but still gradual, depreciation.
Edwards uses a similar model to simulate how much the dollar would depreciate from its 2004
level depending on different assumptions about the foreign demand for U.S. assets. Unlike
Blanchard et al., he projects fairly rapid declines in the CA deficit and dollar in the future. Under
his optimistic scenario, in which he assumes that the U.S. net debt will rise to 60% of GDP by
2010 and then remain constant, the CA deficit would peak at 7.3% of GDP in four years, before
eventually declining to 3.2% of GDP (with most of the decline occurring in the first four years
after the peak). The real value of the dollar would appreciate while the deficit was increasing, but
it would decline 21% in the first three years after the deficit began falling. If net debt were to
decline to 50% of GDP after 2010 instead of remaining at 60% of GDP, which would still be
about double its current level, the decline in the CA deficit and dollar would be greater. Edwards
calculates that the deficit would fall by 5.3% of GDP and the dollar would depreciate by 28% 26
after three years, which would bring both measures close to their long-term projected levels.
Economists Roubini and Setser simulate what would happen to the net foreign debt over the next 27
10 years under three scenarios. In the first scenario, imports and exports continue to grow at
historical rates. The CA deficit would exceed 12.8% of GDP and the net foreign debt would
exceed 80% of GDP by 2012. In this scenario, the authors do not believe it is plausible to assume
that foreigners would be willing to finance borrowing of this magnitude, and use this scenario to
argue that the current path is unsustainable. In the second scenario, the trade deficit stabilizes at
5% of GDP. The CA deficit would still increase to 8.8% of GDP in 2012 (because of higher net
investment income payments), at which point the net foreign debt would exceed 70% of GDP.
Servicing a debt of this size, they estimate, would cost 3% of GDP in 2012. To reduce the trade
deficit to 5% of GDP, they estimate that the dollar would need to depreciate by about 10% from
its 2004 level. This scenario (and the first scenario) is not sustainable in the long run because the
net foreign debt would grow continuously. In the third scenario, the trade deficit declines by the
end of the projection to the point where the net foreign debt stabilizes as a share of GDP. They
estimate that the net foreign debt would stabilize with a trade deficit of 0.8% and a CA deficit of
1.3% of GDP. If the CA deficit gradually declined to this level in 2012, the net foreign debt would
stabilize at nearly 60% of GDP, which would cost an estimated 1.75% of GDP to service in 2012.
To achieve a reduction in the CA deficit of this magnitude, they estimate that the dollar would
need to depreciate substantially and the federal budget would need to be balanced. All of these
scenarios assume that relative interest rates remain similar to past values as the net foreign debt

25 Olivier Blanchard, Francesco Giavazzi, Filipa Sa,The U.S. Current Account and the Dollar,” Massachusetts
Institute of Technology, Working Paper no. 05-02, May 2005.
26 Sebastian Edwards, “Is the CA Deficit Sustainable?” National Bureau of Economic Research, Working Paper no.
11541, August 2005.
27 Nouriel Roubini and Brad Setser, “The U.S. as a Net Debtor: The Sustainability of the U.S. External Imbalances,”
working paper, August 2004,

rises; if foreigners demanded higher interest rates, then this would feed through to a much larger
CA deficit and debt path than simulated.
Economist Paul Krugman estimates that the dollar would need to depreciate by at least 35% from 28
its 2005 value in real terms in the long run for the trade deficit to be reduced to zero. He looks
for evidence of whether this depreciation will happen gradually or abruptly. For the depreciation
to be smooth, Krugman argues that it must be anticipated by rational investors, in which case they
would currently require a rate of return premium to hold U.S. assets (to offset the loss suffered
from the future dollar depreciation). To determine how large the premium would have to be, he
considers two hypothetical paths for the dollar. In one path, the dollar declines by 1.75%
annually, and net foreign debt peaks at 118% of GDP, or at least one-third of the total U.S. capital
stock. If foreigners are unwilling to hold that much U.S. debt, the dollar would have to depreciate
more rapidly. In the other path, the dollar declines by 3.5% annually, and net foreign debt peaks at
58% of GDP. Yet he finds no evidence of a rate of return premium anywhere near the magnitude
of either 1.75% or 3.5%—after adjusting for inflation, U.S. interest rates are very close to foreign
rates. This implies that foreigners do not foresee any significant dollar depreciation in the future.
Therefore, he argues that when investors eventually realize how much the dollar will depreciate,
they will likely sharply reduce their demand for U.S. assets, causing the dollar to plummet.
Recent experience suggests that the dollar depreciation required to put the CA deficit on a
sustainable path may indeed be large. From 2002 to 2006, the dollar depreciated by 16% in
inflation-adjusted terms. Despite the depreciation, the CA deficit continued to rise, both in dollar
terms and as a share of GDP, from 4.5% of GDP in 2002 to 6.1% of GDP in 2006. In response to
the weaker dollar, exports rose rapidly from 2004 onward, but this did not lead to a lower current
account deficit because imports also continued to rise rapidly. This experience points to the fact
that external factors, which can be held constant in the models discussed in this section, also
influence the CA and the dollar in reality. In 2007, the dollar continued to depreciate, and the CA
deficit fell to 5.3%.
The wide dispersion of estimates on the dollar depreciation associated with a decline in the CA
deficit points to the complex and imperfectly understood factors that determine the dollar’s value,
the lack of a consensus exchange rate model that performs well empirically, and the sensitivity of
theoretical models to changes in uncertain empirical parameters. Further complicating model-
based projections, the path of CA adjustment is also subject to non-market forces, such as the
accumulation of foreign reserves by central banks. So far, the dollar and CA deficit has not
closely matched many of the models’ predictions since the predictions were made. Furthermore,
no model reliably answers the underlying question of how much and how quickly the CA deficit
will potentially decline.

In the long run, running a CA deficit at current trend levels (i.e., growing faster than GDP) would
result in net foreign debt continually growing relative to GDP. This is unsustainable if foreigners
have a limited appetite for U.S. assets. Thus, in the long run, the CA deficit will most likely
decline, although it need not decline to zero to stabilize the net foreign debt relative to GDP.
Relative to GDP, the CA deficit declined in 2007 and is projected to continue to decline in 2008.

28 Paul Krugman, “Will There Be a Dollar Crisis?, Economic Policy, July 2007, p. 436.

It is too soon to say whether this decline is being caused by temporary cyclical factors or
represents the beginning of a long-term adjustment process. To date, the net foreign debt has
placed no burden on the U.S. economy because U.S.-owned foreign assets have earned more than
foreign-owned U.S. assets.
Whether policymakers should be concerned about a future decline in the CA deficit depends on
whether the decline were to happen in an orderly or disruptive way. There is little reason to think
that a gradual decline would have a deleterious effect on the overall economy. But a sudden
decline, brought on by a sudden reduction in foreign capital inflows, could be disruptive to U.S.
financial markets, causing negative spillover effects for the broader economy. While a sudden
reduction in foreign capital inflows cannot be ruled out—it has happened to foreign countries—it
seems highly unlikely. The United States is different in a number of ways from the countries that
have experienced CA crises—it is much larger, its financial markets and economy are highly
developed, it has a floating exchange rate, and it is seen as a safe haven in times of financial
turmoil. Nonetheless, even if the risk of a sudden CA reversal is small, it is arguably worth policy
consideration since it could be highly costly to the U.S. economy.
The recent financial crisis in the United States bears resemblance to the sort of scenario
envisioned by economists concerned about a sudden, destabilizing outflow of capital. Yet when
the crisis worsened in September 2008, the dollar began appreciating and heightened demand for
certain U.S. assets, such as U.S. Treasuries, drove their prices up to unusually high levels.
Marc Labonte
Specialist in Macroeconomic Policy, 7-0640