Balance of Payments Imbalances

International Monetary Fund
Published Date:
December 2007
  • ShareShare
Show Summary Details


Considering the nature of the title of my paper, I am incredibly impressed that there are so many people who wish to hear what I have to say on this subject.

Thank you, Sir Andrew and Mr. Van Houtven.

The financial crisis that erupted on August 9 was an accident waiting to happen. Credit spreads across all global asset classes had become compressed to clearly unsustainable levels. Something had to give. If the crisis had not been triggered by a mispricing of securitized U.S. subprime mortgages, it would have eventually erupted in some other sector or market. The candidate of many analysts in recent years has been a dramatic and abrupt unwinding of America’s huge current account deficit, with all sorts of extraordinary aftermaths as a consequence. To date this has not happened. But fear-laden concerns put that deficit on the agenda of virtually every international gathering I attended as Fed chairman and since.

Unless protectionist forces drain the flexibility of the international financial system, I do not view the ultimate unwinding of America’s current account deficit, amounting to 6 percent of our GDP, as a cause for undue alarm. Apprehensions about the U.S. external deficit are certainly not groundless. At some point, foreign investors will balk at increasing the share of dollar-denominated assets in the portfolios they hold. There obviously is a limit to the extent that U.S. financial obligations to foreigners can reach. And perhaps the recent decline in the U.S. dollar and shrinkage of the current account deficit is an indication that America is approaching that limit.

In 2006, the financing of the deficit siphoned off almost three-fifths of all the cross-border savings1 of the 67 countries that ran current account surpluses in that year. Developing countries, which accounted for nearly half the value of those surpluses, were apparently unable to find sufficiently profitable investments at home that overcame market and political risk. The United States a decade ago likely could not have run up today’s near-$800 billion annual deficit for the simple reason that we could not have attracted the foreign savings to finance it. In 1995, for example, total cross-border saving was less than $300 billion.

But the reason I conclude that the persistently growing U.S. current account deficit does not have seriously negative consequences for the U.S. economy is that those deficits are a small part of a far larger but less-threatening, ever-expanding specialization and division of labor that is irreversibly evolving in our increasingly complex global environment.

Pulling together the pieces of evidence—anecdotal, circumstantial, and statistical—strongly suggests, to me at least, that the current account deficit is best viewed as a segment of a broader set of rising deficits and offsetting surpluses that reflect transactions of U.S. economic entities—households, businesses, and governments—mostly within the borders of the United States.

The long-term updrift in this broader swath of unconsolidated deficits and mostly offsetting surpluses of economic entities has been persistent but gradual for decades, probably generations. However, the component of that broad set that captures only the net foreign financing of the imbalances of the individual U.S. economic entities, our current account deficit, increased from negligible in the early 1990s to 6.2 percent of our GDP by 2006.

What data we have suggest that the rise in America’s current account deficit as a percentage of GDP since early this decade is, to a large extent, the result of American business and government’s turning to foreign sources of deficit funding in place of domestic funding, and not predominantly the result of an acceleration in the secular uptrend in economically stressful company or government imbalances. Household borrowing, incidentally, from abroad to finance shortfalls in cash flow has always been negligible.

In my judgment, policymakers have been focusing too narrowly on foreign claims on U.S. residents rather than on all claims, both foreign and domestic, that influence economic behavior and can be a cause of systemic concern. It is the level of debt, not the source of its financing, that should engage us. Our conventional tabulations are often too loosely rooted in the obsession of the mercantilists of the early eighteenth century to achieve a surplus in their balance of payments which brought them gold in settlement, then the mistaken standard of the wealth of the nation.

Were we to measure financial net balances of much smaller geographic divisions, such as the individual American states or Canadian provinces, which we do not, or of much larger groupings of nations, such as South America or Asia, the trends in these measures and their seeming implications could be quite different from those extracted solely from the conventional nation-delineated measure of current account balance.

The choice of the appropriate geographical unit for measurement should depend on what we are trying to find out. I presume that, in most instances, at least in the policy setting, we seek to judge the degree of economic stress that could augur significantly adverse economic outcomes. We should require data on financial balances at the level of detail at which economic decisions are made: individual households, businesses, and governments. Those data are the equivalent of current account balances, but at the level of individual economic entities where leverage and stress are experienced, and hence, where actions and trends that may stabilize economies originate.

National borders, of course, do matter, at least to some extent. Debt service payments on foreign loans ultimately must be funded from exports of tradable goods and services or from capital inflows, whereas domestic debt has a broader base from which it can be serviced. For a business, cross-border transactions can be complicated by legal risks and a volatile exchange rate, but generally these are difficulties not outside most normal business risk.

It is true that the market adjustment process seems to be less effective or transparent across national borders than within them. Prices of identical goods at nearby locations but across borders, for example, have been shown to differ significantly, even when denominated in the same currency.2 Thus, cross-border current account imbalances impart a degree of economic stress that is likely greater than that stemming from domestic imbalances only. But in a flexible economy, are any of these as significant as we tend to make them?

I do not deny that nation-defined current account imbalances do have important implications for exchange rates and terms of trade. But I suspect the measure is too often used to signify some more generic malaise, especially in the context of the so-called twin American deficits, with reference to our politically determined federal budget deficit, which has quite different roots and policy requirements than those of the market-determined current account balance.

This afternoon, I should like, first, to turn to the narrower issue of the current account balance and then proceed to the broader issue of dispersion of unconsolidated economic entities and its implications.

The economic literature of recent years is filled with explanations of the possible causes of outsized U.S. current account deficits or their algebraic equivalent, an excess of domestic investment over domestic savings.3 To me the most persuasive explanations are a major decline in home bias and a concurrent significant acceleration in U.S. productivity growth. Home bias is the parochial tendency of investors to choose to invest their savings in their home country, even though this means passing up more risk-adjusted profitable foreign opportunities. When people are familiar with an investment environment, they harbor less uncertainty, and hence, less risk than they do for objectively comparable investments in distant, less-accessible environs.

A decline in home bias is reflected in savers’ increasingly reaching across national borders to invest in foreign assets. This engenders a marked rise in current account surpluses among some countries and an offsetting rise in deficits of others. For the world as a whole, of course, exports must equal imports, savings must equal investment, and the world consolidated current account balance is always zero.

Home bias was very much in evidence globally for the first half century following World War II. Domestic saving was directed almost wholly toward domestic investment. In that world of exceptionally strong home bias, external imbalances were small. However, starting in the mid-1990s, home bias began to decline perceptibly. The global weighted correlation coefficient between national savings rates and domestic investment rates, a measure of the degree of global home bias,4 declined from around 0.95 in 1993, where it had hovered since 1970, to an estimated 0.74 in 2005 (Figure 1).5

The advance of information and communications technologies that effectively shrank the time and distance that separate markets around the world, and a dismantling of restrictions on cross-border capital flows that significantly reduced the perceived risk of reaching out across sovereign borders, muted uncertainty. Those trends more or less coincided with the boost to competitive market capitalism resulting from the demise of central planning, an issue that I develop at length in my new book.6

Gross domestic income, as a consequence, rose significantly across the developing world. Consumption, inhibited in part by unresponsive financial infrastructures, lagged, propelling the developing world’s saving rate to 33 percent of GDP in 2006, up from 22 percent in 1992. Investment opportunities in the developing world, however, evidently were not adequate to absorb the new surge in saving, and hence investors, now less daunted by the uncertainties of distance, sought investments in the developed world, especially in the United States.

In short, vast improvements in information and communications technologies and rule of law and the enhanced protection of foreigners’ property rights have greatly extended investors’ geographic horizons, rendering foreign investment less risky than it appeared in earlier decades. Doubtless, the worldwide decline in credit-risk spreads, to which I alluded earlier, was also a factor.

Although world trade as a percentage of GDP has been expanding for more than a half century, only since the early 1990s has expanding trade been associated with the emergence of ever-larger U.S. current account deficits matched by corresponding widening of the aggregate external surpluses of many of our trading partners, most especially China.7 To get a sense of how widely cross-border current account balances have dispersed, I calculated the absolute sum of all countries’ current account imbalances, irrespective of sign, as a percentage of world nominal GDP. That ratio hovered between 2 and 3 percent between 1980 and 1996. By 2006, it had risen to almost 6 percent (Figure 2).

Decreasing home bias is the major determinant of wider global surpluses and deficits. But differences in risk-adjusted rates of return, reflecting different rates of productivity growth, seem to have been a contributor as well. Rates of return are clearly a key factor in determining to which countries excess savings are directed for investment. Since 1995, the greater rates of productivity growth in the United States, compared with still-subdued rates abroad, apparently produced correspondingly higher risk-adjusted expected rates of return that fostered a disproportionate rise in the global demand for U.S.-based assets. In addition, U.S. history of more than two centuries of protection of foreign property rights also helps to explain why such a large percentage of cross-border savings has been directed to the United States.8

A far more important question, however, is whether the seemingly inevitable adjustment of the U.S. external accounts will be benign or, as many fear, entail an international financial crisis compounded by a dramatic fall in the dollar. I am far more inclined toward a more benign, market-determined outcome in which financial factors—exchange rates, interest rates, and the prices of assets—change but the real economy—economic activity and employment—is sustained.

My lessened concern rests on the fact that current account balances, as I noted earlier, are only part of a larger set of forces balancing the world’s saving and investment. And that larger set of forces is not exhibiting the degree of economic stress implicit in the current account deficit. The broader context does raise the extent of debt leverage, an issue to which I will return shortly.

The evolution of the world economy during the past century has enabled the scale of sustainable financial surpluses and deficits of individual households, businesses, and governments,9 including those that involve cross-border flows, to persistently increase. Owing to the never-ending expansion of the division of labor, as I document later, the ratios of both financial surpluses and deficits of individual U.S. economic entities relative to their incomes, on average, have been on the rise for at least a century and, more likely, far longer, a process that has apparently not accelerated despite a massive increase in globalization in recent decades.

For most of that period, the rising deficits of most U.S. resident economic entities were almost wholly matched by surpluses of other resident economic entities. Our national current account balances were thus small.10 What is special about the past decade is that the global decline in real long-term interest rates that resulted in significant capital gains on homes and other assets has fostered a large increase in U.S. residents’ purchases of foreign-produced goods and services willingly financed, net, by foreign investors.

Over time, an ever-growing proportion of U.S. households, businesses and governments, both federal and local, have funded their capital investment from sources other than their own household incomes, corporations’ internal funds, or government taxes. In early America, almost all of that financing originated within U.S. financial institutions, and in that case almost all within U.S. commercial banks. The persistent rise in both U.S. household and business assets and liabilities relative to income for more than a half century (Figure 3), as I note later, is also a function of a widening division of labor but in addition irreversible capital deepening.

The evidence of increasing dispersion of surpluses and deficits is impressive, especially in recent decades. A detailed calculation by Federal Reserve Board staff, employing data from more than 5,000 nonfinancial U.S. corporations for the years 1983 to 2004, found that growth in the sum of deficits of those corporations where capital expenditures exceeded cash flow persistently outpaced the growth in corporate value added. The sum of surpluses and deficits, disregarding sign, as a ratio to a proxy for total nonfinancial corporate value added exhibited an average annual increase of 3.5 percent a year.11

Reliable data on the dispersion of the financial deficits of U.S. economic entities, aside from nonfinancial corporations, are sparse. A separate and far less satisfactory calculation of only partly unconsolidated financial balances of individual economic entities, relative to nominal GDP, exhibits a rise over the past half century in the absolute sum of surpluses and deficits that is almost 2 percentage points per year faster than the rise in nominal GDP (Figure 4). The rise, however, is interrupted by a decline between 1987 and 1997, a possible consequence of the credit problems of the early 1990s and the collapse of the savings and loan industry.

The measure charted in Figure 4 estimates saving less investment balances among eight consolidated sectors recorded in U.S. macroeconomic statistics: households; corporations; nonfarm, noncorporate businesses; farms; state and local governments; the federal government; finance; and the rest of the world. I include the rest-of-the-world sector because it measures surpluses or deficits of U.S. residents, even though they reflect the accumulation of net claims on, or obligations to, foreigners. The other seven sectors reflect net claims on, or obligations to, domestic residents only.

Since consolidation generally reduces dispersion, the dispersion of individual economic units presumably has been rising even more rapidly relative to income over the years than the results of this eight-sector model.12 Importantly, these data suggest that although this measure of total dispersion, domestic and foreign, has steadily increased in the past decade relative to GDP, the increase in the dispersion of the imbalance of economic entities within U.S. national borders appears to have slowed. Thus, while, since the mid-1990s, the overall dispersion of imbalances of U.S. economic entities has continued to grow, an increasing proportion of deficits of U.S. households, businesses, and governments has been financed from foreign rather than domestic sources.

This is certainly obvious in the financing of our federal budget deficit and of business capital expenditures.13 In short, the expansion of our current account deficits during the past decade appears to largely reflect the shift in trade and financing from within the borders of the United States to cross-border trade and finance. If so, does this matter? Does it matter importantly, for example, whether a U.S. resident corporation finances its capital outlays from foreign rather than domestic sources?14 With some qualifications, the stress on U.S. economic entities has arguably increased little with the shift in the source of their financing. The rise in the ratio of imbalances—the absolute sum of foreign and domestic—to GDP is a much more modest and less threatening trend over the past decade than that exhibited by its foreign component, the current account only.

Decisions to finance domestic U.S. capital investment by borrowing from U.S. or foreign lenders are often a matter of convenience and can usually be reversed at small cost. It is almost always the level of debt of economic entities, not the geographic location of the lender, that creates stress. Implicit in a widening dispersion of financial surpluses and deficits of individual economic entities is the expectation of increasing cumulative deficits for some entities and, hence, a possible accelerating rise in debt as a share of income or its equivalent, GDP.15

From 1900 to 1939, nonfinancial private debt in the United States rose almost 1 percentage point faster per year on average than nominal GDP. World War II and its aftermath inflated away the real burden of debt for a while. The debt-to-GDP ratio, accordingly, declined. The updrift in the ratio, however, resumed shortly thereafter. And from 1956 to 2006, nonfinancial business debt rose 1.7 percentage points faster at an annual rate than gross nonfarm business product.

The trend toward intracountry dispersion of financial imbalances is likely occurring not only in the United States, but in other countries as well. The existence of such a trend is suggested by the rise in unconsolidated nonfinancial debt of the major industrial economies, excluding the United States, over the past three decades, which has exceeded the growth of GDP by 1.6 percentage points annually. A rising debt-to-income ratio for households or of total nonfinancial debt to GDP is not, in itself, a measure of stress. It is largely a reflection of dispersion of the growing financial imbalance of economic entities that, in turn, reflects the irreversible updrift in division of labor and specialization.

Both non-financial-sector assets and debt have risen faster than income over the past half century worldwide. But in the United States, at least, debt is rising faster than assets. That is, debt leverage has been rising. Household debt as a percentage of assets, for example, reached 19.3 percent by the end of 2006, compared to 7.6 percent in 1952. Nonfinancial corporate liabilities as a percentage of assets rose from 28 percent in 1952 to 54 percent by 1993, but retreated to 43 percent by the end of 2006 as corporations embarked on a major program to improve their balance sheets. The net rise, however, over the whole half century is still quite impressive.

It is difficult to judge how problematic this long-term increase in leverage is. Since risk aversion is presumably innate and unchanging, the willingness to take on increased leverage over the generations likely reflects an improved financial flexibility that enables leverage to increase without increased objective risk, at least up to a point.16 American commercial bankers in the immediate post-Civil War years perceived the necessity to back two-fifths of their assets with equity. Less was considered too risky. Today’s bankers are comfortable with a tenth. Nonetheless, bankruptcy is less prevalent today than 140 years ago.

The same trends hold for households and businesses. Rising leverage appears, in large part, the result of massive improvements in technology and infrastructure, not significantly more risk-inclined humans or arguably objective risk. Obviously, a surge of debt leverage above what the newer technologies can support invites crises, as many analysts now currently fear. I am not sure where the tipping point is, but we can be sure there is one. For example, our subprime mortgage market was clearly seen as overleveraged, as home price inflation came to a halt in the United States.

Globalization is changing many of our economic guideposts. It is probably reasonable to assume that the ratio of worldwide dispersion of the financial balances of unconsolidated economic entities to world nominal GDP will continue to rise as increasing specialization and division of labor, and ever more sophisticated finance and capital deepening, spread globally. Whether the component reflecting dispersion of world current account balances continues to increase as well is a more open question. Such an increase would imply a further decline in home bias. But in a world of nation-states, home bias can decline only so far. Thus, the degree of global current account dispersion would also stabilize, as indeed it may already have done.17

If the current disturbing drift toward protectionism is contained, and markets remain sufficiently flexible, changing terms of trade, interest rates, asset prices, and exchange rates should cause U.S. savings to rise relative to domestic investment without undermining either production or employment. This would reduce the U.S. need for foreign finance and reverse the trend of the past decade toward increasing reliance on funds from abroad.

Thank you very much. [Applause.]

You will excuse me if I put a postmortem on my piece. I stipulated that I am basically of the opinion that this is not a huge problem. But if the pernicious drift toward fiscal instability in the United States and elsewhere is not arrested and is compounded by a protectionist reversal of globalization, the current account deficit adjustment process could be quite painful for the United States and our trading partners. I think this suggests how critically important it is for those of us who are involved in the international community to be acutely aware of how dangerous protectionism is in undermining the flexibility of not only the global system, but our internal economies, as well.

And unless we address that problem, if it arises, I do not think we fully understand how significant a major blow to world economic prosperity, especially in the emerging nations, such a reversal could be. And I truly hope we are acutely aware of what dangers arise, as they arise, and bring whatever power a group such as this can bring to bear to turn such trends around.

Thank you. It has been a pleasure being with you today.

    Other Resources Citing This Publication