A $2 trillion question

Gian Maria Milesi-Ferretti

28 January 2009



With the raging financial crisis and the possible measures to combat it taking centre stage in the policy debate, the focus on the US current account deficit and its implications for the US net external asset position has waned to some extent.

To be sure, the US current account deficit remains large, but with a dramatic decline in domestic demand, plummeting oil prices, and the lagged effects of past dollar depreciation helping US exports, recourse to external borrowing in the US has declined and is projected to fall further the next few years. Yet preliminary estimates suggest that last year featured the most significant deterioration in the US net external position to-date: over $2 trillion dollars. What explains such a large decline? And what consequences will it have going forward?

In order to understand recent developments and put them in a correct perspective, it is useful to look back a few years. At the end of 2002 the US net international investment position (NIIP) was negative to the tune of $2 trillion (some 24% of US GDP), the result of a sequence of growing current account deficits that had reached 4.5% in 2002 (the US had been a net external creditor until the late 1980s). The $2 trillion in net external liabilities reflected a modest net asset position in equity-type instruments (portfolio equity holdings, measuring holdings of stocks, and foreign direct investment, measuring controlling shares in foreign affiliates) and a large net debtor position in debt instruments (primarily bonds).

Over the period 2002-2007 (the heyday of the "global imbalances"), the US continued to borrow heavily − a cumulative total of $3.4 trillion. Ceteris paribus, this should have raised US net external liabilities to some $5.5 trillion (some 40% of GDP).1 It did not. The NIIP deterioration was much more modest – only $400 billion, and as a ratio of GDP it actually improved (Figure 1). So where did those other $3 trillion of US net borrowing go? Were the concerns of global imbalances overstated or were the imbalances themselves a non-entity, offset by mis-measured dark matter?

Figure 1

Actually, those extra $3 trillion are (mostly) accounted for (see Lane and Milesi-Ferretti, 2008 and Curcuru, Thomas, and Warnock, 2008 for further discussion). During this period, US-held overseas assets (predominantly denominated in foreign currency) increased in value much more significantly than foreign-held assets in the US (predominantly denominated in dollars). This occurred for two reasons. First, the dollar depreciated significantly (some 22% in real effective terms between February 2002 and December 2007). Second, there was a much stronger domestic-currency performance of foreign stocks relative to US stocks. A weaker dollar tends to raise the domestic-currency value of US assets denominated in foreign currency, thus strengthening the US external position. And faster growth in foreign equity markets relative to the US stock market imply that the value of foreign equities held by US residents increases in value more rapidly than the value of foreign holdings of US equities. For example, a dollar invested at the end of 2002 in non-US stock markets would be worth $2.9 at the end of 2007, versus $1.8 for a dollar invested in the US stock market. Of course, currencies and stock prices had fluctuated quite significantly also in the past. However, in past episodes cross-border asset holdings were much smaller, and hence the implications of these fluctuations for cross-border positions were considerably less dramatic.

As a result of these developments, by the end of 2007 the asymmetry in the US NIIP between equity and debt had become much more pronounced. The net equity position, boosted by the hefty capital gains on foreign equity holdings, had reached US$3 trillion, while the net debt position had grown to $5.5 trillion. The key question going forward was whether the 2002-2007 developments were to be considered "standard" – that is, whether the US NIIP could "defy gravity" and remain stable despite still hefty current account deficits.

Developments in 2008 suggest not. After a number of bull-market years, stock market retreated across the globe, battered by the financial crisis. Because the US net equity position had grown so large, even an equi-proportionate decline in the US and foreign stock markets would have inflicted significant capital losses on US portfolios. In dollar terms, the decline was actually larger in non-US stock markets, also reflecting the dollar appreciation. These financial market developments account for the lion share in US net capital losses – with preliminary estimates suggesting net portfolio equity losses of $1.2 to1.3 trillion. The US also incurred some losses on their net FDI position – the dollar’s appreciation has reduced the dollar value of US FDI abroad, with the decline in stock prices adding larger losses on the (admittedly more difficult to measure) market-value estimates of FDI.

Of course the turmoil in financial market also affected the value of debt securities, and hence of the US net debt position. However, preliminary estimates suggest that the net effect of these changes is considerably more modest than the effect of equity price declines. Let’s focus first on US debt liabilities. At the end of 2007, they comprised Treasury securities ($2.4 trillion), agency bonds ($1.6 trillion), and corporate bonds ($2.8 trillion). The decline in interest rates implied an increase in the value of long-dated Treasury and agency bonds, but the rise in spreads (and the significant losses on MBS) implied losses for holders of US corporate bonds. On net, these losses likely exceeded the gains on treasury and agency bonds. On the other hand, US residents also incurred losses on their “foreign” bond portfolio, on several accounts:

  • declining emerging-market dollar bond prices;
  • the impact of the dollar appreciation on the value of US-held local-currency bonds;
  • the decline in corporate bond prices in Europe;
  • declining values of asset-backed securities (bonds issued by entities in the Cayman Islands but backed by U.S. mortgages, and bought by U.S. residents).

The net valuation losses incurred by US residents on these debt instruments may well exceed those incurred by foreign residents on US bonds.

Overall, net “valuation losses” on the US external position may well add up to something in the range of $1.5 trillion – and would be even higher if FDI is estimated at market value. To be sure, this is a very large figure, but, as the Figure 1 shows (shaded area), only a partial offset for the large US gains in previous years.

Which countries experienced the corresponding net gains on their net external position during 2008? The global decline in stock prices implies very significant declines in the market value of financial wealth across the globe. Countries that have a larger portion of their stocks held by foreigners (relative to their own holdings of foreign stocks) experienced net capital gains on their external position, even though their aggregate wealth declined. In absolute terms, the capital gains on portfolio equity instruments are going to be largest for the euro area (possibly in the order of $1 trillion), followed by the BRICs (some $200 billion each).

So how are the prospects for the US external position going forward? With regard to external borrowing, the US current account deficit is projected to decline significantly in 2009 and beyond, reflecting significant terms of trade gains (because of sharply lower oil prices), weak domestic demand, and dollar weakness (although the dollar has appreciated sharply over the past 6 months). Nevertheless, net external borrowing will imply – ceteris paribus – rising US net external liabilities over the coming years, and with a much reduced equity cushion the large negative debt position of the US looks more vulnerable, reinforcing the need for a significant correction in the US trade balance.

Editors' note: This column is a Lead Commentary on Vox's Global Crisis Debate page; see further discussion on Vox’s Global Crisis Debate page.

Disclaimer: This column reflects the views of the author and not necessarily those of his employer.


Curcuru, Stephanie, Charles Thomas, and Francis Warnock. 2008. Current Account Sustainability and Relative Reliability. September 2008. NBER Working Paper W14295.
Lane, Philip R and Gian Maria Milesi-Ferretti. 2008. Where Did All the Borrowing Go? A Forensic Analysis of the US External Position. IMF Working paper WP/08/28. IMF: Washington DC.

1. The equity-debt asymmetry would be even more pronounced if FDI were to be estimated at market value.



Topics:  Macroeconomic policy

Tags:  US, global economy

Deputy Director in the Research Department, IMF; Research Fellow, CEPR.