The world economy hangs in a precarious balance. Because of a shortage of domestic saving relative to investment, the U.S. is running large current account deficits of 6% of GDP. These account for three-quarters of the world sum of current account deficits or net global saving. Private capital inflows since 2002 have proved insufficient to finance U.S. external deficits. The shortfall, which has been accompanied by a depreciation of the dollar, has been made up largely by Asian central banks. This arrangement cannot be sustained indefinitely: if the U.S. saving shortage is not corrected, the dollar could depreciate further.
A major counterpart of the huge U.S. external deficits has been external surpluses in East Asian countries. These surpluses emerged since 1997-98, at the same time that large U.S. deficits developed. East Asian surpluses account for about 45% of both the world sum of current account surpluses and of the U.S. global deficit. Whereas the primary responsibility rests with the U.S. to resolve her own large deficit, East Asia could play a significant role in the adjustment process.
The purpose of this policy recommendation is to identify the roles that both the U.S. and East Asian countries can play in resolving the current global imbalances. To do this, we first discuss the nature (Section II) and sustainability (Section III) of the imbalances. We then recommend necessary external adjustment policies for the U.S. (Section IV). Finally, we consider the role that East Asia can play, acting in her own interest, in rebalancing the global economy (Section V) and draw conclusions (Section VI ).
II. The Nature of the New Global Imbalances
A. U.S. Current Account Deficits
U.S. external imbalances were driven by a private investment boom during the late 1990s, a drastic reversal of the fiscal balance from surplus to large deficit after the recession of 2000-01, and high consumer spending from 1997 to the present. All three have contributed to the shortage of domestic saving relative to investment (i.e., the I-S imbalances) and thus to current account deficits now equaling 6% of GDP. Private capital inflows between 1997 and 2001 were sufficient to finance U.S. current account deficits, and the real effective exchange rate of the dollar appreciated 17%. Private inflows since 2002, however, have fallen short of current account deficits and the real effective exchange rate of the dollar depreciated 17% between January 2002 and December 2004. The dollar is now close to its average value over the past quarter century.
B. East Asian Current Account Surpluses
While the U.S. current account balance deteriorated after 1997, current account balances in East Asia have either moved from deficit to surplus (in the case of crisis-hit economies) or from surplus to larger surplus (in the case of non-crisis hit economies). The Asian Capital Account Crisis of 1997-98 was followed by lower investment/GDP ratios than saving/GDP ratios in crisis-hit countries until now, leading to large current account surpluses that stand in sharp contrast to the deficits of the pre-crisis period. Non-crisis Asian economies such as Singapore and Taiwan also exhibited similar drops in investment/GDP ratios relative to saving/GDP ratios, causing current account surpluses to expand further. In contrast, China continued to increase both investment/GDP and saving/GDP ratios, maintaining relatively small surpluses as a share of GDP.
Exchange rates in crisis-hit East Asian economies, after initially collapsing by 50% or so during the crisis, have been kept on average 15-20% below pre-crisis levels through central bank intervention in foreign exchange markets. Exchange rate experience in non-crisis East Asian economies has been similar except for the case of the Chinese currency, which remained unchanged at 8.2 RMB per dollar from 1994 until 20 July 2005. Even larger intervention in the region began in 2002 when the aforementioned shortfall of private capital inflows relative to U.S. deficits triggered the recent dollar depreciation. Since 2002, about 40% of U.S. external deficits have been financed by foreign (primarily Asian) central banks.
The objectives of Asian central banks in accumulating large quantities of reserves have been: 1) to be prepared for another capital account crisis which would drain foreign reserves due to massive reversals of short-term capital flows, and 2) to maintain competitive exchange rates in order to sustain the export-oriented thrusts of their economies. The first objective was more important immediately after the crisis, and the second objective has been more important in recent years.
III. The Sustainability of the New Global Imbalances
The aforementioned new global imbalances are precarious because they may prove unsustainable. The sustainability question can be divided into two parts: 1) the sustainability of large U.S. deficits, and 2) the sustainability of reserve accumulation by Asian monetary authorities.
A. The Sustainability of U.S. Current Account Deficits
As the U.S. runs current account deficits it accumulates debt against the rest of the world. The sustainability of these deficits can be judged by considering whether foreign investors will willingly hold the amount of debt that the U.S. must issue to finance her continued deficits. This analysis should take into account dynamic changes in interest rates and risk premia on U.S. liabilities. Assuming that U.S. current account deficits continue to equal 6% of GDP and that the underlying U.S. nominal GDP growth rate equals 5% and that the net interest rate on net foreign liabilities remains negligible as it is now,  net external debt will eventually reach 120% of GDP (6%÷5%) in the long run. 
These assumptions also imply that after 10 years net external debt will reach 65% of GDP, representing 18% of the wealth of the rest of the world. Such a level might be willingly held by international investors and thus sustainable thanks partly to the key currency status of the dollar.
However, interest rate dynamics get progressively worse over time. At present, despite the fact that U.S. net external liabilities equaled 29% of GDP in 2004, net international investment income remains slightly positive for the U.S. and thus the net interest rate remains negative (hence favorable for the U.S.). This is because, up until now, investment income earned by U.S. residents mainly from equity investments including FDI abroad has exceeded investment income paid by U.S. residents mainly on U.S. bonds held by the rest of the world. Even assuming that the current rates of return on U.S. investments abroad and foreign investments in the U.S. remain unchanged, the massive amount of U.S. liabilities that are projected to accumulate over the next 10 years will raise the net interest rate to almost 1%. In this case the long run debt/GDP ratio would be 150% of GDP (6%÷(5%-1%)), equal to 40% of the net wealth of the rest of the world.
Notwithstanding the key currency status of the U.S. dollar, it is implausible that foreign investors would hold such high shares (as exemplified above at 40%) of their total wealth in U.S. assets without receiving a higher rate of return, especially if they feared that the dollar would depreciate. If after 10 years the required return on U.S. assets rose by only 1%, the eventual debt/GDP ratio would be greater than 300% of GDP (6%÷(5%-3%)). This would exceed the total wealth of the U.S.! Thus current borrowing levels in the U.S.will at some point prove to be unsustainable.
If current account deficits remained below 3% of GDP, on the other hand, the eventual debt/GDP ratio would remain below 60% of GDP in the long run, which should be sustainable, particularly given the key currency status of the U.S. dollar.
B. The Sustainability of Reserve Accumulation by Asian Central Banks
Foreign reserve accumulation by Asian monetary authorities increases base money and hence creates excess liquidity in the banking system. This in turn increases the money supply and exacerbates inflation. To offset this, central banks in the region have engaged in sterilization policies. Sterilization involves selling government bonds or central bank bills to keep the monetary base unchanged and to mop up excess liquidity in the banking system.
Sterilization policy in East Asia has so far been largely successful in preventing money supply growth rates from accelerating. The CPI inflation rate has also remained low, between 1 to 3 percent per year.
However, there are still several difficulties associated with sterilization operations. First, they cause commercial banks to hold more and more central bank bills, eroding bank profitability and interfering with the allocation of credit through the banking system. Second, sustained sterilization may fail in the long run because, as the stock of central bank bills or government securities grows large, interest rates on these instruments may have to rise in order to induce investors to willingly hold them. Higher interest rates would then attract further capital inflows and defeat the purpose of sterilization (the self-defeating hypothesis). Third, continued accumulation of U.S.Treasury securities (external reserves) results in an increasingly inefficient allocation of resources since both private and social rates of return would be much higher for investments in the domestic economy.
IV. Necessary External Adjustment Policies in the U.S.
How should the U.S. respond to her own unsustainable external imbalances? Since the large U.S. current account deficits originated from her own I-S imbalances and account for three-quarters of total world deficits as already mentioned, the primary responsibility for resolving global imbalances rests on the U.S. and her willingness to increase national saving. Increasing national saving would involve fiscal consolidation and also policies to reduce consumption in order to raise household saving rates. However, these absorption-reducing policies would be recessionary, particularly if such policies alone were assigned to external adjustment, and hence should be offset by switching policies to stimulate net exports. Therefore a dollar depreciation, which increases net exports by changing the terms of trade between tradables and non-tradables in favor of the former, together with absorption-reducing policies, could help to both achieve and maintain external and internal equilibria (i.e., a sustainable current account balance and full employment with low inflation). If the dollar depreciated alone without any absorption-reducing policies, inflation could accelerate. Thus the optimal policy mix for the U.S would involve fiscal consolidation combined with a depreciation of the dollar.
If fiscal consolidation were not undertaken, however, market forces could drive the dollar down in response to the rising share of U.S.assets in international investor portfolios (i.e., the aforementioned unsustainable U.S. current account deficits). The weaker the fiscal consolidation initiated, the more the dollar could fall. While this depreciation process might well be benign and gradual, there is a risk that it could be sudden and precipitous.
Many studies indicate that, to reduce the U.S. current account deficit by 1% of GDP, the dollar must depreciate by around 10%. Thus, if fiscal consolidation is not undertaken, a 30% reduction in the value of the dollar would be required to reduce the U.S. current account deficit from its current value of 6% of GDP to the aforementioned sustainable level of 3% of GDP. Thus a 30% depreciation can be taken as a working hypothesis when considering policy recommendations for the risky scenario. This should not be viewed as a prediction, nor can anyone reliably forecast the timing or speed of a dollar depreciation.
V. East Asia's Role in Resolving the Current Global Imbalances
A. Basic Rationales for Concerted Appreciations
How should East Asian countries respond to the risky scenario of a 30% depreciation of the dollar, as implied by the aforementioned working hypothesis?
Concerted action to maintain mutual exchange rate stability in the region would be helpful for the following reasons:
Since intra-regional trade accounts for about 55% of total trade, concerted increases of East Asian currencies against the dollar would attenuate effective exchange rate changes in the region. This in turn would mitigate the recessionary impact if Asian currencies appreciated against the dollar.
Exchange rate stability in the region would encourage continued FDI flows and provide a stable backdrop for regional production and distribution networks which have been the basis for the high intra-regional trade share. By allowing fragmented production blocks to be allocated across countries based on comparative advantage, these networks have acted as an engine of growth. Of course, the comparative advantage of individual countries and thus these production networks themselves will change dynamically over time.
It would help overcome prisoner's dilemma problems that arise because the fear of losing competitiveness relative to Asian trading partners sometimes prevents countries in the region from allowing their currencies to appreciate. Ways of overcoming these coordination problems are discussed in Section V B.
We briefly discuss intra-regional trade patterns and prisoner's dilemma problems below before considering the appropriate policy mix for East Asia.
1. Triangular Trading Patterns in East Asia
A high degree of economic interdependence in East Asia, with intra-regional trade accounting for 55% of total trade, is based upon intricate production and distribution networks in the region that have led to triangular trade patterns.  Japan, Korea, Taiwan, and multi-national corporations in ASEAN produce sophisticated technology-intensive intermediate goods and capital goods and ship them to China for assembly by lower skilled workers. The finished products are then exported to markets throughout the world. These processed exports account for 55% of China's total exports. The lion's share of China's processed exports is from FDI enterprises. Trade within these networks can be classified as vertical intra-industry trade (VIIT).
VIIT differs both from the exchange of final goods emphasized by traditional trade theory for vertical inter-industry trade between the North and the South (e.g., between capital goods and apparel) and for horizontal intra-industry trade between the North and the North (e.g., between two differentiated types of automobiles). VIIT allows the production processes of an industry (e.g., the electronics industry) to be split into fragmented production blocks that can be located in different countries. Production blocks are allocated across developing, emerging, and developed economies in the region based on comparative advantage as determined by relative endowments of capital, skill, and labor and by physical and institutional infrastructure. VIIT has led to large efficiency gains and helped to make East Asia as a whole the manufacturing center of the world.
Because of these trading networks, Chinese value added in processed exports is small (20%) relative to the predominant costs of the intermediate goods imported from the rest of Asia. Thus a unilateral RMB revaluation would not affect much the dollar costs of processed final products and hence China's trade surplus. Furthermore, in the triangular trading patterns, China imports sophisticated intermediate goods essentially from East Asia but not from the U.S. and exports final products all over the world, including to the U.S.Hence, these triangular trading patterns imply that bilateral trade imbalances between China and the U.S. would remain large even if China's global trade surplus disappeared. One policy implication that follows is that it is totally inappropriate to demand a unilateral RMB appreciation simply because America's bilateral deficit with China is large.
2. Complementary and Competitive Trade Relations in East Asia
Because Asian countries compete with each other both domestically and particularly in third markets, they may not follow regional partners in exchange rate appreciations. As implied by the triangular trading patterns and by VIIT, relatively more developed East Asian countries export large amounts of technology-intensive capital and intermediate goods, while relatively less developed countries in the region export large amounts of low skill-intensive consumer goods. Therefore there is essentially a complementary relationship between China and less developed Asia on the one hand and developed Asia on the other hand in sophisticated intermediate and capital goods trade. Furthermore, there is a complementary relationship between China and MNCs located in ASEAN that export sophisticated technology-intensive parts and components there for processing. In contrast, there is largely a competitive relationship between Chinaand less developed Asian countries in labor-intensive consumer goods trade. Existing studies indicate that if one Asian country's exchange rate appreciates relative to other Asian countries' exchange rates, that country's exports (particularly of labor-intensive consumptions goods) to third markets will fall.
All in all, elements of competition exist in any international trading relationship and hence fear of losing competitiveness relative to other countries may prevent individual countries from allowing their currencies to strengthen. This may explain the unwillingness of some countries in the region to allow their currencies to appreciate unilaterally. There could thus be a collective action problem or coordination failure: all countries would be better off if their currency values increased together but individual countries sometimes resist such increases because they are suspicious of whether other countries will allow similar increases.
B. The Appropriate Policy Mix for the Region
In view of 1) the high ratio of intra-regional trade to total trade, 2) the intricate production and distribution networks, and 3) the need to overcome coordination failures, concerted action to maintain mutual exchange rate stability among Asian currencies could be beneficial. One form of concerted action would be for all countries in the region to appreciate more or less jointly against the dollar under the aforementioned working hypothesis. However, this concerted appreciation would not sufficiently take into account individual differences among economies (e.g., policy options, macroeconomic variables, trade/GDP ratios, and structural factors) and would also ignore the actual appreciations of some Asian currencies since 2002. How can the benefits of concerted action be reconciled with the need to accommodate differences in individual economic conditions?
Since concerted action is aimed at mutual exchange rate stability, currencies that have not appreciated while the dollar fell between 2002 and 2005 should catch up with currencies that have already appreciated. This catch up would be achieved more easily if countries in the region adopted more flexible exchange rate regimes.
More flexible regimes could be characterized by two elements: 1) a multiple currency basket-based reference rate instead of a dollar-based central rate, and 2) a wider band around the reference rate.
These two elements would provide policy makers with greater flexibility in managing the speed and magnitude of any necessary appreciation while still taking into account their own individual economic conditions.
A free float would cause exchange rates to more accurately reflect market fundamentals. However, given the shallow and narrow domestic capital markets in some East Asian economies, a free float for some countries would generate excessively volatile exchange rates and harm economies in the region that are highly exposed to fluctuations in international trade.
We thus advocate, neither a free floating regime nor a fixed dollar peg but rather greater exchange rate flexibility in the context of a multiple currency basket-based reference rate with a band. If individual countries adopted greater flexibility in this way, a dollar depreciation under the working hypothesis would tend to produce appreciations across the region and keep mutual real effective exchange rates among Asian currencies relatively stable.
Greater flexibility would benefit China for a couple of reasons. First, China will begin to increasingly face de facto capital account convertibility with the free entry of foreign banks and other financial institutions for renminbi-based financial activities due to the terms of her WTO accession commitments. In this case wider bands would enable her to maintain greater monetary policy autonomy in the face of volatile capital flows. Second, increased flexibility would provide Chinese banks and traders with experience in managing exchange rate risk and allow Chinese forward markets and related infrastructure to develop quickly. At the same time China should purposefully build needed financial institutional infrastructure (e.g., prudential regulation, supervision, accounting, a credit culture, etc.) as quickly as possible so that she can accommodate greater flexibility in the very near future.
If Asian countries allowed their currencies to appreciate instead of accumulating reserves, it would have a recessionary impact. This could be offset by appropriate macroeconomic and structural policies. Switching policies such as exchange rate appreciations thus need to be combined with absorption-increasing policies.
Absorption-increasing policies include employing fiscal and structural policies to build both physical and human infrastructure (particularly in rural areas) and using deregulation to promote competition and productivity growth in the non-tradable sector. These policies could promote production for domestic markets and thus rely more on domestic markets rather than exports to create jobs.
Combining expenditure-increasing policies with expenditure-switching policies would thus be the appropriate policy mix for Asian countries that had previously accumulated massive reserves. On the one hand, without exchange rate appreciations, policies aimed simply at increasing domestic demand could produce overheated economies. On the other hand, without policies to increase domestic demand, exchange rate appreciations would be contractionary. Only by combining these two, namely by implementing an appropriate policy mix, could Asian economies move away from excessive reserve accumulation and simultaneously achieve external and internal equilibria in their own interests. These policies would also contribute to easing global imbalances, thus providing a harmonized way of advancing regional and international interests.
The present global imbalances cannot be sustained indefinitely. The primary step necessary to resolve them is for the U.S. to increase domestic saving. Absent I-S imbalance corrections in the U.S., the dollar could depreciate substantially due to the excessive amount of dollar assets in the portfolios of international investors as explained in Section IV. This decline could be gradual and benign, but it could also be sudden and precipitous.
In the face of an incipient depreciation of the dollar as a risk scenario, Asian countries ought to engage in concerted action to keep exchange rates as stable as possible given their individual economic differences. This could be done if countries with less flexible exchange rate regimes adopted more flexible regimes characterized by a multiple currency basket-based reference rate and a reasonably wide band. The challenge for East Asian policy makers is to harmonize exchange rate policies in this way.
Policy coordination to stabilize effective exchange rates mutually in the region through selecting more flexible exchange rate regimes would provide several advantages. First, concerted exchange rate changes against even a relatively large drop in the dollar's value would reduce the magnitude of “effective" exchange rate increases in East Asia, since about 55% of trade is intra-regional. This in turn would minimize the deflationary effect of a rise in the value of Asian currencies against the dollar. Second, exchange rate stability would facilitate the flow of FDI and preserve production and distribution networks in the region that have exploited comparative advantage and functioned as an engine of growth. These networks will change, though, as the comparative advantage of countries change over time. Third, it should help overcome prisoner's dilemma problems that prevent individual countries from allowing their exchange rates to rise unilaterally, even when joint increases would be mutually beneficial.
Exchange rate increases should be accompanied by absorption-increasing policies. These would offset the recessionary impact of the currency appreciations and also be consistent with the new economic strategy in East Asia of encouraging domestic demand rather than relying excessively on net export expansion.
In coordinating exchange rate policy a regional forum would be helpful. A regional forum could develop the needed surveillance mechanisms and peer pressure in the context of exchange rate coordination policies aimed at stabilizing mutual exchange rates among East Asian economies, which would in turn promote continued economic integration.
The coming East Asian Summit in Kuala Lumpur should be an appropriate venue to begin discussing policy options along the lines spelled out here, in order to advance the interests of countries in the region.
A1. Appendix: U.S. Current Account Deficit Sustainability
As the U.S. runs current account deficits it accumulates debt against the rest of the world. One way to analyze the sustainability of these deficits is to calculate how much U.S. debt must eventually be held in the portfolios of international investors. This analysis should take into account the evolution of interest rates and risk premia on U.S. liabilities held by international investors. In order to understand the dynamic process we start with the following simple equation to calculate the eventual value of U.S. liabilities relative to U.S. GDP in the long run:
(1) n* = c/ (g-r)
where n* is the eventual level of net external debt relative to GDP in the long run, c is the current account deficit relative to GDP, g is the nominal growth rate of the economy, and r is the net nominal interest rate on net U.S. external debt. Assuming that the U.S. current account deficits continue to equal 6% of GDP and that the underlying U.S. nominal GDP growth rate equals 5% and that the net interest rate on net foreign liabilities remains negligible as it is now, equation (1) implies that net external debt will reach 120% of GDP (6%÷5%). In addition, it can be shown that under these assumptions, net external debt will reach 65% of GDP after 10 years.
U.S. net liabilities of 65% of U.S. GDP, which represents 18% of the wealth of the rest of the world, might be willingly held by international investors and thus sustainable. The fact that the U.S. dollar is a key currency implies that there is extra demand for U.S.assets for transactions and precautionary purposes. Given the hegemonic position of the U.S. and the key currency position of the dollar, investors may be willing to hold this much U.S. debt.
The problem is that the interest rate dynamics get progressively worse over time. At present, despite the fact that U.S. net external liabilities equaled 29% of GDP in 2004, net international investment income remains slightly positive for the U.S. and thus the net rate of return (r) on net foreign liabilities remains negative (hence favorable for the U.S.). This is because, up until now, investment income earned by U.S. residents from abroad has exceeded investment income paid by U.S. residents to the rest of the world. This has happened because U.S. assets from abroad are largely in the form of equities (including FDI), while U.S. external liabilities are largely in the form of fixed income assets (including Treasury bills purchased by foreign central banks). Thus the average return received by U.S.investors (3.9% in 2004) exceeds the average return received by foreign investors (2.6% in 2004) because the U.S. is holding riskier and more profitable assets.
Although r has been negative up until now, this will not be the case as the rest of the world accumulates more U.S. liabilities. Even assuming that the rate of return earned by the U.S. on foreign liabilities continues to exceed the rate of return earned by the rest of the world on U.S. liabilities by the existing margins as mentioned above, the fact that the rest of the world is accumulating so many U.S. liabilities implies that after 10 years the net interest rate r on net liabilities will equal almost 1%. In this case the steady state debt/GDP ratio implied by equation (1) would be 150% of GDP (6%÷(5%-1%)). To put this number into perspective, if the rest of the world currently held U.S. assets worth 150% of U.S. GDP they would be holding 40% of their total wealth in U.S. assets.
It is implausible that even the extra demand for U.S. assets due to the key currency status of the dollar would be enough to induce foreign investors to hold very high shares (as exemplified above at 40%) of their total wealth in U.S. assets without receiving a higher rate of return. This would be true a fortiori if investors required a risk premium to guard against the chance that the dollar would depreciate. If after 10 years the required return on U.S. assets rose by only 1% from its current level of 2.6%, the net interest rate r on net foreign liabilities would exceed 3%. In this case the eventual debt/GDP ratio implied by equation (1) would be greater than 300% of GDP (6%÷(5%-3%)). This would exceed the total wealth of the U.S.! Thus current borrowing levels in the U.S. will at some point prove to be unsustainable.
If current account deficits remained below 3% of GDP, on the other hand, equation (1) implies that the debt/GDP ratio would remain below 60% of GDP. 60% of U.S. GDP represents 15% of the wealth of the rest of the world. Foreign investors may be willing to hold this much of their wealth in U.S. assets without requiring an increase in the rate of return on dollar assets, given the hegemonic position of the U.S. and the role of the dollar as a key currency. In this case, the negative interest rate dynamics discussed above would be less of a concern. Thus, U.S. current account deficits below 3% of GDP may prove sustainable.
 In the paper, we take East Asia to include ASEAN countries plus China, Japan, and Korea.
 See Analytical Appendix, Figure 4.
 The net nominal interest rate r is defined as net international investment income divided by net U.S. external debt. Net international investment income is the difference between international income received on gross U.S. assets held abroad and that paid on gross U.S. external debt.
 The following equation can be used to calculate the eventual value of U.S. liabilities relative to U.S. GDP in the long run: n* = c/ (g-nir), where n* is the eventual level of net external debt relative to GDP in the long run, c is the current account deficit relative to GDP, g is the nominal growth rate of the economy, and nir is the net nominal interest rate on net U.S. external debt. Given the assumptions in the text, it implies that U.S. net external debt will reach 120% of GDP (6%÷5%).
 See the Appendix to this report (Section A1).
 See the Appendix to this report (Section A1).
 The importance of increasing national saving is reinforced by the fact that the dollar in 2005, after having depreciated since 2002, is more or less at its long run average and thus not greatly misaligned. Nevertheless, the current account deficit remains huge. This implies that U.S. external imbalances are being driven by shortfalls of saving relative to investment in the U.S. and not by overvalued exchange rates
 Please see the Analytical Appendix, Section IV A, p. 17.
 FDI flows in East Asia have played an important role in strengthening the international production networks, reducing costs, and transferring technological know how (See Analytical Appendix, Section IV B 2, p. 20).
 See Analytical Appendix, Section IV B 4, p. 23.
 The net nominal interest rate r is defined as net international investment income divided by net U.S. external debt. Net international investment income is the difference between international income received on gross U.S. assets held abroad and that paid on gross U.S. external debt.
 The net interest rate r is (rUSROWUS - rROWUSROW)/(ROWUS - USROW), where rUS is the average rate of return that foreign investors earn on U.S. assets, ROWUS is the gross holding of U.S. assets by the rest of the world, rROW is the average rate of return that U.S. investors earn on foreign assets, and U.S.ROW is the gross holding of foreign assets by U.S. citizens. If USROW remains at its current value of 27% of U.S. GDP (YUS), then in 10 years ROWUS will equal 0.92YUS (0.27 YUS + 0.65YUS ). If rUS increases from 2.6% to 3.6%, then r would equal (0.036*0.92YUS - 0.039*0.27YUS )/0.650YUS . This expression equals 3.5%.
 Masaru Yoshitomi is President and Chief Research Officer at RIETI. Li-Gang Liu and Willem Thorbecke are Senior Fellows at RIETI.
 Listed at the end of this report. Extremely valuable and constructive written comments by the Working Group members are incorporated.