In the Wake of the Crisis: Leading Economists Reassess Economic Policy

21. The International Monetary System

International Monetary Fund
Published Date:
March 2012
  • ShareShare
Show Summary Details
Olivier Jeanne

This chapter about the international monetary system is structured around two themes—global financial imbalances and global financial safety nets.

Global Imbalances and Global Safety Nets

Many economists think that improving global financial safety nets will help address the problem of global imbalances. The idea is that global imbalances come in part from the accumulation of reserves by emerging-market countries trying to self-insure against volatile capital flows. Thus, better safety nets, in addition to their intrinsic merits, will reduce the need for self-insurance through precautionary savings and will mitigate the global savings glut.

How important for global imbalances is the precautionary accumulation of reserves? One answer is given in figure 21.1. This figure reports the current-account surplus of all the countries that had a surplus in 2003 to 2005 according to the World Economic Outlook database, excluding oil-exporting countries. (The oil exporters have been removed because they accumulate foreign assets for reasons that are specific to those countries and that we understand pretty well.) The data cover 2000 to 2015, so the last third of the graph is based on a forecast by the International Monetary Fund.

In this group, the “sudden stop” countries either had a sudden stop between 1995 and 2000 or benefited from a swap with the U.S. Federal Reserve or other central banks in 2008.1 These countries can be thought of as the natural customers of the global financial safety nets. The figure shows, in black, the share of the surplus that comes from those countries.

Figure 21.1Current account balances of surplus countries.

Source: World Economic Outlook (2010).

It also shows the contribution of China (in dark gray) and the rest (including Germany and Japan).

Figure 21.1 highlights two striking facts. First, the self-insurance motive accounts for a small and decreasing share of the current-account imbalances. Thus, improving global financial safety nets is important but should not be expected to reduce global imbalances significantly. The second striking fact is the size of China’s contribution to global surpluses, especially looking forward. The World Economic Outlook (WEO) predicts that China will account for more than 60 percent of the global surplus (excluding oil exporters). So if there is one topic to focus on for understanding global imbalances, it is China.

The Chinese Dilemma

In this chapter, I develop the thesis that the capital-account policies of China are key to understanding why the Chinese current-account surplus is very large. My analysis is based on a model I presented in a recent paper (Jeanne 2011), but this simple model is consistent with what many people think China is doing.

To understand how the model works, think of an open economy in which the government uses two policy instruments. First, the government uses capital controls to determine the volume of gross capital inflows, and second, the public sector of this economy accumulates a large stock of foreign assets. It stands to reason—and the model shows—that by playing with these two instruments, the country can control the level of net foreign assets for the country as a whole. Essentially, the controls on capital inflows prevent the accumulation of foreign assets by the public sector from being offset by capital inflows to the private sector.

This simple point has important implications. If the government controls the country’s net foreign assets, it controls the current-account balance, which is the change in net foreign assets. From there, it is easy to see that it also controls the trade balance and that the real exchange rate must be at the level that is consistent with the trade balance. The causality goes from the trade balance to the real exchange rate rather than from the real exchange rate to the trade balance.

This is not a completely conventional way of presenting Chinese policies because it does not involve the monetary authorities in an essential way. The more conventional view is that the Chinese central bank resists the nominal appreciation of the renminbi through foreign-exchange interventions. The conventional view, however, misses an important point: if the Chinese authorities only buy reserves, then they could resist only the nominal appreciation of the renminbi, not the real appreciation, which would still come about through domestic inflation.

Controls on capital inflows come into play by preventing the internal appreciation of the renminbi by repressing domestic demand. Those controls make it possible to maintain a sustained real undervaluation of the renminbi. But when capital controls are introduced into the model, it is possible to abstract from monetary policy and build a model in which the real exchange rate is determined entirely by capital-account policies. The central bank’s foreign-exchange interventions are the way that the Chinese public sector accumulates foreign assets in practice. But this is not essential: the foreign assets could be accumulated through a different channel at the margin without impairing the authorities from undervaluing the real exchange rate. Capital-account policies, in this view, are the linchpin of China’s exchange-rate policy.

If this view is correct, it has several implications. First, consider the vexing question of how to reduce the Chinese saving rate. My hypothesis is that the Chinese authorities induce a form of forced saving in China through their capital-account policies. One way of understanding this is that at the same time that the Chinese public sector is purchasing large volumes of foreign assets, it also is forcing the domestic private sector to buy large volumes of yuan-denominated assets (sterilization bonds and bank deposits) that cannot be sold to foreign investors because of the capital controls. The domestic income that is spent on those assets cannot be spent on consumption; it is a form of forced saving.

There are other explanations for the high Chinese saving rate, as well, including the lack of social insurance, aging, and financial underdevelop-ment.2 But if you believe these other explanations, then you have to believe that not much would happen if China relaxed its controls on capital inflows because the Chinese government is holding the stock of net foreign assets that the Chinese private sector wants to hold. If you think that relaxing the controls on inflows would lead to more inflows on a net basis, then you must believe in some version of the model that I have outlined.

Which explanation you believe has important implications for the structural reforms that would be required to increase Chinese demand. Based on the other explanations for high saving mentioned above, the required reforms should be to “increase social insurance, strengthen corporate governance, and implement reforms to increase access to credit for households and SMEs in China.”3 However, if the high Chinese saving rate is due primarily to its capital-account policies, then those structural reforms are unlikely to increase Chinese demand much. Instead, the focus should be on the capital-account policies of China and, more precisely, on giving foreign investors a larger access to Chinese financial and monetary assets.

The Chinese capital-account policies are closely related to the internationalization of the renminbi. There are important debates (which I do not have space to discuss here) about the transition from a dollar-based international monetary system to a multipolar world with three or four reserve currencies, perhaps including the renminbi (Eichengreen 2011). The Chinese authorities seem to have adopted the objective of internationalizing the renminbi, although the pace and the modalities of the transition are not clear. However, making a reserve currency of the renminbi would require giving foreign investors access to a substantial stock of yuan-dominated financial assets. That is, it would mean relaxing the controls on capital inflows and losing control over the real exchange rate.

The transition from a strictly controlled capital account to a less controlled one is fraught with difficulties. The concerns of the Chinese authorities, as I understand them, are twofold. The first concern is about the pace of this transition—to preserve the benefits of an undervalued renminbi in terms of export-led growth until an alternative engine of growth is up and running. The second concern is about the orderliness of the transition—to avoid a scenario with speculative inflows overflowing the capital-account barriers and leading to an abrupt appreciation of the renminbi, which could be very disruptive for the Chinese economy.

Those concerns are both understandable and legitimate, but there is a tension between the two. On the one hand, the Chinese authorities might want to err on the side of caution and implement a slow transition (“crossing the river by feeling the stones”). On the other hand, the slower the transition, the more foreign assets will be accumulated, the more pressure will build up, and the more difficult it will be to achieve an orderly exit. Therefore, an orderly exit strategy requires a measured—but not too measured—pace. How the Chinese authorities will resolve this dilemma is key for the future shape of the international monetary system.

The Rules of the Game for International Capital Flows

Should China be treated sui generis, or should it be included in a wider concept of the rules of the game for the global economic system? I would argue in favor of the latter—more precisely, that it makes sense to put the case of China in the context of a multilateral framework for the management of capital flows.

The status quo is characterized by an indefensible asymmetry between strong international rules for trade in goods (with the World Trade Organization) and a quasi-absence of international rules for international trade in assets. This asymmetry is indefensible because (as I show in Jeanne 2011) any country can use capital-account restrictions to manipulate its real exchange rate in a way that reproduces the effects of a combination of tariff and subsidy. What sense does it make to discourage trade distortions while putting no restriction on capital-account policies?

This does not mean that all capital controls are bad. There is a strong case for emerging-market economies to use prudential controls to deal with booms and busts in capital flows.4 There have been calls for developing a code of good practices with regard to prudential capital controls that would increase the predictability and reduce the stigma associated with the use of such policies. And the International Monetary Fund has been asked by its shareholders to look into the question of whether there is a need for globally agreed-on rules of the road for the management of capital flows (Strauss-Kahn 2011).

If one is going to develop a code of good practice for prudential controls on capital flows, it would make sense to take an additional step forward and also think about using the code to define, by exclusion, the capital controls that cannot be justified on prudential grounds, are presumably distortive, and should be relaxed (Jeanne, Subramanian, and Williamson 2011).

There are several advantages from focusing the multilateral discussion on capital-account policies. The discussion would be focused on the policy instruments rather than on the outcomes. As I understand it, the multilateral assessment program conducted by the Group of Twenty to reduce global imbalances relies on targets (or indicators) for outcome variables such as current-account balances. This is problematic because economic theory does not give us models that we trust enough to determine the appropriate (undistorted) levels of variables such as current-account balances or the policy instruments that should be used to achieve targets for such variables. This also applies to real exchange rates. A more direct approach that focuses on policy instruments seems preferable.

To sum up, there might be considerable (although not immediate) benefits from trying to agree on rules of the game that recognize the link between free trade in goods and a certain degree of freedom in trade in assets—that full participation in the world trading system should be premised on a certain degree of freedom in trade in assets.

Global Financial Safety Nets

The motivation for reforming global financial safety nets is well known. Emerging-market countries did not rely on the existing multilateral arrangements (such as the IMF or the Chiang Mai Initiative) to obtain international liquidity in the fall of 2008 and relied instead on swaps with the U.S. Federal Reserve and other central banks. The question is how the multilateral arrangements for international liquidity provision can be improved. This question is largely orthogonal to global imbalances (as I argued above), but it is nonetheless important for emerging-market countries and has received a lot of attention under the Korean chairmanship of the G20.

Table 21.1International reserves
Reserves September 2008$240 billion$206 billion$169 billion$99 billion
Federal Reserve swap30 billion30 billion30 billion30 billion

The answer to one question—why did the swaps make a difference?—really matters for how we design global financial safety nets. As shown in table 21.1, the four emerging-market economies that received swaps from the U.S. Federal Reserve—Korea, Brazil, Singapore, and Mexico—already had reserves that were a multiple of those swaps. For example, Korea had more than $200 billion in reserves when it received a $30 billion swap. Why would a $30 billion swap make a difference for a country that has $240 billion in reserves?

I can think of four hypotheses, which are listed below:

  • Psychological threshold The swaps made it possible to maintain the headline level of reserves above a psychological threshold. For example, there were reports that the Korean authorities believed that reserves should be maintained above $200 billion to maintain market confidence.

  • Seal of approval The Fed signaled that it had a positive view of the prospects for the Korean economy, which in turn improved market expectations.

  • Lender-of-last-resort The Fed, as the true lender of last resort for dollars (the creator of the currency), was implicitly committing itself to lend further if necessary. The difference from the previous explanation is that the signal was about the Fed rather than about Korea.

  • Minor part of a policy package The swaps did not really matter. They were components of a larger set of policies that restored a measure of confidence at the center (Wall Street).

It is difficult to distinguish empirically between these views, but I believe more in the final two explanations than in the first two. That is, the Fed swaps mattered more than the $30 billion that was provided to Korea because the Fed is a true lender of last resort and because the support to Korea was part of a larger package that restored a measure of confidence at the center. If this is correct, then the fact that the resources were provided by a central bank—the creator of the currency—was essential to why the swaps had a positive effect. That would mean that for the global financial safety nets to be as effective as possible, they should involve central banks and institutionalize the swaps in some way.5 If it is politically unfeasible for the major central banks to make such commitments, emerging-market countries will have to rely more on prudential capital controls to reduce their exposure to capital-flow volatility.


A sudden stop is defined as a fall in the financial account of more than 5 percent of gross domestic product. The list of sudden stops comes from Jeanne and Ranciere (2011). The list of countries that benefited from swaps can be found in Aizenman, Jinjarak, and Park (2010, table 2).

For the lack of social insurance, see Chamon and Prasad (2010). For financial underdevelopment, see Caballero, Farhi, and Gourinchas (2008) or Song, Store-sletten, and Zilibotti (2011).

This could be done through the International Monetary Fund, as proposed by Truman (2008).

    Other Resources Citing This Publication