The IMF at 75: Reforming the global reserve system

José Antonio Ocampo 08 April 2019



The 75th anniversary of the Bretton Woods agreement is a good time to rethink the role the institutions created at the time should play in today’s world. In the case of the IMF, there are four central issues to consider:

  1. the global reserve system (the way international liquidity is provided); 
  2. managing the macroeconomic linkages among different economies, including the exchange rate system; 
  3. balance-of-payments crisis prevention and resolution, including in the first case rules for capital account management; and 
  4. improving governance of the international monetary system, to develop a more inclusive system and appropriate links between the IMF and regional and interregional monetary arrangements (i.e. the design of the Global Financial Safety Net). 

Reforms efforts in all of these areas should take into account the asymmetries that characterise the global economy, and particularly the vulnerabilities that emerging and developing countries (hereafter just ‘developing countries’) face. In this column, I look at the first of these issues. The broader agenda has been discussed by Eichengreen (2008), Kenen (2001), Williamson (2007, 2010), among others, and in my recent book (Ocampo 2017), from which this column borrows.

As we know, the original Bretton Woods monetary system collapsed in the early 1970s when the US unilaterally abandoned the dual gold-US dollar (hereafter just ‘dollar’) standard, the major developed countries failed to adopt a new stable system of exchange rate parities, and IMF members were unable to agree on a new international monetary system in the Committee of Twenty negotiations that took place in 1972–74 (Williamson 1977). The system that evolved, in an ad-hoc manner (as a ‘non-system’, to use the terminology of the 1970s), had a global reserve system essentially based on an inconvertible (fiduciary) dollar but open in principle to competition from other reserve currencies, and the freedom for each country to choose the exchange rate regime they chose, as long as they avoided ‘manipulating’ their exchange rates – a term that has never been clearly defined.

The basic deficiencies of this system have been identified in a sequential way in the global policy debate. The first, underscored by Keynes (1942-43), is the asymmetric adjustment problem – the strong pressure that deficit countries face to reduce their payment imbalances versus the weak pressure that surplus countries experience to do so. This generates a global recessionary effect during crises. 

The second problem is associated with the use of a national currency as the major international currency. This is widely known as the Triffin dilemma (Triffin 1961, 1968). The essential problem, according to his original formulation, is that the provision of international liquidity requires that the country supplying the reserve currency run balance-of-payments deficits, a fact that could eventually erode the confidence in that currency. Additional implications of the current dependence on the fiduciary dollar is that the stability of the system may be inconsistent with the monetary policy objectives of the major reserve-issuing country (Padoa-Schioppa 2011), and that the currency at the centre of the system has an unstable value.

The third flaw is the inequity bias generated by the need of developing countries to ‘self-insure’ against the volatility in external financing through the accumulation of foreign exchange reserves. This generates an inequity because reserves are invested in safe industrial countries’ assets – i.e. they are nothing else than lending to rich countries at low interest rates. Furthermore, it contributes to the generation of global imbalances.

There are two alternative ways to reform this system. The first and, in a sense, inertial solution is to let the system evolve into what it potentially could be, namely, a multicurrency arrangement. The basic advantage of this route is that it allows reserve holders to diversify the composition of their foreign exchange reserve assets, and thus to counteract the instability that characterises all individual currencies. However, diversification has been limited in practice, as the ‘network externalities’ have continued to favour the use of the dollar as the major international currency, largely as a result of the fact that there is no alternative to the market for US Treasury securities in terms of liquidity and depth (Prasad 2014). Furthermore, aside from diversification, a multicurrency arrangement would not address any of the other deficiencies of the system. 

This arrangement can be improved by introducing elements that enhance the capacity of exchange rates to contribute to correcting global imbalances and to provide a reasonable level of stability. The best suggestion in this regard would be a system of reference rates among major currencies, which was initially suggested by Ethier and Bloomfield (1975) and later on by Williamson (2007), among others. This implies that currencies, and particularly major currencies, would be subject to some form of managed floating around multilaterally agreed parities or bands. Interventions in foreign exchange markets and other macroeconomic policies would reinforce depreciation if the currency is perceived to be overvalued and appreciation if it is undervalued. Such an intervention rule would provide an implicit definition of what ‘manipulating’ the exchange rate means – i.e. encouraging exchange rate movements in the opposite direction to the agreed rate or band.

The second reform route would be to enhance the role of the only truly global reserve asset that the world has created, the Special Drawing Rights (SDRs), at least partially approaching the aspirations that were included in the reform of the IMF Articles of Agreement in 1969 of “making the special drawing right the principle reserve asset in the international monetary system” (Article VIII, Section 7, and Article XXII). 

A more active use of this instrument should preferably make SDR allocations in a counter-cyclical way (Camdessus 2000, Ocampo 2017: Chapter 2). This was the criterion followed in the last and largest issue in history, that of 2009, by the equivalent of 250 billion dollars. The amounts issued through time should, of course, be proportional to the global demand for reserves. Most estimates indicate that average allocations for the equivalent of $200-300 billion a year would be reasonable, but even this size of allocation would only marginally increase the share of SDRs in non-gold reserves. This indicates that SDRs would still largely complement other reserve assets. This is why this alternative would be complementary with a multicurrency standard.

A more active use of SDRs would mitigate several problems in the current system. First, the seignorage would accrue to all IMF members. Second, by issuing SDRs in a counter-cyclical way, it can contribute to reducing the recessionary bias associated with the asymmetric adjustment problem. Third, SDR allocations could reduce the need for precautionary reserve accumulation by developing countries. 

The most ambitious reform in this area would be to finance all IMF lending with SDRs, thus making global monetary creation similar to how central banks create domestic money. This would involve counter-cyclical allocations of SDRs that help fund counter-cyclical IMF financing. This would follow the proposals made by the IMF economist Jacques Polak (1979) of making the IMF a fully SDR-based institution. The simplest alternative would be to treat the SDRs not used by countries as deposits in (or lending to) the IMF that could then be used by the institution to lend to countries in need (Ocampo 2017: Chapters 2 and 7). This would require eliminating the division in the IMF between what ‘general resources’ and the SDR accounts (Polak 2005, Part II).

Following the discussions of the 1960s, there are also ways of including a ‘development link’ in SDR allocations, which would take into account the fact that developing countries have to accumulate reserves as ‘self-insurance’. Williamson (2010) has proposed allocating a certain proportion to developing countries (around 80%), and then assigning the shares in the allocation among individual developing and high-income countries according to IMF quotas. Many authors have also suggested that there could also be a broader use of SDRs in international capital markets (Cooper 2010, Eichengreen 2007). But, as the IMF has recently underscored, the market SDRs are less important than the more active use of them as reserve assets (IMF 2018). For these reasons, it may be better again to think of a mixed system in which national or regional currencies continue to play the major role in private transactions, and the SDRs continue to perform the functions of reserve asset and medium of exchange in transactions among central banks. A mixed system would be politically more acceptable for the issuers of reserve currencies, particularly to the US.

Under a system that mixes SDRs with a multicurrency arrangement, a substitution account should be created, allowing central banks to exchange for SDRs other reserve assets they do not want to hold. This alternative was suggested by Bergsten (2007) before the 2008-09 global crisis, going back to proposals that have been made since the 1970s. This instrument could also be seen as a transition mechanism of an ambitious reform effort (Kenen 2010). The essential issue is how to distribute the potential costs of this mechanism, the problem that blocked its adoption three decades ago.

The use of SDRs to finance IMF programmes would also help consolidate the reforms of the credit lines that have been introduced during the recent global crisis, particularly the creation of contingency credit lines (especially the Flexible Credit Line), including the Short-term Liquidity Swap recently proposed by the IMF (IMF 2017). It would also eliminate the need for the IMF to get financing from its members in the form of “arrangements to borrow” or bilateral credit lines. In fact, it would equally eliminate the need to make additional quota increases – though quotas would still have to be agreed to determine the size of access to IMF facilities as well as voting rights. 


Bergsten, F (2007), “How to solve the problem of the dollar”, Financial Times, 10 December.

Camdessus, M (2000), “An agenda for the IMF at the start of the 21st century”, remarks at the Council on Foreign Relations, February, New York.

Cooper, R (2010), “Does the SDRs Have a Future”, Harvard University. 

Eichengreen, B (2007), Global Imbalances and the Lessons of Bretton Woods, Cambridge: MIT Press.

Eichengreen, B (2008), Globalizing Capital: A History of the International Monetary System, 2nd edition. Princeton, NJ: Princeton University Press.

Ethier, W, and A I Boomfield (1975), "Managing the Managed Float", Princeton Essays in International Finance, 112, October. Princeton, NJ: International Finance Section, Department of Economics, Princeton University. 

International Monetary Fund (2017), “Adequacy of the Global Financial Safety Net –Considerations for Fund Toolkit Reform”, IMF Policy Paper, December.

International Monetary Fund (2018), “Considerations on the Role of the SDRs”, March 6th.  

Kenen, P (2001), The International Financial Architecture: What’s New? What’s Missing?, Washington DC: Institute for International Economics.

Kenen, P (2010), “Reforming the Global Reserve Regime: The Role of a Substitution Account,” International Finance, 13 (1).

Keynes, J M (1942-43), “The Keynes Plan”, Reproduced in J K Horsefield (ed.). The International Monetary Fund 1945-1965: Twenty Years of International Monetary Cooperation. Vol. III: Documents, International Monetary Fund, Washington, D.C., 1969, 3-36.

Ocampo, J A (2017), Resetting the International Monetary (Non)System, Oxford and Helsinki: Oxford University Press.

Padoa-Schioppa, T (2011), “The Ghost of Bancor: The Economic Crisis and Global Monetary Disorder”, in J T Boorman and A Icard (eds.), Reform of the International Monetary System: The Palais Royal Initiative, New Delhi: SAGE Publications, chapter 6.

Polak, J J (1979), “Thoughts on an International Monetary Fund based fully on SDR”, Pamphlet Series No. 28, International Monetary Fund, Washington, DC.

Polak, J J (2005), Economic Theory and Financial Policy: Selected Essays of Jacques J. Polak 1994-2004, edited by James M Boughton, M E Sharpe, Armonk, NY.

Prasad, E S (2014), The Dollar Trap: How the U.S. Dollar Tightened its Grip on Global Finance, Princeton, NJ: Princeton University Press.

Triffin, R (1961), Gold and the Dollar Crisis (revised edition), Yale University Press, New Haven.

Triffin, R (1968), Our International Monetary System: Yesterday, Today and Tomorrow, Random House, New York.

Williamson, J (1977), The Failure of World Monetary Reform, 1971-1974, New York: New York University Press.

Williamson, J (2007), Reference rates and the International Monetary System, Washington, DC: Peter G. Peterson Institute for International Economics, Policy Analysis, No. 82.

Williamson, J (2010), “The Future of the Reserve System”, Journal of Globalization and Development, 1 (2), article 15.



Topics:  Global governance

Tags:  IMF, global reserve system, SDRs, dollars, Bretton Woods, monetary system

Member of the Board of Directors of Banco de la República; Professor of Professional Practice in International and Public Affairs, Columbia University (on leave)


CEPR Policy Research