Dear Governor Velarde, dear Marc Uzan, ladies and gentlemen, 
I would like to thank the organisers for giving me the opportunity to share some thoughts with you on the changes in the international monetary system (IMS) and the role of monetary policy coordination.
History casts a long shadow over the IMS. Over the past three decades, however, the global economy has changed in decisive ways. Three aspects are particularly striking. First, financial factors have become an increasingly relevant source of cross-country interlinkages, with some evidence of cross-country de-linkages since the global financial crisis. Second, emerging economies have raised their clout, as reflected by the impact of their business cycles on global growth and the growing use of their currencies: the global economy has become multipolar. And third, monetary policy in advanced economies has – more recently – entered new territory by resorting to unconventional measures.
I would like to outline these shifts in the global economic landscape, and then elaborate on the implications for the IMS and monetary policy coordination.
Three structural shifts in the global economy
Globalisation, home bias and regionalisation
Clearly, the defining feature of the last three decades has been the dramatic strengthening of globalisation in both trade and finance. Cross-border trade in goods and services as well as investment has soared, driven by trade-liberalising policies, the lifting of capital controls as well as innovations in transport and information technologies.
The global financial crisis has led to a halt in this trend, perhaps only temporarily. But there is also the possibility of a step backwards, i.e. a re-emergence of financial home bias and a stronger drive towards regionalisation. Early evidence of this is the resurgence of the so-called “Feldstein-Horioka puzzle”, that is, a sharp rise in the correlation between savings and investments.  European banks have retreated from dollar-denominated lending.  Banks from emerging economies also have a stronger regional orientation, in particular in South-East Asia, Central America and the Commonwealth of Independent States.  Regulators have been tempted to ring-fence their domestic financial systems. Restrictions on intra-European cross-border liquidity flows or the US rules for foreign banking organisations may be seen as examples.
Such a “de-globalisation” scenario, if confirmed, could become self-validating and have profound effects on the degree of international risk-sharing, spillovers and the global allocation of savings, and therefore on global long-term growth.
Shifting economic centre of gravity
Since the early 2000s the world has also witnessed impressive growth rates in emerging economies, which have exceeded those in the advanced world. In particular the BRICS (Brazil, Russia, India, China and South Africa) have more than doubled their share of global GDP from less than 10% in 2000 to 21% in 2013; similarly, other low and middle-income economies have seen their share increase from 7% in 2000 to 12% in 2013. The rise of emerging economies is also illustrated by the eastward shift of the world’s economic centre of gravity, which is occurring at a faster speed than ever before in human history. And the shift towards a multipolar world is in turn reflected in changes to global institutions and fora, as evidenced by the switch in relative importance between the G7 and the G20, and the ongoing governance reforms at the IMF.
In line with the shift in economic centre of gravity, the currency constellation in the IMS may also be gradually changing. In particular in light of the remarkable growth of China’s economy and recent policy measures adopted by its authorities, the international use of the renminbi has gained momentum.  And China’s plans to liberalise its capital account as well as to increase exchange rate flexibility and develop domestic financial markets will enhance the role of the renminbi as an international reserve currency. Some analysts believe that, over time, it can reach international currency status, alongside the US dollar and the euro.  We would then enter an era of a tri-polar IMS.
Unconventional monetary policy
Before the global financial crisis the widespread adoption of explicit price stability objectives had been conducive to prolonged low inflation outcomes alongside economic growth. Major central banks were content to invoke the Tinbergen principle, aiming to achieve their primary objective of price stability through a single monetary policy instrument: short-term interest rates. The global financial crisis changed this setting abruptly. Pursuing central bank mandates in a zero lower-bound environment has meant that monetary policy needed to broaden the set of policy instruments at its disposal, including forward guidance and asset purchases by central banks. As a result of the latter, the balance sheets of all major central banks have increased significantly.
One of the major challenges in the years ahead will be the differentiated normalisation of unconventional monetary policies in advanced economies. Many have argued that in this process central banks in advanced economies should be mindful of the spillovers to the rest of the world.  There have even been calls for a fundamental strengthening of international monetary policy coordination. I would like to take this opportunity to elaborate on my views on this discussion.
The debate on international monetary policy coordination
There are different proposals on how the IMS should change. In particular, there have been several calls to establish a more formal framework of monetary policy coordination, or some kind of revised Bretton Woods system. Overall, I will argue that the forces which are at play today in the IMS, as described above, call for a strengthening of some of its features rather than an overhaul.
First, while the literature suggests that international monetary policy coordination may be instrumental in achieving a globally optimal solution, it generally concludes that the gains from coordination are small relative to an environment in which national policy-makers pursue optimal domestic policies. 
Second, it is not clear that the benefits from coordination increase monotonically with economic and financial integration.  On the one hand, financial integration intensifies the impact of foreign shocks on the domestic economy. On the other hand, it also improves diversification and insurance opportunities, hence mitigating the impact of foreign shocks. It is difficult to know ex ante which of these effects will dominate both in “normal” and “crisis” times. My suggestion would be that while the diversification effect dominates in “normal” times, in “crisis” times the effects from large adverse spillovers prevail. I will come back to this issue later.
There are also several practical challenges that make more explicit and binding forms of monetary policy coordination difficult. The first stems from political economy. Central banks operate under different mandates, time horizons, objectives and accountability arrangements. They are ultimately backed by their domestic fiscal authority and report to their domestic parliament (the euro area is no exception to this rule, even if sovereignty is shared locally). As a consequence, they cannot act as lenders of last resort other than in a discretionary way beyond the boundaries of their political constituencies.
In addition, economic cycles do not always coincide across countries, which creates tensions between what is needed for domestic purposes and what is optimal from an international perspective (the demise of the Louvre Accord is a classic example).
Finally, given that uncertainty is pervasive in economic policy-making, developing a common assessment of global risks and spillovers will always be difficult. This considerably complicates consensus-building concerning the design and enforcement of policies that countries should implement.
Take, for example, the pre-crisis debate on the risks stemming from global imbalances. The current account deficit in the United States was seen by some as problematic, as they believed it would eventually trigger a balance of payments crisis or a sudden and strong depreciation of the US dollar with global ramifications. Others instead argued that the current account deficit was not of concern given the international status of the US dollar; the positive valuation effects and the positive net foreign income balance were seen as evidence that the current account deficit in the US was sustainable. As a matter of fact, we did not experience a US balance of payments crisis, even though we did experience a spectacular global financial crisis.  What was the connection between global imbalances and the global financial crisis? What were the drivers of those global imbalances? The debate has still not been settled. Similarly, I would claim that the main obstacle today to international monetary policy coordination is the lack of consensus on (and maybe even of understanding of) the degree of slack in the labour and product markets of the major economies.
Reaching a consensus on the nature of spillovers is not straightforward either. Most models of international monetary policy coordination assume that policy-makers are able to identify the types of shock and how they are transmitted to the domestic and global economy. But this is often not the case in practice.
For instance, there have been marked differences in the views on the global impact of Federal Reserve announcements. On the one hand, some have argued that the Fed’s monetary policy is driving a global financial cycle.  On the other hand, others have argued that the key driving forces of capital flows are global risk and uncertainty, with a much looser connection to US monetary policy.  Another example of diverging views concerns the question of whether emerging economies have the tools to neutralise the spillovers originating from other parts of the global economy.
As a result of our poor understanding of the types of shock and their transmission, establishing which policy responses are the best for each country seems to be rather difficult. Internalising the effects of one’s monetary policy on other jurisdictions would be even more challenging. Suppose the Fed had a mandate to internalise the global impact of its decisions, which it does not have, would the financial stability impact of cheap dollar funding in emerging market economies have called for a later, or for an earlier, normalisation of US monetary policy? Ultimately, there is necessarily a high degree of subjective judgement involved. As I already suggested, crisis times such as 2008-2009 are the exception, since the spillovers are at the same time less ambiguous and more visible.
The scope of coordination
This does not imply that regular dialogues between central banks as well as exchanges of views during meetings at the IMF, the BIS, the G20 and at regional level, are not useful. In fact, I believe that it is quite the contrary. Global institutions and fora play an essential role for three reasons.
First, they achieve a better understanding and a common assessment of the global spillovers from domestic policy actions and the potential policy trade-offs. The discussion on capital flow management measures is a good example.  The IMF should continue to lend its analytical and impartial voice to this effort.
Second, even if the scope for coordination is limited in good times, they make coordination possible in bad times. Consider, for example, the establishment of a system of bilateral swap agreements among major advanced economy central banks, or the recent decision to establish such an agreement between the ECB and the People’s Bank of China.
Third, they prompt work at bodies such as the Basel Committee for Banking Supervision, the International Organization of Securities Commission or the Committee on Payments and Settlement Systems, to establish the standards and infrastructures and create a level playing field, which are necessary for global markets to function in a smooth and safe way. In so doing, they foster risk-sharing through decentralised markets, avoiding the need to coordinate in the first place. Against the background of the possible “deglobalisation” forces that I mentioned earlier, and given the ongoing transition to a more multipolar currency system, this work may become even more crucial.
To summarise, I am not convinced that the existing approach to monetary policy coordination should be fundamentally overhauled at the current juncture. Due to knowledge and legitimacy gaps, we should not believe that there can be a “global planner” which sets the rules or prescribes the monetary, prudential, fiscal and structural policies that each country shall pursue. However, countries can and should maintain the foundations for coordination to make global markets a safer place and allow for joint action in times of crises.
The evolution of the IMS
The IMS during the Bretton Woods era was strictly rules-based with fixed exchange rates and a gold standard. While appropriate in a world with capital controls and relatively modest trade flows, it proved to be too inflexible in the presence of large idiosyncratic shocks and growing internationalisation. The end of the Bretton Woods era gave rise to an IMS that is mostly based on principles, which are discussed in fora such as the G7, G20 or the OECD and BIS. In particular, one principle holds that multilateral financial safety nets should be available to countries facing large financial market disruptions caused by major external shocks.
Like any insurance-providing scheme, the IMS is incentives-based. The incentives for any economy to access the IMF’s credit lines, bilateral swap agreements or regional balance of payment support facilities depend on the conditionality embedded in each of these mechanisms, as well as on the degree of self-insurance acquired by this economy, e.g. through foreign exchange reserves. The problem here is that conditionality is not always explicit, nor is it consistent across instruments, providing a fuzzy set of incentives – and therefore a bias towards self-insurance. This may create arbitrage opportunities across facilities (in other words, a “race to the bottom” of conditionality, encouraging moral hazard) and, ultimately, a bias towards self-insurance, a sub-optimal way to deliver global stability. There is scope for improving the consistency of these facilities both horizontally across institutions and vertically between the global and regional level.
A different set of incentives stems from the surveillance carried out by the IMF and other institutions and fora. This has taken place since the demise of the Bretton Woods system at the bilateral level in the context of the IMF’s Article IV consultations and it has expanded recently at the multilateral level with the IMF’s Spillover and External Sector Reports as well as the G20 Mutual Assessment Process, not to mention regional exercises.
The resulting recommendations are however not binding and therefore constitute only a light form of multilateralism. Clearly, countries may be reluctant to submit to closer scrutiny, fearing that the recommended policies may not be in their national interest. Thus, better awareness and maintaining a spirit of multilateralism will be a crucial element in strengthening the effectiveness of surveillance and contributing to global stability. This requires institutions that are both legitimate and effective. In particular, lack of progress in global governance reform (reflecting the shift towards a multipolar world) would fuel a lack of trust in the IMS. It would encourage the ring-fencing of national systems and the re-nationalisation of savings, harming growth and jobs in all economies.
Let me conclude. The belief that the comfortable pre-crisis world could return is long gone. And I also believe that we should not aspire to turn the clock back. The global economy has changed fundamentally, and there is clearly a greater understanding of the role of cross-border linkages. The current IMS leaves countries with more domestic policy options to counter adverse idiosyncratic shocks than the inflexible system of the Bretton Woods era.
At the same time, in a highly interlinked global economy consensus-building concerning the transmission of spillovers is crucial for dealing with shocks that affect substantial parts of the global economy. Moreover, the IMS should also provide incentives to foster appropriate preventive action that renders economies more resilient to shocks, and provide the necessary standards and infrastructures to make global markets work better as well as make coordination possible when warranted. This is already a robust policy agenda.
I would like to thank Michele Ca’ Zorzi and G. Georgiadis for contributing to these remarks. I remain solely responsible for the opinions expressed herein.
For euro area countries the cross-country correlation between the savings and investment-to-GDP ratios fell continuously from 0.6 in the early 1990s to essentially zero in 2005; similarly, for non-euro area OECD countries it even fell from almost one to zero over the same time span. After the global financial crisis the correlation rose sharply to 0.6 for the euro area and 0.75 for non-euro area OECD countries.
See Ivashina, V., Scharfstein, D.S., Stein J.C. (2012), “ Dollar Funding and the Lending Behavior of Global Banks”, NBER Working Paper No 18528, November.
See Bank for International Settlements (2014), “ EME Banking Systems and regional financial integration” , CGFS Papers, No 51.
This has been particularly visible in the rise of the renminbi as an invoicing and settlement currency for China’s international trade: while in 2010 almost none of China’s trade in goods and services was invoiced in renminbi, in 2014 this figure has risen to almost 25%. Similarly, the issuance of renminbi-denominated (“dim sum”) bonds has grown by a factor of about 30, albeit from a very low level.
For a discussion of the international role of the renminbi, see Eichengreen, B., and Kawai M. (2014), “ Issues for renminbi internationalization: an overview," Asian Development Bank Institute Working Paper, No 454, January. On recent trends in the IMS, see the latest ECB Report on the International Role of the Euro, Frankfurt am Main, 16 July 2014.
See Rajan, R. (2013), “ A step in the dark: unconventional monetary policy after the crisis”, Andrew Crockett Memorial Lecture, BIS.
See in particular the seminal paper by Obstfeld, M., and Rogoff, K. (2002), “ Global implications of self-oriented national monetary rules,” Quarterly Journal of Economics, Vol. 117, No 2, pp. 503-35.
This may partly also depend on the model set-up characterised by limited policy trade-offs. See also B. Cœuré (2014), keynote address at the ECB-IMF conference on the “ International dimension of conventional and unconventional monetary policy,” Frankfurt am Main, 30 April 2014, forthcoming in a special issue of the Journal of International Money and Finance.
See B. Eichengreen (2014), “ A Requiem for Global Imbalances”, Project Syndicate.
See H. Rey (2013), “ Dilemma not trilemma: the global cycle and monetary policy independence”, Proceedings - Economic Policy Symposium - Jackson Hole, Federal Reserve Bank of Kansas City, pp. 1-2.
See Forbes, K. and Warnock, F. (2012), “ Capital flow waves: Surges, stops, flight, and retrenchment”, Journal of International Economics, vol. 88(2), pp. 235-251.
See IMF (2012), “ The liberalisation and management of capital flows: An institutional view”, IMF Policy Paper.