Exit strategies: the international dimension
Intervention by Lorenzo Bini Smaghi, Member of the Executive Board of the ECB,
Euro50 Group Meeting
Is there still a paradigm for monetary policy today?
Paris, 20 November 2009
I would like to thank the organisers for inviting me to discuss a very topical issue – exit strategies. It is so topical, in fact, that nearly everything that needed to be said has been said over the last few weeks. Several of my colleagues have recently spoken publicly on both the strategic and technical aspects of this matter.  I myself discussed the issue at a conference in Rome, on 11 September. 
After all these comments, it’s tempting to say that only the implementation remains to be done – but that, in fact, may be a bit more difficult than the talk, as in most cases. To be precise though, the tricky part is perhaps not so much implementing the exit itself as managing the expectations about how the exit strategy will unfold. Given that the exit will be gradual and will involve two dimensions, i.e. the non-standard measures implemented by central banks and the level of interest rate, there is a risk that market participants might interpret a decision about one part of the exit as paving the way for the rest. Markets may then overreact to a specific decision, attributing to it an excessive signalling content. The fear of an overreaction may in turn prompt central banks to delay a specific decision which needs to be taken at a specific time. Communication is thus essential to prepare markets and explain the rationale for each decision.
We are still facing a lot of uncertainty about the underlying conditions, both in the real and financial sectors. So, we cannot lay down a well-defined exit path at the current juncture, either in respect of contents or timing. No pre-commitment can be made to any particular sequence. But it is appropriate to spell out in public all the parameters that the policy-makers will be confronted with in making their choices. I guess that ECB watchers are quite used to this kind of communication, for it has served us well in other circumstances.
Today, I’d like to examine one part of the debate about the exit strategy which has so far attracted less attention – the international dimension. In doing so, I’ll concentrate on the exit from the exceptional degree of monetary and fiscal accommodation, rather than from the non-standard measures implemented in particular by central banks.
The world’s leading economies have recently been called upon to cooperate in, or even coordinate, the design and implementation of their exit strategies. In particular, at the recent G20 meetings in St. Andrews, the IMF defined seven principles for the exit strategy, the last one being about international policy spillovers and cross-border collaboration. It reads as follows:
“Making exit policies credible and consistent, together with communication and consideration of spillovers, will improve outcomes for all countries. Coordination does not necessarily imply synchronization, but lack of policy coordination could create adverse spillovers.”
“…spillovers from differences in monetary policy conditions could pose significant challenges. Interest rate differentials are already generating capital flows into economies with higher yields. Such inflows complicate monetary policy management, requiring difficult choices for authorities…”.
Indeed, not all countries are at the same stage of the cycle – some are already well on the way to recovery, while others are at a very early stage, experiencing more fragile growth. Let me consider, for simplicity’s sake, three main areas or countries: Emerging Asia, Latin America and the US respectively. You could include other countries and regions, such as the euro area, for instance, but it would only complicate matters and not add much to the analysis.
In respect of returning to economic growth, Emerging Asia and Latin America seem to be ahead of the US, whose growth is still fragile. So, logic would suggest that Emerging Asia exits first from the extraordinary post-crisis measures, then Latin America and finally the US. This ‘optimal’ sequence would mean that Emerging Asia tightens its policies as the economy gets back on track, thus avoiding domestic overheating and financial instability. The exit would be accompanied by an appreciation of Asian currencies, and that in turn would favour a rebalancing of growth towards domestic demand, thereby avoiding a large build-up of foreign exchange reserves and excess domestic liquidity. It would also entail a progressive reduction of liquidity flows to the US, which were probably useful as the crisis escalated but are becoming less so as financial markets stabilise in the US. Latin American countries would be the next to exit, as their economies recover gradually, thanks also to exports to Emerging Asia and to stronger domestic demand. In Latin America too, some appreciation of the exchange rates would be expected to accompany the recovery and stem excessive capital inflows. The US and other advanced countries would exit last, as their economies are proving to be the slowest to recover. This would be the ideal sequence.
Unfortunately, the reality appears to be quite different. Emerging Asian economies have not yet exited and do not seem to be in the mood to exit first, despite the stronger pace of their recovery, at least compared with that of the US. They continue to maintain strongly accommodative monetary policies and steadily accumulate foreign exchange reserves. In all Emerging Asia, except Malaysia, foreign reserves are now higher than before the Lehman collapse (see attached table). As a result of these interventions, the strong domestic demand for credit is being met, fuelling potential financial market instability in these countries. Just to give an example, the year-on-year growth rate of domestic credit has reached 32% in China.
The delayed exit in Emerging Asia is affecting other countries. In particular, the interventions in the foreign exchange markets to prevent Emerging Asian currencies from appreciating are shifting capital flows towards other regions, such as Latin America, and pushing up the exchange rates of currencies there, in a disproportionate manner. This may in turn cause Latin American countries to adopt measures to counter capital inflows and also to intervene in the foreign exchange markets and accumulate reserves, thereby further delaying the exit. Emerging Asia’s delayed exit – and possibly that of Latin America too – has also undesirable effects in the US. Preventing the Emerging Asian and Latin American currencies from appreciating, in parallel with the recovery, reduces the export competitiveness of those countries which are recovering more slowly, like the US, and thus further delays their exit. Delaying the exit with an undervalued currency results in a beggar-thy-neighbour policy, which might lead to trade distortions. Furthermore, the reinvestment of foreign exchange reserves accumulated in Emerging Asia contributes to flush the US financial markets with liquidity, making the exit more complicated, once it gets under way.
To sum up, an untimely exit by the countries that are ahead in the cyclical upturn creates distortions and encourages other countries to delay their exit, thus further adding to the imbalances and making the exit more difficult for everybody.
Why is this happening? There might be three reasons.
The first reason is philosophical. And I would formulate it as a question: why should Emerging Asian countries abide by principles different from those which some advanced economies and international institutions are advocating? I have in mind in particular the first principle established by the IMF in the G20 exit strategy document I mentioned previously. I quote it:
“The timing of exit from stimulus should depend on the state of the economy and the financial system, and should err on the side of further supporting demand and financial repair.”
In other words, why should Emerging Asian countries exit in a timely way when international institutions, and some advanced economies, are generally advising to “err on the side of being late”?
In my view, the “ err on the side of being late” paradigm is potentially as dangerous as the “ productivity growth” paradigm of the late 1990s and the “ fear of deflation” paradigm of the early 2000s, which led some advanced economies to implement policy stimuli for too long, sowing the seeds of the subsequent crisis.
Not enough effort has been put into assessing all the risks of the too-early and too-late errors that policy-makers can make in implementing the exit strategy. Analysts are often inclined to believe that the costs of exiting too early are pretty obvious, while those of exiting too late are unclear and can be attenuated over time. As a result, the costs of a premature exit are regarded as being much higher than those of a tardy exit. This leads to the asymmetric recommendation of erring on the side of being late.
The historical examples chosen to illustrate the two risks tend to be analysed in a rather simplistic way, in particular concerning a premature exit, which is generally deemed to be the most dramatic. In fact, it is not so easy to find cases of too early exits. For instance, the monetary and fiscal policy tightening which took place in Japan in 1997 has been considered by some as a case of premature exit, responsible for the Japanese recession of 1998-99. However, careful analysis suggests that that recession was mainly due to the effects of the delayed restructuring of the banking sector and to the collateral effects of the 1997 Asian crisis. 
The consequences of a late exit strategy have not been sufficiently examined. A delayed exit aims at increasing the incentives to lend at (low) fixed rates. But the more the exit is delayed, the greater the adjustment will have to be when interest rates are ultimately increased, in order to catch up. And the greater the interest rate adjustment, the larger the capital loss to fixed income asset holders. In other words, a delayed exit postpones the pain but in the end the pain is greater. Delaying the pain may be sensible if the economy is expected to get stronger and thus become more resistant to it. But experience shows that policy-makers have an incentive to delay the pain to such an extent that most often they will end up inflicting it when the patient starts suffering again. Indeed, experience shows that most often interest rates increases tend to be delayed and thus have to be continued even after the economy peaks.  To put it in other words, late exits seem to be the rule, and possibly also the cause of the crises that follow. This is how Friedman and Schwartz characterise the monetary history of the United States, and I suspect that their theory applies well also to the most recent cycles.
To bring this point to an end, if we want better timed monetary policies across the world, also in emerging markets, we might need to revisit some of the analytical principles and empirical evidence that are currently being proposed.
The second reason why Emerging economies might tend to exit late is tactical.
If emerging market economies, like the ones in Asia and Latin America, exited in a timely way, rather than err on the side of being late, what risks would they incur if advanced economies did instead err on the side of being late? The issue is relevant for emerging markets because over the last few months they have provided the financing needed to match the huge increase in debt issued by advanced economies, in particular by the public sector. In the short term, such funding has provided favourable financing conditions in advanced economies, particularly in the US, and thus facilitated the issuance of public debt. However, if the issuance of public debt continues over time, while the flow of financing by Emerging Asia and Latin America is progressively reduced (as part of the exit strategy), there will be a risk of a fall in the relative value of these assets. This would cause a reverse beggar-thy-neighbour policy, similar to the one which has in the past enabled the reserve currency country to finance its deficit by borrowing at a lower cost than the rest of the world. The risk of such a policy making a comeback is a source of concern for the other countries. Without a credible commitment to a timely exit from the expansionary monetary and fiscal policy in advanced economies, in particular in the US, aimed at reducing the net supply of government bonds, Emerging Asian and Latin American economies might be reluctant to exit in a timely way, as they would suffer a significant loss in the value of their stocks of foreign reserves.
To sum up the argument, emerging economies may have an incentive to exit in a timely way from their expansionary policies if they can be certain that advanced economies, starting with the US, will do the same when their domestic conditions allow it.
The third reason why Asian economies might choose to exit late, rather than early, is prudential.
The measures that they adopted to counter the crisis, a crisis which originated – as they see it – in advanced economies, proved to be effective. Before exiting these measures, including ending the re-pegging of exchange rates and the big build-up of foreign exchange reserves, they might want to be reassured that everything is being done in advanced economies to ensure that the worst of the crisis is indeed over and that no further problems are likely to arise once they have exited. A relapse in advanced economies’ financial markets, in particular, would make any exit premature, even by emerging economies.
This crisis has shown how contagious the failure of systemic financial institutions can be. Concrete progress in reforming the financial system is thus key to reducing the risk of returning to the pre-crisis problems. The question to ask advanced economies is: can the measures taken in the last few months minimise the risk of a relapse?
Substantive work has been done in recent months; it has been witnessed by the emerging markets which now belong to fora such as the G20 and the Financial Stability Board. However, progress at high level is one thing. The hard work of implementation - the devil in the details - is another. Experience has shown that internationally agreed standards, such as Basel II, are not always uniformly implemented in all the signatory countries. Furthermore, pressure from the financial industry to minimise regulatory changes in advanced economies is mounting by the day. That pressure is fuelled by the renewed profitability of market participants, who are benefiting from the very favourable financing conditions. If I may develop that now famous music-and-dancing metaphor used by a banker just before the crisis, financial markets are now giving the impression that the music is back on and that many dancers are up on their feet again. What’s more, regulators and supervisors do not seem to have found a way yet to slow the music down.
Seen from a distance by net creditors and users of the advanced world’s financial infrastructure, a lack of progress in implementing the agreed agenda is an incentive to continue with a strategy of self-insurance, and thus to delay the exit.
Let me conclude. An optimal exit strategy from the extraordinary monetary, fiscal and financial policies implemented by advanced and emerging market economies has several elements. It has to be timed properly, in all countries. This is difficult: not all countries are in the same cyclical position, so some should exit earlier than others. However, those who should exit earlier can do so if they are confident that those who are supposed to exit later will do so in the same timely fashion. Furthermore, they want to be sure that the worst of the crisis is over and that concrete actions are being taken to overhaul the financial system. To make such commitments credible and to build up that confidence, we need a stronger system of international cooperation than the current one, with the major countries undertaking to bear in mind the external implications of their actions. Without a stronger multilateral system we might end up repeating the mistakes of the past, with one difference. The impact on each of us will be greater.
Thank you for your attention.
 See, for instance, Jean-Claude Trichet, “The ECB’s exit strategy”, Frankfurt, 4 September 2009; José Manuel González-Páramo, “Non-standard monetary policy: five questions about the exit”, London, 6 November 2009; Jürgen Stark, “Monetary policy before, during and after the crisis”, Tübingen, 9 November 2009;
 “An ocean apart? Comparing transatlantic responses to the financial crisis”, Rome, Banca d’Italia, 10-11 September 2009.
 D. Leigh, “Monetary Policy and the Lost Decade”, lessons from Japan, IMF Working Paper WP/09/232, 2009.
 See, for instance, M. Friedman and A.J. Schwartz, 1963, A Monetary History of the United States, 1867-1960, Princeton University Press for the National Bureau of Economic Research, Section 7.7, pp. 407-419 (“Why was monetary policy so inept?”); A. Levin and J. B. Taylor, “Falling Behind the Curve: A Positive Analysis of Stop-Start Monetary Policies and the Great Inflation,” in M. D. Bordo and A. Orphanides (eds.), The Great Inflation, University of Chicago Press, forthcoming.