Transcript of a Conference Call on Staff Discussion Note on Multilateral Aspects of Managing the Capital Account

September 7, 2012

Washington, D.C.
Friday, September 7, 2012

Jonathan Ostry, Deputy Director, Research Department
Atish Rex Ghosh, Assistant Director, Research Department

Ms. BHATT: Good morning, everyone. Thanks for joining this conference call. I'm Gita Bhatt. This conference call is on the IMF Staff Discussion Note, a Paper on Multilateral Aspects of Managing the Capital Account. I have here with me Jonathan Ostry who is the Deputy Director of the Research Department. Jonathan will make some opening remarks and then we can go on to questions and answers.

MR. OSTRY: Thanks very much, Gita, and good morning to you all. Thank you for joining the call.

The topic of how to manage capital flows has been a key focal point of the Fund's work over the past 2 years. No fewer than four board papers have been produced on different aspects of the problem and significant analytical work has underpinned these discussions beginning with the Staff Policy Note that we produced in 2010 which looked at the circumstances under which capital controls could be useful to manage inflow surges. Then this was followed-up last year by a Staff Discussion Note that we put out on how capital controls and macroprudential policies could complement each other in managing the financial stability risks posed by volatile flows.

Much of this work has been focused on an individual country’s perspective, asking how a single country could address the challenges posed by volatile flows, taking more or less as given what other countries are doing. In contrast, the question we're asking in today's paper is how policy advice might be affected by spillovers arising from individual country policies and whether coordination of country policies might be desirable to enhance global well-being. An IMF board paper took a first stab at some of these multilateral issues late last year, and what we're doing today builds on some of that work.

Spillovers, as many of you may know, is a hot topic nowadays in particular in the context of the IMF membership's recent adoption of an integrated surveillance decision which shines a spotlight on policy spillovers and how they might affect the stability of the international monetary system. The Fund is also in the process of articulating an institutional view on policies for managing capital flows. And while today's paper is not meant to prejudge that work, it will inform these discussions which of necessity will need to factor in multilateral aspects.

What are the salient multilateral issues as far as capital account management is concerned? We emphasize three issues in today's paper. The first is the possibility that capital flows are the instrument by which countries act to prevent or thwart external adjustments. For example, when inflow controls are used to sustain an undervalued currency. This can raise multilateral concerns because exchange rates and current accounts are inherently multilateral. They add up at the global level. Policy action in one country can therefore impose costs on the rest of the system. Our take is that capital control should not be used as an instrument that substitutes for warranted external or macroeconomic adjustment. In most cases, therefore, and the paper does go through a number of exceptions, an undervalued currency combined with inflow controls will raise red flags particularly when the intent of the controls is to support undervaluation.

A second issue is the possibility that capital controls deflect flows to other countries. If I close my door to inflows, that capital will be diverted elsewhere. There is a lot of debate on how important these diversions are, but the logic is clear: the only thing that's in question is the amount. Whether this is harmful, this deflection, depends on whether or not the country to which flows are diverted welcomes those additional funds. In the case where additional funds are not welcome, for example, because they amplify financial stability risks, the controls lead to a negative spillover for other countries.

A third issue relates to policy from source countries. We don't typically think that source countries impose capital controls and that's more or less right, but it's important to recognize that many other policies, including prudential policies and monetary and fiscal policies, will have material effects on capital flows, and in particular, they have the potential to increase either the level or riskiness of flows. Given all this, and the potential for spillovers to harm others, it's tempting to conclude that the policies that give rise to such spillovers should be proscribed, i.e., they should be disallowed. This logic is flawed, however. Spillovers are an important part of how market systems operate and can be perfectly benign. This is an important point because oftentimes spillovers per se are viewed as a problem in popular debates. To take an example, when there are valid macroeconomic or financial stability reasons for countries to impose controls, they should do so even if this gives rise to spillovers. The answer to the spillovers is for other countries to take measures to protect their economies from risks the high inflows may bring. The result may well be escalating controls in many countries, but such a result would be globally efficient in this example.

The issue though is less clear cut when raising controls is costly for the country. That is to say, there is a benefit from controls, that's why you're imposing them, but also a cost. For example, good flows like greenfield FDI get shut out along with bad flows, say “hot money”. In such a case, if country A raises controls and this deflect flow to country B, the need to raise controls in B is costly for that country. In such a case, there may be gains from coordination in the sense that all countries may be able to manage the risks from inflow surges at a lower cost if they set lower controls than they would choose acting on their own. Getting to a globally efficient outcome may require that source countries participate in coordination. Our sense, as discussed in the paper, is that this may be even more important than coordination across recipient countries. This is reminiscent of Lord Keynes's idea that effectively managing capital control volatility requires that one "operate at both ends of the transaction," that is, in both source and recipient countries.

To conclude, our paper takes the analytics of capital controls further than our own previous papers or the previous institutional papers by the Fund, drawing some possible implications for policy coordination. Our analysis underscores that foreign capital, which has the potential to bring enormous benefits to countries, may require some multilateral principles to ensure its safe absorption by countries in much the same way that motorcars need rules to ensure their safe operation for everybody. The notion that international cooperation can mitigate the severity of boom-bust cycles in capital flows is one that goes back to the IMF's founding fathers. Global financial integration has indeed progressed a long way in six decades, but multilateral oversight of both source and recipient countries to assist in the management of capital flow volatility remains a worthy objective and one likely to be essential to safeguard the stability of the system, a core purpose of the Fund and its members. With those remarks, I'm happy to take your questions and I'm joined here by my colleague and co-author Rex Ghosh.

MS. BHATT: Thanks everyone. Also I want to note that the staff discussion note is the view of the authors and not of the institution. There's that caveat there that I want to emphasize. With that, can we take some questions.

QUESTION: You referred to source country action. Let's take an example. The Fed considering its own policy action to address policy action here, how might or what sort of source country considerations might be taken? Can you give some examples? What's the feasibility of both source and recipient countries coordinating? You seem to indicate in the paper that it would be difficult, that the goals or the objectives are difficult to see. Finally, could you elaborate a little more on the actual gains to coordination? Are they really that significant?

MR. OSTRY: Thank you very much. Those are excellent questions. Let me take them in turn. The issue of source countries, there are many source countries in the system, we often forget China is a potential source country, but we can also think of the Sweden and Baltics example. There are very interesting coordination issues there. But of course the one you raised has to do with U.S. monetary policy. Here as you probably know, the evidence is a little mixed. The data don't seem to support a large effect of quantitative easing on asset booms and inflow booms experienced by emerging market countries. But if you look at a longer historical perspective, I think going back decades, and I’m thinking of the early work of Liederman and Reinhart going back a long time, there is evidence that interest rates in the important financial centers are an important driver of capital flow. What might this mean in practice? Obviously a key goal of any domestic policy must be to secure domestic stability and that's no doubt a key goal of U.S. monetary policy. But I would say what we're underscoring is that, at the margin given that one of the channels through which U.S. monetary policy is going to operate is byencouraging purchases of riskier assets, including risker foreign assets, and at the margin this is going to have an effect on foreign countries, so therefore at the margin the Fed should consider those impacts when it adjusts its own policy setting. So that's the issue that we're calling attention to.

On your second point, what is coordination? What do we mean? How realistic is it? It's important to mention that we don't have ideas here of really formal policy coordination. That is, we don't have the idea of the key players sitting in a room and hashing out targets for exchange rates or current accounts and so on. We simply have the idea, which is inherent in the integrated surveillance decision and indeed before that in the way we consider policies in a multilateral setting, that people will at the margin take into account the effects of their policies on others and that this can yield superior outcomes without significantly compromising domestic objectives. It's a very soft idea of coordination rather than something formal.

On your last question, the gains, we have a box, and I think it's the last box in the paper, that tries to give some feel for the numbers. You'll recall that gone back to at least Lucas's work, welfare triangles tend to be small numbers in the grand scheme of things. So we do give a sense about how much the benefit is if you act to offset your domestic externalities, how much the net benefit changes if you coordinate, how much it changes if people who are coordinating are just the recipient countries or also the source countries and it depends on, as we argue, the nature of the cost function associated with these regulatory interventions. So it gives you a sense of the numbers. I will not sit here and pound on the table saying these are large numbers, but I do think you get a sense of them from that box.

QUESTION: This response is probably based on my ignorance in not having a background in economics. In looking at this, my general sense of the paper is that, yes, you find that there are efficiencies in the global system for coordinated action on capital controls, but doesn't seem like there's a very strong case to be made and that my guess is that the takeaway from this by policymakers is, yes, you're right, we should consider it, but my own economy response is going to be or my own economic policy response is going to be more important. Am I overexaggerating? How would you respond to that characterization?

MR. OSTRY: That's an excellent question and I both agree and disagree with you as you can imagine. Here is my two-handed response. I think one message from the paper is that there are important gains from offsetting your own externalities. If foreign borrowing amplifies the risk of financial crisis in your country or the upward pressure on the currency leads to Dutch disease concerns that are valid, then it's important for you to act to offset that externality and probably that is the more important message f I look at the totality of the paper. That being said, I think some policies have the potential to be quite destructive multilaterally and they have to do with the situation where a country closes its capital account in such a way that it thwarts the external and macroeconomic adjustment process and so imposes costs on the rest of the system because the whole system has to add up. We do talk a little bit in the paper about equivalencies between capital flow taxes and intervention in the foreign exchange markets that is supported by a capital control. You see that when countries accumulate very large levels of foreign exchange reserves under a wall of capital controls, this is almost an order of magnitude larger in terms of its effect than the typical size of capital control taxes that one observes in the real world in response to capital flow volatility. That's the second part. That's the part where I disagree that perhaps the multilateral effects are necessarily small. I think they have the potential to be large. My colleague, Rex, will have something to add.

ATISH GHOSH: Hi, I am one of the co-authors of this paper. Your questions are very good and very pertinent. I would amplify with one comment on what Jonathan spoke about. The example that he briefly mentioned was the lending by the Swedish banks in the Baltics, in Latvia in particular, before the global financial crisis. Initially this was too small of an issue for the Swedish banks and for the Swedish regulators, but it was a big problem for Latvia, and over time the Swedish regulators started getting more worried and taking into account what their banks were doing. So I think in those sorts of specific cases, the source country regulators taking into account what their institutions are doing could have quite large welfare effects.

QUESTION: -- so do I take it that this is another paper to help -- formulate its policy? What's the next step forward?

MR. OSTRY: We often produce papers of this nature on issues that are on the official agenda of the institution. This paper is designed to be background to that work and the Fund has produced four board papers in the past 18 to 20 months on the capital flow management issue. The two other staff discussion notes that I referred to were input into those earlier Board papers and I would see this paper as being input into the upcoming one that is expected shortly that is going to propose an institutional view on those issues, but this paper is not meant to prejudge that work.

QUESTION: I'm just saying in this last part is that this paper is being input on the other notes -- that would propose an institutional view for the Fund?

MR. OSTRY: Yes. As I said, we've had four board papers dealing effectively with the role of the Fund managing inflows, multilateral aspects and managing outflows and liberalization. There is going to be a synthetic paper that is going to draw out the main threads from that previous work and aim to propose for the consideration of our Executive Board a balanced institutional view on all these aspects. And we would expect our analytical work that we are discussing with you today, like our earlier papers, to help in these discussions.

QUESTION: It seems as though the summation of this is pretty much in line with what the Fund already does, do things that help your economy but be careful not to have damaging effects on the global monetary system. In particular, you talked looking at competitive evaluations, foreign exchange intervention and this seems to be smaller but similar plank in that. Am I wrong?

MR. OSTRY: I would say one thing. If you ask people if one country’s capital controls deflects flows to other countries and can make problems in other countriesworse, isn't that an argument against my putting controls to address my own issues? I think a lot of people would say, yes, it's an argument against it. In fact, some very respected academic research has drawn attention to the spillovers and the potential for bubbles and problems that can be foisted on other countries. So one analytical point, and I think it's an important point, is that spillovers themselves don't have normative implications per se.. You need more than the existence of spillovers to draw welfare implications. The appropriate response to the different risks that people face is, as you've said, for people to deal with their domestic problems to the extent they can with their own instruments. The second is to really identify the circumstances under which coordination may be desirable. It isn't always desirable. It isn't always needed to achieve globally efficient welfare maximizing outcomes and I think that's a second important point. The difference between a spillover and an externality, a spillover is normatively benign, an externality is not. These are important considerations when you come to formulate some multilateral principles to guide action, I think it's important to be clear on the distinction between those two concepts. I think the paper has advanced a little bit the analysis, and clarified these issues. Certainly if I look at some of the work in the past I think it wasn't always as clear as it might have been.

MS. BHATT: Thank you. If there are no other questions, we will conclude this conference call. Thank you all for joining us, and thank you, Jonathan.


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6220 Phone: 202-623-7100