Transcript of the IMF’s Press Briefing on World Economic Outlook: Anyaltic Chapters 3 and 4

September 22, 2009

Olivier Blanchard, IMF Economic Counsellor and Research Department Director
Jörg Decressin, Division Chief, World Economic Studies Division
Petya Koeva Brooks, Deputy Division Chief, World Economic Studies
Alasdair Scott, Senior Economist, World Economic Studies
Conny Lotze, Deputy Division Chief, Media Relations
Washington, D.C.
Tuesday, September 22, 2009
Webcast of the press briefing Webcast

MS. LOTZE: Good morning, everyone, and welcome to this press briefing on the analytic chapters of the World Economic Outlook.

I’m Conny Lotze of Media Relations, and we are joined here today for this briefing by Olivier Blanchard, the Fund’s Economic Counselor and Director of the Research Department. To my right is Jörg Decressin, Chief of the World Economic Studies Division. To Mr. Blanchard’s right is Petya Koeva Brooks, Deputy Chief in the World Economic Studies Division and one of the authors of Chapter 4. And, to her right at the end, is Alasdair Scott, Senior Economist in the same division and one of the authors of Chapter 3.

You have had a chance to look at the documents under embargo. The outlook chapters of the WEO, Chapters 1 and 2, will be released on Thursday, in a week, on October 1 at 9:30 in Istanbul as part of the events surrounding the World Bank and IMF Annual Meetings. Those chapters will also be available under embargo about 24 hours before the press conference, but we will send out a media advisory about that event.

The speakers will have some introductory remarks, and we will begin with Mr. Blanchard. Mr. Blanchard?

MR. BLANCHARD: Thank you, Conny.

Welcome. As you know, the World Economic Outlook has two parts. The first is our analysis of current evolutions, current trends, current policy choices, and, as Conny said, this will be released only on October 1. It is not part of this presentation.

The second part, which typically includes two or three chapters, two chapters this time, is an attempt at a more in-depth look at the issues of current importance. The challenge each time is to decide six months or a year in advance what will be hot on the day of the presentation. I think this time we have chosen very well. I think we have two chapters which are both extremely topical and extremely relevant.

The first one focuses on the relation between asset fluctuations and monetary policy. There are some who say that monetary policy is largely responsible for the booming asset prices before the crisis and, therefore, has played a major role. There are some who say that may not be the case, that monetary policy should at least have reacted to the asset price booms before the crisis and would have limited the effect of the crisis.

This is very much the set of issues that we look at in this chapter, and the results, which are very interesting I think, will be presented by Alasdair Scott, to my extreme right.

The second topic that we chose is the effects of the crisis, not on output in the short run but on output in the medium run, on what we call potential output. And, there’s a question of: Are we going to return to the same old trend? When everything is said and done, will we back on the line where we would have been otherwise?

Some say, no, if we will end up lower, maybe with the same growth rate but at the lower level of output, so there will be a permanent, lasting effect.

And, then there are even people who say that maybe the growth rate will be affected, that we will not be able to grow at the rate we had before the crisis, on a sustained basis.

So what this chapter does is look at what happened after a large number of banking crises in the past 40 years. It gives very striking and somewhat disturbing conclusions, and this chapter will be presented by Petya Koeva Brooks.

So let me just give the floor to Alasdair first and then to Petya.

MR. SCOTT: Thank you, Olivier.

The aim of this chapter is to draw out some lessons from asset price fluctuations for monetary policy.

The first part of the chapter takes a look back at history, takes a look at asset price busts in advanced economies over the last four decades, and what we find is that there are some clear macroeconomic patterns leading up to these busts, including the most recent crisis. For example, rapidly expanding credit is often associated with a bust one to three years followed, and, to that, we can also add the phenomenon of deteriorating current account balances and rapid shifts of expenditures into residential investment.

At the same time, we also find that output and inflation are not actually very good leading indicators of ensuing busts. So they’re not good indicators of that kind of overheating, and this also holds for the current crisis.

Now, as Olivier hinted, we also take a look at the role of monetary policy in the lead-up to asset price busts, including the current crisis. Now, we find that monetary policy was not the smoking gun in the sense that monetary policy does not provide an encompassing, systematic explanation of the differences in asset price experiences across countries. There are too many exceptions in countries’ experiences with asset prices for this to be a totally convincing explanation as a cause of the crisis.

However, we do find that many of the same macroeconomic patterns of the lead-ups to asset price busts apply to the lead-up to the current crisis. So those same patterns about the expansion of credit, the deterioration of current account balances, the shift into residential investment also held and correlate well with countries’ different asset price experiences.

So you can say that there were warning signs to policymakers that could have been followed. So, to the extent that policymakers accommodated the relaxation in financial conditions, there were risks that were allowed to build up.

With that historical experience in mind, the final part of the chapter tries to look forward and say, what lessons can we draw from this about how monetary policymakers should act going forward? And, this is a challenging question, and we don’t pretend that we have all the answers here, but we take a first stab.

We look at whether monetary policy could and should be responsible for more than just targeting goods price inflation, in particular, whether there’s a role for leaning against some of these phenomena like the expansion in credit and whether that can produce more stable outcomes.

And, we say that on the assumption that policymakers can correctly understand the drivers of the shocks going on, in the case where, for example, there are financial shocks, there is a good case for reacting more strongly than would otherwise be usual to those underlying phenomena.

Now, we also emphasize that this might require, ideally, extra tools. What you have is a situation where there can be two or more problems facing the policymaker, and for them to have an adequate chance of dealing with this, there is a case in theory at least for them having more than one tool to meet these objectives.

But we also emphasize that this is not going to be easy. For policymakers to take a broadened approach, they will face definite challenges. Now the first of them follows from the empirical work at the beginning of the chapter. Even though we identify warning signs, patterns, macroeconomic patterns leading up to busts, it’s also clear that these leading indicators are, by no means, perfect. So, at the end of the day, policymakers will have to employ very skilled judgment, and they cannot be expected to perfectly anticipate everything, and they certainly cannot be expected to predict a bust.

The second aspect is that if policymakers find themselves in a situation of wanting to, for example, momentarily focus on financial weaknesses and risks, then it will be very important for them to communicate just how this momentary focus is consistent with longer-term objectives, for example, of price stability.

These are difficult challenges, we think, on balance, that they’re worthwhile.

MS. KOEVA BROOKS: Thank you.

The second study, the one that we’ve been working on, looks at output dynamics following 88 banking crises that took place over the past four decades in a very wide range of countries -- advanced, emerging, developing. Our focus was on the medium run, which here we define as seven years after the crisis. Since the first glance at the data told us that there was a wide range of experiences after a banking crisis, we also sought to try to link the various outcomes with initial conditions and policies after the crisis.

So what did we learn from our analysis? Well, let me focus on three things.

The first one was that banking crises tend to have a long-lasting impact on the level of output. And, for the average country, about 7 years after the crisis, the level of output was still around 10 percent below its pre-crisis trend although there was a very large variation across crisis episodes. It’s interesting that this result holds for both advanced and emerging economies.

The encouraging news was that if we look at growth, unlike the level of output, we do see that medium-term growth rates return to their pre-crisis rates. Again, here, there was a large variation in outcomes across countries.

The second finding was that this depressed path of output was mainly due to reductions in all factors of production, which, the employment rate, capital and productivity. In the short run, the output loss was mainly accounted for productivity although its level tended to go back, to come closer to its pre-crisis trend in the medium run.

In contrast, capital and employment suffer enduring losses relative to trend, and one explanation is that the crisis tends to depress investment as the supply of credit becomes more limited.

As far as employment losses are concerned, the typically large increase in the actual unemployment rate seems to last for a long time as the crisis implies a substantial reallocation of resources across sectors and also as working skills may get eroded.

Now the third result which we have in our study was that initial conditions have a strong impact on the size of the ultimate output losses, and also policy efforts, post-crisis, can tend to be associated with lower output losses.

Now let me elaborate on that. As far as initial conditions are concerned, what happens to output in the short run is a very good indicator as to what happens in the medium run, and also a high pre-crisis investment share is a good predictor of large medium-term output losses.

Now turning to policies after the crisis, we find that economies that apply macroeconomic stimulus in the short run following the crisis tend to have smaller output losses over the medium run, and there’s also some evidence that structural reform efforts are associated with better medium-run outcomes. However, here, there’s certainly no one size fits all when it comes to establishing the right policy after a crisis. As a general point, let me also say that much remains to be explained about the dynamics of output after crises.

Now, given our main findings, one may want to ask the question: So what do they imply for the current crisis?

And, in some ways, the analysis has sobering implications for the medium-run output prospects in economies that have suffered recent banking crises. However, the very forceful macroeconomic policy response so far, in the form of substantial fiscal and monetary stimulus, should help mitigate the impact of the crisis on output this time around. Still, to the extent that there are remaining concerns about losses, they underscore the importance of implementing reforms that help raise medium-term output prospects and facilitate the reallocation of resources across sectors.

QUESTIONER: A question on this chapter, I mean the way I understood your analysis is that you have a regular recession. Then after it ends, you have a big bounce in economic growth. So you recover your output loss, right? And, when you have a banking crisis, you don’t have the big bounce. You never recover that output loss. Is that the way to interpret it?

MS. KOEVA BROOKS: What our chapter looks at is what happens after banking crises only, and what we do find is that after banking crises what we see on average, for the average in our sample, is that there is a permanent output loss, which is just that the level of output is not going back to its pre-crisis trend.

But it’s a very fair question. You may want to ask the question of: What happens compared to a scenario in which there’s just a recession rather than a recession that has a banking crisis?

We do not look at this question in our chapter, and one of the reasons this is very hard to do, when we look at the very large sample of countries, is that defining what is a recession in some of the emerging and developing countries is a tricky issue, and we wanted to look at as broad of a sample of experiences as possible.

MR. BLANCHARD: Let me just add to it and be less careful than Petya and go beyond the results of the study. I think you’re right, that we think of a standard recession as a recession with a bounce back. I think what this chapter does is warn us that in the case of banking crises this bounce back may not be very strong, if there at all.

MR. DECRESSIN: May I add one more thing? We had a chapter on this in the April 2009 WEO which basically just looked at recessions in advanced economies for which we can easily define these events, unlike for emerging economies. We find that if a recession has been preceded by a financial crisis, then it takes much longer to revert to the original pre-crisis growth rate than it does in the case of regular recessions.

QUESTIONER: Also on this issue, I look at what you have on the structural reforms, and I wonder if you could elaborate on that because it seemed the evidence was conflicting there. I mean what kind of reforms do you think will happen now?

MS. KOEVA BROOKS: Yes, looking at the structural reform efforts that countries have implemented is something that we did in our analysis. We used a number of indices that have been put together on trade liberalization, capital account liberalization, government efficiency and a number of other indicators that were available for these countries, and what we find is that what has worked in some countries is not necessarily something that would work in other countries. That said, I think there are patterns that emerged which give a sense that, indeed, proceeding with structural reform is a good thing.

Now, in terms of translating this to what needs to be done this time around, this is, in a way, beyond the scope of our chapter. That said, I think there are a number of reforms now in terms of repairing the balance sheets of the financial sector and also making sure that in the area of labor markets the policies that are implemented facilitate this reallocation of resources across sectors, which, and also retooling the skills of workers. These are things that would be beneficial this time around.

And, as was said at the beginning, there will be more on this once Chapters 1 and 2 come around, which is in a little bit more than a week. We’ll be more specific in terms of the reforms in various countries, which would depend on the specific circumstances in the country.

MS. LOTZE: Yes, and following up on that, we have a couple of questions online also on macrofinancial risks—where we see them—and those are also questions that will be answered next week with the first two chapters.

QUESTIONER: I have a question on Chapter 3. On the point you make, central banks’ mandates may need to be expanded, if someone could articulate that, explain it further?

MR. SCOTT: We imagine a situation where central banks might find themselves with good reasons to try to confront growing financial risks. They might see household balance sheets deteriorating and are worried about the potential repercussions.

Now the question is: In a conventional monetary policy framework, as we see a number of advanced economies in, would central bank policymakers in those frameworks feel able if inflation showed no signs of being out of control, would they feel able and entitled to be able to justify policy measures to try to address those risks?

Now I think we don’t want to emphasize the issue of the mandate as being a solution to these problems because what matters at the end of the day is how you execute policy. I can imagine that a number of central bank policymakers would say they already interpret their framework as allowing them to not have to hit inflation each and every quarter. There’s already some flexibility, and they already take a view that shocks that have implications for macroeconomic volatility should be part of their decision-making.

But we can also imagine that there might be some situations in some countries where they feel that their credibility would be bolstered if there were a specific provision in the mandate under which they’re operating, that they could point to in order to communicate why they’re making the policy decisions they are.

MR. BLANCHARD: Just a follow-up, to be concrete, I mean suppose you have a situation in which you have a housing boom and you’re thinking that the rest of the economy seems to be doing fine. Then you might not want to do anything to the interest rate, but you’d like to slow down the housing boom. One of the instruments that you may have—this is one of the macrofinancial tools that Alasdair was referring to—would be a decrease in the maximum loan-to-value ratio.

The question is who is in charge of that? Now you can think of two authorities—the central bank and some other authority—each in charge of one of the instruments, the interest rate and the loan-to-value ratio, but it might make more sense for the central bank to actually have both tools at its disposal. I think that’s the kind of idea that we have.

QUESTIONER: I just wanted to expand on that. I wonder if we could apply that to kind of the real world situation right now because the Fed, the ECB, and the Bank of England, all their proposals for each of those to kind of expand into, to some extent, a macroprudential policy. So I was just wondering, do you have any views on whether those proposals are moving in the right direction or do you have specific proposals for those central banks? Thank you.

MR. SCOTT: The chapter actually stops a long way short of actually getting into a lot of the details simply for time and space reasons.

I’m sure you’re aware that there are a number of proposals circulating along the lines that Olivier just hinted. There are plans for elements such as dynamic provisioning for time-varying capital requirements. Now all of these plans are very much at the stage of early discussion.

One point that has been raised by policymakers, and this is just a follow-up on the point by Olivier, is that if central banks are to adequately address these kinds of monetary, macrofinancial risks, then unless you want a situation where the central bank is trying to accomplish two or more objectives at once with one tool, then equipping them with some of these kinds of tools would be very important.

But, beyond the chapter, we need to be very clear that there are a whole range of practical issues that need to be sorted. It’s not clear which of these proposed instruments would actually have sufficient bite. It’s not clear how they might work in an environment where there are cross-border regulatory considerations. All of these details would have to be worked out, but I think the general sense is that these are under discussion and they are being considered seriously.

QUESTIONER: Regarding Chapter 3, I have a question on monetary policy. You know in the analysis, do we have some kind of mass authority to differentiate the role of monetary policy in different countries? For instance, the monetary policy in the U.S., could it be more important than the monetary policy in Eastern European countries?

So, in that sense, I resolve from your chart we can see that monetary policy is not really that important at all, but it is capital flow and in-flow which control our credit growth from the U.S. to other emerging countries.

So, although at some point, countries—perhaps only the U.S., European major countries or Japan—are dominating the sample. So, in that sense, we can say monetary policy, is not as important, but in fact the role of the key countries can dominate a whole process.

So we cannot say monetary policy in general is fine. It’s (inaudible) point out the key role in major countries. Do we have this kind of analysis, given the current global integration and global capital flow?

MR. SCOTT: You will see in the chapter that there is a footnote buried that makes some nod and glance in the direction of answering this question. I can’t remember the number of the footnote off the top of my head.

But, there have been a number of arguments along these lines, that somehow central banks and smaller economies have less influence under their own monetary policy and somehow the Fed dominates the world.

I think it’s important to distinguish exactly what we’re talking about. It’s obviously clear that conditions in financial markets and asset markets in an economy like the U.S. can have ripple effects around the world. That is well documented.

But it’s another question entirely to say, from that, we conclude that monetary policy in smaller economies lacks independence. Now the economies we are talking about are economies where there is free flow of capital, but there’s also a flexible exchange rate.

So, in order for an argument like this to come together, you would need to somehow argue that, despite that, the central bank somehow doesn’t have control over its own monetary conditions.

There might be a situation in which the central bank actually chooses as one of its short-term objectives to try to smooth exchange rate fluctuations, to try to balance out the kind of tradeables, non-tradeables, fluctuations that Olivier hinted at, but that’s not to say that the central banks somehow lack independence and are somehow under the will of the Fed. But I think all of this awaits a lot more exposition and empirical testing than we were able to accommodate in the chapter.

MS. LOTZE: Any other questions? No? Okay, if we don’t have any more questions, thank you very much. We conclude the press conference here. Thank you for coming.


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