IMF: Transcript of the Press Conference on the Release of the Analytical Chapters of the October 2015 World Economic Outlook
Gian Maria Milesi-Ferretti, Deputy Director, Research Department
Thomas Helbling, Division Chief, Research Department
Daniel Leigh, Deputy Division Chief, IMF Research Department
Oya Celasun, Sr. Economist, Research Department
Olga Stankova, Sr. Press Officer, Communications Department
MS. STANKOVA: Good morning, and good afternoon. Welcome to the press conference on the release of the analytical chapters of the October 2015 World Economic Outlook. For the release of the chapters we have here with us in the studio Gian Maria Milesi-Ferretti, Deputy Director in the Research Department, who oversees WEO production and Thomas Helbling, Division Chief in the Research Department in charge of the World Economic Outlook analytical chapters. And we have two lead authors of the two chapters that we’re releasing today: Oya Celasun, the lead author of Chapter 2 on “Where are the Commodity Exporters Headed?” And Daniel Leigh, the lead author of Chapter 3 on “Exchange Rates and Trade Flows: Disconnected?”
I would like also to remind you that the release of the Chapter 1 of the World Economic Outlook will take place on October 6, in the context of the Annual Meeting events in Lima. Today, we are focusing on the two analytical chapters.
With this I will pass the floor for brief introductory remarks to the authors of the chapters, and then we will take your questions.
MS. CELASUN: Thank you, Olga. Good morning, everyone. So Chapter 2 of the forthcoming World Economic Outlook looks at the growth prospects of commodity exporting emerging market and developing economies. Commodity prices have fallen recently -- some, like oil prices, in a dramatic fashion -- and are expected to remain weak. This raises three key questions for policymakers in commodity exporting emerging market and developing economies. First, how will this decline in prices affect growth in the next few years? Second, is this slowdown a cyclical cooling of the economy or a slowdown in the underlying trend of output? And how should policies react?
So to answer these questions we looked at data for more than 40 commodity exporting emerging market and developing economies over more than 50 years, starting in the early ’60s till 2014. And our analysis suggest that the recent declines in commodity prices could shave off around 1 percentage point annually from the average growth rate across commodity exporting emerging market and developing economies this year and in the next two years as compared to the previous three years. And this is really driven by fuel exporters, where the decline in export prices is more recent and, therefore, the slowdown in growth is still largely ahead of us.
So what drives this variation in growth over a commodity price cycle and to what extent does it reflect potential growth? It’s helpful here to just briefly mention two key mechanisms that our chapter finds. One of them is fiscal policy. So with higher fiscal revenues, government spending tends to grow faster during commodity booms than it does in downturns, and that has effects on the broader economy.
The second channel is the financial channel. Public sector, the governments, firms, and households all tend to have easier access to credit when commodity prices are booming than when they are declining. And that also amplifies a business cycle triggered by commodity prices.
So as a result, investment grows faster during commodity price booms than in downturns. This is largely driven by investment in the commodity sector itself, but it’s not confined to it. And, therefore, this cycle in investment means that potential output, which, as you know, depends on the actual capital stock and the underlying trend in labor effort and productivity, picks up during commodity booms and then slows down during commodity price downturns.
So how large is this effect? Our chapter finds that the variation in potential growth over the cycle is about one-third of that in actual output.
So this finding of lower potential growth during a commodity price downswing, even though it’s less than the variation in actual growth, has an important policy implication. With the variation in growth not just being a cyclical phenomenon, but also having a structural component as well, means that the policy response should go beyond demand-side measures and include structural reforms to help bolster growth potential in the medium term.
Here the priorities, of course, vary greatly across countries, but with investment slowing down, work on efforts to bolster the business climate or alleviate supply-side bottlenecks, such as infrastructure bottlenecks, could be very fruitful at this juncture in commodity exporting economies.
So let me close with a key point. Commodity exporters today are better equipped to deal with a downturn than they were in previous episodes. Fiscal policy has been less procyclical in the latest boom, allowing for greater fiscal savings. There’s more exchange and flexibility and there’s greater financial depths. And our research showed that all of these three factors were associated with less of a growth decline during commodity price downturns.
MS. STANKOVA: Thank you, Oya. And Daniel?
MR. LEIGH: Thanks, Olga. Exchange rate movements have been unusually large and have been making headlines, and they’ve also sparked some controversy as to their likely effects on exports and imports. Some say that exchange rates matter far less than they used to for trade and they have disconnected from trade altogether. The argument that’s made is that production has become fragmented across countries, in global value chains. Think of smartphones assembled in China using parts from multiple economies. So a rising share of a country’s exports now includes a foreign value component, basically an import from abroad. And the currency depreciation should thus give a smaller boost to its exports. That’s one argument.
People also point to the limited movements in trade balances in some countries that have had big changes in exchange rates. Japan is one example where there was 30 percent depreciation of the yen, but exports have not grown as much as expected.
A disconnect of exchange rates from trade would actually really complicated policy-making. It could weaken a key channel for the transmission of monetary policy and it could make it much harder to reduce trade imbalances through the adjustment of exchange rates. So our report sheds new light on this issue by taking a fresh look at exchange rates and the relationship with exports and imports. We look at data for more than 50 advanced and emerging market and development economies over the last three decades. This is a broader sample than is typically looked at in studies. Usually the focus is on advanced economies. And it makes sense given the rising importance of emerging markets in world trade.
So we look and, contrary to the view that I just explained, we find strong evidence of a link between exchange rates and trade. A 10 percent real effective exchange rate depreciation comes with, on average, a rise of net exports, real net exports of about 1.5 percent of GDP. That’s quite a lot and much of this effect -- not all, but much of it -- arrives within one year. And, of course, there is wide variation around that average.
But there is, in fact, little sign of a general breakdown in the relationship between exchange rates and exports and imports. However, there is some evidence that the rise of global value chains has weakened the relationship between exchange rates and trade for some products and some economies. For example, for trade in intermediate products that are used as inputs into other economies’ exports. This is particularly relevant for economies such as Hungary, Romania, Mexico, or Thailand that have really stepped up their involvement in these global value chains more than other economies over time.
But let’s keep this in perspective. Global value chain-related trade has generally increased only gradually over the decades and appears to have actually slowed in recent years.
The bulk, about 75% of all exports, is still sourced domestically, so the foreign value content in exports, on average, is still only about 25%. Total exports and imports, that’s everything, we find little sign of a disconnect with exchange rates.
The responsiveness of trade through the prices and of trade prices to exchange rates has not generally dropped over time. This holds for different groups of economies. A key exception to this pattern is Japan which does display some signs of disconnect, although, this reflects a number of very Japan specific factors. For example, the surge in off-shoring production after the crisis and the 2011 earthquake. But overall, there’s little sign of a generalized disconnect between exchange rates and trade so far.
So these results are important because they mean that recent exchange rate movements probably imply a substantial redistribution of real net exports across economies, for example, from the United States and economies that have appreciated with the dollar to the Euro Area and to economies that have depreciated. To sum up, if a policy maker were to ask us if they should radically downgrade the effective exchange rates in their forecasting model we would say no, exchange rates still matter. Thank you.
QUESTIONER: So I had questions for both of you. Firstly, you singled out energy exporters as having more impact than others is that because the fall in oil prices has been greater than for other commodities or is it something else about the energy exporting countries that makes you think that?
Secondly, for Daniel, on Page 7 of the chapter you’re talking about the substantial cross country variation around the impact. Can you give us some idea of which regions or which countries are more or less affected? Also, in the map that you have on Page 19 about the redistribution effects of exports, I see that -- I mean, you’ve got China as a negative, -0.5 to -3. I guess this is because the figures run until June, right? That didn’t take into account the latest moves in the Chinese currency, but would you expect that pattern to change given the devaluation in August? Thank you.
MS. CELASUN: Thank you. On energy exporters, the key reason why their growth drag going forward is larger is what I said, it’s because the price decline is more recent. It’s happened over the last 12 months, mostly. Whereas, for other commodity prices it began earlier, in 2011 for some of them. That’s the main reason. Other than that the decline in oil prices have been somewhat larger. We look at the terms of trade of each country, so we weight commodities by how important they are for a country, so net energy exporters had a somewhat larger decline in the terms of trade. Among the group of oil exporters there are some economies that tend to be quite intensive in oil production, so that also is a contributing factor.
MS. CELASUN: There’s a great amount of variation there. I mean, we have that table in our chapter. We can look at that later.
Mr. HELBLING: Basically all the countries in the Gulf Region are, for example, very intensive in oil production. You have to think, as Oya was saying, that for other commodity prices sometimes countries that are big exporters of other commodities may be importers of oil. Hence, the net effect is a little bit less.
MR. LEIGH: Thanks. The histogram, the effect of the 10% real depreciation giving a boost in the net export of 1.5% on average. That comes with a range, and the range reflects -- well, first general exchange rate effect on trade will affect the underlying responses of exports and imports and prices to exchange rates. They will also reflect how open an economy is. An economy with a much larger share of exports in its GDP the effect of an exchange rate change will be macro economically more important. So in that particular figure what’s really driving the variation is the openness of the economies, so the range, at the low end, it’s about .5. The United States would be an economy towards the lower end of that range. It will go all the way up to 3.5 with more open economies.
But basically, the other point was about the heat map. You’re absolutely right that we stopped the data, as you noticed, in June for the real effective exchange rates which means that since then whatever’s happened is not reflected. But I know that the reason that, as we show, there was an appreciation of Chinese real effective exchange rate. What’s happened since has not come anywhere close to offsetting this appreciation that you saw in the beginning.
MR. MILESI-FERRETTI: Exactly just in terms of overall figures over the past year despite what has happened in August the RNB has appreciated by over 10%, so it remains clearly a currency that has strengthen over the past year despite the recent depreciation.
QUESTIONER: 10% through July or through August?
MR. MILESI-FERRETTI: Over 10% through August.
QUESTIONER: In August alone?
MR. MILESI-FERRETTI: From last year through August it’s up over 10%. At some point it was up 13%/14%, and then it weakened a little bit in August, but the weakening is only a small fraction of the strengthening over the previous 11 months.
QUESTIONER: Can I ask you, Mr. Leigh, about the exchange rate movement of those three countries you identify in Chile, Canada, Australia that are particularly export dependent on commodities that have really come down in price? How would you access the exchange rate depreciation in those countries, and also in Brazil and Russia in terms of what you would call fundamentals, in terms of the commodity price decline?
MS. CELASUN: One thing I would point out that on exchange rate movements what this chapter does is to look at how exchange rate movements themselves influence the effect of commodity prices on the commodity. What our finding is that we split the example into two groups, those with more flexibility, less flexibility. The finding we have is that the ones that had more flexibility in their exchange rates had a smoother cycle in growth, so growth tended to decline less during commodity price downswings than upswings. So it was a factor smoothing the impact of commodity prices on growth. But the movements of exchange rates themselves reflect factors other than commodity price movements as well.
MR. LEIGH: Right. I was going to add that exchange rates can move for many different reasons, swings in commodity prices, shifts in growth at home and abroad. All of this will effect exchange rates. We don’t dissect why exchange rates have moved in our studies, but our chapter’s job was to assess the direct effect of those exchange rates and the depreciation can definitely, according to our results, support the exports, and the net exports of an economy. When there’s a depreciation they can also lead to a switch from importing, when there’s a depreciation, towards buying domestic goods. This is the sort of switching effect that’s supporting.
Then on the export side, well, even if a commodity is priced in foreign currency well then the earnings in the local currency can go up when there’s a depreciation. So all of this can cushion the effect of shocks. I think that’s where the two chapters come together.
QUESTIONER: I’d like to ask about food prices. Why have they not declined in the same ways of these other commodity prices in this cycle? If you look at that chart that you have, I think in your chapter, they’re pretty much still up from say the inflation adjusted prices going all the way back to the mid-70s?
MS. CELASUN: Okay. One thing to note there is that the global demand boom that drove all commodities up to varying degrees over the last ten years or so, within that there were some parts that was particularly metals intensive. We have some work on that. We have a blog on that and more work is forthcoming next week. The intensity of global demand also matters. Metals saw particularly higher demand, including given China’s infrastructure and construction intensive activities over the last several years. Food prices didn’t have the same, you know, they don’t have the same type of demand, I would say. So the composition of demand matters.
MR. HELBLING: Perhaps one thing to food markets or food prices in general. I think food markets are a bit more elastic in the sense of supply response, take much less time. It can happen with the next harvest season. You can also adjust the intensity with which you use fertilizer, the quality of seeds, and so on and so forth. So food markets have responded to increased demand, higher prices, much more rapidly.
Then I think our index also sort of hides a bit. If you look at the major foods, soy, wheat, rice, prices have come down substantially from the peaks in about 2011 and 2012. The last point to consider is that some of the price spikes we’ve seen in 2008, and again in 2011, 2012 were supply shocks in the sense of draughts and other weather related factors which then we were subsequently in the next harvest season.
MS. STANKOVA: Daniel, anything to add to this?
MR. LEIGH: Well, perhaps if I could make one more point that hasn’t been raised which is about this controversy about disconnect. It’s not the first time we’re having this debate. This goes back decades. In the 1980s, as you may remember, there was a very sharp appreciation of the yen and the depreciation of the dollar after the Plaza Accord in 1985. Then for a few years there wasn’t any obvious movement in trade balances, and people we’re debating, exchange rates have become disconnect from trade, but only a few years went by and trade balances responded as expected. So the real question is whether this time is any different. At least based on the evidence in our chapter, so far, there’s no reason to downgrade the effect of exchange rates in our thinking about how the economy works.
MS. STANKOVA: I will move to questions from online, our audience there. Two questions, in fact. First one is the chapter on (inaudible) warning in a world of global currency depreciations of the risk of a cascade of (inaudible) neighbor policies. Second question, why only measure implied impact, not the real impact?
MR. LEIGH: On the first point, his idea is really speaking to how desirable an exchange rate change is or what should be the policy response to that. That’s an important question, but it’s more complex than what we look into. We’re trying to simply answer the question, do exchange rates, have they lost their effect on trade? Because that’s actually a very important, necessary part of the discussion if we’re then going to talk about what is the right response to exchange rates. We’re finding, again, that they do matter.
What do we mean by they matter? This comes to the second question. This is a real effect we’re talking about when there’s a real depreciation of a currency, for example, by 10%, that real net exports go up by 1.5% of GDP. Most of that within one year, so that’s a real effect. The only thing to note there it’s the direct effect of an exchange rate, and there are many indirect effects also. So when an exchange rate moves it can have an effect on household balance sheets, on corporate -- many different parts of the economy, and that’s a much more complex exercise. So the overall effect on the economy is not something we’re getting into, but we think the export angle is a really important one to get straight.
MR. MILESI-FERRETTI: I would just add that, of course, the inferred overall effect on the economy you have to know what has triggered the change in exchange. So if the exchange rate is triggered, for example, by reduced prospects for domestic growth then overall growth is not going to do as well, but the exercise that the chapter undertakes is saying what does the exchange rate change contribute in terms of growth, holding everything else constant. But clearly, to assess the overall effect when we do general equilibrium analysis we need to take into account the entire constellation of factors effecting the economy, and in particular, what has driven the change of the exchange rate in the first place.
Ms. STANKOVA: Thank you. Then question on Chapter 2: Your report supports that commodity exporting quantum is likely to slow further given commodity price outlook. Do you have any projections on the cost of international shipping industry?
MS. CELASUN: I think on the shipping industry if commodity markets are weak then you would also expect some implications for the shipping industry to the extent that for some parts of shipping commodity shipments are very important. To the extent, as in previous WEOs we have described global growth and global activity has slowed, so the shipping industry in terms of prices has been relatively weak.
MR. HELBLING: And, of course, they also have an impact on cost through the reduced fuel, so you will have a decrease in demand for shipping commodities to the extent that trading commodities has weakened, but lower cost of shipping through lower fuel.
MS. STANKOVA: Thank you. We do have questions, they are more country specific. That’s on the Brazilian real. The Brazilian real is one of the currencies that’s most devalued in (inaudible) terms since mid-last year, and the Central Bank has indicated it will fight against more depreciation. Is it the right policy move considering that Forge’s devaluation can boost exports, as IMF showed?
MR. LEIGH: On that, again, the question is really referring to the right policy response and how desirable that change in the exchange rate is. That goes beyond what we look at. Again, to even have that discussion we need to get straight whether exchange rates can actually have an effect on exports and imports. Beyond what is desirable or why the exchange rate moved we need to get that straight, and that’s why it’s so important that we looked at this in the chapter.
Ms. STANKOVA: I would like also to remind everybody that we will hold a Regional Economic Outlook releases next week in Lima that’s, obviously, a good opportunity to ask country specific questions. Another question on-line is definitely for that type of briefing. What is the position of Egypt currently with the implementation of new economic programs? This is from a newspaper in Egypt.
If we don’t have any more questions from the audience, we don’t have more from on-line. Yes?
QUESTIONER: One question on the United States in terms of this 10% appreciation of the dollar. How would you calculate from what you said earlier the impact on American exports over the next year or two?
MR. LEIGH: Well, actually, it’s all explained in the chapter. What goes into this calculation is, first of all, if you’re focusing on exports you need to know how much the real effective exchange -- was the depreciation of the dollar relative to all its trade partners. Then you need to know how much that depreciation was going to effect export prices that the foreigners, the buyers of U.S. goods and services, pay. That’s one of the points we provide an estimate for. Then you need to know how much, given that change in the prices that foreigners pay for U.S. products, how much the demand response to that.
This sort of chain is what we estimate. Then, of course, the ultimate effect on the U.S. economy is going to depend on, once you know how much are volumes in percent of the total economy. On that, that’s what the calculations are based on. Everybody can compute that based on the numbers that are reported in the table. The histogram simply reflects for all the countries. The U.S., as I said, as everybody knows, across countries is one of the economies where exports and percentage of GDP are on the lower side than other economies, so it’s going to be towards the lower end of this chart, so in the 0.6% or 0.7% of GDP range. That’s basically what I can say.
MR. LEIGH: Did that answer your question? Again, that’s for the direct effect on real net exports.
QUESTIONER: I’d be interested if you have a table or a list of the countries with more specific numbers referred to them that would be helpful.
MR. LEIGH: Well, what I can tell you is that we’ve made one of the points in this chapter which is that estimating these individual effects across countries is actually -- it’s more informative to look at the averages. There is, you know, for these more than 50 countries we look at, so some of them there’s a short amount of data. There’s noise. The best thing is to pool the experience, sort of the wisdom of the crowd, to get an average effect. That’s what pins down these elasticities. They’re reported for the average of groups in Table 3.5.
Then anybody who knows what exports and imports are in percent of GDP can easily compute for any country the effect on net exports and percent of GDP. We gave a sort of illustrative range in the chapter. But we can actually follow that up off line if you have further technical questions on that.
Ms. STANKOVA: With that, thank you very much for coming and joining us for the press conference. Have a good day.