Transcript

Transcript of an IMF Economic Forum - Capital Flow Cycles: Old and New Challenges

November 7, 2003

(View this Economic Forum using Windows Media Player.)

Participants:

Zanny Minton-Beddoes (Moderator), Washington Correspondent, The Economist
Agustin Carstens, Deputy Managing Director, IMF
Jeffry Frieden, Professor, Department of Government, Harvard University
Peter Garber, Global Strategist, Deutsche Bank
Morris Goldstein, Senior Fellow, Institute for International Economics

MR. MODY: My name is Ashok Mody. My task is a simple one. I'm here to welcome everyone on behalf of the Research Department. This is also a joint seminar with the External Affairs Department of the Fund, and Zanny Minton-Beddoes has kindly agreed to moderate this panel, which is the culmination of what I hope you will all think was a fine conference. Zanny writes on the U.S. economy and international affairs for the Economist, and I'm also pleased to say she has been with the Fund in a prior incarnation. Thank you.

MS. MINTON-BEDDOES: Thank you, Ashok, and welcome to this concluding panel of our Research Conference.

For those of you who have been here for the past day and a half, many aspects of the relationship between capital flows and macroeconomic cycles have been addressed. We had discussions on exchange rate management. Just before lunch we had a discussion of how creditors have fared from their investments in emerging markets. And I guess the plan in this concluding panel is really to put many of these thoughts together. We all know that capital flows have huge potential when they're harnessed effectively. They augment domestic savings, boost investment, can boost productivity growth. But we've also all learned that they can be very dangerous. They can be volatile. We tend to have famine or feast. And they're often used inappropriately by the recipient country.

The temptation during a feast is for imprudence, whether it's by the government or by the private sector. Too much capital in many emerging economies can overwhelm badly regulated weak financial systems, and sudden stops, as we've seen, can cause jarring financial crises. And I think in the last decade or so, we've had plenty of experience of those kind of crises. The immediate causes were different, but we saw the painful effect of what happens when capital flows dry up.

In more recent years, we've also got a parallel development going on in global capital markets, which is the huge U.S. external imbalance and the huge current account balance here and the way that it is being increasingly funded by the central banks of East Asia, largely. And I think putting those two big developments together is really what this panel, I hope, will do. We've had a lot of evolution in thinking in the last few years, much of which--some of which you heard about yesterday and this morning. And as the Managing Director said yesterday, as we may be on the cusp of a new upward cycle where we go back to a period of feast, perhaps, it's all the more important that policymakers and analysts understand how better to deal with international capital flows.

So I hope that this panel will help us do that. In particular, I hope we discuss how capital flows can be effectively harnessed, what policy ideas are there at the international level to make them safer, and how should emerging economies themselves act to mitigate the volatility of private capital flows.

This is a really outstanding panel to discuss these issues. We have four people with extremely high levels of experience, both in management of financial crises and in analyzing them. And I think I will introduce all the panelists at the beginning, and if we could start with Agustin Carstens, who has been since, I think, August 1, Deputy Managing Director of the IMF, before that was Mexico's Deputy Secretary of Finance, and before that spent most of his career at the central bank, where in 1995 he had a very close and pivotal role in managing Mexico's crisis.

On my immediate left is Morris Goldstein, who is officially the Dennis Weatherstone Senior Fellow at the Institute for International Economics. I think unofficially he is Mr. Financial Crisis. Basically, he has analyzed--

MR. GOLDSTEIN: Not causing them, just analyzing them.

[Laughter.]

MS. MINTON-BEDDOES: Analyzed if not all, certainly virtually all of the more recent financial crises in great detail. I know of no one who has a more thorough understanding of what happened in each of these crises and often a more provocative view on what ought to have been done differently.

To my immediate right is Peter Garber, who is global strategist at Deutsche Bank. He, too, is one of the top authorities on financial crises and the author of numerous books and articles on the economics of speculative attacks, speculative bubbles, and other such unattractive occurrences.

Finally, on my far right, Jeffry Frieden, professor of government at Harvard University. He specializes in the politics of international economics and particularly international economic crises. He's the author of numerous books that try and help us understand why countries act and react the way that they do. And I'm sure that's what he'll do today.

So, without further ado, I turn the floor to Agustin Carstens.

MR. MODY: Zanny, just--

MS. MINTON-BEDDOES: Yes?

MR. MODY: Comments are on the record.

MS. MINTON-BEDDOES: Oh. Comments--since I'm a journalist, I assumed that. Everything is on the record. I think you're being web cast, actually, so you will immediately be being transferred to the Internet.

Agustin?

MR. CARSTENS: Well, thank you, Zanny, very much for the introduction, and I thank very much the Research Department for having invited me to participate in this panel with very distinguished colleagues.

I had hoped to be the last one in the panel because then I could also benefit a little bit from their comments. But well, anyway, I will have now to jump into the pool and start swimming.

Even though I didn't have time to come and attend some of the very interesting meetings, I had the opportunity to browse through some of the papers. I'm sure that you have here discussed at length what are the main motivations for capital inflows and outflows in different shapes and forms and discuss pretty much the usual collateral damages that these flows might have on the macroeconomic scene.

Therefore, I will try to add some value to the discussion and in a way try to make sort of a laundry list of policy recommendations that I tried to put together in a very informal way, drawing on my previous experience as a policymaker in a country which has a very rich history in capital inflows and outflows, as is Mexico; and also in a way to reflect a little bit of the very rich experience, short but rich experience, that I have had here in the Fund.

When I came here to the Fund, somebody said to me, "Why are you leaving Mexico given that there is never a dull moment in Mexico?" And all I can say, "Well, I'm happy to have come to a place where there is never a dull moment, either."

I would say that the very first recommendation is that countries should have a very strong focus on adopting a strategy to develop internal financial markets and systems. At the end of the day, history tells us and economic evidence tells us that internal sources of finance are the most reliable and the most important. Therefore, if you concentrate in that, you will be ahead of the curve most of the time. And it has also the added benefit that if you do these tasks well, you will be also successful in attracting complementary capital flows from abroad. And, in addition, I would say that not only the capital flows will come along, but they will be what I would call good-quality capital flows.

Of course, to develop the internal financial system, you fall into traditional measures, which is basically the fundamentals, the basics for a solid macroeconomic management, together with a good rule of law, with good financial supervision and regulation, with the emphasis on human capital to supervise and regulate. But, in any case, I think that countries, when addressing these issues of capital flows, they should first turn inside and see what are they doing in their own financial system and see how that system is performing, and in a way leave on the back burner, the considerations about capital flows, because I think that they will follow naturally.

The second point is that at all times--even if you don't have an open economy--you should be extremely aware of what is going on with the health of your financial system, which is a different issue from trying to develop it. You just have to be on top of what is going on there.

What I have learned in my own country and what I have learned in the very few months I have been here at the Fund is that the biggest surprises that countries face oftentimes are weak financial systems in crisis, banking crises, that have the potential of doubling your debt-to-GDP ratio in a matter of months. I really haven't seen a more deterrent to development than having a derailed banking system. And this is important because if you have a surprise there, that will generate immediately huge capital outflows.

And the same thing would happen if you are not careful in seeing how your local financial system is managing the capital flows. This implies that in an open economy, it's paramount to have a very well regulated and supervised financial system. A corollary would be that a country shouldn't open its economy if it is not ready to assess at all times what is going on in that very key market.

The third recommendation would be that the use of derivatives and indexing as an instrument to entice or maintain capital flows can be hazardous for the health of the economy. There are many theoretical arguments that advocate for the development of these instruments if you have incomplete markets. Certainly I am not arguing against this. My point really is that the use by the government of derivatives and indexing as an instrument to entice or maintain capital flows can be hazardous for the health of the country and, in particular, for the public finances. Under certain circumstances, especially where a country is under balance of payment pressure, there is the temptation to substitute gimmicks for good policymaking, and a usual representation of these gimmicks is to put bells and whistles to government debt via derivatives or indexing, or have the government be involved in derivative markets taking explicitly exchange rate risk. Authorities have tried this, thinking that they are buying time and, at the same time, they are sort of relaxing the most important constraints. But at the end of the day, there is no substitute for decisive macro measures, and sooner rather than later, the abuse of indexing and derivatives will come back and eat you alive. So there should be a lot of care being put into the use of these instruments.

The fourth lesson would be that it's a mistake when you are dealing with an open capital account to fool around with not being transparent. Many governments have tried to use the management of information as a policy instrument. And, again, usually it is done trying to gain time to see if you can avoid taking the hard decisions usually on corrective macro measures. I think that markets are now very sophisticated. I think that the space for abusing the management of information is long gone, and I think that it really works against you if you if you try to pull that play in front of the market.

Actually I think that the debate of issues inside international financial institutions can still be enhanced to have a more candid and transparent discussion about many issues. For example, more than once we have had the case that some exit strategies from an exchange rate regime never was discussed just because of the fear that to have that discussion might itself destabilize the regime. At some point, we have to find a way out of this type of trap.

Of course, this becomes a very important issue in the dynamics of capital flows because if you're successful in not providing full information but at some point that information becomes apparent or impossible to hide, you just will have compressed the fury of the market in a very short span of time, and then the situation is really unsustainable and chaotic.

The fifth lesson is that it is appropriate for countries to self-insure against shocks and volatilities in capital markets. There has been a lot of discussion about the appropriate level of international reserves among many countries. Obviously, there is this debate in the context of exchange rate regimes in Asia. I think that many countries have learned that, given the way some policies at the Fund have developed, it is prudent to depend on your own resources, and therefore if you face a lot of uncertainty in the capital markets, it's good that you build up your own reserves. This holds even if you have a floating exchange rate regime. Here, the key motivation would be that even in bad times the country will have the foreign exchange for the private or the public sector to service its debt. This does not mean that the central bank will be the lender of last resort; it only would be the provider of foreign exchange in a transparent and fair basis. This type of policy would make the rollover of debt easier.

I think it's important also that countries make conscious decisions on how are they going to fund themselves in the market. In Mexico, for example, the policy Mexico adopted three years ago has been not to go to the international markets to finance its fiscal deficit. And the Mexican Government hasn't asked for authority to borrow abroad in the last four years. And that obviously has enhanced the credit rating of Mexico in the markets, mostly to the benefit of the corporate sector. And at the same time, you can develop a domestic capital market and rely on it to finance yourself.

The sixth lesson is for countries to pursue FDI, but they should not feel so confident about it. What I mean here is that treasurers of a multinational company run as fast as a hedge fund, especially if they have the opportunity to shorten the local currency, either through derivatives or just borrowing in the local financial market. So it's important to try to attract FDI because it can certainly bring capital, it brings technology, it creates employment, but don't be so sure that they will keep invested in the country in times of uncertainty. The fact that there is a physical plant installed in a country does not imply that the capital will not flow out in case of instability.

So these are six policy lessons for countries to follow so as to avoid capital flows interfering with macroeconomic performance. There is another category of policies that can be promoted by institutions like the IMF The Fund recently has been advocating two branches of policy research that in some way relate to the importance of capital flows, and macroeconomic development. One branch has to do with crisis prevention and the other with crisis resolution. With respect to crisis prevention, the Fund has been advocating for standards and codes, enhanced surveillance, very broad discussion about exchange rate regimes and development of internal financial markets.

Overall, I think that since the late '90s the international financial system and community have had some success in improving the resilience of capital markets. The other day I was reminded that in 1998-99, 30 percent of the countries represented in the EMBI [Emerging Market Bond Index] was investment grade, versus 60 percent currently. So if we think that the rating agencies do a good job in assessing the creditworthiness of countries, there has been something that we have done right.

And where more challenges still exist is in those cases when things go wrong and we have to deal with the resolution of crisis. Even though the Fund has been in the center of the debate, we have at best made partial progress through the countries adopting collective action clauses, but with respect to the more procedural aspect of how to resolve a crisis this is still an open question.

So I'll stop here. Thank you.

MS. MINTON-BEDDOES: Thank you very much.

Morris, let's turn to you.

MR. GOLDSTEIN: Well, thank you, Zanny, for that kind introduction, and it's a great pleasure to be back in the Research Department. And I want to thank Ashok and Rajan and others for inviting me.

I thought I would spend my 12 minutes or so on two issues: first, the trilemma, as Obstfeld and Company call it in their paper, now playing out in China, and, second, what emerging economies can do to reduce their vulnerability to sudden stops in capital flows and to exchange rate volatility.

One of the more interesting sets of trilemma issues is occurring in China. My summary of this situation is as follows: International capital is flowing into China at a very high rate at the same time that bank lending is expanding at a record pace and that the economy is overheating.

In the first half of this year, the surplus in the overall capital account was 7 percent of GDP, much higher than the 1.5 percent of GDP average surplus over the previous four years. International reserves have grown by about $150 billion over the past 18 months. Bank lending in the first three quarters of this year expanded relative to GDP by 35 percent, an all-time high, and this comes on top of a ratio of nonperforming loans to total loans perhaps in the neighborhood of 30, 35 percent.

At the current pace, the increase in bank lending in 2003 will be greater than the cumulative increase in 1998, 1999, and 2000 combined. The investment share of GDP is above 40 percent, also an all-time high. Property prices in Shanghai are going up at a 25- to 30-percent annual rate. In the third quarter of the year, real GDP grew by over 9 percent. Inflation is still running very low but is rising. Interest rates are also low but are rising as the authorities have to sell more and more domestic bonds to sterilize. They're sterilizing hand over fist. They need tighter monetary policy and they need control over bank lending.

By my calculation, the real exchange rate of the RMB is undervalued on the order of 15 to 25 percent. You can get that estimate either by asking what change in the RMB would make the underlying current account position in China equal to normal net capital flows, assuming that they don't liberalize capital outflows significantly anytime soon. Or you can get a similar estimate by asking what the RMB's contribution should be to the fall in the dollar needed to reduce the U.S. current account deficit to a more sustainable level.

Treasury Secretary Snow has asked the Chinese to freely float the RMB and to completely open the capital account. This would be one solution to the trilemma, and I think it's sensible advice for the longer term. But I think it's not good advice for now, and it will not get very much response. Why not? Well, because I think the Chinese are justifiably concerned that if they open the capital account freely to outflows and float the rate, and if there were bad news about the banking system--and I think there is bad news on the horizon--they could get large-scale capital flight and a large depreciation of the RMB. Household saving deposits are equal to 100 percent of GDP.

Well, what would be a better way to solve the trilemma? My IIE colleague Nick Lardy and I have proposed that they reform the currency regime in two steps:

First, they switch from a unitary peg to the dollar to a basket peg with equal weights for the dollar, the yen, and the euro; they revaluate by 15 to 25 percent immediately; and they widen the exchange rate band from less than 1 percent to, say, 5 to 7 percent.

Step two, after they get the banking system on a sounder footing, which may take a few years, they open the capital account and float the RMB. In this way, we don't ask the rest of the world to live too long with a seriously undervalued RMB, and we don't ask China to repeat the lesson of the Asian financial crisis.

I would say that the Chinese case reinforces the message that you need to phase in the liberalization of capital flows to the health and resilience of the domestic financial system. When you have to worry about the banking system, you really have not just a trilemma but a quadrilemma, if there is such a thing.

The Chinese case also suggests that if you do liberalize the account in stages, you have more room for maneuver, I think, in your choice of currency regime. You can fix for a while, but I think you want to do it at an equilibrium exchange rate, not one that is seriously undervalued. If the Chinese don't solve the trilemma in the right way, I think we're going to see a big blowout in the banking system and a large upsurge in inflation.

Next, vulnerability of sudden stops in capital flows and to exchange rate volatility. Here I think there are two things to watch out for, and that is, too much public debt and currency mismatches. A few words about each.

Chapter 3 of the last World Economic Outlook had some sobering thoughts for anybody who'd thought that the debt crisis we've seen in Argentina, Russia, Uruguay, Ecuador, Pakistan, the Ukraine, and the close calls in Brazil and Turkey are a thing of the past. The report notes that the ratio of public debt to GDP now averages about 70 percent in emerging economies, reversing progress in the first half of the 1990s; that over half of debt defaults occur at public debt ratios less than 60 percent; that the typical emerging economy has a public debt ratio about two and a half times larger than its fiscal policy track record suggests is prudent; and that governments usually fail to take corrective policy actions once the public debt ratio climbs above 50 percent.

One of the reasons why I think we've seen such volatility in private capital flows to emerging economies and such sharp upward spikes in interest rate spreads is that public debt and often external debt are just too high. So when you get bad news or adverse shocks, questions arise about whether the debtor will pay.

In the past, I think we've simply not been conservative enough about what's a safe or sustainable ratio of public debt. We used to think that 50 or 60 percent of GDP was not such a big number. I think events have shown that it is a big number. I think we've not paid enough attention to the foreign exchange constraint. We've not paid enough to contingent liabilities that start out in the private sector but don't stay there, to spillovers among currency banking and debt problems, including large-scale holding by emerging market banks of emerging market sovereign debt, and to the frequent resort--too frequent--to exchange rate-linked debt. I don't know if there are any gimmicks to get countries to broaden tax bases or to shoot for fiscal surpluses during upswings, pare down safety nets, and reduce the excessive dependence on foreign currency-denominated and -linked debt.

Some governments seem to muster determination to control public debt and others don't, and sometimes the same ones don't do it for a long time and then get religion.

Brazil's track record in controlling public debt was terrible over the 1994-2002 period, but in the last ten months, there have been some very positive signs that they're doing better, much better than most analysts, certainly myself included, thought they would do.

For its part, I think the IMF has to be tougher than in the past on making debt sustainability a key condition for Fund lending.

Turning to currency mismatches, in my view this is at the heart of emerging market vulnerability to sharp exchange rate depreciation and ought to be higher up on the priority for reform of the architecture. All the prominent financial crises in emerging economies during the past decade have been marked by large mismatches--Mexico in '94-'95, the Asian financial crisis in '97-'98, Russia in '98, Turkey in 2000-03, Argentina 2001-02, Brazil both cases. Currency mismatch variables have proved to be one of the better performing leading indicators of currency and banking crises, and we've seen that output contractions tend to be deepest in those emerging economies with large mismatches and large exchange rate depreciations.

These mismatches also undermine the effectiveness of monetary policy during a crisis. If you've got a big mismatch, the government is going to be very reluctant to reduce interest rates after a contractionary shock because they're worried the exchange rate goes into free fall and then they have a wave of bankruptcies.

Currency mismatches in emerging economies are not, I think, inevitable. The good news is that the aggregate mismatch has declined substantially in emerging Asia since 1997-98. The bad news is that there remain some emerging economies in Latin America and at least one in Asia where mismatch has worsened over the past six to seven years. And the improvement that has occurred elsewhere could prove to be transient if the right policies aren't maintained.

Well, what are the kind of useful things to reduce mismatch? Well, for those emerging economies that have substantial involvement with private capital markets, they should opt for a currency regime of managed floating. The de facto movement in the nominal rate will produce an awareness of currency risk and incentive to keep mismatches under control. A monetary policy framework of inflation-targeting should be employed to provide a good nominal anchor against inflation. Good inflation performance is crucial for developing a healthy local currency-denominated domestic bond market, and here I agree with Agustin that I think this is one of the key things.

Banks in emerging economies should apply tighter credit limits on foreign currency loans to customers that don't generate foreign currency revenues, that is, loans to nontaxable producers, and banking supervisors should strengthen regulations and capital requirements on banks' net open positions in foreign exchange.

The Fund should be publishing data regularly on currency mismatches at the economy-wide and sectoral levels and should comment on mismatches regarded as excessive. They should make reduction of currency mismatches a condition for IMF loans in cases where they're deemed to be too large.

With Phillip Turner of the BIS, I have constructed an index of aggregate effective currency mismatch that tries to get at the currency denomination of assets and liabilities as well as cross-country differences in export openness and the foreign currency share of the total bond market, domestic and external.

Countries that have a high share of public debt denominated or indexed to foreign currency should adopt a medium-run objective of reducing that share. If you have a bad track record on inflation, I think it's better to use inflation-indexed bonds than exchange rate-indexed as a transition device to fixed-rate currency-denominated debt.

Let me stop there.

MS. MINTON-BEDDOES: Thank you, Morris. An extremely trenchant analysis and a list of advice that actually I couldn't keep up with writing. Thank you.

Let's turn to Peter.

MR. GARBER: Well, thank you, Zanny, very much for your kind introduction and also to Ashok for having invited me.

Since this is for the record, I have to disavow all connection with Deutsche Bank and say that this is my own opinion. But I also must say, according to recent regulations, that I actually believe what I'm about to tell you.

[Laughter.]

MR. GARBER: And it's true, I do believe it.

In a previous talk here in the summer, Mike Dooley and I laid out our vision of what the global monetary system is now, and we argued that it's starting to look a lot like the system in the heyday of the Bretton Woods regime in the mid-1950s and through the 1960s in that a large chunk of it is a fixed exchange rate regime or a very closely managed regime with major countries. A large amount of international finance is now going through the official sector. There's accumulation of reserves. And another feature is that the periphery countries of the system, which was then Europe and Japan, the periphery countries of the system now have macro weight in the flow of capital across borders. Whereas, in more recent periods of 15 years ago, 10 years ago, the periphery countries, the emerging markets had very little macro weight, now they're very important and they are the dynamic forces in the movement of capital across borders.

In particular, we divided the world system into three different kinds of zones, one being what we called a capital account zone, where capital flows are driven primarily by the normal risk-return calculations of the private sector. These are open economies with free flows of capital. Europe would fit into this--would be the archetype of this zone, but other countries, such as the commodity producers--Canada, Australia--would also be part. And even some of the Latin American countries which have freely flowing capital would also be part of this, flows in and out, private sector-determined flows.

The other part, a second region is what we call the trade account zone, which Asia is the representative of that region, whose capital flows are driven more by a development strategy which is an export-led development strategy. So capital flows are subordinated to that. These are fixed exchange rate systems, many of which have capital controls, and with heavy intervention in the exchange markets and, in particular, official flows are of equal magnitude to the imbalances and surpluses in current account flows.

And then the third region is the old center region, which is always the U.S., which is kind of a buffer that absorbs--it's open and it absorbs these flows for a price.

This particular world system has developed lately into one in which there are very large imbalances that have emerged, notably between Asia and the rest of the world, and in particular with the U.S. And it's a system in which the dollar has started to weaken, but only vis-à-vis the capital account zones. Since the beginning of this year, the dollar has weakened against the Aussie dollar by 27 percent--or the Aussie dollar has appreciated by 27 percent against the dollar; the euro by 10 percent but over the last 18 months by 25 percent; the British pound over the last 18 months by 17 percent; even the Brazilian real and the Argentine peso have risen by 21 percent and 16 percent from their earlier low levels.

On the other hand, the Asian economies, the renminbi has not appreciated, has moved zero; the Korean won has moved zero; the ringgit has moved zero; the Taiwan dollar has appreciated by 2 percent; the Thai baht by 7 percent; the Philippine peso has depreciated 3 percent. The noteworthy country here is Japan, which since the beginning of the year has appreciated by 10 percent, and by 5 yen down from the former entrenched line around 115 yen. But all of Asia effectively is flying on a fixed exchange rate system vis-à-vis the dollar and vis-à-vis each other.

So those are the zones, and with Asia running large trade surpluses with the rest of the world, and particularly with the U.S., that has led to certain trade frictions and also problems for Europe now that the imbalances have been recognized and the private sector investors in Europe and elsewhere are pulling back from the U.S. because of the rising risk-return--or the declining risk-return ratio in the U.S. as the risks in the U.S. are perceived to grow.

But this has not had any effect, even though it has appreciated the euro dramatically, it has had no effect on yields in the U.S. On the contrary, yields have come down, spreads have come down. The stock market is up. Financial assets are rising in the U.S., not falling. So there is not a lack of capital because of this pullback.

As a result, we must conclude that since savings in the U.S. hasn't risen appreciably, we must conclude that the capital that Europe is being reluctant to put into the U.S. is coming from someplace else, and it's just being replaced by Asia into the U.S..

Asia effectively is underwriting the U.S. capital--the U.S. assets, and the other side of that coin is that they're running huge current account surpluses.

Well, where does this fundamental disequilibrium come form? It's recognized as a problem that's going to grow, and it's got to come to a bad end sometime. Recognition of that, if we do the dynamic differential equation and we move back to the present, the resolution should be soon. But we argue that it really won't be very soon because Asia is willing to underwrite the U.S. demand for its products as part of its development strategy for the indefinite future.

The fundamental global imbalance is not in the exchange rate. The fundamental global imbalance is the enormous excess supply of labor in Asia now waiting to enter the modern global economy. The exchange rate is only the valve that controls that rate of entry.

For China, there's a conflict between external trade pressures and the internal political pressure to increase employment. In China, we reckon that there are 200 million underemployed people who will in the end be employed in the modern economic system, the global economic system, and a quarter of them in the export sector. This is an entire continent worth of people, a new labor force equivalent to the labor force of the EU or North America and certainly more than in Japan. So this is a macro effect.

The speed of employment of this group is what will in the end determine the real exchange rate. And the renminbi will, therefore, not be allowed to appreciate rapidly because, however strong the pressures are from the external--from the rest of the world, there's an internal pressure as well that comes into balance.

Well, where did this global imbalance come from? The analogy I like to use is that 15 years ago there were two isolated planets that were circling the Sun. One of them had three major capital-rich industrial regions--Europe, the U.S., and Japan--and a periphery of small countries more or less in a fully employed equilibrium with floating exchange rates and more or less capital mobility. The other planet was a communist/socialist world that had large amounts of labor and misallocated and valueless capital, valueless to the rest of the world, anyway.

They did not communicate much with each other, and then suddenly they were pushed together to form one large market, global market economy. And this moved the previous industrial world from a fully employed system to being a region of a depression-ridden, disequilibrium, 1930s style Keynesian system with massive global unemployment. There were two ways to correct this imbalance, that is, encourage the growth of internal demand and the importation of capital in the underemployed regions--that is, have huge exports of capital from the industrial zone--and have internal development, internal demand, develop the underemployed zone. That's kind of like the Washington Consensus.

The alternative--and this is basically the way Latin America chose its development path. The alternative is to run an export-led development program with controls, subsidization of the export sector, because at least in that sector you know that there's world-class--there's world demand for the products you're producing, you're producing world-class products, and you will have a capital stock at the end of the process that will actually produce something that people want in the end. Even if the rest of the world doesn't want it anymore at the time that this all falls apart, at least it can be sold internally, unlike the other development strategy which produces mostly recurring crises and failure.

So Asia, in particular China, chose the latter route, undervaluing the exchange rates and exporting capital to the richer zone. In particular, they exported capital to the U.S. in the form of official sector capital exports, and then in amounts that basically supported the exports of goods to the U.S. And, in addition -- [tape ends].

-- million in the export sector a year. This is where 8-percent growth comes from, and exports have to grow in order to do this at 20 to 25 percent a year.

Actually, that only absorbs about 10 million. To absorb 12 has required bumping up the rate of growth to 9 percent this year.

Now, exports generate about 10 percent of value-added in GDP, and the export sector has grown twice as fast as the rest of the economy. But since the export sector is actually more labor-intensive than the rest, this is a very important outlet for employment growth. And this underscores the strategic importance of maintaining growth in the export sector.

Now, if you do this arithmetic and accept the number 200 million that need to be covered, then the conclusion you reach is that this--when you divide 200 million by 12, then you come up with something just--something below 20 years during which this can be maintained with this level of growth, actually producing value. This is a free lunch that's being produced by this policy.

This policy basically adds the equivalent of a medium-size country labor force to output every two years. This is a France or an Italy, not quite a Germany, but every two years this is entering the labor force.

Now, a 10-percent appreciation, which is actually much less than has been suggested elsewhere, of the renminbi we calculate would eliminate half a million jobs in the export sector, which is about 1.5 percent of the jobs. It's not a huge decline, but no more growth. So that means it would cut the knees out from under the growth rate of China and create an enormous political problem. That's why they're not going to do it.

M2 growth is stabilizing now above 20 percent. This is the banking sector growth that Morris was talking about. But 16 percent M2 growth is consistent with price stability, with a 10-percent growth rate, real growth, or a 9-percent real growth rate, and 6 or 7 percent from the trend velocity reduction that we observe. So inflation has not really taken off. It moved from minus 1 percent a year ago to about 1 percent now. There's no real inflation now in spite of the purported overheating. But we reckon it could rise to 4 or 5 percent in the next couple of years with this differential.

But the central bank can raise reserve requirements, which it has started to do, and it can raise the interest rate. The market rate on central bank bills, of which the number is expanding, is still 2 percent. That's the market rate in a controlled market.

Now, there is all kinds of room to soak up this credit, so this can go on for a very long time without a large amount of overheating.

Now, I've talked about China, but it's not really China. This is all of Asia's surplus. China is just an assembly center for the rest of Asia, or is in large part an assembly center for the rest of Asia. The imbalance is between East Asia and the rest of the world. In 2002, the official sector capital outflows from Asia was about $200 billion, all of Asia, net, and that exactly matched the current account surplus. So basically they financed 100 percent of the current account surplus, mostly by buying low-yielding treasury bills, low-yielding dollar claims, effectively having financed all of the Afghan war, all of the Iraq war, all of the homeland defense expenditures at very low interest. So this is somewhat of a mutually beneficial arrangement.

Now, obviously there's a build-up of claims against the U.S. The U.S. is becoming indebted as a result of this, and so there will be a day of reckoning if this goes on for 10 full years or some protracted length of time. But there's a degree of self-financing of the U.S. imports. The claim is that the Chinese are acquiring our low-yielding dollar claims, but a large amount of the exports is generated by U.S. FDI, which is a very high-yielding renminbi value claims in the end. And so there's a service account flow that will emerge to finance part of this. Effectively, they are lending money to the FDI investors. They are likely--again, as Charles de Gaulle and Jacques Rouef (ph) said in the 1950s, they're lending--because of the financial intermediation of the U.S. as the center country, the U.S. gets to borrow at low rates from the periphery countries and reinvest and acquire their entire capital stock. So a large part of the financing--a large part of these goods are basically going to be self-financed.

Now, at the end when there's the day of reckoning and this has to end, which it will once they graduate--once China graduates from the periphery to the center, then the exchange rate has to appreciate. There's no doubt about that. The renminbi will appreciate remarkably. And that will destroy another big chunk of the net claims of China against the U.S. because those are dollar claims.

One last bit, and that is that this is obviously an IMF Research Conference, and so I'd like to leave a research message. In recent years, in a recent decade, the direction of international financial research has been about small, open emerging market economies with no macroeconomic weight in the rest of the world. And so you can look at them through a very, very limited dimensional set of glasses.

In the original days of research, in the heyday of research at the IMF, the central issue was research about how the system operated and how to make the system work. Now we are--and that occurred because the peripheral countries were of macro weight and it was all about the system. Now we are back in a world just like that where the peripheral countries have macro weight and now the focus of research should be much more--should return once again to how the system will work.

MS. MINTON-BEDDOES: Thank you, Peter.

Jeffry, now the politics of this.

[View Mr. Frieden's presentation (56 KB pdf file)]

MR. FRIEDEN: Well, thank you. I'm going to do exactly what Peter says we should not be doing, I guess, if I can get this to...okay. I do have a PowerPoint presentation, or at least I thought I did, which is undoubtedly unnecessary but it has the great advantage of putting a lot of distance between Zanny and me so that when I run out of time, she'll have more difficulty getting me to shut up.

In any event, what I want to do is stand back a bit and look at something, a bit of a longer view, and some enduring regularities about capital flows cycles, and in doing that I want to focus really on two dimensions, which we haven't talked much about. They've been mentioned, but we haven't talked much about them. The first dimension is the domestic aspects of these capital flows cycles; that is, what does it look like from the standpoint of a country experiencing a capital flow cycle; And the second, given my comparative advantage of the political economy factors, none of this is meant to mean that the other global or economic factors are not important. Peter has made a very strong case, and others have as well, about the significance of those factors. But I think that it is worth focusing or at least beginning to focus on what a capital flows cycle implies for individual countries at the domestic level and thinking about some of the political economy of these crises.

I'm going to start with a very stylized picture of this national cycle, the cycle at the national level, and I'm going to run through it very fast because there's nothing at all--well, there may be something controversial, but it's really meant as a backdrop to the main point I want to make, which is about the politics of the political economy, and start the cycle with the capital inflow. For my purpose it doesn't matter where it comes from. There's a domestic economic expansion. That domestic economic expansion is associated with the real appreciation in one way or another, either through a nominal appreciation or in a pegged or otherwise more or less fixed rate real appreciation with relative price effects.

And I do want to focus on the relative price effects because they are of political significance. Some of us believe that much of political economy is about relative prices, so that's worthy of being emphasized.

Component parts of this real appreciation would include a consumption boom, especially in consumption of tradables, whose price, relative price is declining, and all of us know or have experienced stories of the surge of imports of BMWs and Scotch whiskey and television sets and other consumer durables in the process of this; and also an investment boom, especially investment in nontradable goods, again, the relative price effects in finance, real estate, commercial construction, residential construction and other nontradables; and at that point also typically a quickening of the capital inflow because foreign loans are cheap or appear cheap.

I remember talking to a friend, a former student who was a Mexican businessman, in 1994 and asking him things were going. He said, "It's great. The cheapest thing in the country is dollars." And this is a common experience. So capital inflow quickens as lending picks up.

The cycle continues, and some problems begin to surface. Competitive pressures on tradables producers with the real appreciation, the imports surge, there are complaints from the manufacturing sector, from farmers, about their loss of domestic market, perhaps from exporters about their difficulties competing on world markets. The currency mismatch has been mentioned. And eventually perhaps concern about the sustainability of the exchange rate, all of this leading to the inevitable end of a cycle in some instances, in the bad-case scenarios, I suppose, with a crash, continued or exacerbated complaints about overvaluation. This leads to some expectations of depreciation, capital inflows slow as people worry about what's going on in this economy and perhaps even reverse with some capital flight, reserves are drawn down. And, again, in the worst-case scenarios, you get a currency crisis and a banking crisis, maybe both at the same time, and all sorts of bad things that we know all too well.

Now, one point to be made--and, again, I'm not taking--I don't want to take a position on what policies are desirable in response to this kind of cycle. It's not my comparative advantage. Again, I'm simply pointing out that there are policy alternatives that have been taken of one sort or another, sterilization and concerted efforts to avoid appreciations. And I think here Morris has talked about--Morris and Peter both talked about China. In the comparative perspective, there has been a lot of talk about the differences between the Latin American '70, '80s experience and the South Korean and Taiwanese experiences. That's systematically attempting to keep the exchange rate relatively weak, making the exchange rate relatively flexible soon in this cycle, or perhaps later, at least not waiting until there is some sort of massive currency crisis to let the exchange rate go.

The point that I think is striking, at least to me, is that these alternative roads are rarely taken, that delay is common; that often as we move up towards the peak of this cycle, there are continuing and ever greater complaints about the inevitability of an eventual collapse, and the advice from the best minds here and elsewhere is ignored.

The pattern, in other words, is familiar, and here's simply a list of examples that I thought of. The more recent ones I think are very well known, but I think if you look in a historical perspective, there was also a very, very similar cycle in Eastern, Central, and Southern Europe in the late 1920s and early 1930s, with a massive capital inflow, real appreciation, crisis of this sort, eventually collapsing into currency and banking crises, and even in the late 19th century in Brazil, Argentina, and the U.S. in 1893 where very similar phenomena are observed--not universally. There are instances in which countries and regions avoid this.

Again, it reminds me of the comparisons drawn in the 1980s between Latin America, which experienced between the late '70s and the early '80s very substantial real appreciations and then had these debilitating crises in the '82 to '85 period, and East Asia, especially South Korea and Taiwan, which didn't, and other cases in which currency or banking crises were avoided.

Well, the principal--I ran through that as quickly as I could, in any event, because I really want to get to my main point or series of points about why delay is so common. If one were to read or judge simply from the journalistic accounts of these kinds of--not including The Economist, Zanny--journalistic accounts of these crises, one would think that there was a--the commonality was stupidity on the part of policymakers, or at least bloody-mindedness in their unwillingness to take advice. But we all know that stupidity is not an analytically interesting category. So I'm looking for other factors that lead to these recurring crises, and what I want to focus on is what I think is an underlying political tension that affects the making of policy in these cycles in very important, very regular, predictable--perhaps not entirely predictable, but at least recognizable ways, an underlying political tension that is very difficult for policymakers to resolve without contention and contestation.

On the one hand, there is significant resistance to a depreciation in the upswing of this cycle from consumers, especially the middle classes. There is a consumption boom. Consumer durables and other tradables are cheap. Especially in many emerging markets, the urban middle classes--for the urban middle classes, consumer durables and other tradables are very important parts of the consumption basket. There are--many of us, again, know anecdotally stories about how significant segments of the urban middle classes are particularly sensitive to this consumption boom in the upswing of a cycle like this. And then there are those with foreign currency liabilities for whom the prospect of a depreciation is very, very scary, as, of course, it should be.

On the other hand, there is resistance to this real appreciation from tradables producers who are finding, whether they are competing with imports or trying to export, are finding that they are facing ever greater competitive pressures. So as the cycle goes on, policymakers face conflicting political pressures--political pressures that have nothing to do with stupidity or lack of foresight but have to do with regularities in these cycles and regularities in the relative price effects of the cycles, and regularities in the incidence, if you will, of these relative price effects on different groups in society.

One set of political pressures is that consumers vote--not everywhere, actually, but consumers vote in democratic societies, but consumers vote. And so that impending elections are often observed to cause significant delays in the adjustment of exchange rates.

There is the pressure coming from concern about private debts, private foreign currency debts and public foreign currency debts, which causes further delays. And at the same time, there is a drumbeat typically of protests from manufacturers, from farmers, from other exporters, and import competitors, calling for action to reduce the competitive pressures that they're feeling.

So if we were going to put together some sort of empirical analysis of the political variables that would predict or explain a delayed response to a crisis of this sort or to a cycle of this sort, we'd probably want to put in how sensitive the policymakers are to consumer interests. And I think--I don't want to over generalize, but I think it's probably the case that comparatively, going back to my other comparison, East Asian governments on average are a lot less sensitive to the interests of consumers than Latin American governments. This has to do with political structure and social structure and a lot of other factors as well.

Also, we'd want to include elections, pivotal groups in elections. We're not just talking about consumers in general. Again, we're talking often about the urban middle classes. In many societies, the urban middle classes are a crucial swing group in elections. It is certainly the case in Mexico and in many other Latin American societies where this is a group that is in play.

So in the run-up to an election, the last thing you might want to do is engineer a devaluation or a depreciation that's going to harm the interests of these pivotal consumer groups. And I will cite here some work joint with Ernesto Stein--out there somewhere in that black mass in the audience--about the effects of elections or the relationship between elections and these exchange rate movements. One, looking at explicit pegs, and this is Latin American countries over the last 30 years or so, that governments in these democratic societies in Latin America over a 30-year period or so are three times more likely to go off a peg in the four months after an election than they are in the eight months before an election.

Now looking more generally at exchange rate movements, not only explicit pegs but also instances in which the exchange rate is moving, a managed flow or some other regime, that governments are four times more likely to oversee a large depreciation, that is, a greater than 25-percent depreciation, in the six months after an election than they are in the six months before an election. Four times. Big number in my view.

And, on the other hand, they are more than two and a half times more likely to oversee a substantial real appreciation, more than 5 percent of the exchange rate, in the months before an election than at other times in the electoral cycle. So elections have a systematic and predictable effect on these outcomes.

Also, the influence of special interests, whether the foreign currency debtors, import competitors, and exporters--again, I'll cite some work with Ernesto and others. A one-percentage-point increase in the importance of manufacturing in a Latin American society in this period, manufacturing as a share of GDP, is associated with a 10-percent reduction in the likelihood of staying on a peg. So a one-standard deviation increase in the share of manufacturing in the economy associated with two-thirds reduction in the probability of staying on a peg.

So I think there are some clear--there are some clear relationships between these political economy factors, the pressures on policymakers, and the outcomes we observe. And it would be at least an interesting hypothesis to entertain to see the extent to which democratic political institutions, the strength of consumers, of special interests among the debtor community and elsewhere, tradables producers might explain some of the difference between Latin American and East Asian societies or among various societies in the developing world.

So that's the main point that I want to make. Let me sum up quickly. I think I'm talking here really--and this diverges somewhat both from Peter's invocation and from the general tenor of what others have said--about national responses to a global cycle. The political economy rules--I haven't really decided whether that's a noun or a verb, take it either way. But the point here is that I think there is a systematic political economy component to these national responses to the cycle.

I'm not arguing that these national responses have a broad and deep effect on the global cycle, because I do think that these cycles are too systematic, too long-lasting to explain on idiosyncratic national characteristics. But there is a strong sense in which some of the more recent cycles at least are triggered by national experiences, Mexico in 1982, for example. So they're important to understand even from a systemic standpoint. But the national responses vary widely, and policy can either exacerbate or mitigate some of the effects of these cycles.

The policy response I think is engineered within an environment fraught with political tensions. Political pressures pull policymakers in a wide variety of different directions. They face demands on them that cannot easily be met. The character and the impact of these pressures varies with the structure of the economy, with the political institutions, with the environment. And all of this I think has a very clear message, which I think every talk should end with, which is that the political economy of capital flows and their cycles is worthy of further research.

MS. MINTON-BEDDOES: Thank you, Jeffry.

Well, I expected a wide-ranging panel, and indeed we had one. We had Peter giving us a masterful and very provocative view of what the system is right now. We had Morris and Agustin telling countries, giving them a lot of advice on what to do in this system. And we had Jeffry telling us exactly why they won't do it.

I know this is supposed to finish at 3:30. I think since we started late, I'm going to take the liberty of extending it for another 15 minutes because I'm sure there will be many questions. So the floor is open, provided I can see you.

Could you identify yourselves as you ask your question?

MR. : And use your microphones, please, that are located on the left-hand side of your chair.

QUESTION: It's been a very interesting discussion here and a lot of things are known, but there are a lot of new things that were said. One new thing that I found very interesting is the fact that the exchange rate misalignment that we're talking about fundamentally is a reflection of some other imbalance in the labor market, in particular excess labor supply in some--in particular, in China, or in Asia.

Would that imply that allowing migration, allowing more labor mobility would smooth capital flows? I mean, the argument is very simple, that, you know, if you allow people to migrate, there will be no need for overvaluing--sorry, undervaluing the exchange rates. That [inaudible] argued.

The second question I wanted to ask is countries borrow because they need to. I get a sense from here that--you know, several speakers tended say countries don't need to borrow, I mean, or they should not borrow. No, it's not that. It is right, they shouldn't borrow if they can avoid it, but they need to. So are there any instruments out there which can allow countries to borrow and yet reduce risks of overshooting, excess or--I mean, sort of cycles in the capital flows?

Thank you.

MS. MINTON-BEDDOES: Thank you.

Peter, perhaps you could start on the first part of that.

MR. GARBER: Well, yes, if you move the labor to the capital instead of capital to the labor, then the capital flows move in a different way. But in the case of China, it's really Chinese savings that are employing Chinese labor through the intermediation of the foreigner. So it's not that you're moving U.S. capital, I mean real physical capital to China. It's more like Asian capital that's moving to China, and Chinese capital is moving to China but being intermediated through the U.S.

But, yes, obviously, if you move labor to the capital, then, again, international flows are going to be different.

MS. MINTON-BEDDOES: Agustin?

MR. CARSTENS: The second part of the question probably was directed to me. I think that the decision has to be between borrowing internally or externally. And, first of all, you also have to pay attention to debt sustainability, as Morris mentioned.

Countries could increase internal savings, make it more efficient. Obviously, that's a whole challenge in itself. But at the end of the day, even then it might be optimal for countries to borrow. The issue, really, is what is the most efficient way to borrow and to what market do you have access to. So it's not the fact that to borrow by itself is wrong. To over borrow is wrong. Once you have policies in place to prevent that from happening, you then need to have a comparable policy mix in place to help develop the most efficient way of borrowing.

MR. GOLDSTEIN: Just a quick addendum. I would certainly agree with regard to the internal-external. I think there is a case that you can make for GDP-indexed bonds. I think that would be helpful in conditioning the kind of ability to pay to performance in these countries. Under the present system, you know, if you have dollar debt and the country has deep difficulty, well, then, it doesn't--the creditor gets paid--doesn't get paid in the currency of his choice. He gets less than the face amount. Or if you do it with local currency, he gets paid the face amount, but in depreciated currency.

So I think there's something to be said if we could get it in for GDP-indexed bonds. The Fund is in a sense doing that with the restructuring of Argentina's debt, except it's one-sided, on the downside but not on the upside.

MS. MINTON-BEDDOES: [inaudible]?

QUESTION: I have a question about China, to the panelists, I suppose, but mostly to Morris. It seemed to me that the proposal that you and Nick have made makes eminent sense, except that the Chinese are just not going to do it. And the question is whether there are alternative policies that could be undertaken in order to ease the burden of the misalignment of the exchange rate and the imbalance in the world economy.

I suppose that my question is at the end your 10- to 15-percent or maybe 20-percent misalignment of the yuan, of course, is the real exchange rate. And one possibility is instead of moving the exchange rate proper or the nominal exchange rate is to implement policies that would move the equilibrium exchange rate. And what I have in mind is for China to accelerate its trade liberalization. And we have all the old theories that say--all the way back to Harry Johnson and before--that say that countries as they open up on their trade side, they will require an equilibrium real depreciation.

Now, if you read the report on China by the U.S. Trade Representative Office, there are lots and lots of complaints that, in spite of all the commitments that China made when it accessed and it joined the WTO, there are still significant nontrade barriers and so on and so forth.

So my question is whether it wouldn't make more sense from a political point of view to try to move in that direction than in the direction that you have suggested of a one-step change in the exchange rate and then pegging to the basket.

MR. GOLDSTEIN: Well, there's two parts to your question. First of all, will they do it, that is, 15 to 25? I think not right away, but I think well before what I took was Peter's answer, which was ten years.

I think the rest of the world will not allow that to go on. There are already charges of manipulation, I think justified ones, at least what I interpret as currency manipulation. You have bills in the Congress, you have other things.

So I don't know, right away, but a year down the road, particularly if they don't get the banking system under control, it may be different.

In terms of whether the trade route might work, maybe. I'm a little bit skeptical. It may be that imports can go up. Tariff rates are already down. They can go down a little further. But exports can go up as well because the Multi-Fiber Agreement expires, and there are some people who believe that when that happens or a little bit further down the road, China's share of the world textile market will go from 18 percent to 40 percent.

So it's not clear to me that trade liberalization will do the job.

MS. MINTON-BEDDOES: [inaudible].

QUESTION: The question relates to Malaysia, really. Malaysia went ahead and imposed capital controls post-'97-'98. Is the verdict out yet that it has worked? And if it has, does it mean that non-IMF prescriptions might work?

MS. MINTON-BEDDOES: Agustin, I think that's one for you.

[Laughter.]

MR. CARSTENS: I think it's important to take the whole picture into account. As you say, Malaysia adopted capital controls, but then relaxed them. So what does it mean, the fact that they relaxed the capital controls?

So, in a way, the use of that instrument at the time of the crisis probably worked in favor of the country. But then the country saw the costs of that instrument, and when they had the opportunity to remove it, they removed it capital controls.

Certainly Malaysia's experience is evidence that needs to be taken into account when evaluating capital controls. In principle, capital controls should not be a favorite instrument, but that doesn't mean that there are no occasions where they might be useful.

MS. MINTON-BEDDOES: Someone else on the panel want to comment?

[No response.]

MS. MINTON-BEDDOES: Okay. Next question.

QUESTION: To Peter Garber, a fascinating analogy between the Atlantic economy of the '60s-'70s, and the Pacific economy of the '90s and--and what do we call it now? In any case, one fundamental difference strikes me, and that is that you did not have in the Atlantic story of the '60s-'70s, or '50s-'60s even, a massive difference in the need for one side to absorb labor and, therefore, to grow much faster than the other side of the ocean. Here you have that need. That's what's driving it, the enormous need in the case of China and the other Asian countries to absorb labor.

The question is then how does this get resolved. In the Atlantic story, it was a monetary crisis in '71, and they moved to floating exchange rates. What I see looming here as the possible crisis is -- [Forum ended at this point due to fire alarm.]