Transcript

October 13, 2006

Participants:
Paolo Mauro
Research Department, IMF

Nathan Sussman
The Hebrew University of Jerusalem and CEPR

Yishay Yafeh
The Hebrew University of Jerusalem and CEPR

Interviewer:
Prakash Loungani
External Relations Department, IMF

Politics, Institutions and Sovereign Bond Spreads

LOUNGANI: This is a fascinating comparison of the two waves of financial globalization, the present, since 1990, and the pre-World War I era. One of your main findings reminded me of the saying among journalists: "If it bleeds, it leads"-I'm referring to your finding that events like a political assassination or the start of a war are reflected very quickly in asset prices.

Paolo Mauro

MAURO: Yes, wars and political instability do seem to be the main driver of changes in sovereign bond spreads. So even though today we place a lot of emphasis on institutional reforms, such as introduction of a new constitution and so on, those types of reforms don't seem to be immediately reflected in a lower cost of borrowing. [For the modern period, spreads are defined in the book as the U.S. dollar interest rates at which emerging market governments are able to borrow, over and above long-term interest rates on U.S. government bonds. For the historical period, all emerging market bonds are denominated in pounds sterling and British Consols are the benchmark.]

Of course, you could view wars and political instability as reflecting some types of institutions. So in some ways we're redefining a bit what one means by institutions. If good institutions are reflected in the fact that a country doesn't have a lot of political instability, then that certainly is shown to have an impact on bond spreads.

Yishay Yafeh

YAFEH: One reason may be that investors care about what countries actually do-a country may produce a nice-looking constitution today, but what really matters to investors is how that constitution is implemented in practice over subsequent years.

SUSSMAN: And all three of us have shown in separate work that there's a case to be made that institutions matter over a longer term for growth and for asset prices. But, in the approach taken in this book, one doesn't capture that because it takes time for the impact of institutional reforms to be reflected in bond spreads.

LOUNGANI: So what are the implications of this finding for countries today?

MAURO: I think that one of the messages for emerging markets is not to expect that desirable reforms will give an immediate payoff in terms of lower cost of borrowing. There are a few instances in which we see that but they are the exception. For example, one of the studies that Yishay and Nathan did that I thought was very impressive, and led me to be interested in this issue, was their study on the impact of the introduction of the gold standard in Japan in the early era of globalization, which did reduce the cost of borrowing for Japan.

YAFEH: And that was successful because at the time Japan needed credibility and this was one reform in the monetary area that was seen as being a focal point for commitment to macroeconomic stability.

MAURO: A similar instance in modern times may be with the introduction of the currency board in Bulgaria. But these are very much exceptions, as I mentioned, and they are confined to the monetary or sometimes the fiscal area. The immediate impact of broader types of institutional reforms on asset prices is much harder to capture in the data.

Nathan Sussman

SUSSMAN: These days you have various rankings of the quality of institutions for different countries. There were no equivalents for the pre-World War I era with the exception of the gold standard, which was supposed to be a good institution for monetary stability. An affiliation with the British empire was also supposedly good for rule of law and so on. So in our book we don't really follow this modern approach of ranking the quality of institutions. We just want to see whether there was an immediate reward for setting up an institution that was at least nominally considered favorable to investors, such as an independent judicial system or a constitution. And our finding is that it's very rare that financial markets reward it immediately with lowering cost of capital and thereby fostering economic growth through that channel.

LOUNGANI: But the modern era in your study is fairly short, about fifteen years.

SUSSMAN: Yes, obviously fifteen years is not enough to assess the very long term impact. But there have been enough institutional changes and reform undertaken over this time that we can establish that countries were not rewarded swiftly. The result from the modern data is also confirmed by historical results.

Role of international institutions

LOUNGANI: You find that there is greater co-movement of spreads now than in the past. Why is this the case?

YAFEH: One reason is that today the fundamentals of emerging markets are more similar to each other-the main emerging markets now have fairly diversified production and export structures. We document in the book that in the earlier era of globalization countries were more specialized in production than they are today-one would produce coffee and the other wool.

SUSSMAN: But today you also sometimes observe very high co-movement of spreads of countries that have really nothing in common in terms of their economic fundamentals. Our conjecture is that the behavior of investment funds today may explain this co-movement. For example, because these funds are all emerging market funds or organized in a way that forces them to liquidate assets in response to a crisis in one emerging market and to sell some assets in another, they spread the disease, so to speak.

LOUNGANI: Then why did we not find co-movement or contagion after the Argentine crisis of 2001-2? I've heard people say it showed that markets were discriminating better across countries rather than behaving in herd-like fashion to avoid all emerging markets.

SUSSMAN: It could be because it was a widely anticipated crisis and so funds were able to get out of Argentina gradually before the crisis. Also, the weight of Argentina in the Emerging Markets Bond Index (EMBI) portfolios was reduced in advance. Or may be markets have begun to differentiate better; I think it's too early to tell.

LOUNGANI: What are the implications of these findings about greater co-movement of spreads today, and therefore the possibility of greater contagion, for the design of IMF loan facilities?

MAURO: We know that there's interest on the part of several countries, and in particular on the part of emerging markets, in IMF arrangements that would provide large scale financing, quickly, on a precautionary basis. The idea is that these types of programs would help protect emerging markets against financial flow reversals. I think we show this type of interest is warranted by events that have taken place in these first few years of the new era of financial globalization. Now, how exactly one goes about designing these new types of facilities is very much under discussion.

LOUNGANI: Were there institutions that in the early globalization period played the role that the IMF plays today?

YAFEH: In the historical period there was no potential international authority like the IMF. But during the Baring crisis, which is the most severe financial crisis during the early era of globalization, we look at the role that the Bank of England played in terms of mitigating the consequences of a severe financial crisis in Argentina. The Argentinean default caused for a couple of days something which we could view as maybe as contagion-all the stocks and bonds in the London market reacted sharply, but the Bank of England intervened immediately in this event with the help of the other banks in England and provided liquidity and so people did not have to liquidate portfolios. The effect on the non-Argentinean bonds was reversed within days. With the exception of the Argentinean bonds, which were in default for a good reason at the time, the rest of the market stabilized within days. So a key role in these situations is the rapid provision of liquidity.

SUSSMAN: In fact, something which is often said, may be with the benefit of hindsight, about the international financial crisis of 1998, is that the provision of adequate liquidity to the markets could have stabilized the situation even faster.

LOUNGANI: You say that institutions aimed at resolving debt crises faster are needed to keep the expansion of international bond markets in the modern era going. As you know, the IMF's former first deputy managing director, Anne Krueger, had proposed the sovereign debt restructuring mechanism (SDRM). Is that the kind of proposal that comes out of your book?

MAURO: We have a chapter in the book that documents how debt restructurings were conducted in the past. Our goal was not necessarily to say something directly related to the SDRM proposal, but to provide information on how restructuring took place in the past.

Our focus was on a fascinating institution called the Corporation of Foreign Bondholders (CFB), essentially a club of British loan holders. This was a relatively small group of individuals-they could probably all meet in one very large room. This was a permanent institution that was in place on a continuing basis for many, many years. What happened in times of default was that the institution would name a committee of representatives who would go to the country and discuss with the authorities what they were going to do to get the economy back on track. So there are some loose similarities with IMF missions today, but essentially what the CFB would do was to try and renegotiate the debt.

There are some observers who have proposed a return to these types of permanent creditor clubs today. So our goal was to try to figure out to what extent these types of permanent creditor clubs would help in today's environment. Our conclusion was that while there might be some benefit, historically this institution wasn't particularly successful. It was successful in some cases, but it failed to bring about restructurings for several decades in some other cases.

Today, the environment is much more difficult for this type of institution. For example, the loan holders are far more dispersed today, and the ability to coordinate them would be much smaller. So our conclusion was that, yes, this is an intriguing historical institution that gives us some lessons, but in practice today it would probably be less effective.

SUSSMAN: In the pre-World War I period, when a country would default, it was very common for loan holders to take over certain assets which were used as collateral for debt, something that is not observed today. You needed the cooperation of other bondholders to jointly operate, say, a mine in Peru. They had very strong reasons to cooperate, but even under these circumstances, the track record of the corporation of CFB was mixed, as Paolo said.

Today there is nothing like collateral that brings bondholders together, and in fact there are more options for them to try to take all kinds of different legal actions against defaulting governments. So we would conjecture that the success observed then is probably the upper bound of what we can expect of similar institutions today.

LOUNGANI: So you are saying that it remains very important to have some mechanisms for debt restructurings, but it is just that it has gotten more difficult.

SUSSMAN: Yes, exactly.

Comparing the eras of globalization

LOUNGANI: This discussion leads me to ask more generally about the similarities and differences between the two periods of financial globalization that you study.

YAFEH: The pre-World War I era was really the only other period prior to the present where you had global capital flows through bond finance. If you look at volume of capital flows, relative to say British GDP, the similarities with the present period in terms of the magnitude of the flows are very striking. The freedom of capital movement in that early period was maybe unparalleled.

MAURO: To complement what Yishay said, clearly the market was very large at the time. It was very liquid. Investors were very well informed. For example, the data on spreads were reported in the newspapers on a daily basis. And there was plenty of information about events in emerging markets. All of these things give us a sense that this was a sophisticated market, probably as sophisticated as the one today.

SUSSMAN: Of course, there were differences. One may be that the transfer of information, and the speed of transactions, is much faster today than it was back then. But in the earlier period too the telegraph was already in place and so at least information could be transmitted very quickly. Not as fast as pressing a key on your keyboard, but it nevertheless was quite fast. Another obvious difference relates to the geopolitical situation, with empires and colonies and so forth in the past, and many more sovereign countries today.

YAFEH: There are also some financial differences. At that time there was much more investment in bonds, especially sovereign bonds, which were issued and traded in London. Nowadays direct investment, also through the stock market, in the various emerging markets is far more prevalent.

MAURO: Of course, you're looking almost a century apart so there will be differences. But those are probably not preventing a comparison. The one difference that I would emphasize, as already mentioned, is the fact that individual investors at the time were much more important compared with institutional investors that seem to be predominant today. And practices like benchmarking are much more widespread today than in the past.

LOUNGANI: So returning to your findings, how did you actually go about measuring things like institutional changes and political turmoil and then showing the impact on spreads?

SUSSMAN: There are three ways in the book in which we do that. One way is to use historical case studies, such as the example of nineteenth century Japan that was mentioned earlier.

The second is to look for years or months in which the bond spread changed substantially and then search contemporary newspapers to see what events were associated with these sharp changes. Similarly, we looked at big news items and then saw whether bond spreads changed.

The third approach is econometric analysis, where you run regressions trying to explain how spreads depend on various things like political instability-we measure that by the number or articles in the month that mentioned rebellion, wars and the like.

Summing-up

LOUNGANI: Any closing thoughts on the contributions of your book?

MAURO: One contribution that we hope will be of interest to researchers is that we've put together a very systematic data set on bond spreads and macroeconomic variables but especially on news for the emerging markets, both for the pre-World War I collected from the London Times, and for the modern period, collected from the Financial Times. These news were collected systematically using categories such as wars and instability or institutional reforms or trade agreements. So, for example, one can go back to our data set and find that in Uruguay in May 1870 there were, say, three news items regarding wars and political instability.

The other thing of more general interest that we would like people to take away from the book is that there is a genuine business need to look at previous historical episodes. This is not just an intellectual curio. The environment that emerging markets operate in today has only been in place for 10-15 years, and it is still very much a new environment. It is really necessary to go back to lessons from history, and those that ignore lessons from history do so at their own peril. That is true for policymakers; it is also true for investors. So we really think that the early era of financial globalization is a gold mine in terms of information that can be obtained, and maybe that is the broader lesson that one should take away from this book.

MR. LOUNGANI: That's a good note to end on. Thank you, Paolo in D.C., and Nathan and Yishay in Jerusalem.