Last Monday the President Cristina Fernández de Kirchner announced that a group of banks (working with some of the holders of bonds that did not participate in the 2005 debt exchange) has submitted a proposal to the Argentine government to consider a potential second exchange. The group of holders of non-exchanged debt that structured this new proposal seems to have been far more subtle and realistic in their strategy, eschewing the heavy-handed approach followed by other holdouts, and deciding that it was far more productive to present their own proposal rather than try to browbeat Argentina into submission.
As a background to this potential second exchange it may be useful to recapitulate some facts (as I see them) about the original debt exchange, given widespread misunderstandings about what exactly happened. I start with a statement that I think summarizes the facts (as I see them), and then I expand on the statement.
A. The Argentine debt re-structuring has been the most complex in history of emerging markets (and until Lehman’s bankruptcy probably not only for emerging markets). The debt re-structuring was done through the only practical approach given the complexity of the deal.
1. The Argentine debt re-structuring has been far more complex than any other of the cases of emerging debt restructuring with which it is usually compared: it encompassed 152 bond issues in 7 separate currencies under 8 distinct governing laws. Impartial observers at the time (like Nouriel, who wrote on the subject) recognized that the only workable approach was what the Argentine government did: to hold numerous meetings (over 70 altogether) with different groups in many countries and then define and present an exchange offer that considered the different interests and possibilities.
2. It was simply not possible to negotiate with each group. It was even less appropriate to negotiate with just some specific group of bondholders who claimed to represent this or that group (sometimes with no clear mandates). Either it had to be negotiated with all, or it had to be done the way it was done: to hold as many consultations as possible and then come up with a single offer. Any other approach opened serious legal and operational difficulties.
B. Argentina’s payment capacity was determined by objective factors such as the collapse of Argentina’s GDP, the fact that the country was making net payments to international organizations, and the need to use realistic assumptions about Argentina’s potential future performance in order to avoid further restructurings in the near term (which would have hurt both creditors and the country). The GDP-linked coupon was another demonstration of good faith as it was designed with the intention to share with bondholders the potential upside in case actual performance proved to be better (as it has been) than what history would have suggested.
3. Argentina payment capacity (and, therefore, the level of the discount or “haircut” in the debt exchange) was defined by the combination of three separate objective factors.
4. First, the collapse of the Argentine economy was far worse than other cases of debt restructuring. The next Chart compares the decline in the dollar value of total GDP in the case of Argentina and for the average of other 5 countries that renegotiated their debts in the late 1990s and early 2000s (Ecuador, Pakistan, Russia, Ukraine, and Uruguay). It utilizes IMF data (from the World Economic Outlook database). The index shows the evolution of the GDP in current dollars, starting 3 years before the year of the default/rescheduling (T) and ending 3 years after that. The starting year (T-3) is normalized to 1 in all countries to facilitate comparisons. The countries considered are Dominican Republic, Ecuador, Pakistan, Russia, Ukraine and Uruguay. This country is part of the average (the line labeled Rest) but it is also shown separately (because it is usually utilized as more direct comparison of a “friendly” debt restructuring as opposed to an “aggressive” one). The collapse of the Argentine economy in current dollars is far larger at T (the year of the default) with a drop of about 2/3 with respect to the dollar GDP at T-3. Also, Argentina’s economy by the time of the debt restructuring had recovered far less than other countries, including Uruguay, all of which after 3 years were almost back to the level of 3 years prior to the default (T-3).
5. This drastic decline in the dollar GDP has been a combination of two factors. First Argentina suffered an economic depression and increases in unemployment that were comparable to the US Great Depression of the 1930s. It was accompanied by widespread social unrest that led to tens of death and hundreds of injured, and by institutional disruption, with the Republic going through 5 Presidents in a matter of weeks. Second, the dollar value of the Argentine economy was vastly exaggerated by the 1:1 peso/dollar exchange rate, suggesting a debt-carrying capacity that it did not have. When the crisis forced an overshooting in the exchange rate, GDP in dollars dropped and the debt carrying-capacity was initially only 1/3 of what appeared before.
6. Besides the decline in the GDP the second important objective element that defined the Argentine debt offer was the lack of international support, under the “market based approach.” In this regard, it is important to notice the existence of two general paradigms for debt resolutions. The previous traditional model had been the public-sector approach funded and led by international public institutions. The competing approach was called the market-based (or “moral hazard”) approach, which postulates that the private creditors and the debtor government have to work out a solution by themselves without money from the international community (which in any case was running out of money). Argentina was placed under the latter paradigm while the majority of the other cases were treated with new public money (substantial at times) and, therefore, with the involvement of public institutions.
7. In April 27, 2004, John Taylor, then Under Secretary of Treasury for International Affairs, addressing the Fitch Ratings’ Latin America Conference (http://www.ustreas.gov/press/releases/js1476.htm), explained clearly the logic of the market based approach in the case of Argentina: “The IMF program was designed specifically to give Argentina the flexibility it needs to negotiate a debt restructuring with its creditors. The program leaves Argentina’s primary surplus targets for 2005 and beyond unspecified above a 3 percent of GDP floor, so that the increment to this floor can be decided upon as Argentina determines the level that will be needed to achieve a debt exchange that achieves broad creditor support. Once the amount of resources from the international financial institutions is determined, specification of a primary surplus path in the IMF program would have largely determined the amount that private creditors got paid. It would not be appropriate for the IMF to be in the middle of negotiations between Argentina and its creditors in this way (Note: this is my emphasis, not the original’s). The amount of IMF exposure is to be held constant during the current three year program and should then decline significantly. There is no need for exposure to the World Bank and Inter-American Development Bank to decline in this way if viable projects and programs continue to merit new lending.”
8. In the end, and for different reasons, international organizations not only did not facilitate the debt exchange agreement with new public money, but they did not even maintain neutral flows (as originally agreed with Argentina at Dubai in September 2003), and eventually became net recipients of payments from Argentina (see below). The IMF staff kept on bringing up in the program revisions the issue of the primary surplus going forward, which, as indicated by John Taylor, was an interference with the “market based” negotiations (i.e. with neutral flows with the IFIs, if Argentina had agreed with the IMF the size of the total primary surplus, the negotiations would have been with the IMF and not, under the “market based” approach, with the bondholders). Argentine authorities manifested publicly several times their disagreement with the IMF staff in this regard. Finally, Argentina had to suspend in late 2004 the program in order to be able to close the debt exchange within the “market based” approach. These problems were part of the confusions between the two paradigms, which greatly complicated the negotiation for Argentina. The pressure from some bondholders to interject the IMF in what was to be a market-based negotiation also hurt those same bondholders because Argentina’s negative flows with the IFIs reduced the amount of money that could have been available for the private debt offer. All in all Argentina paid net to the international organizations about 12 billion dollars (capital and interests) since the beginning of the crises in 2002 until the debt exchange was completed in early 2005 (and more net payments continued to take place since then). Therefore, given the government’s budget constraint it follows that the haircut to the bondholders had to be larger, and therefore acceptance lower, than the other cases treated under the approach that utilized new public money.
9. The next Chart shows net flows from the international financial institutions (IFIs) to different countries as percentage of the GDP, during the period around the time of the respective debt restructurings. It is clear the important difference between the large amounts of money that most of these countries received from IFIs to help their process of debt restructuring, while Argentina was making net payments to those institutions. The data on net flows with IFIs other than the IMF comes from the Global Development Finance (GDF) data base of the World Bank, while the IMF data comes directly from that institution (converted from SDRs into dollars) for the year of the default/restructuring plus the next 2 years (for instance, in the case of Argentina the period considered was 2002-2004; for Pakistan 1999-2001, and so on; in the case of the Dominican Republic, the numbers are only for 2005 and include just the IMF). These are net payments of principal, not counting payments of interests (which would obviously increase the flows of net resource transfers towards the IFIs by each country). The Chart shows that while Argentina was making net payments to the IFIs of about 1% of the GDP on average, all other countries were operating under the paradigm of “bailout with public money,” receiving net flows from IFIs (Russia also got net flows but in a very small amount of the GDP, of about 0.01% which barely register in the scale of the Chart).
10. To put the annual net payments from Argentina to the IFIs in perspective they are equivalent, given the size of the respective economies, to the combined yearly expenditures of the US Federal Government for Education and for General Government.
11. The fact that there were two paradigms and that there was a debate about the adequacy of one or the other approach, was a source of frustration and misunderstandings, with Argentina being blamed for implications that were inherent in the paradigm of debt restructuring to which the country was assigned. The government of Argentina permanently asked, as a matter of simple consistency, that the requirements to the country and judgments about Argentina’s performance be framed within the parameters of the approach under which it was operating, and not by the results in countries that were treated with the other approach (which included substantial public funds to help them).
12. The third objective factor to define Argentina’s payment capacity was the use of realistic projections according to the historical performance of the economy, instead of overly optimistic future values for key variables such as growth or fiscal surpluses. In the case of the fiscal accounts, some of the demands for Argentina to run very large primary surpluses were based on inadequate comparisons with other highly indebted countries (Brazil, Turkey), which appeared to have higher surpluses as percentage of the GDP. The key issue is that while Argentina had modified the Social Security system and privatized many public companies (including the national oil company), these other countries have not. Therefore, given the lost fiscal income on both accounts, a primary surplus of, say, 3% of GDP in Argentina, would have been about 5% without those policy changes, which was comparable to the countries used as benchmark. In summary, using overly optimistic projections (considering Argentina’s past history) and by ignoring different structural issues, some people insisted on parameters which would have produced a higher offer but with an important downside risk of future rescheduling. Argentina permanently maintained that good faith meant promising what can be realistically delivered given what it was known then. While some argued the need for high acceptance rates as proof of a good-faith negotiation, the Argentine government constantly indicated the need to ensure the sustainability of the debt going forward, so that there would be no need for further restructurings as occurred frequently in the previous decades. The short-term fixes of the past created uncertainty and affected investment and growth, to the detriment of all stakeholders.
13. But while the projections were based on realistic numbers according to past performance, good faith also meant that bondholders were offered the possibility to share in the future benefits through the GDP-linked coupon, if the considerable upside potential of the economy werevalued as nearly worthless by many analysts during the debt exchange, have added significantly to the net present value of the exchange from the point of view of the creditors.
14. In summary, the massive collapse in the previously overestimated GDP in dollars, large net payments to international organizations, and the need to use realistic assumptions determined the level of the debt exchange offer, which given all these objectives factors, necessarily implied a larger discount (“haircut”) than in the case of other debt restructures.
C. Still the debt exchange had a high level of acceptance (76.15%), as shown in the Table attached, where the amounts are divided by original legislation.
15. It should be noted that, contrary to some claims, the New York jurisdiction, which includes most of the US creditors, had a higher acceptance than the average (about 82%).
16. After the debt exchange was closed the government of Argentina assumed, following precedent, that those that did not accept the offer wanted to litigate. Therefore, it was odd to hear some groups of holdouts to claim that the government was infringing the law by not reopening the offer and negotiating with them. Reopening the offer by the Argentine government would have unraveled the whole agreement, which had been closed with a large majority of the bondholders. Now it is different, because the current proposal has been presented to the government by some bondholders (and not the other way around).
We need to learn the right lessons from Argentina’s painful experience. First, it is important to distinguish between the market-based approach (without new additional public money) and a public sector-led approach (with new public money). We can debate the merits of either one, but they should not be mixed when analyzing specific cases. The relative marginal role of the IMF, the larger haircut and the lower acceptance rate are consequences of the approach followed, and not an Argentine invention. Second, we need to separate what is specific to the Argentine case, from issues that are more generally applicable. Even within a market-based approach, the Argentine economic collapse was especially deep and I do not think that anyone in its right mind would postulate that experience as a model to emulate.
The initiative presented by a group of the holders of debt in default offers an opportunity to close this painful episode for everyone affected, starting with the Argentine people who suffered the most during those terrible days of social and economic collapse. Unfortunately, it is impossible to bring back to life those that died. Regarding the economic consequences we can only hope that burdens are equitably shared, considering specially the poor and vulnerable.