Bradies, crises contagion and the architecture
Gustavo H. B. Franco
(Former Governor of the Central Bank of Brazil)

Through the 1990s, and in contrast to the previous decade, emerging economies’ indebtedness took place mostly in the form of bonds. The securitizations engineered under the Brady Plan, the last of which being the Brazilian one in 1993, issued US$ 154 billion in new securities and in so doing reopened and reinvented international borrowing by such economies. The new indebtedness cycle resulted in some US$ 330 billion of new securities (globals, euros and corporate bonds) issued during 1990-1996 by Asian countries and some US$ 190 billion by Latin American countries. Since current account deficits and international borrowing have been an integral part of the economic landscape in these economies, the well functioning of international capital markets has always been an important concern, invariably revived when things go wrong.

Yet, in the several blueprints and reform proposals drafted after the crises in 1997 and 1998 around the idea of a new international financial architecture very little attention was devoted to the workings of markets for emerging market sovereign bonds, an important avenue through which crises waves travelled from one place to the other. There have been many complaints on the volatility of capital in general, and against the behaviour of hedge funds in particular, even though the alleged villainy of the latter could very well be found in other financial market players with better reputation. In this connection, a study conducted under the IMF auspices would dismiss a large part of the usual wrong doings thrown at hedge funds, but would broadly agree that there is a case for regulation on grounds of market integrity, a case that goes well beyond hedge funds. Even though sympathetic to these conclusions, however, regulators in the G-7 area would claim that a heavier set of requirements, prudential or relative to trading limitations, would do nothing but prompt the business offshore. Therefore, the crucial market in which emerging economies’ risk spread was priced, and borrowing contracted, was left mostly unregulated. This means that the trading with Bradies and other sovereign bonds issued by emerging economies takes place offshore, in a decentralised or "virtual" fashion, out of any organised exchange and without the supervision of any authority. The obvious question to ask is whether this environment was such as to generate market practices that could destabilise prices and amplify contagion in times of strain. The Brazilian experience is worth examining in this respect.

Following the Russian moratorium and devaluation in August 1998 one could see a broad reassessment of risks in emerging markets’ investments, which was made even worse in September in the failure of LTCM. Brazilian securities would suffer very strong selling pressure, made even stronger as hedging strategies for long positions in Russian instruments would very commonly be constructed with a short position against the EMBI, or against the Brazilian "C bond", which although represented only 4 % of the EMBI, it was by far the most traded emerging market instrument at the time. More specifically, short selling the "C bond", and to a lesser extent the "IDU", became widespread, thus leveraging the fall in prices of these securities and serving to reduce losses made in long positions in Russia

It is interesting to note that, in the US Securities Law there are restrictions as to the repeated use of short selling by a broker. Rule 10a-1, known as the "Tick Rule", establishes that a second short sale of a given security cannot be done at a price inferior to the one of the tick, or the exchange reference price. The rule would be difficult to implement in the context of Brady bonds, for which trading goes on all over the planet, in the cyberspace, without a "reference price" fixed in any exchange or jurisdiction. In the absence of any regulatory safeguards, the aggregate amounts short sold in "C bonds" and "IDUs" appeared out of proportion with amounts available in the market place, approximately US$ 6,o billion for the former, and somewhat less for the latter. Short sellers, usually market makers, when faced with delivery demands, would offer surrogates or "synthetics" of the security (and with a better rating!), as if issuing the security or by passing its scarcity in the market place. As if Brazil was much more indebted than it really was, and reinforcing similarities with Russia that were only apparent. No doubt, these procedures were crucial to let contagion hit Brazil in a magnified fashion.

At a point, delivery difficulties could be seen everywhere, producing the highly unusual phenomenon of negative interest rates for "repo" operations with "C bonds" and also with "IDUs". When this happens, it means that someone needs the security so badly that borrows it, i. e. lends money against this security, at a negative rate. The underlying security is said to be "rich", i. e. bound to go up shortly. If delivery was forced on short sellers, they should not be able to sustain their positions for very long. Yet, nothing happened, even after an attempt by EMTA (Emerging Markets Traders’ Association) to put together a clearing and settlement facility. In fact, for securities traded off shore the only procedure available to force delivery would be the "buy in" provisions under the ISMA (International Securities Market Association) Rules and Recommendations (Rule 451 of section 450), which is subscribed by trading houses on a voluntary basis. The procedure is somewhat cumbersome but could certainly work if pursued to the end. Yet, even though there was considerable amounts of sales of "C bonds" technically on default, very few "buy in" procedures were actually taken to advanced stages. The common complaint of those on the long side of the deal was that the procedure almost invariably would produce threats and retaliations on the part of the market makers, usually large international banks no one wanted to confront. Clearly, a regulatory "asymmetry" favouring the large market maker in detriment of the smaller player challenging a short-sale, and even the issuers, resulted detrimental to market integrity and amplified contagion.

These examples are only to highlight the mechanics of contagion in one specific situation in which the lack of regulatory constraints to certain market practices produced a leveraged fall in Brazilian securities. It is true that there are market remedies for that: the "attack" on Brazilian Bradies dragged a considerable amount of funds from Brazil to take the arbitrage opportunity. Bradies, globals and also corporate bonds were all incredibly cheap. In September 1998 alone Brazil lost US$ 22 billion in reserves, and a substantial part of it was looking for the benchmark bonds and there were many corporations buying back their own bonds. One should expect this movement to recover prices, given that securities were very "rich", but the upward movement was barely visible. The contraction on the demand side was certainly strong, but in the presence of unrestricted short selling and with the difficulties in starting "buy in" procedures, it was made even stronger. The arbitrage outflows from Brazil, in spite of their considerable volume, were not sufficient to change prices, let alone bring normalcy back to the market, eventually reviving primary issues. Interest rates were then sharply raised in Brazil, to stem the outflow, and the crisis transmission mechanism was then completed.

The episode highlights the fact that for capital account convertibility to be advantageous to countries like Brazil, or for capital flows to move in a stabilising fashion, some innovative regulatory background should be set in place. The reality of offshore markets has to be dealt with by regulators, issuers and market players. The latter have been involved in efforts, mostly through EMTA, to increase market transparency and to improve practices and institutions. Clearing and settlement has been a key concern, and the basis for the launching of the EMCC (Emerging Markets Clearing Corporation) initiative. Yet, much more could be done if market players were willing to discuss market integrity themes with regulators and the issuers. The former would certainly be willing to advance on extending home rules to offshore trading, and the latter would have much to propose in order to create new avenues for active liability management strategies.

In addition, if emerging economies’ bonds are traded in organised markets with acceptable regulatory oversight, it may be feasible to conceive ways through which official and concerted intervention in these markets in moments of strain in order to prevent unduly high fluctuations. A new facility could be created at the IMF, for instance, not to support a given country’s adjustment and crisis resolution, but to act on capital markets through bond swaps, for instance, exchanging Bradies for US Treasuries on a temporary basis. If unduly large fluctuations in spreads could be avoided by this mechanism, it would succeed in preventing a crisis by reducing the interval of time markets remain inaccessible to economies under pressure. This may sound anathema after so many failures in intervention efforts, mostly with currencies. But it may be worth thinking about; especially if we note that, in hindsight, a mechanism like that could have been much cheaper than the ones actually engineered, and would be represent an alternative to the "burden sharing" approach to involve the private sector in the process of preventing crisis.