I. Latin America, victim of a structural, financial lack of confidence

After the pause following the Mexican crisis, Latin America was one of the first areas to benefit from the return of private origin capital flows toward developing countries during the period from the middle of 1996 to the middle of 1997. The primary borrower in an increasingly speculative market, in spite of everything its situation was paradoxical. While market quality was degrading (drop in the quality of the issuers, increasingly reduced risk premiums), Latin America, which on the contrary constituted a zone of progress, has suffered from a persistent lack of trust. In the period prior to the Asian crisis, the principal Latin American issuers (Brazil, Argentina and Mexico), despite everything, benefited in absolute figures from the reduction in risk premiums, obtaining less onerous financing than in the past. But when the financial crises were unleashed, they were the first affected by the problem and the flight to quality.

From the end of 1995, private flows toward emerging countries returned to the marked progression that had characterized them in the 1991-1993 period. The debt flows in obligations (domestic and international) have been occupying a growing proportion of these. With 46 billion dollars in 1996 and an equivalent amount reached in just the first quarter of 1997, those flows represented more than 20% of total flows prior to the financial crisis. (Table 1)

The primary and secondary international debt markets at that time experienced considerable expansion, to which was added the equally dynamic development of the restructured debt markets (the Brady bonds market). In turn, those markets became increasingly more speculative. High profitability, with low historic international yield, attracted an increasingly more sustained demand and led to the progressive erosion of risk premiums (Graph 1), to the extension of the average term, and development of fixed income securities. Simultaneously, in search of volume, the market opened up to ever more risky issuers, among which were, in the first place, sovereign issuers like Russia, but also numerous banks and emerging companies. In the case of just national securities, the risk quality of the issue dropped two positions in general terms (on a 16-level risk scale) when compared with 1993.

In absolute terms, Latin America benefited significantly from this process. During this period, its issues accounted for more than 50% of placements; in addition, the countries of the region could resort to longer duration issues. At the same time, interest rates dropped significantly (Graph 2). Latin American issues were so popular in the markets that Argentina, for example, was able to obtain its annual financing through that procedure.

A more in-depth analysis however shows that the entire area appears to have suffered from an excess of a lack of trust when compared with other issuers¹. Calculations on a selection of Euro-obligations effectively show that the implicit risk level deducted from the risk premiums used was, overall, with respect to Latin America, much higher than the effective risk, calculated on the basis of country-risk indicators (Graph 3 and Box 1).

That excess mistrust, very evident in the case of Brazil and Argentina, has been fed by the significant volume of Latin America’s financing needs. But it also appears to be the result of a traditional view of the risks, dating back to episodes of a lack of liquidity and particularly, the debt crisis of the 80s. Even in the presence of abundant liquidity, international investors (influenced by agency ratings) preferred countries with low indebtedness levels. In the new context, however, countries that were already indebted provided the best advantages. The need to reduce internal demand in order to confront the debt reimbursement maturities, imposed maintaining restrictive monetary policies. Simultaneously, that external adjustment policy led to exchange policies geared to keeping the same levels of competitiveness (actual anchoring) or to drastically adjusting the productive sector (increase in productivity and salary adjustment), in the case of nominal anchoring. In all cases, these countries appeared to be the ones in the best condition to contain their trade deficits and the progression of their external debts. (Graph 4)

This configuration contrasted resoundingly with the Asian growth systems that, in retrospective, proved to be the most dangerous. In Asian economies, the absence of initial restrictions led to reheating processes associated with expectations of increasingly higher growth. Credit explosion, increase in unproductive investments and loss of control over capital income were definitely reflected in an increase in trade deficits and the explosive dynamics of indebtedness (particularly in the private sector) that led to the crisis.

The financial crises in Thailand in July 1997 (which provoked the Hong Kong crack of October 1997 and the Korean crisis), and then later in Russia in August 1998, share the same brutal turnaround in confidence exhibited by investors. Both translated into a rupture in external financing flows and inability to honor short-term commitments (lack of liquidity crisis). Loss of confidence was proportional to the previous lack of foresight, leading to a phenomenon of overreaction. That phenomenon was quickly corrected after the Asian crisis, but appeared again in spectacular proportions in the case of the Russian crisis². (Graph 1)

Moreover, both episodes evolved into a contagious phenomenon that affected the set of emerging securities. This phenomenon may be partially explained by the panic that spurred investors to take refuge in safer securities, not as subject to unfavorable risk valuations (flight to quality). But it is fundamentally the result of a strategy of assigning portfolios of international investors that compensate their losses with the sale of assets, which at that moment offer the best yields. In both cases, Latin America was severely affected.

Because Euro and Brady bonds were an important part of the obligation stocks (Euro and Brady) benefited from high yields during previous months, Latin American securities were naturally the first assigned. Given that in addition they were victims of a structural mistrust and classified poorly on the emerging risk scale (agency ratings were not too favorable), these, more than others, suffered from the flight to quality phenomenon.

During the most acute phase of the Asian crisis (end of October and beginning of November 1997), spreads of sovereign securities of Latin American countries, with less than 10-year maturities showed an increase of 130 base points (thus reaching 300 base points above U.S. Treasury bonds). On the other hand, securities with maturities exceeding ten years experienced an increase of 240 base points in their risk premiums. (Graph 2) The most affected were Ecuador, characterized by premium levels that were quite high before the crisis, and Argentina and Brazil, countries that had to deal with very elevated external debt servicing payments. (Graph 2)

Table 2 - Latin America : sovereign Euro-obligations spreads

The evolution of liquidity conditions and pes of international interest was not unfavorable. Therefore, the Euro-obligation market slowly reopened between January and July 1998. This evolutionary path was associated with a moderate decrease in spreads. But the Russian crisis marked the beginning of a new phase. In effect, the vertiginous drop of the ruble and the moratorium on the Russian public debt payment (August 1998), the threat of a generalized contagion, with possibilities of deteriorating into successful speculative attacks against Latin American currencies all became a reality. Meanwhile, the economies of the region had effectively worsened because of the poor evolution of the international environment and its foreign trade (see sections later in this paper). Bewildered, markets and some observers grew concerned about Venezuela, weakened as it was by the drop in oil prices (which seemed to draw it nearer to the Russian example) and by the lack of political visibility, and Argentina and its currency board, although the latter was less vulnerable than in 1995, the year in which it was subjected to strong pressure by the «tequila effect». Moreover, in early September, when Colombia decided to expand the fluctuation band for the peso in comparison with the dollar, some thought that this was the beginning of mass devaluations in Latin America. However, the overreaction of the markets was aimed above all at Brazil, perceived to be the weak link in the South American chain. Thus, Brazil experienced an attack on its currency sign, the "real".

Within a period of a few days, spreads in the sovereign securities of Latin American countries, securities with less than 10-year maturities, rose almost 750 base points to more than one thousand base points above the U.S. Treasury bonds. As a result, the spreads of the short and medium-term bonds reached a level similar to that achieved by securities with maturities exceeding ten years, which increased their risk premium by 575 base points. (Graph 2)

Reflecting investors’ risk perception, Ecuador –affected by the mistrust of markets at a time when growing internal macroeconomic imbalances were converging with the open political crisis in 1997, Brazil and Argentina recorded the highest spreads when the Russian crisis exploded. (Table 2) Colombia, Mexico and Venezuela also experienced strong hikes in their external financing costs.

The convergence of a series of internal factors in the Latin American region (particularly the firm but costly political defense of the "real" by Brazilian authorities, as well as the results of elections in that country, along with the announcement of an agreement with the IMF), and external factors (declarations of the G7, drop in U.S. interest rates) explains the moderate drop in the spreads in October – November 1998. However, risk premiums stabilized at a much higher level than they had reached before the Asian crisis.

Latin American risk premiums experienced an important rise during these two episodes. The analysis of confidence indicators shows that the rise was comparatively higher than for all the other emerging securities, particularly when the Russian crisis exploded. Only Venezuela escaped this phenomenon: the effective degradation of the risk far outstripped the perceived degradation. (Graph 3)

Besides risk premium level, throughout the financial crisis and particularly as of the Russian crisis, Latin American countries experienced a reduction in external financing volume. In effect, during the first three quarters of 1997, Latin American countries went ahead with gross issues of Euro-obligations on the market, amounting to more than 50 billion dollars. On the other hand, issues during the last quarter, did not even reach 4 billion dollars (Graph 5). Interrupted by the Russian crisis, recovery during the first half of 1998 was limited. For 1998 in its entirety, Latin American issuances reached 38 billion dollars (of a total of 73 billion), as compared with 58 billion in 1997 (of a total of 110 billion). Although they continue to dominate, their amount, as a relative portion, has significantly dropped.

Most assuredly, Latin American countries have experienced the self-realizing phenomenon of mistrust. Extraneous to the immediate causes of the Asian and Russian crisis and rather protected from the perverse effects of excess liquidity, they have seen their volume and financing conditions deteriorate, a posteriori, justifying the mistrust that penalized them.

NOTES

¹ For a detailed presentation of the methodology used to measure excess lack of confidence, see Ricoeur-Nicolaï, 1998.

² Regarding the effect of the contagion unleashed by the Asian and Russian crises and the impact on Latin America, see: ECLAC, 1998a: Laurent, 1998; Miotti and Quenan, 1998; Perry and Lederman, 1998.

http://www.sela.org
sela@sela.org
   SELA,  Permanent Secretariat
Torre Europa, Fourth floor, Urb. Campo Alegre, Av Francisco de Miranda,
Caracas 1060- Venezuela
Tlf: (58) (212) 955.71.11 Fax: (58) (212) 951.52.92