The events in the second half of 1997 in Southeast Asia and their effects on the global economy are still far from fully understood. However, because of their immediate impact on Western financial markets, concerns about the integrity of financial markets in a global context became a top priority issue in almost every regulator’s agenda. The extent of spillover effects from one market to another, the impact of the information-revealing process on markets in different time zones, the effectiveness of circuit breakers, and the role of derivatives markets are issues still being evaluated by regulators and market participants, in a world where markets are no longer as segmented as they were just a few years ago.

In Brazil, stock markets enjoyed an upward trend during the whole first semester of 1997. From January 1^st to July 8^th, when it reached its annual high, the Ibovespa, the main Brazilian market index, rose 93.4% (see Figure 1 below). This impetus was mostly fueled by the perceived strengths of the Brazilian stabilization plan along with an aggressive privatization programme, including the public telecommunications system, Telebrás, whose shares account for well over half of the daily trading volume in Brazilian markets.

July 15^th marked the beginning of the crisis. Thailand abandoned the fixed exchange rate system against the US dollar and the baht quickly fell more than 20%. Other economies of the region followed close behind, all their currencies suffering abrupt devaluations during the next few days. The contagion effect reached Brazil immediately. By the end of the week, on July 18^th, the index had lost 15%. As shown in Figure 1, the next seventy days were a period of very high volatility, during which Ibovespa had many up and downturns around the average of roughly 12,000 points. The crash on October 23^rd caught most players in the domestic financial institutions off-guard. The shock came after three weeks of upward momentum, as fundamental prospects for the Brazilian economy seemed to be improving.

Futures markets reflected the crisis accordingly. Figure 2 depicts the effect of the crisis on the supposed efficiency of the Brazilian stock and index futures markets. It shows both the Ibovespa and the relative difference between the Ibovespa futures basis and its cost-of-carry. It is clear that after October 23^rd arbitrage conditions worsen with an unusual narrowing of the basis.

There is no question that such high volatilities represent a threat to the integrity and efficiency of domestic markets. A deeper understanding of the mechanisms of transmission of information and volatilities between markets across the globe is of paramount importance to all regulators. This understanding is a prerequisite for the design of new regulations and the amendment of current market rules, wherever they are needed.

In this paper we investigate the existence of an eventual spillover in returns and volatility from equity markets in Southeast Asia to those in Brazil — used as a proxy for the Latin American region. The time frame chosen is the year of 1997 and the key events are the Thailand currency crisis and the Hong Kong crash. The analysis of the spillover effect during that year is important not only to increase our understanding of what happened at that time but to assess the way relevant information is revealed between non-overlapping time zones. This enlarges regulators’ awareness of the consequences that financial events on the other side of the world have for their respective markets.

Statistical tests are performed in order to measure the general increase in volatility and the spillover effect itself. Section 1 aims at the former, using two daily regional indices and several subsamples in 1997. Fundamentally the question asked is whether the perceived increase in volatility is supported statistically. The latter problem, the analysis of an eventual spillover effect, is addressed in Section 2. Causality tests between the Asian regional index and the Brazilian main market index are performed, both for the series of returns and for a proxy of the volatilities, the series of squared returns. As will be seen, there is very strong statistical evidence for the existence of a spillover effect.

Próxima Página

Índice

Fale com a CVM