RDP 2001-03: The Response of Financial Markets in Australia and New Zealand to News About the Asian Crisis 1. Introduction

As financial markets become more integrated, shocks can be transmitted quickly between them. In times of market turmoil, this implies that the effects of negative shocks might be felt in markets far removed from the originating market. In this paper, we investigate the spillover of financial-market volatility, specifically the impact of news from Asia (Korea, Thailand and Indonesia, as well as Malaysia and the Philippines) during the 1997–1998 financial crisis, on financial markets in Australia and New Zealand. We examine the initial impact of key events and announcements in the Asian crisis period and the spillover of these effects, as measured by both financial prices and proxies of their volatility.

We find that news – both positive and negative – that came out of Asia during the crisis clearly had repercussions for financial markets that were not directly affected by these events. But these effects must be put in perspective: developments in the US market generally had a much greater influence on Australasian price movements and volatility than cross-market shocks originating in the Asian crisis economies. This result is in line with previous work on the importance of overseas returns in Australian markets (Kortian and O'Regan 1996). We also find evidence indicating that stock markets reacted to developments in Asia with a lag, after the US reacted, rather than reacting directly to the news itself.

Our results indicate that the volatility in Australian and New Zealand financial markets was generally as great or greater in late 1998 – which we term the ‘world crisis’ period – than in the 1997–1998 period, when the main news events of the Asian financial crisis occurred. We also find that the apparent spillover of financial market shocks from Asia to Australia and New Zealand was small and – for some asset classes – smaller in the Asian crisis period than previously. This implies that the shocks originating in Asia were less important for Australian and New Zealand markets than were the global ‘common’ shocks affecting all of these markets simultaneously.

The evidence suggests that the volatility seen in Australian and New Zealand markets was not affected by the different stances of monetary policy, or the differing natures of the monetary policy regimes in the two countries. The effects of developments in Asia on volatility in Australian and New Zealand financial markets were remarkably similar, despite the distinctly different methods used to conduct monetary policy over that period. These results may reflect the short-run measure of volatility that we adopt in this paper, however. The levels of the financial-market variables in Australia and New Zealand display differing profiles: there were large divergences in stock and bond prices over the period. On the other hand, the exchange rates of the two currencies against the USD moved together, reflecting that these currencies are generally traded as a bloc.

This paper draws on the literature on contagion (Eichengreen, Rose and Wyplosz (1996) is a key empirical paper; see Dungey (1999) for a survey). Calvo and Mendoza (1999) show that contagion of financial-market volatility might increase as world markets become more integrated. In certain circumstances, the costs of gathering and analysing information about unfamiliar foreign markets may outweigh the perceived benefits. This can result in investors choosing to act on the basis of rumours unrelated to market fundamentals, instead of on complete information. In addition, fund managers may face incentives that encourage herd behaviour in portfolio allocation decisions. Both of these effects can result in contagion of financial volatility from markets in one country to those in other countries.

Masson (1998) has defined contagion as the portion of financial-market volatility that cannot be explained by normal factors such as domestic fundamentals and global common shocks. However, much of the contagion literature focuses on the propagation of exchange rate crises and does not deal explicitly with the transmission of volatility outside of crisis periods (Dungey and Martin (2001) is an exception). This paper bears greater resemblance to the literature on ‘meteor showers and heat waves’, which studies geographic (time-zone) patterns in the volatility of particular securities (Engle, Ito and Lin 1990; Fleming and Lopez 1999). We seek to identify the effect of ‘meteors’ – as measured by news events or volatility in one market – on returns and volatility in other markets.

Previous work on the effects of macroeconomic ‘news’ on Australian financial market prices and volatility has focused on announcements made at pre-scheduled times, such as Australian CPI releases (Campbell and Lewis 1998; Kim 1996). In these cases, the content of an announcement may be a surprise, but its timing is not. Therefore, it is possible for market participants to plan their contingent trading strategies in advance. If the timing of an announcement is not known in advance, however, traders have less opportunity to plan for its effects. Previous empirical work for other countries has suggested that unscheduled announcements tend to have more persistent effects on financial returns than do scheduled announcements (Almeida, Goodhart and Payne 1998), although the difference can be measured in hours. In general, studies of this kind examine the impact of economic announcements on ‘own’ financial markets. The present paper, however, focuses on the effects of unscheduled (though potentially anticipated) announcements relating to one group of countries on the financial markets of other countries.

The paper proceeds as follows. In Section 2, we discuss the reasons why financial markets in Australia and New Zealand might have been affected by the financial crisis in Asia. We also discuss our measure of news events and the financial-market data to be analysed. Section 3 contains the empirical evidence on the response of financial markets in Australia and New Zealand to these news events, both in terms of volatility and price movements. In Section 4, we examine whether the spillover of financial-market returns is greater in times of crisis than in more-normal times, using results from vector autoregressions (VARs). Section 5 contains some concluding remarks.