The liberalization of capital flows has facilitated high integration between international financial markets, increasing interdependence among the developed economies in the East Asian region. The investigation into this interdependence among financial markets has been a significant focus throughout literature, where understanding the behaviour of international financial markets’ inter dependencies is crucial for making asset allocation and risk management decisions. Assessing the changing interdependencies is also critical for determining the nature of financial crises.
For example, the experience of recent financial crises suggests that the interdependence among the financial markets during tranquil periods is different from that of crisis periods, where often, during financial crises, we observe that the interdependence tends to break down. Consequently, we can observe a strong increase in the co-movements (correlations) of the returns between markets. It is argued by some that a structural break in the correlations demonstrates that the international propagation mechanisms of financial shocks are discontinuous (Monica Billio and Loriana Pelizzon, 2003; Giancarlo Corsetti, Luca Dedola, and Sylvain Leduc, 2005; and Toni Gravelle, Maral Kichian, and James Morley, 2006). Indeed, this break is owing to financial panics, or the herding or switches of expectations across multiple equilibria (equilibrium with speculative attacks vs. equilibrium without speculative attacks; Paul Robert Masson, 1999).
Although there is no consensus among specialists (Carlo A. Favero and Francesco Giavazzi, 2002), this phenomenon has often been described as “contagion” (Taimur Baig and Ilan Goldfajn, 1998; Kristin Forbes and Roberto Rigobon, 2002; and Roberto Rigobon, 2003). Forbes and Rigobon (2001) discuss crisis-contingent theories, qualifying this phenomenon as “shift-contagion”. The authors assume that investors behave differently after a crisis, implying the generation of new temporary channels of propagation, in addition to the permanent channels. This behaviour characterizes the interdependence between the economies. By contrast, in non-crisis-contingent theories, there is no difference in the transmission mechanisms between crises and stable periods. In that vein, the shocks are propagated through strong linkages between countries, such as trade links (Stefan Gerlach and Frank Smets, 1995; and Giancarlo Corsetti, Paolo Pesenti, Nouriel Roubini, and Cedric Tille, 1999), financial links (Graciela L. Kaminsky and Carmen M. Reinhart, 2000; and Caroline Van Rijckeghem and Beatrice Weder, 2003) or common shock (Masson, 1999; and Forbes and Rigobon, 2001). Forbes and Rigobon (2002) used the term “interdependence” to refer to this situation.
The objective of this paper is to investigate the presence of shift contagion in the context of the Asian crisis. Our aim is to study the stability of the international propagation of financial shocks across various stock markets. More specifically, we test for a structural break in the correlation of asset returns across countries during periods of high turbulence. In contrast to previous studies on financial contagion, we allow for a time-varying correlation. There are extensive empirical studies investigating the stability of the international propagation of financial shocks by a correlation analysis. In the empirical literature, the contagion is measured by the significant increase in the correlation between financial markets (Forbes and Rigobon, 2002). Mervyn A. King and Sushil Wadhwani (1990) are the pioneers who used this methodology to test for the presence of contagion. They found that the correlation between the stock markets of the United States, the United Kingdom and Japan had increased after the U.S. crash of 1987. Other studies have extended this test of correlation into other types of financial markets (markets of the sovereign debts, exchanges and the interest rate) and other episodes of crises (Sara Calvo and Carmen Reinhart, 1996; and Baig and Goldfajn, 1998).
According to Forbes and Rigobon (2002) these tests, based on cross market correlations, have reached the same conclusion of contagion occurring. However, tests based on the analysis of conditional correlation admit several limitations. The use of the high frequency financial series affects the test through three types of bias: heteroskedasticity, simultaneous equations and omitted variables (Ehud I. Ronn, 1998; Forbes and Rigobon, 2002; Rigobon, 2003; and Ga-won Yoon, 2005). Forbes and Rigobon (2002) tested the increase in the correlation coefficients adjusted from only a heteroskedasticity bias, where no. structural break was detected. Thus, they concluded that the propagation of the Asian crisis resulted from the interdependence between the financial markets and not from contagion. Moreover, Forbes and Rigobon (2002) showed, by simulations, that their tests are biased when the data suffer from simultaneous equations and omitted variable problems. In order to correct these problems Rigobon (2003) has proposed an original methodology to test for a structural break in the correlation across financial markets. He applies a structural change test (determinant of the change in the covariance matrix test) using a limited information estimation based on an instrumental variable (IV) method, which is constructed by splitting the sample into two windows (a window of the stability and a window of the crisis). Rigobon (2003) studies the stability of the international propagation mechanisms between 36 stocks markets during three recent international financial crises (Mexico 1994, Asia 1997 and Russia 1998).
The results illustrate that the increase in the correlation between these stock markets does not result from instability in the mechanisms of propagation, but rather is the consequence of a strong interdependence during the crisis periods, as well as during the stability periods. Although the conclusions of Rigobon (2003) are interesting, these results have been considered not robust as the size of the crisis window has an important influence on the sensitivity of the results (Mardi Dungey and Diana Zhumabekova, 2001; and Billio and Pelizzon, 2003). Another important consideration, as Gravelle, Kichian, and Morley (2006) point out, is the subjective and arbitrary choice of the structural change points, which define the beginning and the end of the crisis window.
Download
PDF Ebook The Asian Crisis Contagion: A Dynamic Correlation Approach Analysis
