بازارهای نوظهور و بحران مالی: شوک های جهانی یا منطقه ای و یا ایزوله؟

This paper investigates financial contagion of three emerging market crises of the late 1990s, as well as the subprime crisis of 2007, focusing on financial markets of emerging economies, USA and 2 global indices. Conventional cointegration and vector error correction analysis show long and short run dynamics only among emerging stock markets during the Russian and the Asian crises, for both stock and bond markets during the subprime crisis, while the Argentine turmoil has no impact on any of the examined markets. Further analysis into a multivariate time-varying asymmetric framework provides evidence on the global impact of the Russian default, the contagion effects of the subprime crisis, the regional aspect of the Asian crisis and the isolated nature of the Argentine turmoil. Moreover, stock markets seem to constitute a stronger transmission mechanism during the three contagious crises. Our findings have crucial implications for international investors, policy makers and multi-lateral organizations.

A central issue in asset allocation and risk management is whether financial markets become more interdependent during financial crises. This issue has acquired great importance among academics and practitioners, especially since the appearance of several emerging market crises of the 1990s (Mexican currency crisis in 1994–1995, Asian crisis in 1997, Russian default in 1998, Argentine crisis in 1999–2001, and Brazilian stock market crash in 1997–1998). Until then, financial crises models were developed with regard to crises as events occurring in individual countries. However, those crises episodes refocused the empirical research on the examination of contagion effects and the inter-regional or inter-continental nature of the shocks. Common to the majority of these events was the fact that the turmoil originated in one market extended to a wide range of markets and countries, in a way that was hard to explain on the basis of changes in fundamentals (Rodriquez, 2007). There is an extensive literature on financial contagion during several crises of the 1980s and 1990s (see Dornbusch et al., 2000 and Kaminsky et al., 2003, for excellent surveys). To measure volatility spillovers and contagion among markets, early research used a range of different methodologies, such as the principle components model (e.g., Calvo and Reinhart, 1996), spillover models (e.g., Glick and Rose, 1999), cointegration and vector error correction models (e.g., Sheng and Tu, 2000), models of asymmetries and non-linearities (e.g., Baur, 2003), and models of interdependence (e.g., Bekaert et al., 2005). Furthermore, the existence of financial contagion has been studied mainly around the notion of “correlation breakdown” (a statistically significant increase in correlation during the crash period) (e.g., King and Wadhwani, 1990 and Calvo and Reinhart, 1996). However, since the thought-provoking paper by Forbes and Rigobon (2002), a number of limitations to the literature on financial contagion have been highlighted (e.g., a heteroskedasticity problem when measuring correlations, a problem with omitted variables and the need for a dynamic increment in the regressions, affecting at least in the second moments correlations and covariances). To overcome those restrictions and provide sufficient evidence of contagion, researchers have been already using more sophisticated approaches, such as models of conditional asymmetries and correlations (e.g., Chiang et al., 2007 and Kenourgios et al., 2011), regime-switching models (e.g., Pelletier, 2006), and dynamic copulas with or without regimes (e.g., Rodriquez, 2007 and Okimoto, 2008). This paper investigates the existence of a correlated-information channel, through which contagion can be viewed as the transmission of information from more-liquid markets or markets with more rapid price discovery to other markets, focusing initially on three major emerging market crises (Asian crisis, Russian default and Argentine turmoil).1 The analysis covers both equity and bond markets in selected emerging market economies (EMEs) of various regions (Latin America, Asia, Europe, Middle East and Africa), as well as USA and 2 global indices for equities and bonds, for comparative reasons. To expand the scope and the contribution of our research, we also investigate the contagion effects of the subprime crisis of 2007–2008 in the examined EMEs.2 Our purpose is to identify in a broader framework the propagation mechanism of crises with different characteristics occurred in emerging (Asian currency crisis, Russian and Argentinean government defaults) and developed countries/regions (U.S. subprime crisis), and elucidate how vulnerable emerging financial markets are to both emerging and global shocks. To serve this purpose, we maintain, following similar studies in the literature (e.g., Forbes and Rigobon, 2002 and Bekaert et al., 2005), an equivalently strict definition of contagion as the increase in the probability of crisis, beyond the linkages in fundamentals, and the rapid increase in co-movements among markets during a crisis episode. Understanding the nature and the differences in crises dynamics has crucial implications for international investors, portfolio managers, policy makers and multi-lateral organizations. Initially, we employ a conventional empirical analysis (Johansen cointegration tests and vector error correction model), and find significant long and short-run market dynamics only among emerging stock markets during the Russian and the Asian crises, for both stock and bond markets during the subprime crisis, while the Argentine crisis has no impact on any of the examined financial markets. However, to overcome the limitations of the contagion literature, we also apply a recently developed GARCH process, the asymmetric generalized dynamic conditional correlation (AG-DCC) model developed by Cappiello et al. (2006), who generalized the DCC-GARCH model of Engle (2002). Average correlations between the crisis country and all other countries during stable and crisis periods are estimated, which allow capturing conditional asymmetries in both volatilities and correlations in a multivariate framework and determining whether cross-market correlation dynamics (contagious effects) are driven by changes in macroeconomic fundamentals or by behavioral reasons. Results provide evidence on the global contagion effect of the subprime crisis and the Russian default, the intra-regional aspect of the Asian crisis and the decoupling of the examined national and global market indices to the Argentine turmoil. This paper contributes to the existing literature in the following aspects. First, we provide new evidence on financial theory of contagion by examining the existence of an asymmetric propagation mechanism for financial crises which have different characteristics and origins. The AG-DCC GARCH model applied in this paper is well suited to examine asymmetric responses to negative shocks (stronger contagion during negative shocks) and, to the best of our knowledge, has not been used before to test the contagion hypothesis for all four financial crises. Second, we differentiate our analysis from previous studies, focusing on emerging stock and bond markets from various regions around the world, instead of individual mature economies, given the limited research on EMEs. In this framework, we identify which of the two asset markets are more prone to financial contagion in EMEs. The propagation of contagion into bond markets of emerging economies with heavily exposure on debt as primary financing source (IMF International Capital Markets, September 1998, and IMF Global Financial Stability Report, September 2006) constitutes a topic of little empirical investigation, since the majority of existing studies focus mainly on the transmission of shocks across foreign exchange and stock markets. Third, the analysis of contagion of the subprime crisis is also of great importance, given the existing debate on whether the contagion effect on EMEs has been muted and uneven (decoupling hypothesis) or not. Fourth, examining differences in crises dynamics reveals several explanatory factors and contributes to the debates regarding the resilience and sustainability of emerging-market policy performance and the construction of a new international (or regional) financial architecture. The structure of the paper is organized as follows. Section 2 presents the methodologies applied to examine financial contagion. The data used for the empirical analysis is presented in Section 3. Section 4 reports the empirical results. Finally, concluding remarks are stated in Section 5.

This paper investigates the correlated-information channel as a contagion mechanism for three emerging market crises of the late 1990s, as well as the subprime crisis of 2007, using data from equity and bond markets of EMEs from various regions around the world, USA and 2 global stock and bond indices. Through conventional cointergation and VEC analysis, we report long and short run dynamics only for stock markets during the Russian and Asian crises, for both stock and bond markets during the subprime crisis, while the Argentine crisis has no impact on any of the examined financial markets. To provide a more robust analysis of financial contagion, we also examine conditional correlation dynamics into a time varying asymmetric framework, applying the recently developed AG-DCC model. This empirical analysis elucidates how vulnerable emerging financial markets are to both emerging and global shocks and displays differences in crises dynamics that could be attributed to several factors. Results confirm the existence of asymmetric contagion to EMEs and globally only for the Russian default and the subprime crisis. Both crises hit those economies regardless of their economic integration, since cross-market correlation dynamics are driven by behavioral reasons, due to shifting investor sentiment (increased risk aversion), causing significant changes in the emerging countries’ financial structures. The other government default episode (Argentine crisis) examined in this paper has an isolated nature, due to the lower macroeconomic and financial vulnerability to shocks of emerging economies with close trade and financial ties to Argentina and the shifts in the emerging economies investor base due to earlier emerging market crises. On the other hand, the asymmetric contagious effects of the Asian crisis display strong intra-regional characteristics, due to the growing share of investment and trade in the region and the more common monetary policy followed by the Asian countries after the October crash of 1987. Our findings have important implications for international investors and portfolio managers. Evidence on contagion implies that diversification sought by investing in multiple markets from different regional blocks is likely to be lower when it is most desirable. As a result, an investment strategy focused solely on international diversification seems not to work in practice during turmoil periods. On the other hand, evidence provided on the lower spread of all four crises through bond markets indicate that, in times of distress, any potential benefits from international diversification are greater for the bond investors than the stock investors. Finally, since countries and financial markets react differently to sovereign shocks, combining bonds and stocks from different emerging economies could provide advantages over debt-only or equity-only portfolios. The results also provide useful implications regarding the ability of policy makers and multi-lateral organizations to insulate or at least attenuate an economy from contagious effects. The intra-regional contagious effects of the Asian crisis indicate the effectiveness of the co-ordinated rescue packages from the International Monetary Fund (IMF) to the most affected Southeast and East Asian (Thailand, Indonesia and South Korea). The rescue packages were of a considerable sum, as of July 1998 the combined value of the IMF rescue packages amounted to a total of $100 billion. Regarding the Argentine crisis, policy initiatives by both the IMF and EMEs (e.g., increased EME data dissemination, floating rate regimes) following previous crises have led to improvements in country surveillance. In the case of the Russian crisis, the high interest rates both in the U.S. and other economies after the shock led to widespread fears of a liquidity crash and raised questions about policy performance and coordination, despite the bailout of LTCM organized by the U.S. Fed and the followed expansionary monetary policy on a global base. Finally, the subprime crisis raised the need for a revamped international financial architecture. The global contagious effects of this crisis and the rejection of the decoupling hypothesis for the EMEs question the resilience and sustainability of emerging-market policy performance. It seems that strong economic indicators in many EMEs before the crisis (high growth rates, massive foreign exchange reserves, balanced budgets) were not enough to decouple them from the crisis, because of their cyclicality and endogeneity. A consequence of the contagion on EMEs would be the redirection of development loans by the World Bank, the IMF, and the regional development banks to the public sector, since those funds had been crowded out by private-sector lending throughout the boom decade (World Bank, Global Development Finance, 2008).