حرکت جمعی بازار سهام بین المللی: اصول اقتصادی و یا سرایت بیماری؟

We investigate the return comovement in international equity markets with a focus on the distinction between economic fundamentals and contagion. We examine the potential macro news effect based on a comprehensive data set of macroeconomic news announcements made in the U.S., U.K., and Japan. Our results show that the bulk of the observed comovement in the intraday and overnight returns of the international equity markets cannot be attributed to public information about economic fundamentals. In contrast, foreign market returns exert a dominant influence on the subsequent domestic market returns. Overall, our findings suggest that future inquiry on market comovement may focus on the distinction between contagion and trading on private information, rather than public information.

In this paper, we compare the extent to which return comovement between international equity markets may be explained by economic fundamentals versus contagion. Earlier studies establish the economic significance of equity market comovement across national borders (e.g., Becker et al., 1990 and Hamao et al., 1990), but finding the source of the comovement has been difficult for academics and practitioners. Understanding the determinants of the comovement has profound implications for international diversification, market integration/efficiency, and the cost of capital for multinational firms. The traditional view stresses the role of common fundamental factors Solnik, 1974a, Solnik, 1974b, Stulz, 1981 and Adler and Dumas, 1983, but it is difficult to account for the large comovement around the globe in extreme circumstances often without apparent new information. King and Wadhwani (1990) propose an alternative explanation for the market linkage. They argue that trading of stocks in one market per se affects stock prices in other markets, even if the source of the trading is purely noise. They call this the “market-contagion” hypothesis. To distinguish between the two competing explanations, economic fundamentals and contagion, we separate the influence of foreign markets on the domestic markets into two components: one that is driven primarily by economic fundamentals and the other by foreign market returns. Specifically, we examine the market comovement between the domestic intraday (and overnight) returns and foreign intraday returns for the U.S., U.K., and Japan, conditional on both the macroeconomic news announcements and the most recent foreign intraday returns. In a linear return-generating model, we measure the marginal effect of the foreign market returns and the macro news effect. If the foreign returns are redundant after controlling for the macroeconomic news announcements, then evidence for the contagion hypothesis is weaker. On the other hand, if the foreign returns do have a separate and substantial influence, then the evidence favoring the contagion hypothesis is stronger. King and Wadhwani (1990) argue that the return correlation between markets increases with the volatility in each market, and they interpret this as evidence supporting contagion hypothesis. Karolyi and Stulz (1996) also find evidence consistent with this interpretation. On the other hand, Ross (1989) argues that market volatility is related to the underlying information flow including public information. Public information flows may then be associated with higher volatility and more pronounced comovement, all in the context of a rational approach to asset pricing. In this paper, we try to distinguish between these two views by examining the return comovement between market conditional on the volatility of foreign news in a nonlinear return-generating model. Using our approach, we gauge the extent to which return comovement changes as the volatility of the macro news varies. The contagion hypothesis implies that the return comovement should change very little after controlling for the volatility of public information. This model also captures the intuition of McQueen and Roley (1993), who stress that macroeconomic news may have nonlinear effect on stock returns. Our sample includes stock index returns and 14 different macroeconomic news announcements from U.S., U.K., and Japan for the period from January 1, 1985 to December 31, 1996. In the study, we mitigate the “stale open quote problem” in both the U.S. (Stoll and Whaley, 1988) and Japanese markets (Hamao et al., 1990) by choosing the “opening” price as the price index quoted at 10:00 am in each market. Furthermore, we account for the stylized volatility patterns using the Glosten et al. (1993) asymmetric GARCH model and estimate both conditional mean and volatility equations simultaneously. Our empirical results show that the bulk of the observed return comovement in the international equity markets cannot be traced to public information about economic fundamentals. We show that our macroeconomic announcement data do convey valuation-relevant information to the capital markets, a result in the spirit of Flannery and Protopapadakis (2002). However, foreign intraday returns play a dominant role in influencing domestic market returns even after we control for the impact of macroeconomic news. This result is robust for both linear and nonlinear news models and across all three countries. Our findings suggest that foreign market returns convey information distinct from the public information flows about economic fundamentals (as measured by the macroeconomic news announcements). In this case, domestic market traders may have incentive to infer the unobservable information from the previous foreign market returns and incorporate valuation information into their subsequent domestic trading. At the same time, the noise embedded in foreign market returns may also be transmitted into the domestic market, thereby connecting returns in the two markets. We find evidence, however, that the return effect of foreign macroeconomic news increases when the announcements are accompanied by large foreign intraday returns and when the volatility of the announcements is large. This result is consistent with the conjecture of McQueen and Roley (1993) that the effect of macroeconomic news announcements depends on the context in which investors interpret them, not just the announcements themselves. Nevertheless, the nonlinear news effects measured here are still too small to have a material impact on the basic return comovement between market. Overall, we find for each country in our sample that returns from the two preceding foreign market sessions are far more important than the preceding domestic returns and macroeconomic news announcements taken together in explaining the domestic intraday and overnight returns. This suggests an interesting path to pursue in future work on international equity market comovement: focus on distinguishing between the contagion hypothesis (e.g., Bae et al., 2001, Forbes and Rigobon, 2002 and Connolly and Wang, 2002) and the hypothesis of trading on private information (e.g., Craig et al., 1995). Our comprehensive data set of macro news events (2902 announcements in total from three countries) distinguishes our paper from the earlier studies (e.g., Cornell, 1983, Pearce and Roley, 1985, Hardouvelis, 1987 and Ito and Roley, 1987), which tend to examine one macro event or news from one country. Moreover, to isolate the potentially short-lived news effect (Ederington and Lee, 1993), we focus on intraday and overnight returns. This setup distinguishes our paper further from the earlier studies, which often explore news effects using longer horizon data (e.g., monthly returns in King et al., 1994). This paper is related to Connolly and Wang, 1998 and Connolly and Wang, 2002 on stock market comovement between countries. Connolly and Wang (1998) demonstrate that the macro news shocks play a more important role in explaining volatility linkage between markets than in explaining return linkage. Connolly and Wang (2002) study the effects of market volatility, dispersion of beliefs and extreme returns on equity market comovement. The present paper distinguishes itself from the two papers by focusing on the role of public information and in using specific tests designed to distinguish news versus trading effects as sources of equity market return comovement. The paper proceeds as follows. Section 2 describes the data and summary statistics. Section 3 introduces the conditional mean and volatility models used in our empirical investigation and the specific test hypotheses. Section 4 reports our empirical results, and Section 5 concludes.

Extensive empirical evidence establishes a strong comovement in international equity markets particularly for the contemporaneously correlation between the foreign intraday returns and the domestic overnight returns in the U.S., U.K., and Japanese markets. In this paper, we investigate two competing explanations of the comovement: economic fundamentals versus contagion. Specifically, we explore a comprehensive data set of the macroeconomic news announcements made in the U.S., U.K., and Japan from 1985 to 1996. We separate the influence of the foreign markets on the domestic markets into two components: one that is driven primarily by economic fundamentals and the other by foreign market returns. In our empirical analysis, we distinguish the unconditional news effect from the conditional news volatility effect of the macroeconomic announcements. The most important finding is that the macro news effect is too small to account for any economically sizeable part of the return comovement among the three national equity markets. By contrast, returns from the previous foreign market session exert an economically significant impact on subsequent domestic market returns and apparently contain information distinct from economic fundamentals. Our results suggest that the bulk of the observed comovement in the intraday and overnight returns of the international equity markets cannot be attributed to public information and, in particular, economic fundamentals. Consequently, further work on the sources of comovement in international equity markets might encompass the distinction between contagion and trading on private information, rather than public information.