روابط متقابل بازارهای جهانی سهام، بازار پول و بازار ارز خارجی

This article investigates different aspects of global financial markets, specifically relationships among equity markets, money markets, and foreign exchange markets across countries. To represent the three major financial markets of the world, Japan is the proxy for Asia, Germany is the proxy for Europe, and the United States is the proxy for North America. Strong evidence exists that international money markets and international equity markets are becoming increasingly integrated over time. This article incorporates foreign exchange values as partial determinants of equity returns and money market returns and investigates the interactions among these three asset markets from a global perspective.

The world economic order appears to be evolving into three primary geographic regions: Asia, Europe, and the Americas. Financial markets are leading the charge due largely to their ability to transmit and incorporate information rapidly. Linkages between and among countries continue to increase in importance, and the recent Asian financial markets crises have emphasized the need to better understand the interactions. Shocks, such as stock market disruptions, attacks on currency values, or interest rate changes to protect exchange rates spill over not only into other financial markets within the originating country but also into financial markets of other countries around the world. The effects are felt through changes in basic economic fundamentals and often result in political instability as trade flows are disrupted, production costs and sales of multinational firms change, and social unrest begins to appear. These issues affect investors as they attempt to evaluate returns; domestic, foreign, and multinational businesses as they try to maintain profit levels; and governments as they strive to sustain political stability while accomplishing domestic economic goals. Extensive research has focused on cross-country relationships of equity markets, money markets, and foreign exchange markets, separately. The studies have investigated international correlations, cause/effect, and mean/volatility spillovers for each of the three financial market segments. However, in the domestic setting, the interaction of equity markets, interest rates, and exchange rates within national boundaries has received appreciable attention. This study combines the domestic and the international analysis and (1) investigates relationships among national equity markets, national money markets, and foreign exchange markets and (2) examines linkages among the three sets of financial markets. An extensive literature exists for each of these markets. In recent years, relationships between national stock markets seem to have received the most attention. The crash of October 1987 made clear that none of the world's major equity markets operates autonomously, and the jolt to global stock markets spurred new studies relating to how closely markets are tied, which markets influence other markets, and whether identified relationships are intertemporally stable. A primary method of studying linkages and measuring the degree of integration between national stock markets has been to analyze the correlation structure between returns of indices in the different markets. Findings have been somewhat mixed. Ripley (1973) found that well-developed markets displayed a high degree of comovement during the decade of the 1960s, while Dwyer and Hafer (1988) found low correlations when studying the 1973–1987 period. Meric and Meric (1989) concluded that the comovements were more stable when measured over long as opposed to short-time horizons, and Medewitz, Abdullah, and Olson (1991) found the correlation coefficients of national market index returns to be increasing over time. Increased integration has been supported by additional studies, which focused on the behavior of equity market relationships before and after the October 1987 crash, such as Arshanapalli and Doukas (1993), Jeon and von Furstenberg (1990), Koch and Koch (1991), Lee and Kim (1993/1994), Longin and Solnik (1995), Malliaris and Urrutia (1991), and Tang (1995). Theodossiou, Kahya, Koutmos, and Christofi (1997), however, discovered little change in correlations for the United States, Japan, and the United Kingdom as a result of the October 1987 crash. Researchers have also studied cause/effect relationships. Eun and Shim (1989) and Knif and Pynnonen (1999) found innovations in the U.S. equity market to be transmitted to other markets with no significant feedback effects, while Koch and Koch (1991) found U.S. market influence decreasing over time. Arshanapalli, Doukas, and Lang (1997) and Theodossiou and Lee (1993) found that both mean and volatility spillover effects were stronger from the U.S. market to the other markets than the reverse. Focusing only on the United States and Japan, Becker, Finnerty, and Gupta (1990) concluded that the United States strongly affects Japan but the Japanese influence on the United States is insignificant. Darbar and Deb (1997) found significant permanent and transitory covariance between major equity markets except the United States and Japan where only transitory covariance was significant. Thus, recent studies generally conclude that national equity markets are linked, that those linkages have been strengthening over time, and that U.S. dominance is decreasing. Relationships in the second set of markets, money markets, were among the earliest relationships investigated. In these works, integration was measured by linkages between interest rates in different geographic markets. Early studies (based on Hendershott, 1967) assumed the dominance of the U.S. money market thereby, ignoring the possibility of a reverse transmission mechanism or feedback effects. Later works, such as Hartman (1984), Kaen and Hachey (1983), and Swanson (1987), explicitly allowed for feedback effects, acknowledging impacts of foreign market behavior on the U.S. market. These works concentrated on the relationship between yields on dollar-denominated assets in the domestic U.S. market and in the Eurodollar market, the latter typically being represented by dollar yields in London. Lin and Swanson (1993) and Thornton (1992) broadened the investigation to include currencies other than US dollars and found feedback effects to be significant for the currencies included. More recently, the recognition that time series of most currency denomination yields are nonstationary but cointegrated has led to interrelationships being studied in the framework of error correction models (ECMs). These studies include Fung and Isberg (1992), who use U.S. and Eurodollar certificates of deposit, Fung and Lo (1995) and Tse and Booth (1995), who analyze U.S. interest rates and Eurodollar futures, Fung, Lee, and Pan (1996), who examine the London Interbank and Europound markets, Karfakis and Moschos (1990), who focus on linkages within the European Monetary System, Arshanapalli and Doukas (1994) and Bremnes, Gjerde, and Saettem (1997), who investigate Eurocurrency deposit rates, and Lin & Swanson, 1993 and Lin & Swanson, 1997, who study domestic, European, and Asian currency markets. All found two-way linkages of the interest rates. The literature for the third set of markets, foreign exchange markets, is more diverse and, in many instances, secondary to some other primary focus. Exchange rates have been investigated in such contexts as purchasing power parity, interest rate parity, portfolio diversification, central bank intervention, etc. Some of the identified relationships include the work of Van de Gucht, Dekimpe, and Kwok (1996), who found significant persistence measures (long-term effects) between Exchange Rate Mechanism currencies. When studying seven of the world's major currencies, Diebold and Nerlove (1989) found commonality in volatility movements across exchange rates. Baillie and Bollerslev (1991) and Engle, Ito, and Lin (1990) reached the same general conclusions for different currencies and different periods. Based on an interest rate parity framework, Moosa and Bhatti (1997) provided evidence of integration between Japan and six other Asian countries. Studies of relationships between exchange rates and stock markets include Bhandari and Genberg (1990), who found no stable long-run relationship between nominal stock prices and nominal exchange rates, Bahmani-Oskooee and Sohrabian (1992), who found no long-run relationship between the S&P 500 and the effective exchange rate of the US dollar, and Solocha and Saidi (1995), who found only weak evidence of linkages between U.S. and Canadian equity markets and the foreign exchange market. Ajayi and Mougoue (1996) analyzed eight advanced economies and found differing directions of effects between long and short run. They found that increases in domestic stock prices negatively affect domestic currency values in the short run but positively in the long run. Conversely, a weakening domestic currency creates a negative stock market effect both in the short and the long run. No studies were found that combined equity markets, money markets, and foreign exchange markets. This omission is somewhat surprising given the role of exchange and interest rates in equity market relationships. Vast quantities of capital flows make exchange rates important factors in determining equity market values, and interest rates are significant determinants of capital flows. Understanding the role of exchange and interest rates becomes increasingly important in light of current world events. As the Euro takes its place among world currencies, how will the European and U.S. equity market relationships be affected by the changing value of the US dollar relative to the new Euro and by common monetary policies required for economic union? What impact on equity market, money market, and foreign exchange market relationships will the Southeast Asian currency crises have? Can identifiable relationships be helpful in determining the potential effects of the 2001 World Trade Center attack? The intent of this study is to investigate relationships within and between these three sets of financial markets: equity markets, money markets, and foreign exchange markets. To represent the three dominant financial markets of the world, the countries selected are the United States, Germany, and Japan.

This article provides a preliminary investigation of cause/effect and lead/lag relationships among financial markets within countries and among financial markets across countries. Based on the three dominant regions of the world, North America, Europe, and Asia as proxied by the United States, Germany, and Japan, relationships among equity markets, foreign exchange rates, and interest rates are investigated for the period 1993–1997, inclusive. The following comments are generalizations derived from the three sets of causality tests: bivariate (Table 4), multivariate (Table 5), and combined models (Table 6). The strongest relationships identified are among the equity markets, with causality emanating from the U.S. and the German markets. Each significantly affects the other two markets. The Japanese equity market, however, displays no significant effects on the others. These findings must be interpreted in light of the financial market conditions in Japan during the period covered. The Nikkei did not recover from its drastic fall from its high of 38,916 (which occurred at the end of 1989) during the study period. Instead, the index moved around 20,000, reaching a low of 14,517 by the end of 1995. In addition, the lingering recession in Japan was affecting both interest rates and the value of the yen. Although signs of improvement occurred several times during the 1993–1997 period, none of the improvements could be sustained. A relationship is identified between the yen and Nikkei returns and between the mark and DAX returns, with the German relationship being stronger. Interest rate differentials contribute to explaining U.S. equity returns but they reflect no effect on any of the other variables. The Eurodeposit returns results identify no relationships involving the Euroyen, but bidirectional causality is suggested between the Eurodollar and Euromark. The direction of the Granger causality effects provides further insight. Based on the combined models results, the U.S. equity market affects the German and Japanese equity markets in a positive manner, that is, they move in the same direction. Causality from the DAX to the S&P and to the Nikkei, however, is negative, with bivariate tests providing support for the negative relationship between the DAX and the S&P but not between the DAX and the Nikkei. The U.S. equity market influences are stronger than those of Germany. These German results may be reflecting inflation concerns created by reunification and possible effects on German financial markets from the removal of Gorbachev and the threat of Russian bond defaults. These equity market results support the findings of Theodossiou and Lee (1993) and Knif and Pynnonen (1999), who discovered stronger spillover effects from the U.S. equity market than from other markets, and those of Becker et al. (1990), who found that the U.S. equity market strongly affects the Japanese market but that reverse effects are insignificant. The findings, however, do not agree with those of Eun and Shim (1989) who, using VAR methodology and studying the period 1979–1985, found no feedback to the United States; this study finds feedback from Germany. A part of these discrepancies probably relate to the different time periods (November 1987 to June 1994 vs. January 1993 to December 1997), differing length observation periods, the unique economic conditions in Japan, and reunification effects in Germany. Interest rate effects are significant in the U.S. equity market only. Both the dollar/yen and dollar/mark differentials show significant causality to the S&P in the bivariate tests; in the combined model tests, only the dollar/mark differential is significant. Since the differentials are computed as Eurodollar minus Euromark (yen), the indicated direct relationship means that higher dollar returns (lower foreign currency returns) strengthen the U.S. equity market. This finding does not match the conventional wisdom. An increase in interest rates usually is bad news for stock markets as higher interest rates raise the cost of capital for firms. However, other influences were being played out in the United States. During the study period, the Federal Reserve increased real federal funds rates as the U.S. economy began to recover from the slowdown of the early 1990s. This interest rate change was a reaction to a rebounding economy and concurrent recovery in corporate profits, which aided U.S. equity prices. In this situation, economic conditions rather than interest rates were helping to drive equity prices. Relationships within the Euro yield markets themselves provide more traditional results. Bidirectional causality exists between Eurodollar and Euromark returns in the bivariate tests but not in the combined models. Even in the bivariate case, the direction of impact is opposite. An increase in the Eurodollar yield causes a decrease in the Euromark yield. This suggests a substitution effect between dollar- and mark-denominated short-term assets. However, an increase in the Euromark yield causes an increase in the Eurodollar yield. As Germany raised interest rates to combat expected inflation, concerns for the German economy may have created a safe haven effect for dollars, driving interest rates in the same direction. No relationship is identified for Euroyen returns. Purely comparable literature is limited. The most relevant study is probably that by Ahmad and Sarver (1994) who analyze 3-month money market returns for 10 of the world's major financial markets and find less interdependence among money market yields than had been found among stock market yields. In addition, while they found the U.S. market to be less influential in money markets than in equity markets, they concluded that the United States plays a leading role in that after effects of U.S. shocks are stronger and last longer than shocks in other markets. They found Japan to be among the least influential. Their results are consistent with the results of this article. Exchange rates, like interest rates, exhibit a minor role in the markets of this study. The results concerning the relationship between the foreign exchange value of a currency and its respective equity market are mixed. There are no significant relationships in the combined models except an inverse relationship between the mark and the German equity market. A stronger mark (an increase in the exchange rate expressed as dollars per mark) is associated with a weaker German stock market. The same inverse relationship is found between the yen and the Japanese equity market in the bivariate tests. For the U.S. equity market, however, the opposite is true. The U.S. equity market strengthens as the dollar strengthens against both the mark and the yen (bivariate tests). This means a stronger dollar Granger causes a stronger U.S. equity market. This latter relationship of currency values and equity markets moving together is the more usual finding in the literature. The results do not allow broad generalizations about the relationships among equity market returns, interest rates, and foreign currency values. In bivariate and multivariate analyses of relationships of the markets, numerous significant causality relationships are identifiable. The same is true for paired asset markets. However, when all three markets are combined, most of the significant lead/lag relationships disappear. These combined markets tests are the most relevant because they allow simultaneous determination of values in the three asset markets across countries. Because of nonsynchronous trading hours, however, no pure contemporaneous measure exists. In summary, the day starts in Japan with other equity markets closed. In this study, the close of the Nikkei does not significantly affect the opening of the DAX or the S&P. However, the close of the DAX has a negative effect on the close of the S&P and the open of the Nikkei. Continuing westward, the S&P close has a positive effect on both the Nikkei open and the DAX open. The U.S. effects are stronger than the German effects. Although shocks in the U.S. equity market (and, to a lesser degree, the German market) will be transmitted to the Asian markets, the Asian markets appear to react and adjust to the impact, but that adjustment does not get passed on to other markets when viewed in an east-to-west time sequence. The Japanese market appears to be entirely reactive. This interpretation is supported by findings related to foreign exchange effects. A stronger dollar strengthens the U.S. equity market while a stronger yen and mark have the opposite effect on their respective equity markets. Interest rate effects appear to be internalized in the Japanese and German markets, but the U.S. market is affected by U.S. and German interest rate differentials. This again suggests the dominance of the United States' role in international financial markets. To further define the included relationships, two extensions of the current study should be explored. The first deals with volatility issues. The inclusion of volatility spillover effects, in addition to the mean spillover effects studied here, may help clarify the related behaviors of these markets. Further, an analysis of intertemporal stability may yield new information, especially relative to the role of the Japanese financial markets.