ارتباط نوسانات محدود جهت نفوذ بین بازار 225 نیکی ژاپن و دیگر بازارهای مالی مشخص
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|14164||2014||12 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Physica A: Statistical Mechanics and its Applications, Volume 393, 1 January 2014, Pages 364–375
Recent financial crises have shown the importance of determining the directionality of the influence between financial assets in order to identify the origin of market instabilities. Here, we analyze the correlation between Japan’s Nikkei stock average index (Nikkei 225) and other financial markets by introducing a volatility-constrained correlation metric. The asymmetric feature of the metric reveals which asset is more influential than the other. As a result, this method allows us to unveil the directionality of the correlation effect, which could not be observed from the standard correlation analysis. Furthermore, we present a theoretical model that reproduces the results observed in empirical analysis.
The availability of an increasingly large amount of financial data is providing opportunities to understand how macroscopic effects in markets emerge from interactions between microscopic system components. Financial markets can be regarded as an evolving complex system, where the assets prices change by traders sentiment which is also affected by multiple interactions among different financial assets. The emergence of the self-organized behavior of the market has been studied using a variety of approaches from statistical physics and non-linear dynamics to computer science. A lesson learnt from the recent financial crisis initiated by the announcement of the bankruptcy of Lehman Brothers in 2008 is that the financial markets are getting more interconnected around the world. As a consequence of this interdependence, a perturbation in a single market may have a high impact in other financial markets, leading to domino losses. The Euro crisis is believed to be a systemic crisis involving many financial institutions which are closely connected to each other through capital flow and political linkages between countries. The financial institution network is strongly affected by the dynamics and correlation of financial assets such as bonds, stocks, foreign exchange rate, and vice versa. This urges us to understand and examine in depth the correlations among different financial markets, from a complex system perspective, in an effort to determine the directionality of the influence between financial assets. This knowledge could then lead to identify the origin of market instabilities and prevent future global crises. On the other hand, the dynamics and correlations of financial products have attracted the interest of asset managers in the field of portfolio risk management, in which data analyses have led to the development of financial management techniques such as the tail risk and market crash risk strategies. For price dynamics, several stylized facts are well known, such as fat tailed return distributions , , ,  and , the absence of autocorrelations of returns  and , volatility clustering , ,  and  and aggregational normality  and . Recently, the resistance (support) and breaking-acceleration effects in foreign exchange markets have also been reported . Moreover, memory effects in the resistance and support lines for stock price dynamics have been studied in Ref. . Research on the correlations between financial assets has been carried out in several works, mainly in terms of risk and portfolio management. International asset correlations increase in a period of high volatility of the market . The average correlation between daily stock returns predicts subsequent stock market excess returns . High volatile markets are directly related to strong correlations between them . The average correlation among US stocks scales linearly with market stress and is reflected by normalized Dow Jones Index Average (DJIA) returns . These results imply that when the market volatility drastically increases in response to any shock to the system (e.g., a bankruptcy that leads to a financial crisis), the correlation between assets could be stronger than that in a normal state. As a consequence, the dispersion effect of the portfolio could decrease significantly. On the other hand, several studies have investigated correlations among financial markets from different view points. For example, the Index cohesive effect on stock market correlations was investigated in Ref. . The results supported the view of the stock market as a Complex Adaptive System, CAS, as was proposed previously in Ref. . Their findings implied that the Index is more than a weighted average of all the stocks. In other words, it represents a self-reference or a feedback loop between the market as a whole and the individual stocks . Moreover, the analysis of the daily correlation present among market indices of stock exchanges located all over the world in long time periods revealed that the correlation among market indices presents a combination of both fast and a slow dynamics . In addition, a partial correlation analysis of the stock market uncovered a dominating clasp of the financial sector . A correlation-based framework was also proposed for quantitative assessments of the coupling and interdependences between different markets in the global financial village. This methodology was used to analyze the feedback between the micro (intra market) and the macro (inter markets) levels . In Wang et al. , a study for 48 world indices was presented, one for each of 48 different countries. The results revealed a long-range power-law cross correlations in the absolute values of returns that quantify risk. In addition, their findings also indicate that cross correlations in the absolute values of returns decay much more slowly than cross correlations between the returns. Podobnik et al.,  reported on the long-range cross-correlations in time series of absolute price fluctuation of NYSE members. The interplay between network dominance and node influence was recently introduced to study financial markets . It is worth mentioning that a recent empirical study of stock returns in the Nikkei 225 Index for the period of Jan. 2007–Dec. 2009 showed that market-wide price co-movement becomes prominent before and after a large price decline such as an endogenous market crash . However, although correlations among markets, using the specific frameworks have been extensively investigated, metrics that combine correlations with the directionality feature have not been studied before. In this work, we focus on the correlation of returns for Japan’s daily Nikkei 225 financial index and other markets: the Dow Jones Index Average (DJIA) and the foreign exchange rate of the US Dollar and Japanese Yen (USDJPY). We empirically analyze how the correlation between these assets relates to their volatility by introducing new metrics to capture the asymmetric correlation phenomena. By definition, the standard correlation coefficient (Pearson’s correlation coefficient) of returns for two financial assets does not vary by exchanging the two asset returns, because the concept of standard correlation coefficient Cor(A,B)Cor(A,B) between AA and BB is symmetric by exchanging the two AA and BB variables. Here, we propose a measure of the correlation of returns, which is computed from a collection of pairs of returns constrained by its standard deviation. This constrained correlation metric has an asymmetric feature with respect to exchanging the two asset returns (AA and BB). As shown later, this asymmetric feature allows us to determine which asset (AA or BB) affects more to the other (BB or AA), respectively. Our real-data driven approach using the asymmetric property of the constrained correlation determines which asset is more influential to the other. This is our main result and it indicates that we can measure not only the importance of the transmission channels among markets but also the directionality of the effects during a contagion event. The economic implications of this result are also discussed. At the same time, as a second result, our analysis reveals that when the volatility of returns increases, the correlation between the pair of assets becomes stronger. Namely, higher volatility implies a higher correlation and, conversely, lower volatility leads to a lower correlation between Japan’s Nikkei stock average index (Nikkei 225) and other financial markets. A theoretical model is also proposed to investigate the origin of the observed directional correlation effects. The computer simulated results using the model shows a fair agreement with the empirical data, suggesting a way to explore the directional correlations using theoretical approaches.
نتیجه گیری انگلیسی
In this work, we have analyzed the correlation between the Nikkei 225 and other financial assets, in terms of the volatility of the base assets. We have introduced the volatility-constrained correlation metric and examined the asymmetric feature of this metric with respect to the base asset. Our main finding shows that the Nikkei 225 is influenced by the USDJPY and the DJIA more than the reverse case. This reflects the international economic power balance between Japan and the US, namely, the economy of the US is more influential than that of Japan. Furthermore, our results also show that, in general, when the volatility of base asset returns increases, the correlation between Japan’s Nikkei stock average index (Nikkei 225) and other financial markets gets stronger. On the other hand, in the low volatile period, the correlation is weak. Unlike the standard correlation coefficient, our metric allows us to identify which asset is more influential than the other. In future research, it could be important to study the correlation among stocks, bonds, and currency markets with respect to this asymmetric feature and analyze the direction of the correlation between them. Large-scale financial data analysis across assets is becoming highly important for asset managers, banks and even monetary authorities to develop quick responses to fluctuating market conditions. Moreover, in many cases, it is a priority to identify the origin of fluctuations. The presented method using volatility-constrained correlation could be an additional technique to determine the directionality of the influence between financial markets.