Research Highlights : SAI Gyousei

■ Research Field:

International Financial Measurement and Time-Series Analysis

■ Research Activities:

2020-Present: An approach known as DCC-Mixed Data Sampling is being used to improve the accuracy of the traditional economic model of analyzing data at the same frequency. DCC-Mixed Data Sampling consists of short- and long-term components, the long-term correlation between two assets being determined.

2018-2019: It was found that directly observable data from financial markets showed that several components, ranging from short-term fluctuations to long-term fluctuations, were superimposed upon one another. Thus, when considering investment strategies, it is necessary to grasp the relationship between short-term, medium- and long-term periodic movements and different assets. To address these issues, we used a method of analysis called the wavelet analysis to clarify the relationship between cyclical financial market movements and different assets.
Paper: “Multi-Horizon Dependence between Crude Oil and East Asian Stock Markets and Implications in Risk Management” (with Hamori, S., Yang, L., and Tian, S. 2020)

2016-2017: Using what is known as the extended asymmetric DCC correlation model of the DCC model and the copula model, we aimed to capture the concerted behavior of the Chinese stock market. We showed that the dependence structure of the Chinese stock market was asymmetric, and that the lower tail dependence was particularly strong.
Paper: “Dynamic correlation and equicorrelation analysis of global financial turmoil: evidence from emerging East Asian stock markets” (with Tian, S. and Hamori, S. 2016）

2014-2015: It has been found that the conventional method of analyzing the bivariate correlation of markets is not applicable to actual investment strategies and portfolio risk management, which require correlation with multiple variables and temporal changes. For these problems, the correlation coefficient of temporal changes using a correlation model known as the dynamic conditional correlation (DCC) was estimated, making it possible to discover an unusual movement in the financial crisis. In addition, a multivariable correlation model known as dynamic equicorrelation revealed dynamic correlations among nine stock markets.
Paper: “Modeling dependence structures among international stock markets: Evidence from hierarchical Archimedean copulas.” (with Yang, L., Li, M. and Hamori, S. 2015)