A New Multiple Time Series Approach for Data Analysis
utF L woϊw
F 2012N809() 16:30`17:30
ꏊF RwHwPّ7u`
yuTvzWhat caused what between multiple time series is the central and classical question in data analysis. For the theoretical and empirical contributions on cause and effect in the macroeconomy, C.W.J. Granger and C.A. Sims earned Nobel Prize in Economics Sciences in 2003 and 2011, respectively.In this talk I will show the Wald test of one-way effect causal measure for cointegrated process presented by Yao & Hosoya (2000) and Yao (2007). This new approach for the non-stationary multiple time series analysis was considered as an extension of Grangers non-causality theory. I also show how to apply the new approach to the analysis of causal relationships of the spotlighted stock market composite indices of the United States and East Asian countries in time domain and frequency domain. Based on error correction model for daily observations in the mentioned stock markets around 2008 financial crisis, we showed the stock market causal characterizations between the United States and Japan, China, Korea, Hong Kong as well as Taiwan.

jpeg jpeg jpeg jpeg