透過您的圖書館登入
IP:3.133.108.241
  • 期刊

因素分析運用於股票與基金市場相關性探討

Factor Analysis of Multivariate Time Series Data on Stocks and Funds Market

摘要


本研究運用Error Corrected Model,將財務時間序列資料間存在的關聯性,如自我相關,共整合先予以抽離,這樣使得資料做初步分析後,能排除這些特殊意義的關聯成份,讓訊息更顯明確,也讓模型的殘差序列資料穩定且合乎因素分析假設。進而對殘差序列資料做因素分析,萃取出資料短期的關聯因子。本研究運用EM演算技術在因素分析的參數估計上,這樣的方法相當便捷及具效率。實證的資料分析結合股票與基金市場,找出兩市場間的互動關聯因子,發現台灣股票市場及開放式股票型基金幾乎只存在一個主要的關連性,也就是所謂的短期伴隨效應,自我相關及共整合效應並不顯著。這種分析的結果,將多變數時間序列資料之間的不同關聯結構完整地分析出來。這是之前大多數的研究者所無法釐清的事,乃是本文主要之貢獻。更爲此類資料分析提供一個完整的範例程序。

並列摘要


This thesis presents the process of research into the use of factor models for multivariate time series data on stocks and funds Market. Time series data of financial exist distinctive behavior including autoregresstion and cointegration, to apply Error Corrected Model (ECM) to extract the correlation to gain more simple information. Using the process is quite applicable to agree the hypothesis of factor analysis about the residual series. And through the residual to analyze the correlation of short run movement for multivariate time series. This thesis present an expectation-maximization (EM) algorithm to estimate parameters .It is computationally very efficient in space and time. The real data of this research shows that almost one relationship between the stocks market and funds market is concurrent effect. To express more completely the multivariate time series data about correlation structures is main contribution by this thesis and provides a complete procedure for this kind of data analysis.

參考文獻


林建義(2008)。股價相關性共整合分析研究─以台灣股市與國際股市爲例。天主教輔仁大學金融研究所在職專班。
楊奕農(2005)。時間序列分析─經濟與財務上之應用。台北市:雙葉書廊。
Bartlett, M. S.,"The Statistical Conception of Mental Factors",British Journal of Psychology, Vol 28, 1937, pp.97-104.
Bartlett, M. S., "Further Aspects of the Theory of Multiple Regression", Proceedings of the Cambridge Philosophical Society, Vol 34(1), 1938, pp.33-40.
Bentler, P. M.,Tanaka, J. S.(1983).Problems with EM for ML Factor Analysis.Psychometrika.48,247-257.

延伸閱讀