美國油價期貨報酬與股市報酬率之非線性關係

機構典藏 > College of Business and Management > Graduate Institute & Department of Banking and Finance > Thesis > Item 987654321/31681

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/31681

Title:	美國油價期貨報酬與股市報酬率之非線性關係
Other Titles:	Nonlinearity of oil price and stock price returns in US market
Authors:	陳隆昌;Chen, Lung-chung
Contributors:	淡江大學財務金融學系碩士班聶建中;Neih, Chien-chung
Keywords:	單根檢定;門檻共整合檢定;門檻誤差修正模型;Unit Roots Test;Threshold Cointegration;Threshold Error-Correction Model
Date:	2005
Issue Date:	2010-01-11 01:07:39 (UTC+8)
Abstract:	隨著國際間對於能源需求量愈趨殷切，而全球能源中又以石油最為國際市場所重視，因此探討油價報酬對於股價報酬率之重要性自然不在話下。而油價波動對於一國所受到最直接之影響反應在財務市場上則為股票市場。本研究以美國紐約商品交易所之最近月份期貨價格及道瓊工業指數作為樣本，利用非線性模型研究1995年1月1日至2004年12月31日間，探討兩財務市場間之相互非線性關係及價格傳遞效果與因果關係。為了解油價期貨及股價指數報酬所可能存在的非線性關係，本文首先以NP(Ng and Perron,2000 ) 、PP (Philips and Perron, 1988) 、KPSS ( Kwiatkowski, Philips, Schmit and Shin,1992)、 KSS單根檢定法(Kapetanios et al.,2003)等之線性與非線性單根檢定以測試時間序列資料定態(Stationary)關係，再利用Enders and Granger(1998)之門檻自我回歸模型(Threshold Autoregressive Model；TAR)及動差門檻自我回歸模型(Momentum-Threshold Autoregressive Model；M-TAR)進行門檻共整合檢定(Threshold Cointegration Tests)，並利用非對稱之誤差修正模型(Error-Correction Model)來捕捉長期關係並與傳統之誤差修正模型(Error-Correction Model, ECM)之估計結果進行比較。實證結果則發現無論利用線性或非線性單檢定模型，兩時間數列在近十年之時間序列原始資料皆為非定態，但經過一階差分後均呈現定態。而在運用門檻共整合檢定方式則得知股價指數報酬及油價期貨報酬在1%之顯著水準下呈現拒絕無共整合及對稱效果之虛無假設，表示兩變數之間具有長期均衡關係及非對稱關係。並再進一步利用門檻誤差修正模型以測試價格傳遞之效果則發現，單向的因果關係短期間存在於股價指數報酬影響油價期貨報酬之中，此一結果亦與Cetin(2002) 對油價期貨與S&P500指數研究之結論相呼應。而在長期，本研究亦發現單向的因果關係存在於油價報酬影響股價指數報酬之間。 With the coming of increased needs on international commodities, oil price movement has surely been focused on because it is certainly one of the most important and popular commodities in the world. It goes without saying that the importance of the relationship between oil price returns to stock price returns . In this paper, we are trying to use Nymex nearest oil futures prices and Dow Jones Industrial Index as the samples during the periods of 1995-2004 to discuss nonlinearities and causality relationship between these both financial markets. In order to understand the possible nonlinear relationships between oil price returns and stock price returns, we tried to use nonlinear and linear unit roots tests as the comparison to test the stationary of all the data bases, then we further used threshold cointegration and threshold error-correction model to understand long-term equilibrium relationship and causality while it is up or down of the pre-tested threshold through the choice of optimized module of MTAR or TAR. Through experimental results, we found that these time series of both financial markets are stationary after first differential however they are non-stationary for the original samples no matter by traditional or nonlinear unit roots tests. And both of the variables have asymmetrical long-term equilibrium after the testing of nonlinear threshold coin-integration. We further used threshold error-correction model and found that one way causality relationship, in the short-run, existed on stock price returns to oil price returns, which is the same result shown in Cetin’s (2002) paper. However, in the long-run ,one way causality relationship existed on oil price returns to stock price returns.
Appears in Collections:	[Graduate Institute & Department of Banking and Finance] Thesis

Files in This Item:

File	Size	Format
	0Kb	Unknown	311	View/Open

Loading...