[[alternative]]美國次級房貸危機前後美國股市與國際股市的不對稱連動性之實證研究

博士[[abstract]]近來在高度的全球化與自由化之下，各國貿易及金融交易往來頻繁，其金融、證券市場之間的連動性與日俱增。因此，探討國際間經濟衝擊之傳遞 (transmission) 或蔓延 (contagion) 的課題，相當受到學者及實務專家的重視。尤其近廿年來，全球發生經濟、金融危機的次數越趨於頻繁，每一次危機事件的發生，國際間股市皆呈現顯著的危機傳遞、蔓延的共同移動 (co-movement) 或是共同趨勢 (common trend) 效果，不僅導致各國股市的重創，也使得投資人欲藉由國際證券的投資組合 (portfolio) 以達到分散風險的效果大為降低。 過去在蔓延現象的實證研究中，常以共整合檢定比較危機發生前後共整合關係的改變與否，檢測蔓延現象是否存在，但是許多的實證研究卻表示，傳統共整合模型的對稱性調整假設會忽略股票市場在上漲、下跌時，其調整的速度是有所差異的，意即股市之間的共整合關係，在漲與跌時的調整方式是具有不對稱性或是非線性的特徵。 2007 年 3 月初，發生於美國的所謂「次級房貸危機」，除重創了美國與西歐的金融、證券及信貸市場，並迅速地蔓延到世界各國的金融市場，更對各國的實體經濟面造成巨大衝擊，企業倒閉所引發的失業潮，以及資本支出、消費、投資與進出口的減少，導致了全球自 1930 年代經濟大恐慌以來最嚴重的 經濟衰退。而另一方面，近年來中國金融市場的快速崛起，以及對亞洲各國金融市場日益增加的影響力，吸引了許多投資人與研究者的關注，而他們所關心的其中一項議題，便是中國對亞洲的影響力是否已能與美國並駕齊驅，甚至是超越美國，尤其是在次級房貸危機之後。 由於過去關於金融危機所導致之國際股市蔓延效應的文獻多是著重源於新興市場的危機事件，影響範圍有限。而次級房貸危機，卻是源於先進的美國與西歐各國，這使得我們有再去探究次貸危機影響性的必要。在本篇論文之中，我們將利用適當的時間序列模型分析並探討在次貸危機的前後，美國股市與亞洲、歐洲與美洲各國股市的連動性是否具有差異？以及在次貸危機前後，美國股市與中國股市對亞洲各國股市影響性的改變。 本文的實證分析可分為三個主要部分，茲將此三個部簡要分述如下： 第一部份的主題為 “次級房貸危機前後美股對亞洲、歐洲及美洲股市的不對稱性蔓延效果─門檻共整合模型之應用”。在此部分中，我們將採用 Enders and Siklos (2001) 的不對稱門檻共整合模型，分析在次貸危機的前後，美股與亞洲、歐洲與美洲各國股市之間的共整合關係是否有所差異？探討次貸危機是否會使美股對亞洲、歐洲及美洲各國的股市產生不對稱的蔓延現象。第二部分的主題則為 “次級房貸危機前後，誰對亞洲股市具有較強的影響性─美國或是中國”。在此部分中，我們同樣將利用上述的方法，分析在次貸危機的前後，美國股市與中國股市兩者與亞洲各國股市之間共整合關係的改變，探討兩國股市對亞洲各國股市影響力在次級房貸危機前後的變化。而第三部份的主題則為 “次級房貸危機前後美股對亞洲、歐洲及美洲股市的不對稱性蔓延效果─對數平滑轉換共整合模型之應用”。在此部份中，我們將進一步採用以 Granger and Teräsvirta (1993)及 Teräsvirta (1994) 所提出之 logistic smooth transition regression (LSTR) 方法擴充後的 Enders-Siklos 門檻共整合模型，探討次貸危機是否會使美股對亞洲、歐洲及美洲各國的股市產生不對稱的蔓延現象。[[abstract]]In recent years, because of globalized and deregulated financial markets and frequent international trade and financial transactions all over the world, the relatedness among security markets has been steadily on the increase. Therefore, exogenous events, shock transmission, and crisis contagion are the issues that researchers and government officials have been interested in, especially when frequent economic and financial crises took place all over the world in the past twenty years. Every time when a financial crisis occurs, there are significant co-movement effects or common trend effects in crisis transmission or contagion among international stock markets. The co-movement or common trend not only inflicts heavy losses in the international stock markets, but it also weakens the effect of risk diversification for investors who utilize international security portfolios. Past literature about the transmission and contagion effect often employed the co-integration method to examine whether the co-integration relationship around the crises had changed and whether the contagion effect had existed. However, many empirical research pointed out that the assumption of “symmetric adjustments” in the traditional co-integration model ignored that the adjustment speeds were different when the stock market was in an upward status or in a downward status, which demonstrated the “asymmetric” or “nonlinear” features which are often unconsidered in the traditional co-integration method. In early March 2007, “the subprime mortgage crisis,” which occurred in the United States, seriously affected the financial, security, and credit markets in the U.S. and Western Europe, quickly spread to the financial markets in other countries and also hit the real output in the economy in these countries. Corporate bankruptcies accompanied by unemployment import and capital expenditure cuts led to the decline in every country in consumption, investment, import, and export. The scholars thought the chain reactions incurred by the subprime mortgage crisis had never been seen since the Great Depression. Furthermore, the China financial market has been growing fast in recent years, and the influence from China to Asian financial markets has also been steadily on the increase, which is what investors and researchers have been interested in. One of the issues they are concerned with is the influence of China and whether it can keep pace with the U.S., even surpass the U.S., especially after the subprime mortgage crisis. Because past literature about the contagion effect due to the financial crisis mostly emphasized the crisis from emerging markets, the scope of their influence was limited. The subprime mortgage crisis was from the U.S. and Western Europe, which makes it necessary for us to further explore its impact. This article employed the suitable time series model to investigate whether the relatedness between the U.S. stock market and Asian, European, and American stock markets around the subprime mortgage crisis had changed, and whether the relatedness between the U.S. and China’s stock markets and Asian stock markets had also changed. The empirical analysis in this article is classified into three main parts. The topic of the first part is "The Asymmetric Contagion Effect from the U.S. Stock Market to Asian, European and American Stock Markets around the Subprime Mortgage Crisis-An Application of the Threshold Co-integration Model." We employed the Enders and Siklos (2001) asymmetric threshold co-integration model to investigate whether the co-integration relationships between the U.S. stock market and Asian, European, and American stock markets around the subprime mortgage crisis had changed, which analyzed whether the asymmetric contagion effects between the U.S. stock market and these markets around the subprime mortgage had existed. The topic of the second part is "Who has more influence on Asian Stock Markets around the Subprime Mortgage Crisis－the United States or China?." In this part, we also employed the above-mentioned methods to investigate the changes in the asymmetric co-integration relationships between the U.S. and China’s stock markets and Asian stock markets, which analyzed the changes of influence between the U.S. and China’s stock markets and Asian stock markets. The topic of the third part is "The Asymmetric Contagion Effect from the U.S. Stock Market to Asian, European and American Stock Markets around the Subprime Mortgage Crisis－An Application of the Logistic Smooth Transition Co-integration Model." Furthermore, in this part, we employed the Granger and Teräsvirta (1993) and Teräsvirta (1994) logistic smooth transition regressive (LSTR) method, which was the expansion of the Enders-Siklos threshold co-integration model, to investigate whether the asymmetric contagion effects between the U.S. stock market and Asian, European, and American stock markets around the subprime mortgage had existed.[[tableofcontents]]Chinese Abstract iii English Abstract vi Table of Contents viii List of Tables x List of Figures xii Chapter 1 Synopsis of the Dissertation 1 Chapter 2 Introduction Remark 4 Chapter 3 Literature Review 11 3.1 The Definition of Contagion 11 3.2 The Examine Methods for Contagion Effect 12 Chapter 4 Methodologies 16 4.1 Advanced Nonlinear ESTAR Unit Root Test 16 4.2 Threshold Co-integration and Asymmetric Adjustment 17 4.3 The Logistic Smooth Transition Co-integration Model 20 Chapter 5 Empirical Investigation 24 Part I The Asymmetric Contagion Effect from the U.S. Stock Market to International Stock Markets around the Subprime Mortgage Crisis - An Application of the Threshold Co-integration Model 5.1.1 Data Description 24 5.1.2 Empirical Results and analysis 26 Part II Who has more influence on Asian Stock Markets around the Subprime Mortgage Crisis - the United States or China? 5.2.1 Data Description 42 5.2.2 Empirical Results and analysis 42 Part III The Asymmetric Contagion Effect from the U.S. Stock Market to International Stock Markets around the Subprime Mortgage Crisis - An Application of the Logistic Smooth Transition Co-integration Model 5.3.1 Data Description 58 5.3.2 Empirical Results and analysis 58 Chapter 6 Conclusions and Discussion 62 List of Tables Part I Table 1 Summary Statistics for Return on Stock indices 32 Table 2 Results of Various Unit Root Tests 34 Table 3 Results of the Nonlinear Unit Root Test – the KSS Test 35 Table 4 ARMA(p,q)-GARCH(1,1) Modeling and Results of the ARCH-LM Test for the Volatility of Return 36 Table 5 Results of Non-conditional Correlation Coefficient of Return 37 Table 6 Results of Correlation Coefficient of Volatility of Return 38 Table 7 Results of the Engle-Granger Test for Co-integration 40 Table 8 Results of the Enders-Siklos Test for Threshold Co-integration 41 Part II Table 1 Summary Statistics for Return on Stock Indices 46 Table 2 Results of Various Unit Root Tests 49 Table 3 Results of the Nonlinear Unit Root Test – the KSS Test 49 Table 4 ARMA(p,q)-GARCH(1,1) Modeling and Results of the ARCH-LM Test for the Volatility of Return 50 Table 5 Relationships between the U.S. and China 51 Table 6 Results of Non-conditional Correlation Coefficient of Return 52 Table 7 Results of Correlation Coefficient of Volatility of Return 53 Table 8 Results of the Engle-Granger Test for Co-integration 56 Table 9 Results of the Enders-Siklos Test for Threshold Co-integration 57 Part III Table 1 Results of the Logistic Smooth Transition Co-integration Test 61 List of Figures Figure 1 Flow Chart 7 Part I Figure 2 Logarithms of the Stock Price Index in 19 Stock Markets 33 Figure 3 The Volatility of Return in 19 Stock Markets 39 Part II Figure 4 Logarithms of the Stock Price Index in 8 Stock Markets 48 Figure 5 The Volatility of Return in 8 Stock Markets 55[[note]]學號: 895530110, 學年度: 9