本研究應用Diebold and Yilmaz (2012)所提出的外溢指標(spillover index) 檢視台灣股票集中市場八大產業指數之報酬率與波動度的 外溢效果, 藉此探討台灣產業間的連動性因果關係。在去除共同因 子的影響後, 本研究實證結果顯示: 台灣產業指數間的外溢效果相 當顯著; 換言之, 台灣各產業間存在著密切的連動性。此外, 各產業 影響其他產業以及受其他產業影響的程度並不一致。就全樣本下的 淨外溢效果而言, 營建業為其他產業的主要影響者, 而電子業與金 融業對其他產業的外溢效果並不如預期地高。實證結果亦發現: 在 不同子樣本期間內, 各產業間的連動性結構亦不盡相同。以美國次 級房貸危機引發全球金融風暴的期間為例,金融產業對其他產業的 外溢效果最為顯著。最後,本研究利用滾動視窗估計法(rolling window estimation) 描繪外溢指標的動態行為,結果發現外溢指標的確 具備隨時間改變的動態特性,且指數波動度對於衝擊發生時的反應 程度較指數報酬率更為劇烈。

In view of the important role played by industrial factors in explaining the cross-sectional stock returns, we apply the spillover index of Diebold and Yilmaz (2012) to investigate the return and volatility spillover effects among industrial indices in the Taiwan stock market. Our empirical results indicate that there exist significant spillover effects among industrial indices; that is, the comovement of industries in Taiwan is evident. We also find that while most industries are receivers of spillover effects, a few industries are the main exporters of return and volatility spillover effects. In particular, the construction industry in Taiwan is the key transmitter of spillover effects. In contrast, the electronic industry and financial industry do not have as significant net spillover effects as we expected. Moreover, based on the robustness analysis, we find that key exporters of spillover effects are actually quite different within sub-samples. For instance, the financial industry has the most significant net spillover effects on other industries during the period of the subprime debt crisis in 2008. Finally, we utilize the method of rolling window estimation to illustrate the dynamics of industrial spillover effects. In fact, the results reveal apparently dynamic behavior of industrial spillover effects in the Taiwan stock market.