系統風險是否可以預測台灣經濟與股票市場的衰退

Abstract

本論文運用Adrian and Brunnermeier (2011)所提出的系統風險衡量方法CoVaR與衡量極端風險的VaR來估計出樣本中每家公司對系統風險的貢獻，從而累積出國家的系統風險。結果顯示兩種衡量方式不管是CoVaR或VaR都具有預測台灣經濟衰退或股價指數下跌的能力。我們發現累積的非金融機構極端風險(VaR)與台灣GDP呈現負顯著。與美國市場比較，台灣的系統風險與非金融機構的極端風險也非常相關。連續前兩期的金融機構CoVaR與連續前五期的非金融機構VaR對台灣股價指數皆具有負顯著的解釋能力，而相對於衡量單一機構極端風險的VaR，衡量機構對系統風險影響力的CoVaR能擷取到金融機構的影響力。

Our study builds on the CoVaR proposed by Adrian and Brunnermeier (2011) and VaR methodology, which allows us to aggregate the time-varying estimates of the systemic and catastrophic risk contribution for each firm in our sample. The result shows that both measures of systemic risk and catastrophic risk have predicted power for the economy and stock market in Taiwan. We find that the aggregate nonfinancial VaR is negatively significantly related to Taiwan GDP. Compared to the impact of the catastrophic risk from the financial sector in the U.S., the effect of the catastrophic risks in Taiwan is influenced more by nonfinancial sector. Furthermore, the lagged one-month and two-month financial ∆CoVaR and the lagged one- to five-month nonfinancial VaR are negatively significant in explaining Taiwan weighted index. Although the significance of the aggregate systemic risk measures is different for financial and nonfinancial sectors, the aggregate ∆CoVaR measure still captures the influence of financial sector in Taiwan.