COVID-19疫情對美國股市的影響真的比全球金融風暴時期嚴重嗎?

Abstract

COVID-19疫情影響全球的經濟，造成全球股市動盪不安，在2008年發生了全球金融風暴，也造成全球經濟衰退，因此本研究應用財務計量模型擬探討COVID-19期間和全球金融風暴期間十六家公司的個股，就估計波動率而言，哪個時期對美國股市的影響最嚴重？本研究資料來自於Yahoo Finance 資料庫，並以八種產業，十六家公司的個股每日調整價格和報酬率進行分析，資料從2006年1月3日至2021年10月14日，共3974筆。並將2006年1月3日至2009年12月31日作為全球金融風暴的時間；2018年1月2日至2021年10月14日作為COVID-19疫情期間。我們使用變異數異質性Generalized autoregressive conditional heteroskedasticity (GARCH模型)、Integrated Generalized autoregressive conditional heteroskedasticity (IGARCH模型)、Glosten-Jagannathan-Runkle Generalized Autoregressive Conditional Heteroskedasticity (GJR-GARCH)模型配適財務時間數列。也使用具有變異數異質性市場模型評估「系統性風險」並計算風險係數。診斷分析涵蓋的方法有採用Ljung-Box test檢定時間序列自我相關性，Jarque-Bera test檢定樣本偏態與峰態是否服從常態假設、其他分配假設之檢定，ARCH effect 檢定變異數異質性，Joint effect 檢定模型波動性的不對稱性。我們使用全球金融風暴時期和COVID-19期間的日報酬波動來比較，在金融風暴時期震盪幅度較大的是好市多、亞馬遜公司、美國銀行、康卡斯特集團和台積電。而在COVID-19時期波動較劇烈的公司有嬌生公司、沃爾瑪、NIKE、棒約翰、麥當勞、百勝餐飲集團、華特迪士尼、微軟以及英特爾，而在兩段時期皆受到很大波動的公司是美國航空和蘋果公司。我們發現在全球金融風暴時期十六家公司的個股中唯獨美國銀行的風險係數大於1，屬於高風險的金融資產。

The COVID-19 pandemic has greatly affected the global economy and caused turbulence amonginternational stock markets. In 2008 the Global Financial Crisis (GFC) also hit financial markets caused a global economic recession. Therefore, this research uses econometric models to compare 16 stocks listed in the U.S. equities market during the COVID-19 and the GFC time periods. With an aim to find which period has seenthe most severe impact on the U.S. stock market in terms of estimated volatilities, we consider those 16 stocks within eight industries and download their daily adjusted closing prices from the Yahoo Finance database. For GFC, we use the timeframe from January 3, 2006, to December 31, 2009 and take January 2, 2018, to October 14, 2021 as the period of the COVID-19 epidemic. We examine the ARCH effects and utilize GARCH-type models for model fitting, including the integrated GARCH model, IGARCH model, GJR-GARCH model. We also employ market models with heteroskedasticity error to capture time-varying conditional variances.Regarding the diagnostic checking, we use the Ljung-Box test for testing autocorrelation, the JB test for testing the normality assumption, the ARCH effect for testing the heteroskedastic variance, and the Joint effect for testing the asymmetry of model volatility. We compare the estimated volatility of each stock based on a suitable econometric model during the GFC and COVID-19 periods. The findings indicate that Costco, Amazon, Bank of America, and Taiwan Semiconductor Manufacturing Co. experience the most significant fluctuations during the GFC period, while Johnson & Johnson, Walmart, Nike, Papa John’s International Inc., McDonald’s, Yum Brands Inc, Walt Disney, Microsoft, and Intel have the largest volatilities during the COVID-19 period. The results also show among these sixteen stocks during the GFC period that only Bank of America had a risk factor greater than 1, implying a high-risk financial asset.