【摘要】The pandemic of COVID-19 has caused severe public health and economic consequences around the world. It is of great importance to evaluate the impact of COVID-19 to the economic， especially the stock market. To this end， we proposed to use several state-of-art sparse principal component analysis (PCA) methods to the stock data of the CSI300 index from February 1st 2019 to February 1st 2021. To show the influence of the outbreak of COVID-19， we divide this period into two periods， i.e.， before and after January 1st 2020. Based on this division， we attempted to extract the principal components and construct portfolio accordingly. The results show that the proportion of principal components representing the market declined after the COVID-19 epidemic. For the constitution in the first two principal component， the important stock sets are substantially different after the outbreak of COVID-19. The stocks from the health care sector start to play important role in the portfolio of the CSI300 index after COVID-19. Compared with the CSI300 index， the first two principal components from the sparse PCA methods can obtain higher returns with a much smaller set of stocks in portfolio. In conclusion， the outbreak of COVID-19 led to changes in both proportion and constitution of the principal component of the stocks in the CSI300 index.