This paper takes the internet public opinion of Sino-US trade friction events as the research object, tracks the change trend of Baidu Index by searching the keyword“trade friction”, and takes the duration of Baidu Index peak value as the“event window”to construct GARCH (1, 1) model based on virtual variable regression. The model studies the impact of the network public opinion of Sino-US trade friction on the volatility of stock price excess return of three representative enterprises in China’s textile industry chain: Xinfengming, Bailong Orient and Jiangsu Guotai from March 2018 to April 2020. The purpose of this paper is to consider the complete market information contained in internet public opinion from the perspective of behavioral finance, and to test the impact of Sino-US trade friction on the industrial chain of Chinese textile enterprises through the influence of the information on investor behavior. The empirical study shows that the important events in each stage of Sino-US trade negotiation process represented by the peak value of Baidu Index have a significant negative impact on the abnormal return rate of stock prices of three sample enterprises in China’s textile industry chain, but the marginal effect of the impact tends to decrease with time. For specific enterprises in the industrial chain, the early impact of trade friction is transmitted to the industrial chain first. Through the transmission of the industrial chain, the upstream enterprises are gradually impacted. When the enterprises participate in the division of labor in the industrial chain, the impact and diffusion effect of trade friction on the upstream and downstream enterprises of the industrial chain is stronger.