The stock market is viewed as a complex dynamic system, and investor sentiment has an important impact on index fluctuation. This study constructs investor sentiment from margin trading business perspective and investigates its impact on the Chinese stock market index fluctuation in multiple time scales. First, we utilize 12 indicators and two-stage PCA to construct a composite investor sentiment index of the margin trading business(ISMT). Second, based on TEI@I complex system theory, we use the VMD algorithm to decompose and reconstruct the ISMT, Shanghai Securities Composite Index(SSEC), and Shenzhen Securities Component Index(SZI), and obtain multiple time scale measurement sequences that reflect short-term, medium-term, and long-term fluctuation of each index, respectively. We provide evidence that the ISMT has an asymmetric impact on stock market index fluctuations. Specifically, for long-term trend, the ISMT has a significant positive impact on the SSEC, and a significant negative impact on the SZI. For medium and short-term trends, the ISMT has a significant positive impact on both the fluctuations of SSEC and SZI, and the impact degree on SSEC is greater than SZI. We also find that the impact degree of ISMT on SSEC and SZI decreases from short to long-term trend. In addition, we measure the fluctuation periodicity of ISMT in multiple time scales based on Fast Fourier Transformation, investigate the impact result during the COVID-19 pandemic, discuss the impact of ISMT on the other nine indexes commonly used in the Chinese stock market, and evaluate the predictive power of ISMT for 11 stock market index returns. This paper takes a new perspective and technology to investor sentiment research, and the results enrich relevant financial theories. The findings are crucial for investor decision-making and financial department regulation. •Extend the research scenario about investor sentiment to the margin trading business, and construct corresponding sentiment index(ISMT).•Investigate the impact of ISMT on index fluctuation in multiple time scales.•Measure the fluctuation periodicity of ISMT in multiple time scales, discuss the results during the COVID-19 pandemic, and analyze the impact on more stock market indexes.•Evaluate the predictive power of ISMT for 11 stock market index returns.