Analysis of the Multifractal Characteristics of the Chinese Stock Market Based on Deep Wavelet Transform

Through in-depth analyses of stock market data and trends, we explore its indicative role in reflecting the overall state of the economic system, thereby laying the foundation for reducing financial risk and providing a reliable basis for decision-making by investors and government agencies. By leveraging the latest deep wavelet learning framework, we combine wavelet analysis and deep wavelet transforms, focusing on both single and multifractal theories. Our primary investigation centers on the fractal structural characteristics of the Chinese stock market in recent years, with particular attention to the evolution of multifractal features. Using the CSI 300 Index, Shanghai Composite Index and Shenzhen Composite Index as case studies, we find that while these indices exhibit similar overall trends, they differ in their detailed fluctuations. Further analysis reveals significant multifractal characteristics within the Chinese stock market, and we demonstrate that deep wavelet transform can decompose stock market closing prices at multiple levels, enhancing the analysis of stock price volatility trends. Additionally, we observe that although retail investors may influence short-term market fluctuations, their impact on long-term market trends remains relatively limited. These findings offer fresh insights into the complex dynamics of the Chinese stock market and provide valuable theoretical and practical guidance for policymakers in optimizing financial regulation and for investors in adjusting their investment strategies.