Research On Stability Of Chinese Stock Market Under The Viewpoint Of Evolutionary Network Analysis

Stock market is an important part of the financial market,which can reflect the stability and health of the industrial economy.During last two decades,many researchers have well utilized complex system theory to guide their research on financial market modeling and analysis.There are fruitful theoretical modeling,analysis and dynamic characterization studies have been published.Among them,by using geometric topological properties and dynamic evolution behavior analysis of complex financial systems,the quantitative characteristics related minor changes of the market structure and stability can be easily found based on complex network analysis methods.Based on the inherent statistical correlation of stock data,this paper carries out the modeling of the correlation network,the construction of the measurement index system and the stability analysis of the Chinese stock market from the perspective of building a single-layer and multi-layer the staged spatiotemporal evolutionary network that reflects the development of the financial market.First of all,according to the internal correlation characteristics between stock entities,this paper constructs a single-layer and multi-layer correlation network of stock assets with "linear correlation layer","non-linear correlation layer"," tail correlation layer" and"partial correlation layer;Then,by employing the information filtering tools to obtain the minimum spanning tree,the plane maximum filter graph and the triangular maximum filter graph,of the association matrix,the important nodes and edges of the backbone network of a specific stock market are analyzed;Furthermore,this paper also incorporates the historical memory effect of the market into modeling and analysis,and studies the potential secondary impacts of events such as "Sino-US trade friction","pandemic of the new crown epidemic" and"derivative effects of Russia-Ukraine conflict" on the stability of China’s stock market;By introducing some quantitative characteristics measures,such as global efficiency of the associated stock network,fragmentation index,the network fractal dimension,and the network entropy,the internal relationship between the occurrence of the above-mentioned events and the structural characteristics of the stock network was investigated.In particular,this article selects the textile and apparel submarket for in-depth stability research;Finally,much attention was paid to the analysis of the robustness and fragility of the stock correlation network.The study found that: China’s A-share stock market generally maintains good robustness in the face of random disturbances;While suffering from deliberate attacks(such as major business mistakes of key enterprises or individuals,and attacks by malicious capital,etc.)the stock network behaves obviously vulnerability.On one hand,the ‘non-homogeneity’ of China’s A-share stock market network revealed in this article further reflects the remarkable characteristics of the unbalanced development of industries of all stripes in China.On the other hand,our research also reveals that the economic supporting effects brought about by relative policy have enhanced the stability of the association network.Our conclusion can provide some references for enterprises,government,and all investors involved in the Chinese stock market.