Study On Stock Price Volatility Based On Fund Networks

There are extensive network connections in financial markets.In recent years,research on investor networks has shown that individuals or institutions are not isolated in the capital market.They not only participate in the market through the network,but also communicate information through the Internet and learn from each other.In the overall structure of the social investors,institutional investors are in an important position.Whether from the regulatory point of view or from the market point of view,the existence of institutional investors,to reduce the market irrational fluctuations,to guide the rational return is of great significance.In this paper,the stocks which constitute 5%or more of the portfolio of at least one mutual fund are included in the return analysis.And hypothesize that two fund managers allocating 5%or more of their portfolio to the same stock are connected to each other.The mutual fund data in this paper is from January 2012 to December 2016.On this basis,this paper describes two important indicators of network structure:network density and network concentration,in which the network concentration is expressed by the coefficient of variation and standard deviation of the number of network connections.Subsequently,this article uses these network indicators to examine three hypotheses:1.after controlling other stock characteristics known to affect price delay,the price delay will increase with decreasing network density or decreasing speed of information diffusion.2.The information-driven volatility component will increase with increase in network centralization.3.The lagged overall network concentration has a causal relationship with the average idiosyncratic volatility in the time series.The empirical results show that the lower the density of the network,the more obvious the price delay effect.Also,this paper finds that centralized information networks lead to higher volatility of individual stocks in cross-section and also explain the variation in average stock idiosyncratic volatility over time.This provides evidence supporting the predictions of recent theoretical models that study the effect of information networks on stock prices.Moreover,this conclusion will help to better understand and grasp the behavioral characteristics of institutional investors,and further understand the diffusion effect of asset prices.