Volatility persistence in stock market

•We propose an empirical methodology to construct a Granger-causality financial network.•We find asset volatility increases as the financial network density increases.•The financial network density can be a probe to the systemic risk.

By using the unique dataset that consists of all brokers' daily trading information, I propose an empirical methodology to construct a Granger-causality financial network to study the relationship between dynamics of volatility and information diffusion in a stock market. The financial network I proposed not only provides a new way to describe mutual interconnectedness of brokers in the market, but the empirical results also show the financial network density is positively correlated to the realized volatility of the market.