Price Volatility on Investor’s Social Network

Based on an Ising model proposed by Harras and Sornette, we established an artificial stock market model to describe the interactions among diverse agents. We regard these participants as network nodes and link them with their correlation. Then, we analyze the financial market based on the social network of market participants. We take the random network, scale-free network, and small-world network into consideration, and then build the stock market evolution model according to the characteristics of the investors’ trading behavior under the different network systems. This allows us to macroscopically study the effects of herd behavior on the rate of stock return and price volatility under different network structures. The numerical simulation results show that herd behavior will lead to excessive market volatility. Specifically, the greater the degree of investor’s trust in neighbors and their exchange, the greater the volatility of stock price will be. With different network synchronization capabilities, price fluctuations based on the small-world network are larger than those based on the regular network. Similarly, price fluctuations based on the random network are larger than those based on the small-world network. On the other hand, price fluctuations based on both the random network and the small-world network firstly increase and then decrease with the increase of the average node degree. All of these results illustrate the network topology that has an important impact on the stock market’s price behavior.