Emergence of Scale-Free Networks in Markets

Abstract

Financial markets are complex systems; all the information scattered around the market is fairly and dynamically reflected in the current prices. It is difficult to understand the dynamics of markets merely by traditional analyzing methods. We propose here a new concept inspired by complex networks to study the trading behavior and the dynamics of markets. A web-based platform for prediction market which trades the political futures contracts is built to monitor the trading behavior among the human players. Two experiments are conducted on this platform in parallel for 30 days. From the accumulated transaction data, we reconstruct the so-called cash-flow networks. By examining the degree distributions of these networks, we observe that the network structure is scale-free with a power-law exponent of 1.15, which means that there must be a non-trivial mechanism governing the network growth in our markets. Through carrying out a post-simulation modelling by a continuous double auction market with "zero-intelligence" traders, we demonstrate that such a simple model is capable of generating scale-free networks. We thus suggest that the scale-free nature of the cash-flow networks should rely on the institutional design and the structure of markets rather than on the traders' strategies.

Keywords: Prediction market, complex networks, political futures