It's a dog eat dog world: predator prey models of the financial markets.

Abstract

In this thesis, a model of interacting agents for financial markets was discussed. In the mean field limit of the model, the system of agents resembles a predator prey relationship described by a pair of coupled Volterra equations. Simulations done for 1000 agents revealed three different phases when the parameters of the system were changed. Analysis showed that the transitions between these phases are first order in nature and regions of coexistence between the phases were found on the phase diagram. Interestingly, part of these regions exhibits coexistence of all three phases. Extrapolation of simulation results for different number of agents suggest that all the observed features persist in infinite system size limit.