The ideas presented in this paper are a good start, as are discussions around agent-based models or econophysics. Building a useful model of systemic risk is not trivial, but its not impossible ...