Conclusion

We found similar behavior between all three of our models. This is both expected and a good sign, as systematic risk does exist in real financial systems, thus any model we create to represent these system should contain it as well. For almost all of the parameters we found a tradeoff between the mean number of defaults and the likelihood and severity of a systemic event. We came up with our own assessments of these trade offs and possible recommendations, but without a quantifiable cost function, either created from some underlying theory or empirically derived, exact parameter specifications or system structures were not found. We ultimately concluded that the USBC model was superior to the other two, at least in their current state. We concluded this because the UCSB model clearly displayed systemic failure while also allowing the underlying graphs to be manipulated to better model the real world.