Catastrophe loss modeling and its application for enterprise risk management and land-use planning

The global increase in the cost of natural disasters can be explained by the increasing concentration of assets and wealth in hazard-prone areas, a situation unlikely to change for a very long time. This state of affairs is aggravated by poor land-use planning practices, with the 2009 Victorian bushfire and 2011 Brisbane flood events two recent Australian examples of such planning failures. This paper will argue the benefits of both stochastic and scenario loss modeling to help businesses and governments prepare for future large events. The use of such modeling is already well developed in the insurance sector, where, for example, Risk Frontiers' Multi-hazard platform is used to simulate the impact of hundreds of thousands of plausible natural hazard events on company portfolios, and estimate the annual probability of experiencing losses in excess of cer tain dollarthresholds. Perils include earthquake, tropical cyclone, hailstorm, flood and bushfire events. Model outputs are used to inform negotiations in regard to the purchase of reinsurance and, in conjunction with geo-spatial tools, risk-adjust premiums. Outside of the insurance sector the use of such modeling remains primitive although many businesses would benefit from its use and by exploring a number of large event scenarios could quickly ascer tain enterprise-specific vulnerabilities, including business interruption. In the government sector, the models could be used to under take cost-benefit analyses, prioritize mitigation measures, make risk-informed decisions in respect to land-use planning, whilst emergency managers could also them to plan and stress test evacuation operations.