A Hurricane Risk Index for the U.S

Preparing adequately for a potential hurricane landfall on the U.S coast several days ahead can be a challenging task, given the uncertainty in the future path and strength of the storm. Homeowners, businesses, as well as insurers all brace to face the potential damage and loss from such storms, including direct physical property damage and indirect forms of loss such as business interruption. Therefore, there is a need for products that allow for better planning strategies that would potentially mitigate losses or allow for improved emergency response. The ClimateCast® U.S. Hurricane Risk Index developed by AIR Worldwide can capture, several days in advance, the range of potential loss to the insurance industry as a whole, or for a particular portfolio of properties. The index relies on the expertise of the AIR U.S. Hurricane catastrophe model and an ensemble of operational forecast models.

The AIR U.S Hurricane model is a catastrophe model with three main components: hazard, engineering and loss estimation. The hazard component computes storm associated wind speeds which are subsequently used by the engineering component to estimate the damage to property. The loss component assesses the insured losses by applying policy conditions to the total damage estimates. By means of different stochastic modeling techniques, the ensemble of operational forecast models is augmented to produce up to 500 unique storms with different tracks and intensities. The AIR U.S Hurricane model uses these scenarios to develop a more complete spectrum of damage and loss for the industry. The estimated losses are then indexed against the AIR U.S. Hurricane model's loss probability curve to capture the potential risk.

The ClimateCast® U.S. Hurricane Risk Index provides decision makers throughout the insurance management chain a consistent and transparent means of evaluating the loss potential to U.S. onshore properties from active Atlantic storms. The performance of the risk index, which is provided four times daily in real-time, has been verified for recent active hurricane seasons. The present paper discusses the index calculation, performance, and real-time applications.