Probabilistic Wind Risk Assessment Induced by Hurricanes on Economically Vulnerable Households in Mexico

Abstract

The economically vulnerable households in Mexico are used as an illustrative case study to present wind risk assessment induced by hurricanes, in which the risk parameters are assessed in an event-based probabilistic framework using a set of hazardous events integrated probabilistically, including all uncertainties related to hazard, vulnerability, and risk assessment. The hazard is defined as a stochastic set of events that describes the spatial distribution, the annual frequency, and the randomness of the hazard intensity. The structural vulnerability is defined, for specific dwelling types, in terms of the expected damage and its corresponding standard deviation. The risk is expressed in four different economic terms: (1) expected loss for a single event, (2) average annual loss as a percentage of the reconstruction cost of the exposed asset, (3) pure premium, and (4) loss exceedance curve. It is observed that the expected annual loss for this type of dwelling in Mexico is approximately USD 39.6 million, and the probable maximum loss associated with a return period of 1,5 years is approximately USD 1,181 million. It is concluded that these quantitative metrics are of particular importance, especially to define risk transfer or retaining schemes, such as that employed by the Mexican Insurance Commission, which uses a 1,5-year return period to fix the solvency of insurance companies in Mexico, the Natural Disaster Fund, FONDEN, or a Catastrophe bond (CAT bond), used by the Mexican government.