Developing a New Underwriting Strategy Model for Natural Catastrophe InsuranceEric Figueroa According to research by Munich Re, natural catastrophe losses (both insured and uninsured) have been increasing exponentially since about the 1990s. The problem facing insurers and reinsurers of natural catastrophe insurance is the difficulty they have in quantifying the risk of event occurrence and setting an appropriate price for a premium. Natural catastrophes do not follow characteristic probability curves but rather open up insurers to right-tail risk. Even though catastrophe insurance is essential to rebuild after a large devastating event like an earthquake, the price is often prohibitively high, especially in emerging markets, precisely because insurers must add a risk premium to compensate for the grave levels of uncertainty involved. CFAR-m is a new algorithmic method based on the artificial intelligence of a neural network that can analyze large sets of data objectively, without external manipulation. What’s more, it can also provide the contribution of each variable to the final risk index to allow for subsequent simulations. CFAR-m has produced a case study showing it can outperform traditional statistical techniques like PCA in predicting which country is most at risk and which type of disaster will be the most likely to occur there. This case study has been linked to an overarching Underwriting Strategy model built using Strategy Foresight’s General Morphological Analysis. General Morphological Analysis, or GMA, is a form of visual extended typology analysis that allows small groups, guided by impartial facilitators, to identify and link all the parameters of a problem complex in a multi-dimensional matrix. This matrix can then be used to produce “what if” scenarios and provide decision support. The Strategy Underwriting model already produced uses a combination of these two organizations’ methodologies (both quantitative and non-quantifiable modelling) and hopes to provide reinsurers with an expert tool for developing catastrophe insurance underwriting strategy at the macro level. This study aims to test this preliminary model in the field to determine its real value in the field of risk management and the insurance industry as a whole. |
