Actuarial Risk Approach to the Insurance Protection Gap

Catastrophic events stemming from natural disasters are a major factor of systemic risk. Yet, only a quarter of the losses related to extreme weather events observed accross Europe during the period from 1980-2021 were actually insured, a trend that still persists, and creates a material insurance protection gap in the European Economic Area [EIOPA-22/507, EIOPA-BoS-22/505]. Climate projections indicate that the frequency of climate-related extreme events (e.g. heat waves, droughts, flood, storm, and wind speeds) is likely to increase in the future in all main European regions. [EEA No 1/2017]. As a means to describe the insurance protection gap, EIOPA proposes to adopt a science-based approach to protection gap management and decision-making, explicitly dependent on risk drivers and regions at-risk. Under the EIOPA science-based approach, the insurance protection gap is described by means of a current and a historical view. The current protection gap is described by means of a composite measure of insurance coverage and accepted risk. The historical protection gap is based on historical data on economic and insured losses. We develop in this presentation a scientific-based approach to the management of Nat Cat risk exposures within an actuarial risk theoretical framework, based on estimations of the exceedance probability (EP) of Nat Cat losses, defined as the probability that a loss random variable exceeds a certain amount during a predefined future period of time. We assume that EP curves can be modelled by some probability trend, e.g. Exponential, Lognormal, or Pareto-type. Estimations of the parameters of the stochastic representation of Nat Cat losses can be derived from either statistical fittings or expert knowledge. Theoretical estimations can be thus obtained of the fair-actuarial price of Nat Cat risk, and the optimal insurance retention for given levels of insurance protection and exposure to risk. The consequences of shocks of Nat Cat risk, as measured by shocks to the parameters of the underlying stochastic representation, can then be described by means of the induced variations in the theoretical price of Nat Cat insurance and optimal insurance retention. Estimations of the levels and variations of the Nat Cat risk parameters can be based on the climate projections provided in broad European studies [EEA No 1/2017]. Publicly available catastrophe data such as the EM-DAT Public disaster database can be used to create a benchmark of actuarial fair prices and optimal insurance retentions, corresponding to actual historical trends (EMDAT Public DB). We thus propose a complementary view of the insurance protection gap based on the actuarial characterisation of Nat Cat losses, and Nat Cat risk insurance. We propose two measures to assess the business impact of variations in the stochastic representation of Nat Cat losses, namely: deduced financial costs, as measured by the mathematical expectation of excess-losses, and deduced capital requirements, as measured by the inverse probability distribution evaluated for a given confidence level. These criteria can be related to relevant Solvency II criteria.