Key points are not available for this paper at this time.
Abstract. Insurance is an important element of flood risk management, providing financial compensation after disastrous losses. In a competitive market, insurers need to base their premiums on the most accurate risk estimation. To this end, (recent) historic loss data are used. However, climate variability can substantially affect flood risk, and anticipating such variations could provide a competitive gain. For instance, for a year with higher flood probabilities, the insurer might raise premiums to hedge against the increased risk or communicate the increased risk to policyholders, encouraging risk-reduction measures. In this explorative study, we investigate how seasonal flood forecasts could be used to adapt flood insurance premiums on an annual basis. In an application for Germany, we apply a forecasting method that predicts winter flood probability distributions conditioned on the catchment wetness in the season ahead. The deviation from the long term is used to calculate deviations in expected annual damage, which serve as input into an insurance model to compute deviations in household insurance premiums for the upcoming year. Our study suggests that the temporal variations in flood probabilities are substantial, leading to significant variations in flood risk and premiums. As our models are based on a range of assumptions and as the skill of seasonal flood forecasts is still limited, particularly in central Europe, our study is seen as the first demonstration of how seasonal forecasting could be combined with risk and insurance models to inform the (re-)insurance sector about upcoming changes in risk.
Dũng et al. (Fri,) studied this question.