The University of Oxford and Amlin, one of the largest independent insurers operating in Lloyd’s of London, will jointly study the reliability of existing models that predict natural disasters.
Improved forecasts could reduce insurance companies' losses. The study will last for three years, informs the Financial Times (FT).
Since the advent of catastrophic models in the 1990s, they have played an increasingly important role in determining the amount of losses an insurance company will have to bear to cover damage from natural disasters.
Forecasting models include scientific disciplines such as climatology and seismology. Assessing the value of properties at risk from natural disasters such as hurricanes and earthquakes has forced insurers to spend millions a year on such surveys.
The huge capital investment in making such models does not guarantee an accurate risk assessment, as was the case with the floods in Thailand 3 years ago, which costs more than insurers' preliminary estimates, partly because they were not accurately prepared.