Last week Risk Management Solutions (RMS) announced the release of a new version of their Europe Windstorm Model. The new version increases the number of countries covered, enhances the dataset and offers a much improved view of European windstorm risk according to RMS.
They held a webinar event this week which we attended, in which they explained the impact of the model on the expected loss of European windstorm catastrophe bonds. It revealed that unlike their updated U.S. hurricane risk model which has significantly increased expected losses for many cat bond transactions, the enhanced Europe Windstorm Model will not have the same level of impact.
RMS say that the Europe Windstorm Model has been ‘rebuilt from the ground up’. It contains a new and improved hazard model with significant additional wind speed data and a new approach to clustering. Overall, they say, it offers much stronger and better view of the risk of damaging windstorm losses across Europe.
Overall, the total change in expected loss is slightly down due to a decline in hazard which offsets a rise in vulnerability up to a 10 year return period. The reason for this is the inclusion of significantly more high frequency events in the wind field model. The one exception to the rule is for certain parametric deals, in particular aggregate deals which see an increase in expected loss. RMS also explained that clustering causes larger increases on aggregate instruments and also changes the regional distribution of risk.
RMS tested their new model using benchmark cat bonds which they have created for the purpose of checking the impacts of model changes on the cat bond market. The benchmark cat bonds for European windstorms consist of standard exposure and parametric benchmark cat bonds. They then run these benchmarks through the model in different configurations on per-occurrence and aggregate terms to find out the impacts versus their previous model versions.
The results of the benchmarking can be found below. As you’ll see the trend is for a decrease in expected loss percentage using V11.0 of the model except for a station parametric bond which increases. Issuers of European windstorm catastrophe bonds will be pleased with the results as they now have a much improved risk model available to use in new transactions which gives a lower expected loss. The market as a whole will be pleased that this new model won’t cause the same uncertainty in the cat bond world as resulted from the updated U.S. hurricane model.
|Benchmark bond||V10.0 Europe Windstorm Expected Loss – %||V11.0 Europe Windstorm Expected Loss – %|
|Fixed Exposure Bond – Occurrence||1.71||1.36|
|Fixed Exposure Bond – Aggregate||2.0||1.89|
|Interpolated Parametric Bond – Occurrence||2.97||1.16|
|Interpolated Parametric Bond – Aggregate||3.44||2.30|
|Station Parametric Bond – Occurrence||2.04||1.41|
|Station Parametric Bond – Aggregate||2.47||3.13|