Why "Traffic Light" Risk Scores Are Flawed

[00:00:01]There seems to be a trend in the industry of relying quite heavily on standardized dashboards for climate risk in particular, which often it boils complex perils down to a simple and perhaps generic score, maybe a one to five score.

[00:00:17]Why is this oversimplification fundamentally flawed? And how does it break down when an underwriter tries to evaluate the same peril in maybe vastly different climates?

[00:00:28]>> Any deterministic prediction, like a risk score or like saying, "No, it's going to be this amount of rain or it's going to be this temperature in 5 years, in 10 years." is wrong by default.

[00:00:40]Yeah?

[00:00:41]So, when you use a traffic light, you're using a qualitative deterministic prediction, so it's wrong by default.

[00:00:49]Indeed, for the AG, for reporting, it was enough, but for underwriting risk, it's not enough. You need complexity.

[00:00:56]You need to embrace the complexity. You need indeed climate data, you need climate predictions, like the ones that we do, but we are not just delivering a number or we are not just delivering, let's say, a a qualitative view of the future, like orange, red, or 1 2 3 4 5.

[00:01:11]We are delivering a probabilistic approach. We are delivering uncertainty.

[00:01:16]We are focusing on the transparency. We are sharing the models that we are using. Let's say, everything. So, this complexity is what allows underwriters or is what allows any business at some point to really measure and to really have a an adequate view of the impacts of climate change, of the climate.