Daphne Douglas / Chief Data Officer
Perspectives from a life in leadership.
The latest essay
Canadian insurers are licensing wildfire models faster than they are correcting where their policyholders' homes actually stand. A model run on postal-code locations produces confident numbers about the wrong houses. Exposure data should come first, and the model's limits should travel with every price it sets.
Read the essayCanadian insurers are licensing wildfire models faster than they are correcting where their policyholders' homes actually stand. A model run on postal-code locations produces confident numbers about the wrong houses. Exposure data should come first, and the model's limits should travel with every price it sets.
Quebec's private-sector privacy law requires a business to tell people when a decision about them rests solely on automated processing. Haverlock now gives an equivalent notice in every province where it writes. It costs money and removed some rating variables, and it made our models and our data better.
Insurers put their pricing and reserving models through independent challenge, yet the chain of systems and mapping tables that turns policy and claims records into a capital return is usually treated as a reporting routine. Its judgements deserve owners, change control and validation.