I write you as director of finance, insurance and trade policy at the R Street Institute, a nonprofit, nonpartisan public policy research organization (“think tank”). We appreciate the opportunity afforded by the Casualty Actuarial and Statistical Task Force to offer input on the revised draft of the Regulatory Review of Predictive Models White Paper.
We commend the Task Force for its work and are heartened by several updates in the revised paper. For example, in the section on relevance of variables and relationship to risk of loss, we welcome the revised paper’s substitution that a rate-filing narrative ought to explain a variable’s “rational” relationship to cost, rather than original “logical and intuitive.” Actuarially credible and statistically significant variables may, in fact, prove to be counter-intuitive.
However, we remain concerned that the paper exceeds the scope of its stated purpose and that it could be interpreted to recommend more stringent reviews of existing models that have served consumers well for decades. In response to earlier comments we filed, the Ad Hoc Team asserted “the fact that predictive models have been reviewed in depth by regulators for many years under the current confidentiality provisions…is prima facie evidence that the negative impacts that R Street is alleging will not arise.” However, the team elsewhere notes that it “would be unfortunate if a model is withdrawn from the market because the modeler is unwilling to share information with regulators.”
Our position is that it would be not just unfortunate, but disruptive, and threatens to reverse progress made over three decades toward more competitive insurance markets that better serve consumers. We will continue to recommend that regulators exercise to avoid such market disruption.