Predictive modeling -- how much is too much?

Speaker(s): Robert Brown (University of Waterloo), Mary Hosford (Commonwealth of Massachusetts)

Is there such a thing as too much predictive modeling? Might we reach a point at which predictive models become so good that certain individuals or businesses become uninsurable? If so, does that point undermine the whole risk-sharing concept of insurance? Or should we continue to refine our predictive models using all available data and technology? And how about price optimization, charging higher prices to people who are not as sensitive to price and price change -- how fair is that? Please join us for a lively debate on these topics which are currently facing actuaries in many markets and different types of insurance.
This session will be in the format of a debate among actuaries taking opposing viewpoints on the issues.