Personal Lines Pricing

Personal lines pricing is becoming an arms race in analytical sophistication -- and the gap between leaders and followers is widening.

The carriers at the front of this race have deployed generalized linear models, gradient boosted trees, and increasingly neural network architectures to identify pricing segments that simpler models miss. The result is that they can offer competitive prices on the best risks in every segment while correctly pricing or declining the worst.

For carriers without the same analytical depth, the market dynamic is adverse: they win the risks the leaders priced up and lose the risks the leaders priced down. Over time, that selection pattern degrades the book quality of less sophisticated carriers even if their overall rate levels appear adequate.

The response for mid-market carriers is not necessarily to build competing data science capability internally. Vendor partnerships, InsurTech collaborations, and bureau plan enhancements can provide access to sophisticated rating variables without fully internalizing the capability.