Bristol West, a subsidiary of Farmers, is introducing a new private passenger auto program in Maryland, effective June 24, 2026. The rollout reflects a structured expansion into the state with a pricing framework built around granular risk segmentation and modern actuarial techniques.

At the core of the program is an “average driver factor,” which aggregates risk across all drivers within a household. When the number of drivers exceeds the number of vehicles, the model prioritizes higher-risk individuals through a weighted ranking approach tied to key coverages, including bodily injury, collision, and comprehensive. The result is a pricing structure that leans toward the highest-risk exposure within a policy rather than a simple average.

The program also introduces underwriting tiers based on prior insurance limits and lapse history, effectively grouping risks according to behavioral and coverage signals. This is complemented by detailed driver class factors that vary by age, driving record, and other attributes, reinforcing a highly segmented pricing architecture.

From a modeling standpoint, Bristol West relies on Generalized Linear Models (GLMs), supported by a 70/30 train-test validation split. While GLMs are standard in actuarial pricing, the validation approach mirrors machine learning practices, signaling a continued convergence between traditional actuarial methods and data science workflows.

Notably, the model is built on nationwide data, excluding California. This means Maryland pricing is partially informed by loss experience from other states—an approach that may draw scrutiny from a regulatory perspective, particularly around rate adequacy and fairness at the state level.

Distribution insights

Bristol West distributes through both exclusive agents (EA) and independent agents (IA), with clear performance differences between the channels:

• Independent agents (IA) show higher loss frequency and severity:
  • Bodily injury frequency: 2.45% (IA) vs. 1.80% (EA)
  • Property damage frequency: 8.10% (IA) vs. 6.56% (EA)
  • Bodily injury pure premium: 230 (IA) vs. 192 (EA)
  • Property damage pure premium: 267 (IA) vs. 225 (EA)

These results indicate that IA-sourced business is both more loss-prone and more expensive on a per-policy basis.

• Selected pricing response:
  • BI relativity: 1.0555 (~+5.5%)
  • PD relativity: 1.0393 (~+4%)
  • EA remains the baseline at 1.000

The selected relativities reflect a moderated response to the indicated differences, suggesting Bristol West is balancing actuarial signal with competitive positioning.

Bottom Line: Bristol West is entering Maryland with a tightly engineered pricing model that emphasizes household-level risk, behavioral segmentation, and channel differentiation—supported by data science-style validation and multi-state loss experience.