Part 16: Presenting results to the pricing committee
Part 16: Presenting results to the pricing committee¶
This is the part most technical courses skip. The optimiser produces a solution; getting a pricing committee to accept it requires a different skill.
The structure of the conversation¶
The pricing committee will ask five questions, approximately in this order:
"Can we just see the headline number?"
Yes. The factor adjustments are 3.6-3.7% on each shared factor, zero on the tenure discount. The combined effect is a 15.1% premium increase for the average customer. The expected LR at new rates is 72.0%, and we expect to retain 97.3% of current volume.
"What if lapses are worse than the model predicts?"
Show the stochastic result. At 90% confidence (planning to a distribution, not just the mean), the factor adjustments would need to be 4.3-4.5% rather than 3.6-3.7%. The committee can decide whether to price to the expected value or to the 90th percentile.
"What if we wanted to do less rate?"
Show the frontier table. Relaxing the LR target from 72% to 73% reduces the factor adjustments to approximately 2.8% and improves expected volume retention to 97.9%. The committee can choose any point on the frontier; you are presenting the full trade-off rather than a single recommendation.
"Why can we not take more rate on [specific factor]?"
This is about the factor movement caps. If the underwriting director approved [0.90, 1.15] movement caps, the optimiser cannot exceed them. If the committee wants to take, say, a 20% movement on the vehicle factor, they need to approve a wider mandate and re-run the optimiser. This conversation is healthy: it surfaces the implicit constraint that previously existed only in the underwriting director's judgment.
"Are we treating any customers unfairly?"
Show the cross-subsidy analysis. The percentage change is uniform across all customer segments. Young drivers see a larger absolute increase (£91 vs £46) because their base premium is higher, not because the rate action targets them disproportionately. This is the Consumer Duty evidence.
What to put on the slide¶
The one-slide summary for the pricing committee:
| Item | Value |
|---|---|
| LR target | 72.0% |
| Expected LR at new rates | 72.0% |
| Expected volume retention | 97.3% |
| Factor adjustments (shared) | +3.6% to +3.7% |
| Tenure discount adjustment | 0.0% (ENBP constraint) |
| Customer impact (mean) | +15.1% / +£72 per year |
| ENBP compliance | Verified per-policy (0 violations) |
| Solver converged | Yes |
Below the table, the efficient frontier chart. Below that, the constraint binding summary: "LR constraint is binding at the optimum. Volume constraint is not binding (expected volume 97.3% vs 97.0% floor). ENBP constraint is binding (tenure discount cannot move above 1.0). Factor bounds are not binding."