500 submissions scored · prepared by Plumbline Risk · July 29, 2026 · decision-support (assist) output
0.713 AUC (score vs. loss) · Gini 0.43 | 14.8% Book loss frequency | 46% of losses sit in the RED band (23% of book) | 113% RED-band loss ratio vs. 62% book average |
| Band | Risks | % of book | Loss rate | % of all losses | Loss ratio | |
|---|---|---|---|---|---|---|
| GREEN | 205 | 41% | 5% | 15% | 23% | |
| AMBER | 179 | 36% | 16% | 39% | 65% | |
| RED | 116 | 23% | 29% | 46% | 113% |
The RED band is 23% of the book and carries 46% of the losses, running at a 113% loss ratio against a book average of 62%.
It also carries 26% of the premium ($5,289,548). That is the part a headline number usually hides: removing the band removes its premium too, so blanket declinature is rarely the right answer and rarely available in practice.
A more useful read: bringing the RED band alone to a 65% target loss ratio implies roughly a 74% rate increase on those risks — achievable through pricing, deductibles, sub-limits on the flagged peril, or selective declinature. The value of the score is knowing which risks warrant that attention before binding, not declining a fifth of your submissions.
For reference, the theoretical upper bound — removing the RED band entirely — takes the book from 62% to 44%. Treat that as a ceiling on the opportunity, not a recommendation.
| Decile | Risks | Loss rate | Lift vs. book | |
|---|---|---|---|---|
| 1 | 50 | 44% | 3.0x | |
| 2 | 50 | 20% | 1.4x | |
| 3 | 50 | 12% | 0.8x | |
| 4 | 50 | 24% | 1.6x | |
| 5 | 50 | 20% | 1.4x | |
| 6 | 50 | 8% | 0.5x | |
| 7 | 50 | 2% | 0.1x | |
| 8 | 50 | 6% | 0.4x | |
| 9 | 50 | 6% | 0.4x | |
| 10 | 50 | 6% | 0.4x |
Risks carrying at least one high-severity flag had a 18% loss rate vs. 11% for unflagged risks (53% of the book was flagged).
| Flag | Risks | % of book | Loss rate | Lift vs. book |
|---|---|---|---|---|
| CAT_WILDFIRE | 62 | 12% | 31% | 2.1x |
| UNPROTECTED_WILDFIRE | 85 | 17% | 27% | 1.8x |
| CAT_HAIL | 62 | 12% | 26% | 1.7x |
| AGED_ROOF | 156 | 31% | 21% | 1.4x |
| FIRE_LOAD_MISMATCH | 31 | 6% | 19% | 1.3x |
| CAT_FLOOD | 31 | 6% | 16% | 1.1x |
| UNDER_INSURANCE | 159 | 32% | 14% | 0.9x |
| CAT_EARTHQUAKE | 34 | 7% | 12% | 0.8x |
Lift above 1.0x means the flag predicts loss frequency — those risks lost more often than the book average. Note that UNDER_INSURANCE is a premium-adequacy flag, not a frequency predictor: it identifies risks where premium is charged on too small a base, so a lift near 1.0x is expected and does not make it less valuable. Other flags sitting below 1.0x should be retuned on your loss data.
159 risks under-insured (<80% of rebuild) | 32% of the book | $251.2M total insured-value gap identified |
Each under-insured risk means premium is charged on too small a base — a silent margin leak the engine flags at intake, before it reaches the rating engine.
Illustrative run on synthetic data with a built-in offline provider. In production, enrichment comes from Opta/iClarify, Ecopia and provincial hazard layers, and weights are fit to the MGA's own loss experience. Output is decision-support (assist), not a bind decision.