Glossary of Terms
The following definitions establish the canonical meaning of key terms as used throughout the Market Terminal and Risk Data Reporting Standards.
Property Policy
An insurance policy whose primary purpose is to insure physical real property and/or tangible property against direct physical loss or damage.
Liability Policy
An insurance policy whose primary purpose is to cover the insured’s legal liability to third parties for bodily injury, property damage, or personal and advertising injury.
Package Policy
A single insurance policy that combines multiple distinct coverage parts, typically property and liability, under one policy form.
Rate on Line (RoL)
A normalized measure of insurance pricing calculated as the ratio of premium to coverage limit. RoL expresses the amount of premium paid per unit of insured capacity and is typically reported as a rate per $1,000 of coverage.
Formula: RoL = (Premium / Coverage Limit) × 1,000
Coverage Percentage
A metric used in pricing comparables to represent the percentage of policies within a defined population that evidence or include coverage for a selected peril.
Comp Group / Peer Group
A set of policies whose data falls within defined filter criteria, used for like-for-like pricing comparison.
Pricing Factors
Quantified explanations of how property characteristics, coverage options, and market factors influence Rate on Line (RoL). Factors are computed using machine learning attribution methods to identify which variables drive pricing variation within a peer group.
Importance Score
A normalized metric ranging from 0 to 1 that represents the proportion of RoL variation attributable to a specific factor. Importance scores sum to 1 across all factors, allowing direct comparison of factor influence.
SHAP (SHapley Additive exPlanations)
A machine learning interpretability method based on cooperative game theory (Shapley values). SHAP assigns each feature a contribution value that explains its impact on model predictions. Used in Pricing Factors to attribute RoL variation to specific characteristics.
Value Breakdown
A detailed view of how a factor’s impact on RoL varies across different values. For categorical factors (like state), shows impact per category. For numeric factors (like unit count), shows impact by quartile. For boolean factors (like coverage flags), shows impact for true/false values.