Loss Cost Pricing

Learn to price catastrophe risk using loss cost analysis

Overview

A loss cost analysis calculates the loss cost of a location for a catastrophe event based on an RMS peril model. Loss cost represents the sum of the projected annualized losses for a location. Using loss cost projections, you can determine whether the premium on a location is sufficient to cover a risk.

Location Intelligence APIs enable loss cost analysis using seven different peril models: flood, windstorm, earthquake, terrorism, convective storm, winterstorm, and cyclone.

Each peril model characterizes the vulnerability of a geolocation to a specific peril based on Detailed Loss Module (DLM) analysis. Leveraging pre-compiled catastrophe simulations, the platform returns an annualized loss for a location that is based on the projected losses from every simulated catastrophe event multiplied by the annual rate (probability) of the hazards causing that loss.

The sum of the annualized losses at each location produces the lost cost for that location, which is normalized by the exposure to yield the annualized loss rate (ALR). The model returns ALR calculations for buildings, contents, and business interruption coverages.

Step 1: Geocode the location

A location must be "geocoded" before it can be enriched or analyzed using the Location Intelligence APIs. Geocoding is the process of enriching a location exposure with geolocation data that precisely identifies the global coordinates of a site. To learn more, see Geocode Addresses.

Step 2: Loss cost the geocoded location

The POST eu_fl_loss_cost/latest request performs lost cost pricing analysis on a geocoded address using the latest version of the Europe Flood Loss Cost data product.

All parameters are specified in the request body.

curl --location --request POST 'path.to.service/eu_fl_loss_cost/latest' \
--header 'Content-Type: application/json' \
--data-raw '{
  "layer": "EU flood loss cost",
  "peril": "FL",
  "version": "1.0",
  "tenant": "rms",
  "location": {
    "address": {
      "countryCode": "FR",
      "countryRmsCode": "FR",
      "countryScheme": "ISO2A",
      "rmsGeocodingResolutionCode": 19,
      "rmsGeoModelResolutionCode": 1,
      "latitude": 42.69703826,
      "longitude": 2.897703698,
      "postalCodeGeoId": 12726668,
      "ugid": ""
    },
    "characteristics": {
      "construction": "RMS0",
      "occupancy": "ATC1",
      "numOfStories": 3,
      "yearBuilt": 2015,
      "basement": "YES"
    },
    "coverageValues" : {
      "buildingValue": 640000,
      "contentsValue": 100000,
      "businessInterruptionValue": 10000
    }
  },
  "layerOptions" : {
    "deductibleType": 1,
    "deductibleAmount": 0.01
  }
}'

The request body must define the location and layerOptions objects. Within the location object, define the address object attributes. The characteristics and coverageValues objects are optional.

The layerOptions object enables you to define loss cost pricing layer attributes. A successful request returns a 200 status code a lost cost analysis using the specified model.

{
  "buildingAlr": 0.000863401691354315,
  "contentsAlr": 0.000358529682823589,
  "businessInterruptionAlr": 0.000559604863156912,
  "grossLoss": 750.16,
  "groundUpLoss": 901.43
}

A loss cost analysis returns the following statistics:

  • The buildingAlr attribute is a number that represents the annualized loss rate for buildings covered at the location.
  • The contentsAlr attribute is a number that represents the annualized loss rate for the personal property covered at the location.
  • The businessInterruptionAlr attribute is a number that represents the annualized loss rate for a business covered at the location.
  • The grossLoss attribute is a number that represents the loss to the insurer, accounting for the application of all insurance terms but without consideration for any reinsurance recoveries.
  • The groundUpLoss attribute is a number that represents the total loss to the exposure, independent of any insurance or reinsurance terms format.

Did this page help you?