Skip to Main Content
Menu
Contact Us Search
CA Department of Insurance
CA Department of Insurance
CA Department of Insurance

Wildfire Risk Report Format and Methodology

(A Fire and Wildfire Exposure Risk Manual - Updated June 23, 2023)

Format of the Report

The 2020 report, which includes 2018 and 2019 experience years, summarized the data for all companies that provided both a fire risk classification and PPC score. For the 2022 report, which includes 2020 and 2021 experience years, there are two components of the report. The first summarizes the data for all companies that provided a fire risk classification. The second version summarizes the data for all companies that provided a PPC score.

There are several worksheets for each of the reports. The first worksheet – the Summary worksheet – provides a statewide summary of the data, broken out by the six different policy forms. The second worksheet – the Loss Data (All Companies) worksheet – transposes the loss data on the “Summary” worksheet to increase the ease of comparing incurred losses and claim counts across policy forms and calendar years.

The remaining worksheets are then specific to the companies that provided a fire risk classification or PPC score (depending on the report). The third worksheet is identical to the second worksheet but only includes the subset of companies that provided a fire risk classification or PPC score. The following 12 worksheets provide ZIP code level detail for each of the six policy forms in each of the two experience years, and the final worksheet describes the methodology about how the data was collected and summarized.

All of these worksheets are explained in detail in this document.

The following definitions are utilized in the report:

  1. Policy Type: The report contains data from 6 different policy types.
    • Homeowners’ (HO) type policies, which cover 1-4 dwelling-units in which the owner lives in one or more units (e.g., ISO HO-1, 2, 3, 5, 8, or equivalent).
    • Renter's/Tenant's (RT) type policies, which cover the personal property of persons renting or leasing apartments, condominium-units, or dwelling-units (e.g., ISO HO-4 or equivalent).
    • Condominium Unit Owner (CO) type policies, which covers a specific unit within a condominium building or co-op (e.g., ISO HO-6, or equivalent).
    • Mobile Home (MH) type policies, which covers the structure and contents of a manufactured or mobile home (e.g., ISO HO-7, or equivalent).
    • Dwelling-Fire Owner Occupied (DO) type policies, which cover 1-4 dwelling-units in which the owner lives in one or more units (e.g., ISO DP-1, 2, 3 or equivalent).
    • Dwelling-Fire Tenant Occupied (DT) type policies, which cover 1-4 dwelling-units that are rented to a tenant for a whole or part of a year (e.g., ISO DP-1, 2, 3 or equivalent).
  2. Coverage: The report contains the average coverage amount and the amount of losses for two coverage types. The methodology section of the workbook describes how an average coverage amount was derived.
    • Coverage A covers damage to the structure of the dwelling-unit, excluding any additional structure coverage (e.g., extended replacement cost coverage or other structures).
    • Coverage C covers damage or loss of personal property, excluding any additional contents coverage (e.g., additional coverage for specific valuables such as jewelry or furs).
  3. PPC Scores and Fire Risk Classification:
    • PPC scores refers to the Public Protection Classification program, a Verisk product, that scores an address between 1 (best) and 10 (worst). The scores reflect municipal fire protection efforts available to that property (e.g., distance to primary responding fire station).
    • Fire Risk Classification refer to the output of a commercial tool that models a property’s exposure to fire risk, typically measured by environmental/community level factors around a property such as fuel, slope, and road access. The output may be a numerical score within a range, or a category (e.g., negligible, low, moderate, high, very high). The methodology section of the report provides more detail about how we calculated an average of these scores.
  4. CAT and Non-CAT: A Fire Catastrophe (CAT) loss is a severe event which results in losses or claims which are larger than usual. CDI did not designate the events considered to be Fire Catastrophes. Instead, each insurer made that determination based on either the number of claims or dollar amount of losses associated with each event. However, insurers were advised to use the Insurance Service Offices catastrophe coding when possible, which defines a catastrophe as an event that causes $24 million or more in insured losses to a large number of policyholders. The remaining losses not designated as Fire Catastrophe (CAT) losses were coded as Fire Non-Catastrophe (Non-CAT) losses.
     
  5. Policy Records vs. Loss Records: The report combines data from two separate records provided by the insurers. The policy record provides the details of an individual policy, including the policy type, the effective dates of the policy, the coverage amounts, ZIP code of the insured exposure, the amount of premium earned for the policy, and (if available) the PPC and fire risk classification. This policy record is unique for each policy, and cannot repeat within a calendar year. These policy-specific details are provided on a calendar-year basis, and are to be reported as the most recent amount in-effect during that calendar year. In other words, the insurer should report the most recent amount of coverage available at the end of the policy period or the end of calendar year (whichever is first).

    In contrast, a separate loss record collects the losses for each policy. The loss record is linked to the policy record by a unique policy ID. However, since a policy can report multiple claims during a calendar year, a given policy ID will repeat in the loss records for every claim that occurred during the calendar year, and a policy ID may not occur at all if the policy reported no claims. These claims are reported on an accident/occurrence-year basis, which means only claims that occurred during the given calendar year were reported. For each claim, insurers estimate the amount of loss incurred for the claim based on the coverage (Coverage A or C). Incurred losses include the total amount of paid losses and case loss reserves, are gross of subrogation recoveries, and exclude incurred but not reported (IBNR) losses and loss adjustment expenses. All losses were evaluated as of January 31, of the reporting year, which means the losses from the calendar year closer to the report year have had less time to develop than the older year losses.

Summary Worksheet

The “Summary” worksheet provides a statewide summary of all the data items responsive to the report. These items include the average coverage amounts, the average PPC score or average fire risk score (depending on the report), the number of exposures rated as having negligible, low, moderate, high, or very high fire risk (only in the fire risk class report), the total amount of earned exposures and earned premiums, and the number and amount of losses. The loss information is based on the combination of three dichotomous variables: claims resulting from a catastrophe as defined by the insurer (CAT vs. Non-CAT), coverage type (Coverage A vs. Coverage C), and cause of loss (Fire vs. Smoke).

Not all insurers provided both a PPC score and a fire risk classification. Therefore, for the 2022 report, the data is summarized separately for the subset with fire risk classifications or PPC scores, and the total for all insurers. The 2020 report summarized insurers with a fire risk classification and PPC score. The total for all insurers is identified as a “Total” row, followed by the particular policy form (e.g., “Total – HO” summarize the entire HO data, regardless of whether the insurers provided PPC scores or fire risk classifications). The totals on the “Summary” worksheet can be used to calculate the size of the sample described in the other worksheets.

Loss Data (All Companies)

This worksheet is only available in the 2022 report. The “Loss Data (All Companies)” worksheet transposes the loss data on the “Summary” worksheet to increase the ease of comparing incurred losses and claim counts across policy forms and calendar years. As on the “Summary” worksheet, the losses and claim counts are differentiated based on the combination of three dichotomous variables: claims resulting from a catastrophe as defined by the insurer (CAT vs. Non-CAT), coverage type (Coverage A vs. Coverage C), and cause of loss (Fire vs. Smoke). It is also important to note that all losses are evaluated as of January 31, of the reporting year, which means the losses from the calendar year closer to the report year have had less time to develop than the older year losses.

The loss data is broken into four sections. The first section summarizes the claim count and losses for non-catastrophic fire and smoke losses reported under Coverage A. The second section summarizes the claim count and losses for non-catastrophic fire and smoke losses reported under Coverage C. The third section summarizes the claim count and losses for catastrophic fire and smoke losses reported under Coverage A. Finally, the fourth section summarizes the claim count and losses for catastrophic fire and smoke losses reported under Coverage C.

Loss Data (Has Fire Risk Class)/Loss Data (PPC)

The second “Loss Data” worksheet is identical to the first loss data worksheet, but only includes the experience of insurers that provided PPC scores or fire risk classifications (depending on the version of the report).

ZIP Code Data

There are 12 worksheets with ZIP code level detail. These worksheets are identified by the combination of the calendar year and the policy form. For example, ZIP code level homeowners’ data for 2020 is in the “2020HO” worksheet. The same data elements available in the “Summary” worksheet are also provided in the ZIP code worksheets. However, only insurers that provide PPC scores or fire risk classifications (depending on the report version) are included in the ZIP code data.

Google Translate