This write-up explains a simple rating algorithm for a homeowners’ policy. From this example, a general ratemaking structure for general insurance products is developed.
Homeowners’ insurance covers damage to the:
- Outstanding structures
- Loss of use
- Medical coverage
Underwriting basis is that the company writes one home per policy. Below we cover an exercise to rate the premium for a homeowners’ policy.
Rating variables included in the calculation of premium are:
- Base Risk
- Amount of Insurance
- Protection class and Construction type
- Underwriting Tier
- Miscellaneous credits
- Additional optional coverages
- Expense Fees
Rate relativity is the simply the loading applied on top of the base premium.
Description of rating variables
A base risk is the specific risk profile pre-defined by the insurer. It often represents a set of risk characteristics that are most common.
AOI or Amount of Insurance is one of the key rating variables for homeowners’ insurance. AOI represents the amount of coverage purchased to cover damage to the dwelling and is the maximum amount the insurer expects to pay to repair or replace the home.
The territory or the location of the home is a major determinant of homeowners’ insurance risk. Insurers typically group similar geographic units like zip codes together to form rating territories.
The protection class is a ranking based on the quality of fire protection and availability of water in the district. Class 1 indicates highest quality protection. Within each protection class, there is a separate relativity based on construction type of frame and masonry.
Frame construction relies more on lumber and wood products and hence is more susceptible than masonry for some types of loss such as fire or hail loss.
For underwriting tiers, the company uses various underwriting characteristics and lumps them together into tiers based on the overall riskiness of the exposure to loss.
Deductible is that amount of starting loss which the company will not cover. This is so that small claims are excluded, saving unnecessary operating expenses.
Miscellaneous credits cover the various discounts offered on the policy like discount for new homes insured that are 5 year claim free and insured with multiple policies with the company.
Additional optional coverages: Homeowners’ policies place a limit on the amount of coverage though the policyholder can opt to purchase additional coverage. This policy considers optional purchases of higher limits for liability: higher limits for medical coverage. In addition, the limits applying to jewelry contents can be extended to include higher amounts of coverages.
Expense Fees is an explicit expense fee that is used to cover fixed expenses incurred in the acquisition and servicing of the insurance policy.
Homeowners’ Rating Algorithm
Total Premium = All-Peril Base Rate* AOI Relativity*Territory Relativity*Protection Class and Construction Type Relativity*Underwriting Tier Relativity*Deductible Credit* (1- New Homes Discount- Claims Free Discount)* (1- Multi Policy Period) +Increased Jewelry Coverage Rate+ Increased Liability/ Medical Coverage Rate+ Policy Fee
All numerical examples given here are hypothetical. The tables used for ratemaking purposes are shown as follows:
Additional Optional Coverages has two tables as shown below:
Than a policy is taken up by the underwriter and the risk characteristics identified and classified as follows:
This policy embeds the following features of the risk rating manual as shown below:
In the excel a sensitivity analysis of each of the rating factors have been taken as leading to either an increase in ultimate premium when increased, or vice versa.
Similarly, the process can be repeated for different types of policies that may emerge. It is important to have comprehensive variations included in the analysis including the ideal policy and the worst possible policy to determine the maximum and minimum amounts of premium.
The target of our exercise here is to have a general framework available for ratemaking of any general insurance products. A few vital observations follow from this:
- Listing rating variables is fundamentally important
- Quantifying rating variables in the form of rate tables is necessary for a premium to be generated
- A reasonability test of the premiums has to be carried out which includes changing policy risk features to see how premiums change. All of our exercise is fruitless unless the premiums generated are marketable and financially sound. This will require considerable knowledge of market features as well. The pivot of change will be changing the rating tables until a range of reasonable premiums are generated for that line of business.
- Sensitivity test to check that those rating variables designed to have a direct relationship with premium should have so and those designated with inverse relationship should have so.