What is a Rating Engine

The rating engine is a tool used to calculate the policy premium for an insured based on their responses to input questions. It evaluates these answers to assess the insured’s risk level and determines the appropriate premium to charge based on that assessment.

Imagine an insured named Ben, who lives and works in San Francisco and is seeking medical insurance from ABC Insurance Company. Ben is 26 years old and has been smoking since he was 15, but he has no family history of heart disease. How much would his policy premium be?

To calculate his premium, we need to consider two key elements:

• Base Rate: The base rate is the starting cost to calculate an insurance premium before adjustments like surcharges, discounts, or other rating factors are applied. It represents the standard cost of insuring a particular risk, in this case, Ben, under typical conditions without accounting for individual risk characteristics. This base rate can vary by state or country. In Ben’s case, since he lives in California, let’s assume his base rate starts at $100.

• Modifier Factor: This adjustment factor is applied to the base rate to reflect specific characteristics, behaviors, or conditions that impact the insured’s risk. Modifier factors help tailor the base rate to the individual’s risk profile, making the premium more accurate and fair based on their unique circumstances.

While Ben’s base rate starts at $100 due to his location in California, his smoking history likely makes him riskier to insure than some, though less risky than others. The modifier factor helps adjust the base rate, ensuring that his premium is more customized and fair for his situation.

For this documentation, assume the following modifier factors for the rating questions:

• Rating Question: How old are you?

Age range	Modifier factor
0 - 20	0.8
21 - 60	1.5
61+	2.5

• Rating Question: Do you smoke?

Do you smoke?	Modifier factor
Yes	2.0
No	1.0

• Rating Question: History of Heart Disease in family?

History of Heart Disease or family?	Modifier factor
Yes	2.0
No	1.0

How is a base policy premium calculated?

A base policy premium is calculated thus:

Base policy premium = Base rate * Modifier factor

So, for 26-year-old Ben, who lives in California and started with a base rate of 100 dollars. Who smokes but does not have any family history of heart disease.

We can calculate his policy premium as:

Base premium = 100 * 1.5 * 2.0 * 2.0
Base  premium = 600

So Ben’s base premium before adding credits, debits, fees, or taxes is 600 dollars.

Rating engine with AI Insurance

It’s up to you, the underwriter, to define and provide your base and modifier factors. When a potential insured applies for a policy through you, AI Insurance asks relevant questions (questions you have sent us), gathers additional information, and calculates the policy premium based on what we learn.

The questions we ask depend on the insured and their entity type. For example, if you’re determining the right premium for a group of doctors, the questions will differ significantly from those you’d ask when insuring a group of pilots.

Similarly, the questions will change depending on whether you’re insuring a group of farmers, a farm, or a tractor.

You need to know entity types to fully understand a multi-level rating engine, how they relate to policies and premiums, and the difference between policy-level and risk-level rating questions.

What are entity types?

An entity type refers to the classifying risks covered under an insurance policy based on their roles, characteristics, or risk profiles. These classifications help determine the nature of coverage, risk assessment, and premium calculation.

For example, consider a company providing comprehensive insurance to a hospital. This insurance covers not only the hospital itself but also the doctors working within it and potentially the hospital’s vehicles. In this scenario, the hospital might have an entity type of company, the doctors an entity type of healthcare provider, and the vehicles an entity type of auto.

When issuing a policy to the hospital, insurers use policy-level rating questions to evaluate the organization’s overall risk. These questions help determine how likely it is for the hospital to be involved in incidents involving patients, doctors, or others. Examples of such questions include:

• What is the average length of employment across all hospital locations?
• Does the hospital have a negotiated risk agreement?

Although covered by the same policy and influenced by the policy-level rating questions, doctors working at these hospitals have varying degrees of individual risk. For instance, an allergy doctor poses a much lower risk than a heart surgeon and would pay a lower premium. Additionally, a doctor working in California might pay a higher premium than one in Ohio due to higher lawsuit payouts in California. Questions that assess the individual risk of a doctor are called Risk-level rating questions, and the doctors themselves are referred to as Risks.

Similarly, hospital vehicles, classified under the entity type auto, are also considered Risks. Insurers use risk-level rating questions to determine the level of risk associated with these vehicles.

To clarify this hierarchy, here’s an example of how policy and risk-level rating questions are structured for a medical insurance policy:

• Policy-Level Rating Questions (evaluate the organization as a whole):

  • Does the hospital have a negotiated risk agreement?

  • What is the average length of employment across all locations?

• Risk-Level Rating Questions (evaluate specific entities within the policy):

  • Risk A (Doctor): Questions about specialization, location, malpractice history, etc.

  • Risk B (Truck): Questions about usage, maintenance, accident history, etc.

How the customer and AI Insurance work together to get you your rating engine

As a customer, your first step is to define the rating questions and their corresponding modifier factors. You need to clearly outline your calculations so we can assist you in automating the underwriting process for potential insureds.

To begin, contact the AI Insurance team to request a Rating Engine Requirements Doc. This template outlines the rating information you require to automate the underwriting process.

Following the instructions outlined in this template, you must specify some information, such as:

• The line of business you’re insuring, such as General Liability, Business Interruption, Commercial Property, Cannabis Liability, Medical Malpractice, or more.
• The entity types you are insuring under each line of business—such as Company, Healthcare Providers, Property, Auto, and others.

For each combination of line of business and entity type, you’ll need to replicate and customize the Base Rate Calculator spreadsheet template accordingly.

In this template, you will outline the base rates for an insured based on factors such as the state in which they reside. You will then define the modifier factors used to adjust these base rates and calculate the base policy premium, as demonstrated in the example with Ben.

Line of Business	Entity Type	Base rate spreadsheet
General Liability	Company	Link
General Liability	Healthcare Provider	Link

In addition to the base rate spreadsheet, you’ll complete a Rating Questions, Adjustments, Taxes, and Fees spreadsheet. Here, you’ll list the questions you’d like AI Insurance to use during the underwriting process—these are the rating questions—and define the corresponding modifier factors for each.

Furthermore, you’ll outline any custom adjustments to the premium policy, including credits, debits, warnings, caps, fees, and taxes. All of these details will also be documented in the spreadsheet to ensure a comprehensive and automated underwriting process.

Rating Questions, Adjustments, Taxes, and Fees

In the Rating Questions, Adjustments, Taxes, Fees spreadsheet, you will do four things:

• Define Underwriting Questions: You should specify what questions you want AI Insurance to ask an insured, broker, or underwriter during policy creation. This section of the spreadsheet is organized into six columns, which are:
  • Question: What is the question?
  • Possible Values: What type of answer are you expecting (e.g., long-form text, date, or dropdown option)?
  • Display Logic: Based on the answers, what tasks should be performed related to the policy?
  • Location: Does this question apply at the policy level or risk level?
  • If Risk Level, Entity Type: If the question pertains to the risk level, specify the entity type it relates to.
  • Question ID: What is the number for this question?
• Credits and Debits: Specify discretionary adjustments— additions or deductions—to the original base policy premium rates initially calculated. This section also includes discounts, which are always subtracted from the original amount (unlike factors, which are multiplied). This section is divided into seven columns:
  • Adjustment Name: What is the name of the credit or debit affecting the calculated base policy premium?
  • Increase or Decrease: Does this adjustment increase or decrease the policy premium? Indicate the direction.
  • Entity Type or Policy: Does this adjustment apply at the policy or the risk levels? If risk level, specify the entity type.
  • Percent of Base Premium or Fixed Dollar Amount: Is this adjustment a fixed amount or a percentage of the original base premium?
  • Logic/Calculation for Applying the Adjustment: How was the adjustment figure determined, and why is this insured receiving the adjustment?
  • Relevant Files for Calculation/Logic: Are any files associated with this adjustment?
• Warnings and Caps Define conditions to flag specific scenarios as warnings. Additionally, set caps on discounts to ensure they do not exceed certain limits. This section is divided into three columns:
  • Warning or Cap: Specify whether this is a warning or a cap. Warnings are used to display notifications, such as when discounts exceed a certain percentage. Caps specify the maximum discount amount an insured can receive, regardless of the number of credits or debits.
  • Name: What is the name of this warning or cap?
  • Logic: What is the logic applied to this condition?
• Fees and Taxes
  • Fees: Add administrative fees that carriers charge, unrelated to the insurance coverage.
  • Taxes: Include taxes, which vary by state and must be applied accordingly.

After completing the data entry and premium calculations, you can make final discretionary adjustments to an insured’s premium. These adjustments can be applied at the policy or risk levels as a percentage or a fixed cash amount.