Rate standards: excessive, inadequate, unfairly discriminatory

Where the standards come from

The NAIC's two property and casualty model rating laws both state the rule in the same words:

• the Prior Approval Version (Guideline #1780) states the standard and leaves the specific definitions to regulation, in a drafting note;
• the File and Use Version (Guideline #1775) defines each term, and ties the excessiveness test to whether the market is competitive.

Both models also tell ratemakers what to consider: past and prospective loss experience within and outside the state, catastrophe hazards, a reasonable margin for profit and contingencies, dividends and similar returns to policyholders, past and prospective expenses, special assessments and other relevant factors. Both say investment income must be considered in judging the reasonableness of profit, and both say expense provisions should reflect the insurer's own operating methods and expected expenses.

State laws built on these models, or on the earlier All-Industry Bills, often follow the same wording. The details differ, so always read the state's own definitions.

Excessive

A rate is excessive if it is too high for the coverage provided. The NAIC handbook describes it as a rate that would exceed what the company needs to achieve an acceptable level of profit.

The file-and-use model makes the test depend on competition:

• In a competitive market, a rate is not excessive. Competition is presumed to discipline price.
• In a noncompetitive market, a rate is excessive if it is likely to produce an unreasonably high profit for the insurance provided, or if expenses are unreasonably high in relation to services rendered.

Under that model, a market is presumed competitive unless the commissioner, after a hearing, finds that a reasonable degree of competition does not exist. The handbook notes that some state laws likewise say no rate in a competitive market is deemed excessive.

How regulators test it: comparing the requested change with the indication; checking that loss development, trend and catastrophe provisions are supported and not overstated; reviewing the expense and profit provisions, including the treatment of investment income; and, in a competitive rating system, looking at whether the market is in fact competitive.

Inadequate

A rate is inadequate if it is too low to sustain the costs it is meant to cover. The handbook describes it as a rate the company could not sustain for long without threatening its solvency, and notes the assessment sometimes also considers competition, such as one company pricing low to gain market share.

The file-and-use model sets a narrow test: a rate is not inadequate unless it is clearly insufficient to sustain projected losses, expenses and special assessments in the class and its use substantially lessens competition or tends to create a monopoly. An optional alternative in the model also treats a rate as inadequate when it is unreasonably low and its continued use endangers the insurer's solvency.

How regulators test it: the same indication review in reverse. Inadequacy questions arise when a filing selects well below its indication, makes large decreases without support, or prices a new program with thin data.

The handbook adds a point both sides should remember: actual results that come in better or worse than expected do not by themselves show that a rate was excessive or inadequate. A rate is an estimate of future costs, and is judged by the support available when it was set.

Unfairly discriminatory

Insurance pricing discriminates by design: different risks pay different prices. The standard prohibits unfair discrimination, which in the cost-based tradition means price differences that do not reflect differences in expected cost.

The file-and-use model states it this way: unfair discrimination exists if, after allowing for practical limitations, price differentials fail to reflect equitably the differences in expected losses and expenses. The handbook gives an example. Charging a man and a woman different prices is discriminatory, but it is unfairly discriminatory only if differences in expected costs cannot reasonably explain it.

Risk-based differentiation

Both NAIC models allow risks to be grouped by classifications for rates and minimum premiums, and allow classification rates to be modified for individual risks under rating plans. Such standards may measure any differences among risks that can be shown to have a probable effect on losses or expenses.

The actuarial standards line up with this. The Casualty Actuarial Society's Statement of Principles, adopted in 1988 and reinstated in 2021 for reference in U.S.-regulated ratemaking, treats a rate as reasonable and not excessive, inadequate or unfairly discriminatory if it is an actuarially sound estimate of the expected value of all future costs of an individual risk transfer. ASOP No. 12 on risk classification asks actuaries to select characteristics related to expected outcomes. It does not require proof of cause and effect, but the characteristics must be objectively determinable, and applicable law can significantly limit the choice.

Protected characteristics and restricted factors

Cost-based support is necessary but not always sufficient, because some characteristics cannot be used whatever the data show:

• Both NAIC models say no risk classification may be based on race, creed, national origin or religion.
• The NAIC handbook notes that a few states constrain rating criteria such as gender or marital status in private passenger auto, and that factors such as education or occupation can sometimes be found contrary to state law or regulation.
• The handbook also notes that some statutes limit how long violations, loss history or insurance scores can be used, and some allow credit history to be used for re-rating only at the policyholder's request.

The specific restrictions vary by state and line and change over time. Check the state's rating statute, unfair trade practices act and regulations for the line before relying on a factor that has been accepted elsewhere.

How regulators test it

Reviewers look for support for each rating factor and classification change, not only the overall rate level. They ask how factors were derived, whether multivariate methods accounted for interactions, how credibility was handled, whether tier placement rules are objective and auditable, and whether any variable is a prohibited characteristic or a close proxy for one. The handbook says regulators are focused on identifying rating or underwriting characteristics that may violate state law as predictive models become more common, and points to the NAIC white paper Regulatory Review of Predictive Models for the questions reviewers ask about generalized linear models.

What the standards mean for an actuarial memorandum

Because the filing must show the rates meet all three standards, the memorandum and exhibits need to support both the level and the structure of the rates:

• Overall level (excessive and inadequate). The data, adjustments, development, trend, catastrophe load, expense and profit provisions, and the resulting indication, with the reasons for any selected change that differs from it.
• Profit and investment income. How the profit and contingency provision was set and how investment income was considered, since the models require it.
• Expenses. That expense provisions reflect the company's own operations and expectations.
• Rating factors and classes (unfair discrimination). The derivation and support for each changed factor, its relationship to expected losses or expenses, and the treatment of credibility and interactions.
• Compliance with restrictions. A clear statement of the variables used, with definitions, so the reviewer can confirm none is prohibited for the line.
• Policyholder impact. The distribution of changes and any capping, which reviewers use to judge reasonableness.

The actuarial memorandum in a rate filing and Rate filing exhibits cover how to present that support.