Two drivers with identical loss expectations get different renewal quotes. One shopped around last year; the other never leaves. Is charging the loyal one more a smart business practice or unfairly discriminatory pricing?
Traditional ratemaking loads expected losses and expenses to reach an indicated rate. Price optimization layers on a demand-side adjustment: how much can this insured absorb before shopping or lapsing? Insureds who rarely switch (price-inelastic) get charged more; sensitive shoppers get held down.
KEY: The regulatory objection is simple. If two insureds have the same expected cost, a premium gap driven only by their willingness to pay violates the "not unfairly discriminatory" standard.
States moved first, using existing rating law. Maryland released the first bulletin against price optimization in October 2014, and numerous states followed (15 states plus the District of Columbia by late 2015). The NAIC Casualty Actuarial and Statistical Task Force then adopted a white paper in November 2015 that defined the practice, catalogued the state actions, and recommended regulatory responses.
Common mistakes
- Thinking price optimization means any use of data. It specifically means pricing on elasticity or willingness to pay, not cost. Legitimate cost-based classification is untouched.
- Assuming statistical significance alone validates a model factor. Regulators also require a cost-rational link and absence of proxy discrimination.
- Believing a bulletin bans predictive models outright. States restrict price optimization and demand documentation for models; they do not forbid GLMs.
Bottom line
- Price optimization sets premium using non-cost factors, mainly price elasticity of demand and retention, rather than expected loss and expense.
- Regulators treat pure price optimization as unfairly discriminatory because identical-risk insureds pay different premiums for reasons unrelated to cost.
- State bulletins led the response: Maryland acted first in October 2014, and 15 states plus DC had acted by late 2015.
- The NAIC task force's November 2015 white paper then documented the practice and the state actions.
Exam shortcut
For any pricing question, ask whether the premium factor ties to expected cost. If the driver is retention, elasticity, or propensity to shop, flag price optimization and unfair discrimination. For a predictive-model factor, run a three-part gate: statistically significant AND cost-rational AND not a proxy. Significance alone never saves it, and cost-rational means a rational explanation, not causal proof.
The full lesson (about 1,479 words, 10 min read) adds 2 worked examples, all 7 common mistakes, a self-check, free in the app.
Learning objectives
- A5
Browse all free Exam 6U lessons or jump into free Exam 6U practice questions.