From Rate Selection to Rate Recommendation
A rate analysis is not complete when a selected change has been chosen. Decision-makers also need to know what action is proposed, what effect it is expected to have, why that action is reasonable, and what must happen before the change can be implemented. A recommendation brings those points together without presenting an estimate as a guarantee.
The recommendation should distinguish the indicated change, the result of the selected analytical method, from the selected change, the change proposed for adoption. The selected change may differ from the indication; if it does, the recommendation should state the difference and give a defensible rationale. It should also identify the basis for any estimated premium impact, including the assumed timing and the extent to which the existing business is expected to receive the change.
A concise recommendation can be organized around four questions:
- What is proposed? State the selected change, the affected business, and the effective timing or rollout approach.
- What is the expected effect? Give an estimated premium or rate-level effect, identify its measurement basis, and state the assumptions used.
- Why is this action reasonable? Summarize the selection rationale and relevant evidence without confusing the indication with the decision.
- What must be considered to implement it? Identify timing, system, communication, approval, transition, or monitoring matters that could affect delivery.
Quantifying the Expected Effect
The effect calculation should match the decision being described. A rate-level change is not automatically the same as the change in a particular policyholder’s premium or in total company premium. Premium can also change because of exposure, coverage, limits, deductibles, classification, or business mix. For a simple estimate, state the assumptions that hold those other influences fixed.
Here, \(c\) is expressed as a decimal: a 7.5% change is \(c=0.075\). The premium base must be defined. It might be an estimate of premium at current rates for a specified book of business, rather than total premium from a different period. When implementation is phased, only the portion of the premium base receiving the change, and the time for which it receives it, should be reflected in the estimate.
Worked Example: A Uniform Change on a Defined Premium Base
A hypothetical insurer is considering a 7.5% selected increase for a product. The indicated increase is 10.0%. The estimated annual premium base at current rates is $6,400,000. Assume the selected change applies to the entire base for a full year, with exposures, coverage, and business mix unchanged. Prepare the core quantitative statements for the recommendation.
Method. Apply the selected change—not the indication—to the stated premium base. The constant-portfolio and full-year assumptions make the simple proportional calculation appropriate.
First convert the selected percentage to a decimal: \(7.5\%=0.075\). The estimated premium increase is:
The estimated premium after the change is:
The selection is 2.5 percentage points below the indication: \(10.0\%-7.5\%=2.5\%\). A recommendation could therefore state: “Adopt a 7.5% increase for the specified product. On an estimated annual premium base of $6,400,000, the increase is expected to add approximately $480,000 for a full year, assuming unchanged exposure and mix. The selected change is 2.5 percentage points below the 10.0% indication; the selection rationale should explain that difference.”
Check. The estimated post-change premium is \(1.075\) times the base: \(\$6{,}400{,}000\times1.075=\$6{,}880{,}000\). The difference from the base is exactly $480,000. The common mistake avoided is using the 10.0% indication as though it were the proposed action; that would imply a $640,000 increase, not the selected change’s $480,000.
Representing Phasing and Partial Implementation
If the change is phased, the expected effect depends on how much of the premium base receives each level of change and for what portion of the measurement period. A useful calculation treats the base as portions with different effective changes. Each portion’s contribution is its share of the base multiplied by its effective change; the contributions are then added.
For example, if a premium-weighted share \(w_j\) receives effective change \(c_j\), the estimated overall change is \(\sum_j w_jc_j\), provided the shares sum to 1 and the same reference premium base is used. A half-year effective change is represented as half the annual change only under an assumption that the affected premium is earned evenly through the year. Actual policy effective dates and earning patterns can make that approximation unsuitable.
Worked Example: Estimating a First-Year Effect Under Phasing
A hypothetical portfolio has an estimated annual premium base of $10,000,000. A 12% increase is selected, but policies renew throughout the year. Assume that 40% of the premium base receives the full 12% change for the year, while the remaining 60% receives an average half-year of the change, equivalent to a 6% first-year effect. Assume exposures and mix remain unchanged. Estimate the first-year premium increase.
Method. Weight each effective change by its share of the premium base. The full-year portion contributes \(0.40\times12\%=4.8\%\) of the total base. The half-year portion contributes \(0.60\times6\%=3.6\%\).
The estimated first-year change is 8.4%. The estimated dollar increase is:
The estimated first-year premium is therefore \(\$10{,}000{,}000+\$840{,}000=\$10{,}840{,}000\). A recommendation should describe this as a first-year estimate under the stated renewal and earning assumptions, not as the full-year effect of a uniform 12% change.
Check. The contributions to the increase are $480,000 from the full-year portion and $360,000 from the half-year portion; together they equal $840,000. The common mistake avoided is applying 12% to the entire base immediately, which would give $1,200,000 and ignore the stated rollout.
Explaining the Rationale and Implementation
A sound rationale connects the selected action to the analysis and the decision context. It should say whether the selection matches the indication or differs from it. If it differs, state the size and direction of the difference and identify the evidence or constraints that support the choice. A bare statement such as “management selected a lower change” does not explain why the selection is reasonable.
The rationale need not repeat every calculation underlying the rate analysis. It should identify the decision-relevant points, such as the principal supporting indications, any important sensitivity or reasonableness considerations, and material assumptions already documented. If uncertainty or conflicting evidence affects the choice, describe that limitation in measured terms. Do not imply that an estimate is certain or that a selected change eliminates all rate risk.
Implementation considerations are part of the recommendation because they can affect the difference between a selected rate change and the effect actually realized. Relevant matters may include the proposed effective date, whether the change applies at renewal or another policy event, any approved cap or phase-in, system readiness, customer communications, and applicable review or approval processes. State these as conditions or action items when they are not yet complete; do not present a planned step as already accomplished.
Worked Example: A Partial First-Year Rollout
A hypothetical insurer proposes a 5.5% increase for a product with an annual premium base of $2,400,000. For the first-year estimate, 25% of the base is expected to receive the increase for the full year, 50% for an average half-year, and 25% not to receive it during that year. Assume the stated shares are measured on a comparable premium basis, and exposures and mix do not otherwise change. Estimate both the first-year and mature full-year effects.
Method. The first-year effective change is the selected 5.5% multiplied by the weighted portion of the year receiving it. The full-year portion contributes 25% of the base at 5.5%; the half-year portion contributes 50% of the base at an effective 2.75%; the remaining portion contributes zero during the first year.
The estimated first-year increase is 2.75%, so its dollar amount is:
For a mature full-year estimate, assume that the entire base eventually receives the full 5.5% increase. The increase is then \(\$2{,}400{,}000\times0.055=\$132{,}000\). Thus the estimated first-year and mature full-year increases are $66,000 and $132,000, respectively.
A recommendation could state the proposed 5.5% change, identify the $66,000 first-year estimate and its rollout assumptions, and separately note the $132,000 mature full-year estimate. It should also identify the renewal timing and system changes needed to support the rollout. Check: the first-year weighted share is \(0.25+0.50(0.5)=0.50\), so half of the mature $132,000 effect is $66,000. The common mistake avoided is reporting the mature effect as though the entire portfolio receives the change for the full first year.
Common Recommendation Errors
The arithmetic is usually straightforward; the difficult part is making sure that the number answers the question being asked. The following errors can make an otherwise polished recommendation misleading.
- Substituting the indication for the selection. The indication describes an analytical result; the selected change is the proposed action. Quantify the effect using the proposed change.
- Confusing percentage points with percent change. A selection of 7.5% versus an indication of 10.0% is 2.5 percentage points lower. It is not a 2.5% reduction to the indication.
- Leaving the premium basis unstated. A dollar estimate without a defined premium base, time period, and implementation assumption cannot be interpreted reliably.
- Calling a rate-level estimate an individual premium prediction. A particular policy’s premium may change by a different amount when its exposure, coverage, rating characteristics, or effective date differs.
- Presenting an estimate as guaranteed. Label the result as expected or estimated and state material assumptions. Future business and implementation may not match them.
Check Your Understanding
For each problem, state the proposed-effect calculation using the assumptions given.
- A hypothetical product has a $3,200,000 annual premium base and a selected 4% increase. Assume full-year implementation and unchanged exposure and mix. What are the estimated increase and post-change premium? (Answer: $128,000 increase; $3,328,000 post-change premium.)
- An indicated change is 9%, but the recommendation proposes 6.5%. The stated annual premium base is $8,000,000, with full-year implementation and no other changes. What is the expected dollar increase, and by how many percentage points is the selection below the indication? (Answer: $520,000 increase; 2.5 percentage points below.)
- A $5,000,000 base receives a selected 10% increase. Assume 30% of the base gets the full-year change and 70% gets an average half-year change. Estimate the first-year percentage and dollar increase. (Answer: 6.5%; $325,000.)
- A $1,600,000 base is subject to a selected 7% increase. In the first year, 50% receives the full-year change, 25% receives an average half-year change, and 25% receives no change. What is the estimated first-year dollar increase, and what is the mature full-year dollar increase if all business eventually receives the full change? (Answer: $70,000 first year; $112,000 mature full year.)
- A recommendation proposes an 8% increase on a $4,000,000 base, but assumes only 75% of the base receives the change for an average half-year in the first year; the remaining 25% receives none. Under unchanged exposure and mix, calculate the first-year increase and the full-year mature increase. (Answer: $120,000 first year; $320,000 mature full year.)