What Premium Shows About Rate Adequacy

Introduction

A book of insurance can produce more premium than another book and still be less adequate. The first book may cover more exposures, provide broader coverage, or simply have a different mix of risks. Premium volume describes the dollars collected or earned; by itself, it does not reveal whether those dollars are sufficient for the expected costs of providing the coverage.

To interpret premium, an actuary needs a meaningful comparison: premium for a defined period of coverage, measured against the expected losses and expenses associated with that same coverage. The comparison is an assessment of expected costs, not a guarantee of the actual result. Claims and expenses will vary, and a fuller pricing assessment may also consider items such as profit and contingencies.

Definition: Premium volume is the total premium for a specified group and period. Rate adequacy asks whether the premium for the coverage provided is sufficient in relation to the expected losses and expenses for that coverage. An expected underwriting margin is premium less expected losses and expenses, on a consistent basis. A positive expected margin means expected premium exceeds those included costs; a negative margin means it does not.

Make the Comparison on a Consistent Basis

Premium must be paired with the coverage to which it relates. In particular, specify the policies or exposures being considered, the coverage period, and whether premium is earned for that period. Earned premium is associated with coverage already provided; written premium can include coverage that will be provided in a later period. The earlier material on matching premium to coverage and exposure provides the basis for making that distinction.

Expected losses and expenses must refer to the same policies, coverage terms, and period as the premium. For example, comparing a year of premium with expected losses for six months would give a misleading result. Comparisons also need consistent financial bases: if one amount includes recoveries or particular expenses and another does not, their difference may not answer the intended question.

For a simple portfolio, let $P$ be earned premium for the period, $E[L]$ expected losses for that period, and $E[C]$ expected expenses included in the comparison. Then:

\text{Expected underwriting margin} = P-E[L]-E[C].

This calculation is deliberately limited to the costs specified. It is not automatically a complete rate indication or a guarantee that the rate is adequate under every pricing objective. If profit, contingencies, or other costs are relevant but omitted, the margin is only a partial comparison.

Why Premium Volume Is Not Enough

Total premium combines the amount charged per exposure with the number and type of exposures. A larger total can arise simply because there are more insured risks. It can also reflect higher limits or broader coverage, which may come with higher expected costs. Consequently, comparing two premium totals without considering exposure and coverage can reverse the apparent conclusion about adequacy.

A useful way to organize the calculation is to evaluate expected margin by group and then add the group margins. If group $j$ has earned premium $P_j$, expected losses $E[L_j]$, and expected expenses $E[C_j]$, then the total expected margin is:

\sum_j \left(P_j-E[L_j]-E[C_j]\right) = \sum_j P_j-\sum_j E[L_j]-\sum_j E[C_j].

The equality holds because addition distributes across the subtraction: each group's premium and costs contribute to the same portfolio totals. This also shows why a total can hide important differences among groups: positive margins in some groups can offset negative margins in others.

Worked Example: A Larger Premium Book With a Larger Expected Deficit

Consider two hypothetical books of one-year policies. All policies have completed their coverage periods, so the stated premium is earned for the same period as the expected costs. Assume no discounting, with effective annual interest rate $i=0\%$; mortality is not relevant. In Book A, each exposure has claim count $N\sim\operatorname{Poisson}(0.08)$, and every claim costs exactly $10,000. In Book B, $N\sim\operatorname{Poisson}(0.10)$, and every claim costs exactly $10,000. For both books, expenses are 20% of premium plus the stated fixed expense per exposure.

Book A: calculate premium. There are 100 policy-years at $1,200 each:

P_A=100(1{,}200)=\$120{,}000.

Calculate expected losses. The expected claim count per exposure is $0.08$. With a fixed $10,000 cost per claim, expected loss per exposure is $0.08(10{,}000)=\$800$. For 100 exposures, expected losses are $100(800)=\$80{,}000$.

Calculate expenses and margin. Variable expenses are $0.20(120{,}000)=\$24{,}000$, and fixed expenses are $100(150)=\$15{,}000$. Total expected costs are $80{,}000+24{,}000+15{,}000=\$119{,}000$. The expected margin is $120{,}000-119{,}000=\$1{,}000$.

Book B: calculate premium and costs. Premium is $200(900)=\$180{,}000$, which is more than Book A's $120,000. Expected loss per exposure is $0.10(10{,}000)=\$1{,}000$, so total expected losses are $200(1{,}000)=\$200{,}000$. Variable expenses are $0.20(180{,}000)=\$36{,}000$, and fixed expenses are $200(100)=\$20{,}000$. Total expected costs are $200{,}000+36{,}000+20{,}000=\$256{,}000$. The expected margin is $180{,}000-256{,}000=-\$76{,}000$.

Check. Book A's premium exceeds its included expected costs by $1,000; Book B's costs exceed its premium by $76,000. The larger premium volume in Book B does not indicate greater adequacy. The calculation avoids the mistake of treating total premium as a measure of sufficiency without comparing it with the corresponding exposure and expected costs.

Coverage Changes What Premium Means

Two policies with different coverage terms should not be compared using premium alone. A broader limit, for example, can increase the insurer's expected payment when a claim occurs. Even when the expected claim frequency is the same, the expected cost per exposure can differ because the coverage changes the amount paid.

For a simple claim model, suppose claim count $N$ has a Poisson distribution with parameter $\lambda$, and every claim produces a fixed covered payment $s$. Then the expected loss per exposure is $\lambda s$. This follows because the expected number of claims is $E[N]=\lambda$, and the total payment is $sN$, so $E[sN]=sE[N]=s\lambda$. This model is a simplifying assumption, not a claim that all insurance losses have fixed severity.

Worked Example: Comparing Premiums for Two Coverage Limits

A hypothetical insurer offers two one-year coverage options to otherwise comparable risks. Each option has 100 policy-year exposures. The lower-limit option charges $700 per exposure; its claim count is $N\sim\operatorname{Poisson}(0.05)$, with a fixed covered payment of $8,000 per claim. The higher-limit option charges $900 per exposure; its claim count is also $N\sim\operatorname{Poisson}(0.05)$, but its fixed covered payment is $15,000 per claim. For each option, expenses are $180 per exposure plus 15% of premium. Assume all premium is earned for the period, $i=0\%$, and mortality is not relevant.

Lower-limit option. Premium is $100(700)=\$70{,}000$. Expected loss per exposure is $0.05(8{,}000)=\$400$, so total expected losses are $100(400)=\$40{,}000$. Variable expenses are $0.15(70{,}000)=\$10{,}500$; fixed expenses are $100(180)=\$18{,}000$. Total expected costs are $40{,}000+10{,}500+18{,}000=\$68{,}500$, giving an expected margin of $70{,}000-68{,}500=\$1{,}500$.

Higher-limit option. Premium is $100(900)=\$90{,}000$. Expected loss per exposure is $0.05(15{,}000)=\$750$, so total expected losses are $100(750)=\$75{,}000$. Variable expenses are $0.15(90{,}000)=\$13{,}500$; fixed expenses remain $100(180)=\$18{,}000$. Total expected costs are $75{,}000+13{,}500+18{,}000=\$106{,}500$, giving an expected margin of $90{,}000-106{,}500=-\$16{,}500$.

Check. The higher-limit option has $20,000 more premium but $38,000 more expected costs, so its margin is $18,000 lower: $-16{,}500-1{,}500=-\$18{,}000$. A higher premium does not, on its own, mean a more adequate rate. The common mistake is to compare $900 with $700 and conclude that the higher-limit option must be better priced. The correction is to include the expected cost of the coverage being sold.

Portfolio Totals Can Hide Class Differences

An aggregate margin is useful, but it does not show which groups contribute to it. If groups differ in premium, expected loss, or expense, calculate their margins separately before interpreting the total. This is especially important when the portfolio mix changes: total premium can rise or fall because the proportions of the groups have changed, even if each group's premium per exposure has not.

Worked Example: Combining Two Risk Classes

A hypothetical book has 80 policy-year exposures in Class A and 20 in Class B. Class A earns $1,000 per exposure; its claim count is $N\sim\operatorname{Poisson}(0.07)$, with a fixed $10,000 payment per claim. Class B earns $2,000 per exposure; its claim count is $N\sim\operatorname{Poisson}(0.10)$, with a fixed $17,000 payment per claim. Expenses are 20% of each class's premium plus $100 per exposure. Assume all premium is earned for the same one-year period as the costs, $i=0\%$, and mortality is not relevant.

Class A. Premium is $80(1{,}000)=\$80{,}000$. Expected loss per exposure is $0.07(10{,}000)=\$700$, so total expected losses are $80(700)=\$56{,}000$. Expenses are $0.20(80{,}000)+80(100)=16{,}000+8{,}000=\$24{,}000$. The expected margin is $80{,}000-56{,}000-24{,}000=\$0$.

Class B. Premium is $20(2{,}000)=\$40{,}000$. Expected loss per exposure is $0.10(17{,}000)=\$1{,}700$, so total expected losses are $20(1{,}700)=\$34{,}000$. Expenses are $0.20(40{,}000)+20(100)=8{,}000+2{,}000=\$10{,}000$. The expected margin is $40{,}000-34{,}000-10{,}000=-\$4{,}000$.

Aggregate check. Total premium is $80{,}000+40{,}000=\$120{,}000$; total expected losses are $56{,}000+34{,}000=\$90{,}000$; and total expenses are $24{,}000+10{,}000=\$34{,}000$. Therefore, total expected margin is $120{,}000-90{,}000-34{,}000=-\$4{,}000$, equal to the sum of the class margins. The overall deficit comes entirely from Class B in this example. A total alone would conceal that Class A's expected margin is zero while Class B's is negative.

Common Mistakes and Practical Interpretation

Using premium volume as the conclusion. More premium may mean more exposures or broader coverage, not a larger margin. Calculate expected costs on the same basis before judging adequacy.
Comparing mismatched periods. Premium for a full year cannot be compared directly with expected costs for half a year. Match the coverage period and exposure measure.
Ignoring expenses. A premium that exceeds expected losses may still be below expected losses plus the expenses included in the pricing comparison.
Assuming a positive expected margin guarantees a profit. Expected values are averages under stated assumptions. Actual claims and expenses can differ, and omitted costs or pricing objectives can change the interpretation.
Looking only at an aggregate. A portfolio total can combine groups with different margins. Review relevant segments when the mix or coverage varies.

A sound conclusion is therefore specific: identify the coverage and period, state the expected losses and expenses included, and report the resulting expected margin. For instance, “the expected margin is $1,500 for these 100 earned policy-years under the stated claim and expense assumptions” says more than “premium is $70,000.” The first statement makes clear what the premium is being compared with and which assumptions support the assessment.

Key takeaway: Premium shows the amount associated with a portfolio and its coverage; it does not establish rate adequacy by itself. Assess expected adequacy by matching earned premium, expected losses, and included expenses to the same exposures, coverage terms, and period.

Check Your Understanding

For each hypothetical problem, compare premium with expected losses and the stated expenses on the same coverage basis. Assume no discounting, $i=0\%$, and no mortality unless stated otherwise.

A portfolio has 50 earned policy-years at $800 per policy-year. Claim count per exposure is $N\sim\operatorname{Poisson}(0.06)$, and every claim costs $10,000. Expenses are 15% of premium plus $100 per exposure. Find the total expected margin. (Answer: $-\$1{,}000$.)
A book has 120 earned policy-years at $950 each. Claim count per exposure is $N\sim\operatorname{Poisson}(0.04)$, with a fixed $15,000 payment per claim. Expenses are 20% of premium plus $90 per exposure. Find the expected margin. (Answer: $+\$8{,}400$.)
Class A has 40 exposures, premium of $1,200 per exposure, claim count $N\sim\operatorname{Poisson}(0.05)$, and fixed payment $12,000 per claim. Class B has 60 exposures, premium of $1,000 per exposure, claim count $N\sim\operatorname{Poisson}(0.08)$, and fixed payment $9,000 per claim. Expenses are 20% of premium plus $100 per exposure in each class. Find each class margin and the total margin. (Answer: Class A $+\$10{,}400$; Class B $-\$1{,}200$; total $+\$9{,}200$.)
A portfolio has 200 policies, each covered for six months and earning $450 for that coverage. Express exposure in policy-years, so there are 100 policy-years. Claim count per policy-year is $N\sim\operatorname{Poisson}(0.08)$, with a fixed $10,000 payment per claim. Expenses are 15% of premium plus $120 per policy-year. Find the expected margin for the six-month policies in total. (Answer: $-\$15{,}500$.)

What You'll Learn