Sensitivity to the Premium Base

Why the Premium Base Matters

An indication compares the premium required for a prospective period with the premium at the current rate level for the same business and basis. The projected cost selection is important, but so is the current-level premium used as the comparison base. If that base changes while required premium is held fixed, the indicated factor changes even though the projected costs have not.

This sensitivity analysis isolates the denominator. It helps explain whether a difference in an indication comes from the premium base selection or from a changed cost estimate. The comparison is meaningful only when the alternative premium bases refer to the same intended coverage, exposure population, and current rate level.

Definition: Let $B$ denote the selected premium at the current rate level for the experience being evaluated, and let $T$ denote the fixed required premium for the corresponding prospective period. The indicated factor is $I=T/B$, for $B>0$. An indicated factor of 1.06 corresponds to a 6% indicated increase; a factor of 0.97 corresponds to a 3% indicated decrease.

Here, $T$ is treated as fixed for the comparison. It represents the premium amount required under the selected projected costs and any applicable provisions. This does not require every component of $T$ to be a loss: for example, the required premium can reflect expenses and profit provisions as well. The key assumption is that those components do not change between the premium-base alternatives being compared.

The premium base $B$ is not simply any premium figure available in the records. It must represent premium at the current rate level and correspond to the costs and exposures in the indication. The earlier material on on-level premium and a consistent basis addresses how to establish that correspondence. This tutorial considers the effect of selecting among bases that are treated as comparable.

The Inverse Relationship

Suppose two candidate premium bases, $B_1$ and $B_2$, are evaluated against the same required premium $T$. Their indicated factors are $I_1=T/B_1$ and $I_2=T/B_2$. Dividing the factors gives an exact sensitivity relationship:

\frac{I_2}{I_1} = \frac{T/B_2}{T/B_1} = \frac{B_1}{B_2}.

Thus, if the premium base rises, the indicated factor falls in inverse proportion, provided $T$ remains fixed. A 5% rise in the base does not produce exactly a 5% decline in the factor: the exact factor ratio is $1/1.05$, not $0.95$.

A short differential argument gives the local sensitivity. For fixed $T$, differentiating $I=T/B$ with respect to $B$ gives:

\frac{dI}{dB}=-\frac{T}{B^2}, \qquad \frac{dI}{I}=-\frac{dB}{B}.

The negative derivative confirms that a larger base lowers the factor. The second relation says that small relative changes are approximately equal in magnitude and opposite in sign. For finite changes, use the exact ratio $I_2/I_1=B_1/B_2$, rather than treating the approximation as exact.

The indicated percentage change is $100(I-1)\%$, while the relative change in the factor between two selections is $100(I_2/I_1-1)\%$. These are different comparisons. For instance, a factor moving from 1.05 to 1.008 moves from a 5% indicated increase to a 0.8% indicated increase, a decline of 4.2 percentage points. The factor itself has declined by 4% relative to its original value.

Worked Example: Comparing Two Current-Level Premium Bases

A hypothetical insurer is reviewing a small commercial package indication. Its selected projected costs and provisions imply required premium $T=\$1{,}260{,}000$. The analyst is comparing two otherwise comparable current-level premium bases: $B_1=\$1{,}200{,}000$ and $B_2=\$1{,}250{,}000$. Calculate both indications and describe the change. Use the exact factor relationship, then round factors to four decimal places.

Method. Hold the required premium fixed and divide it by each candidate premium base.

For the first base:

I_1=\frac{\$1{,}260{,}000}{\$1{,}200{,}000}=1.05.

This is a 5.00% indicated increase. For the second base:

I_2=\frac{\$1{,}260{,}000}{\$1{,}250{,}000}=1.008.

This is a 0.80% indicated increase. The factor difference is $1.008-1.05=-0.0420$, so the indicated change declines by 4.20 percentage points. To calculate the relative change in the factor, use the exact relationship:

\frac{I_2}{I_1}-1 = \frac{1{,}200{,}000}{1{,}250{,}000}-1 = 0.96-1 = -0.04.

The factor has declined by 4.00% relative to its original value. The larger base produces the lower indication, although the required premium is unchanged.

Check. Multiplying the second factor by its base gives $1.008\times \$1{,}250{,}000=\$1{,}260{,}000$, recovering $T$. The comparison avoids the common mistake of assuming that a 4.17% increase in the base must cause a 4.17% decrease in the factor.

When Required Premium Is Built From Fixed Costs

Sometimes the fixed quantity in a sensitivity comparison is initially expressed as projected dollar costs rather than required premium. If $K$ is a fixed dollar amount of projected costs and $q$ is a variable expense-and-profit proportion of premium, then, under the assumption that $q$ is fixed, required premium satisfies $T(1-q)=K$. Therefore $T=K/(1-q)$, and the factor is $K/[(1-q)B]$. This conversion makes explicit that the base sensitivity still operates through $B$, while $K$ and $q$ stay unchanged.

This expression does not apply unchanged if the expense or profit provision varies with the premium base, or if the dollar cost selection changes between the alternatives. In those cases, the numerator is not fixed and the comparison must recalculate the required premium for each alternative.

Worked Example: Converting Fixed Costs to Required Premium

For a hypothetical specialty liability program, projected losses and fixed expenses total $K=\$960{,}000$. Variable expenses and profit provisions together are 20% of premium, so $q=0.20$. Compare current-level premium bases of $\$1{,}000{,}000$ and $\$1{,}040{,}000$. Assume $K$ and $q$ are unchanged, and round factors to four decimal places.

Method. First solve for the required premium that leaves 80% of premium available for the fixed dollar amount:

T(1-q)=K, \qquad T=\frac{\$960{,}000}{1-0.20} =\frac{\$960{,}000}{0.80} =\$1{,}200{,}000.

Using the first base:

I_1=\frac{\$1{,}200{,}000}{\$1{,}000{,}000}=1.2000.

Using the second base:

I_2=\frac{\$1{,}200{,}000}{\$1{,}040{,}000} =1.153846\ldots \approx 1.1538.

The indications are 20.00% and approximately 15.38%, respectively. The exact relative change in the factor is:

\frac{I_2}{I_1}-1 = \frac{\$1{,}000{,}000}{\$1{,}040{,}000}-1 \approx -0.0384615,

or a 3.85% decrease relative to the original factor. The base increased by 4.00%, but the factor decreased by about 3.85%, not 4.00%. As a check, $\$1{,}200{,}000(1-0.20)=\$960{,}000$, so the calculated required premium still funds the stated fixed costs after the 20% provision.

Common mistake avoided. Dividing $K$ directly by $B$ would omit the variable provision and give factors 0.9600 and 0.9231. Those are not the required-premium factors under the stated 20% provision.

Selections That Combine Candidate Bases

An analyst may select a premium base by combining two or more candidate bases, for example by assigning weights to alternative estimates. If the selected base is a weighted average, calculate that base first and then divide the fixed required premium by it. In general, the factor for the weighted base is not the weighted average of the factors calculated separately, because division is nonlinear.

Worked Example: Applying Weights to Candidate Premium Bases

A hypothetical farm equipment program has a fixed required premium of $T=\$2{,}400{,}000$. Two comparable current-level premium estimates are $B_A=\$2{,}400{,}000$ and $B_B=\$2{,}520{,}000$. The analyst selects 60% of $B_A$ and 40% of $B_B$. Calculate the indication from the selected base, compare it with the indication from each candidate alone, and round factors to four decimal places.

Method. Calculate the weighted premium base before applying the indication formula:

B=0.60(\$2{,}400{,}000)+0.40(\$2{,}520{,}000) =\$1{,}440{,}000+\$1{,}008{,}000 =\$2{,}448{,}000.

The selected-base indication is:

I=\frac{\$2{,}400{,}000}{\$2{,}448{,}000} =0.980392\ldots \approx 0.9804.

This corresponds to an indicated decrease of approximately 1.96%. For comparison, using $B_A$ alone gives $I_A=\$2{,}400{,}000/\$2{,}400{,}000=1.0000$; using $B_B$ alone gives $I_B=\$2{,}400{,}000/\$2{,}520{,}000=0.9524$.

A weighted average of the two factors would instead be $0.60(1.0000)+0.40(0.95238095)=0.98095238$, or 0.9810, which is not the selected-base indication of 0.9804. The correct sequence is to select the premium base, then calculate its factor. The result also passes the direct check: $0.980392\ldots\times \$2{,}448{,}000=\$2{,}400{,}000$.

Exam Traps and Practical Interpretation

A useful way to interpret any premium-base sensitivity is to separate three questions: what is being held fixed, what is changing, and what is the reported measure of change? State the fixed required premium, identify the alternative current-level premium bases, and specify whether the answer is a factor, an indicated percentage change, a percentage-point difference, or a relative change in the factor.

Applying the base's percentage change directly to the factor. The relationship is inverse. If $B_2=1.05B_1$, then $I_2/I_1=1/1.05$, not 0.95.
Confusing percentage points with relative percent. A move from a 5% indicated increase to a 0.8% indicated increase is a 4.2 percentage-point decline. The factor's relative decline is 4.0%.
Changing the numerator without acknowledging it. The inverse sensitivity formula assumes $T$ is the same under both choices. If the alternative base corresponds to different exposures, coverage, or selected costs, that assumption may fail.
Averaging factors instead of selecting the base. When weights define a selected premium base, apply them to the bases first. Do not assume that a weighted average of separately calculated factors gives the same result.

The analysis matters because a seemingly modest difference in the selected premium base can move an indication toward or away from 1.00. It does not, by itself, show that prospective costs have changed or that one base is more appropriate. It quantifies the denominator's effect, allowing that effect to be distinguished from changes in the projected cost selection.

Key takeaway: When required premium is fixed and the current-level premium base changes from $B_1$ to $B_2$, the indicated factor changes by the exact ratio $I_2/I_1=B_1/B_2$. Confirm that the bases are comparable and that the required premium truly remains fixed before interpreting the result.

Check Your Understanding

For each problem, assume the stated alternative bases are comparable and hold the required premium fixed unless the problem says otherwise. Round factors to four decimal places where needed.

A hypothetical program has required premium of $\$840{,}000$. Compare current-level bases of $\$800{,}000$ and $\$824{,}000$. What are the two indicated factors, and by what percentage does the factor change relative to the first? (Answer: 1.0500 and 1.0194; a 2.91% decrease.)
Projected dollar costs for a hypothetical product are $\$1{,}080{,}000$, with variable provisions of 10% of premium. The selected current-level premium base is $\$1{,}150{,}000$. Calculate the required premium and indicated factor. (Answer: required premium $\$1{,}200{,}000$; factor 1.0435.)
For a hypothetical equipment portfolio, candidate bases are $\$3{,}000{,}000$ and $\$3{,}120{,}000$. The selected base weights them 25% and 75%, respectively, and fixed required premium is $\$3{,}000{,}000$. Calculate the selected base and indication. (Answer: selected base $\$3{,}090{,}000$; factor 0.9709, a 2.91% indicated decrease.)
A required premium of $\$1{,}440{,}000$ is fixed. The current-level base is $\$1{,}200{,}000$. What base would produce an indicated factor of 1.08, and what is its percentage increase from the original base? (Answer: $\$1{,}333{,}333.33$; an 11.11% increase.)
A hypothetical program's required premium is $\$1{,}900{,}000$, and its selected current-level base rises by 6% from $\$2{,}500{,}000$. Calculate the new base and the relative percentage change in the indication factor. (Answer: new base $\$2{,}650{,}000$; the factor declines by 5.66%.)

What You'll Learn