Components of an Experience Rating Plan

How the Parts of an Experience Rating Plan Fit Together

An experience rating plan uses a risk’s past loss experience to adjust a premium for a future period. The calculation is only meaningful when its parts fit together: historical losses must be paired with the exposures that produced them, the losses and benchmark must use a compatible basis, and the resulting adjustment must be applied to the intended part of the prospective premium.

There is no single formula that defines every plan. Plans can differ in which losses and exposures are eligible, how historical experience is brought to a common basis, and how much of the prospective premium can vary. The framework below makes those design choices visible. Its numerical formulas are illustrative plan structures, not universal rules.

Definition: The experience period is the historical interval whose eligible losses and exposures are used in the rating calculation. The prospective period is the future interval for which the plan determines a premium. Eligible losses are the losses included under the plan’s stated rules. Eligible exposures are the corresponding units of risk used to measure the amount of coverage or activity that generated those losses. An expected-loss benchmark is the loss amount expected for those exposures on the same stated basis as the eligible losses.

A plan generally needs to specify at least the following components:

Period and timing: which historical dates count, when the losses are valued, and the future period being priced.
Experience data: which losses and exposures qualify, and how they are matched by time, coverage, and unit.
Benchmark: the expected losses for the eligible historical exposures, calculated on a basis comparable to the losses.
Adjustment parameters: the rules translating the actual-to-expected comparison into a premium adjustment. These can include a credibility weight and the share of premium subject to adjustment.
Premium base: the prospective premium amount to which the plan applies the adjustment, including a clear statement of which components can change.

A General Calculation Framework

Let $L$ be eligible historical losses after any specified adjustments needed to place them on the plan’s chosen basis. Let $E$ be the expected-loss benchmark for the same historical exposures and coverage. For $E>0$, define the experience ratio $Q=L/E$. A ratio above 1 indicates losses above the benchmark; a ratio below 1 indicates losses below it. This ratio is not itself a premium modification unless the plan says how to translate it into one.

One illustrative plan applies a credibility weight $Z$, with $0\leq Z\leq1$, to the percentage deviation $Q-1$. It then applies the resulting deviation to a fraction $w$, with $0\leq w\leq1$, of the prospective base premium $B$. Here $w$ represents the share of that base that is adjustable under this illustrative formula.

M=1+wZ(Q-1), \qquad P=BM

In these equations, $M$ is the illustrative premium modification and $P$ is the resulting premium. The formula assumes that $B$ and $wB$ are stated in compatible premium units, that $L$ and $E$ refer to the same experience basis, and that the plan permits this linear adjustment. A plan with different adjustment rules requires its own formula. If $E=0$, $Q$ is undefined; the ratio formula cannot be used without an explicitly specified alternative.

The structure also makes an important distinction: the expected-loss benchmark $E$ is usually formed for the historical exposures, whereas $B$ is the premium for the prospective period. They are related through the plan’s rating basis, but they are not interchangeable amounts. A historical loss deviation is measured relative to its historical benchmark; the plan’s selected adjustment parameters then determine how that signal changes the future premium.

Why the Adjustment Formula Has This Form

The formula follows directly from the intended interpretation of the two adjustment parameters. The experience ratio $Q=L/E$ expresses the observed loss amount as a multiple of the benchmark. Its percentage deviation from expected is therefore $Q-1$. Credibility scales that deviation to $Z(Q-1)$: if $Z=0$, the experience has no effect under this formula; if $Z=1$, the full deviation is used before applying the adjustable share.

If only fraction $w$ of the base premium is subject to that percentage change, the dollar change is $wB \times Z(Q-1)$. Adding this change to the original base gives

P=B+wB Z(Q-1)=B\left[1+wZ(Q-1)\right].

Thus the formula does not multiply every premium component by the experience ratio. It changes the designated adjustable share by a credibility-weighted percentage deviation. This justification depends on the illustrative plan’s linear rule; it is not a claim that all experience rating plans use this structure.

Worked Example: Building the Calculation From Matched Experience

A hypothetical equipment-liability plan rates a prospective one-year period beginning January 1, 2026. Its experience period is the three calendar years 2022–2024. Assume all figures below have already been placed on the plan’s common coverage and loss valuation basis. Eligible earned exposure is measured in machine-years, and the expected loss rate on this basis is $500 per machine-year. Eligible losses were $42,000, $55,000, and $68,000; corresponding exposures were 90, 100, and 110 machine-years. The prospective base premium is $90,000, with $Z=0.40$ and $w=0.60$. Using the illustrative formula, calculate the prospective premium.

Method and basis check. First aggregate the eligible historical losses and their matched exposures. The data use the same three years, coverage, valuation basis, and exposure unit, so the benchmark can be computed using the stated expected loss rate.

Step 1: Total eligible losses.

$$ L=42{,}000+55{,}000+68{,}000=165{,}000. $$

Step 2: Total eligible exposure and expected losses.

X=90+100+110=300\text{ machine-years}, $$ $$ E=300\times 500=150{,}000.

Step 3: Experience ratio and deviation.

Q=\frac{165{,}000}{150{,}000}=1.10,\qquad Q-1=0.10.

Step 4: Apply the adjustment parameters.

wZ(Q-1)=0.60\times0.40\times0.10=0.024, $$ $$ M=1+0.024=1.024.

Step 5: Apply the modification to the prospective base.

P=90{,}000\times1.024=92{,}160.

The indicated prospective premium is $92,160.00. As a check, the experience is 10% above expected, but the plan applies only $0.60\times0.40=0.24$ of that deviation to the base, producing a 2.4% increase. The result is greater than $90,000, as the direction of the loss deviation implies.

Common mistake avoided: multiplying the whole base by $1+0.10$ would give $99,000 and ignore both the credibility weight and the adjustable share. The stated formula applies $wZ$ to the deviation, not the full 10% to all premium.

Keeping Losses, Exposure, and Benchmark Aligned

An actual-to-expected comparison is informative only if numerator and denominator describe comparable risk. For example, losses for one coverage should not be divided by a benchmark for a broader package unless the plan explicitly reconciles those bases. Likewise, exposures must correspond to the period and population generating the losses. The units also matter: a rate per machine-year cannot be multiplied by exposure measured in vehicle-years without a justified conversion.

The expected-loss benchmark is not simply a premium total. It is the plan’s expected loss amount for the historical eligible exposures. The prospective premium base, by contrast, is the future premium amount to which the plan applies its adjustment. Keeping those roles separate prevents a frequent conceptual error: using historical expected losses as though they were the premium being adjusted.

Worked Example: Checking the Exposure Denominator

A hypothetical data-center property plan uses an experience period with 160, 200, and 180 eligible site-years in three successive years. The respective eligible losses, already adjusted to a common basis, are $39,000, $51,000, and $45,000. The plan’s expected loss rate on that basis is $250 per site-year. The prospective base premium is $150,000, with $Z=0.50$ and $w=0.80$. Calculate the premium modification and premium.

Method and common mistake. Aggregate the matched loss and exposure totals before comparing them. Do not use the number of years, or an unweighted average of annual loss ratios, as the exposure denominator when the plan’s benchmark is defined per site-year.

Step 1: Aggregate experience.

L=39{,}000+51{,}000+45{,}000=135{,}000, $$ $$ X=160+200+180=540\text{ site-years}.

Step 2: Calculate the expected-loss benchmark.

E=540\times250=135{,}000.

Step 3: Compare actual with expected and apply the plan parameters.

Q=\frac{135{,}000}{135{,}000}=1.000, $$ $$ M=1+0.80\times0.50\times(1.000-1)=1.000.

Step 4: Calculate the prospective premium.

P=150{,}000\times1.000=150{,}000.

The premium is $150,000.00. The losses equal the expected benchmark, so the experience adjustment is zero. As a check, the aggregate actual loss rate is $135{,}000/540=250$ dollars per site-year, exactly the stated expected loss rate.

Common mistake avoided: dividing $135,000 by three years gives an annual average loss amount, not an actual-to-expected ratio. The benchmark must be calculated for the 540 site-years, yielding $135,000.

What the Premium Base Includes

A premium base should be defined before a plan is applied. It may include several components, but not every component necessarily varies with experience. For a particular hypothetical plan, the base might consist of an adjustable loss-related component and other components held fixed. The parameter $w$ in the illustrative formula expresses the adjustable fraction of the total base; alternatively, a plan can state an adjustable dollar amount directly. Those two descriptions should not be mixed: if the formula uses $w$, the amount subject to adjustment is $wB$, not an additional amount added on top of $B$.

A plan specification should state whether the adjustment applies to the entire premium, only a defined component, or some other expressly defined base. Any charges or components excluded from adjustment should not be accidentally modified through the calculation. The formula also needs to state its adjustment parameters and any applicable limitations. If the plan places a floor or ceiling on the result, that limitation is part of the plan rule; do not assume it from the experience ratio alone.

Worked Example: Applying the Adjustment to the Stated Base

A hypothetical municipal equipment plan has a prospective base premium of $240,000. The plan states that 50% of this base is adjustable, uses $Z=0.75$, and has an experience ratio $Q=1.80$. Assume the illustrative linear formula applies without any additional limitation. Calculate the adjustable dollar base, the modification, and the prospective premium.

Method and common mistake. Compute the experience-based percentage change first, then apply it only to the stated adjustable share. The remaining share stays in the base premium.

Step 1: Find the adjustable and non-adjustable portions.

wB=0.50\times240{,}000=120{,}000, $$ $$ (1-w)B=0.50\times240{,}000=120{,}000.

Step 2: Calculate the credible experience deviation.

Q-1=1.80-1=0.80, $$ $$ Z(Q-1)=0.75\times0.80=0.60.

The adjustable share increases by 60% of its original $120,000 amount, or $120{,}000\times0.60=72{,}000$. The modification to the whole base is therefore $72{,}000/240{,}000=0.30$, giving $M=1.30$.

Step 3: Calculate the premium and check by components.

P=240{,}000\times1.30=312{,}000, $$ $$ P=120{,}000+120{,}000(1+0.60)=120{,}000+192{,}000=312{,}000.

The prospective premium is $312,000.00. Both calculation routes agree. The non-adjustable share remains $120,000, while the adjustable share rises from $120,000 to $192,000.

Common mistake avoided: increasing the entire base by 60% would produce $384,000. That applies the credible deviation to the whole base, despite the plan specifying that only half is adjustable.

Common Design and Exam Traps

When reviewing a plan or solving a calculation, trace the quantities in order: period, eligible experience, benchmark, adjustment, and premium base. This sequence makes it easier to identify where an apparently plausible number stopped matching the plan definition.

Using mismatched periods. Losses from one set of dates cannot be compared casually with exposures or expected losses from a different set. The plan must establish a consistent basis.
Confusing actual loss with expected loss. The experience ratio requires both. A loss total alone does not say whether experience is above or below expected.
Mixing exposure units. Confirm that the exposure unit in the benchmark rate matches the unit used to total eligible exposures.
Adjusting the wrong premium amount. Apply the formula to the defined prospective base and the designated adjustable share, not to an assumed amount.
Treating an illustrative formula as universal. Credibility, adjustable shares, and any plan limitations are design choices. Use the rules stated for the plan in question.
Using a ratio when its denominator is zero. If $E=0$, $L/E$ is undefined. A separate plan rule is needed; the ratio formula does not resolve that case.

A good plan description therefore lets a reader answer, without guessing: What historical experience is used? What benchmark makes that experience comparable? Which parameters translate the deviation into an adjustment? What future premium amount receives that adjustment? This component-by-component interpretation is aligned with the CAS Exam 8 (Advanced Ratemaking) syllabus.

Check Your Understanding

For each problem, identify the relevant plan components before calculating. Dollar answers are to the cent.

A hypothetical greenhouse-glass plan has $72,000 of eligible losses and 240 eligible building-years. Its expected loss rate is $275 per building-year. Find the expected-loss benchmark and experience ratio. (Answer: $E=$$66,000; $Q=1.0909$.)
A hypothetical delivery-fleet plan has $Q=0.90$, $Z=0.60$, $w=0.50$, and prospective base premium $B=$$80,000. Using the illustrative formula, find $M$ and $P$. (Answer: $M=0.9700$; $P=$$77,600.00.)
A hypothetical solar-installation plan records eligible losses of $34,000 and $46,000 over 100 and 140 eligible site-years. The expected loss rate is $300 per site-year. Find $L$, $E$, and $Q$. (Answer: $L=$$80,000; $E=$$72,000; $Q=1.1111$.)
A hypothetical library-property plan has a $180,000 base, $w=0.40$, $Z=0.50$, and $Q=1.25$. Find the dollar amount of the base subject to adjustment, the modification, and the premium. (Answer: adjustable base = $72,000; $M=1.0500$; $P=$$189,000.00.)
A hypothetical cold-storage plan reports $40,000 of eligible losses, but its expected-loss benchmark is zero. Can the experience ratio in this tutorial be calculated, and what must the plan provide to proceed? (Answer: No; $L/E$ is undefined when $E=0$, so a separately specified alternative plan rule is required.)

What You'll Learn