Premium Weighting Across Experience Years

Why Premium Weighting Matters

A multi-year loss ratio combines losses and premiums from several experience years. Each year may have a different premium volume and a different loss ratio. The combined result must reflect both. If one year has much more on-level premium than another, its experience represents a larger share of the total premium base and should receive more weight in the combined ratio.

This tutorial derives that weighting directly from the definition of a loss ratio. It applies when losses and premiums are measured on a consistent basis across years—for example, when losses have been adjusted to the selected basis and premiums have been brought to the same rate level. The mechanics of those adjustments are separate from the weighting calculation.

Definition: For experience year $j$, let $L_j$ be the loss amount used in the indication after any applicable loss adjustments, and let $P_j^{OL}$ be the earned premium for that year on the selected on-level basis. If $P_j^{OL}>0$, the annual adjusted loss ratio is $R_j=L_j/P_j^{OL}$. The combined loss ratio is the total adjusted losses divided by the total on-level premiums.

The phrase on-level premium is important: the denominator should represent premium expressed at the rate level used for the indication, rather than simply adding earned premiums written at different historical rate levels. The loss amounts and premium basis must also be defined consistently across the experience years. Once those inputs have been established, premium weighting follows from the ratio definition.

Deriving the Premium-Weighted Average

For $n$ experience years, the combined ratio is the sum of the annual losses divided by the sum of the annual on-level premiums. Since each annual loss amount equals its annual ratio multiplied by its premium, substitute $L_j=R_jP_j^{OL}$ into the combined ratio:

R_{\mathrm{combined}} = \frac{\sum_{j=1}^{n}L_j}{\sum_{j=1}^{n}P_j^{OL}} = \frac{\sum_{j=1}^{n}R_jP_j^{OL}}{\sum_{j=1}^{n}P_j^{OL}} = \sum_{j=1}^{n}w_jR_j, \qquad w_j=\frac{P_j^{OL}}{\sum_{k=1}^{n}P_k^{OL}}.

The weights are each year’s share of the total on-level premium. They are nonnegative when premiums are nonnegative, and they sum to one:

\sum_{j=1}^{n}w_j = \sum_{j=1}^{n}\frac{P_j^{OL}}{\sum_{k=1}^{n}P_k^{OL}} = \frac{\sum_{j=1}^{n}P_j^{OL}}{\sum_{k=1}^{n}P_k^{OL}} =1.

This derivation explains why years with more on-level premium contribute more: their premium shares, $w_j$, are larger. It also shows that the combined ratio is not a new kind of annual ratio; it is exactly the ratio formed from the aggregate losses and aggregate premiums. The weighted-average expression is an equivalent way to calculate and interpret it.

Theorem: If the annual on-level premiums are positive and the annual ratios use the corresponding losses and premiums, the combined loss ratio is a weighted average of the annual ratios, with weights equal to each year’s share of total on-level premium. Consequently, the combined ratio lies between the smallest and largest annual ratios.

The last result follows because the weights are nonnegative and sum to one. A weighted average cannot be below every value being averaged or above every value being averaged. A year’s weight does not by itself determine whether the combined result rises or falls: that depends on whether its ratio is above or below the current combined ratio. A larger premium weight means a larger influence, not necessarily an increase.

Set the basis.
Identify the loss amount and on-level premium used for each experience year. Do not mix premium bases.

Calculate annual ratios.
For each year with positive on-level premium, divide its loss amount by that premium.

Combine the years.
Either divide total losses by total on-level premium or multiply each annual ratio by its premium share and sum.

Check the result.
Confirm the weights sum to one and the combined ratio lies between the smallest and largest annual ratios.

Worked Calculations

In the examples, dollar amounts are stated in thousands unless otherwise indicated. Each scenario is hypothetical. The annual loss amounts are assumed to be the amounts selected for the calculation; the focus is how the years are combined.

Worked Example: Combining Two Years With Different Premium Volumes

A hypothetical insurer is combining two experience years. Year 1 has $1,200 thousand of on-level premium and $720 thousand of adjusted losses. Year 2 has $2,800 thousand of on-level premium and $1,960 thousand of adjusted losses. Calculate the annual ratios, premium weights, and combined loss ratio.

Method: aggregate losses divided by aggregate on-level premium, checked using the weighted-average form.

Year 1 ratio: $R_1=720/1{,}200=0.6000$, or 60.00%.

Year 2 ratio: $R_2=1{,}960/2{,}800=0.7000$, or 70.00%.

Total on-level premium is $1{,}200+2{,}800=4{,}000$ thousand. The weights are $w_1=1{,}200/4{,}000=0.3000$ and $w_2=2{,}800/4{,}000=0.7000$. They sum to $0.3000+0.7000=1.0000$.

The weighted result is $0.3000(0.6000)+0.7000(0.7000)=0.1800+0.4900=0.6700$, or 67.00%.

Check by aggregation: total adjusted losses are $720+1{,}960=2{,}680$ thousand, and $2{,}680/4{,}000=0.6700$. The two calculations agree. Common mistake avoided: the unweighted average, $(60.00\%+70.00\%)/2=65.00\%$, gives each year equal weight despite their different premium volumes. It does not reproduce the aggregate loss ratio.

Worked Example: Weighting Premiums After On-Level Adjustments

A hypothetical book has earned premiums of $900 thousand in Year A and $1,500 thousand in Year B. The supplied on-level factors are 1.10 for Year A and 0.96 for Year B. Selected adjusted losses are $594 thousand and $1,008 thousand, respectively. Calculate the combined ratio using on-level premiums.

Method: first express each premium at the selected rate level, then calculate ratios and premium weights. The factors are given for this problem; apply each factor to its corresponding earned premium.

Year A on-level premium is $900(1.10)=990$ thousand. Its annual ratio is $594/990=0.6000$, or 60.00%.

Year B on-level premium is $1{,}500(0.96)=1{,}440$ thousand. Its annual ratio is $1{,}008/1{,}440=0.7000$, or 70.00%.

Total on-level premium is $990+1{,}440=2{,}430$ thousand. The weights are $990/2{,}430=0.407407\ldots$ for Year A and $1{,}440/2{,}430=0.592593\ldots$ for Year B. Using unrounded weights, the weighted ratio is $0.407407\ldots(0.6000)+0.592593\ldots(0.7000)=0.244444\ldots+0.414815\ldots=0.659259\ldots$.

Check by aggregation: selected losses total $594+1{,}008=1{,}602$ thousand, and $1{,}602/2{,}430=0.659259\ldots$. Rounded to two decimal places as a percentage, the combined ratio is 65.93%. Common mistake avoided: weighting by the original earned premiums would use the wrong denominator basis. The premium weights must come from the on-level premiums used in the combined ratio.

Worked Example: Seeing the Influence of a Large-Premium Year

A hypothetical insurer has the following three years of experience, with amounts in thousands: Year 1 has on-level premium 500 and adjusted losses 400; Year 2 has premium 2,000 and losses 1,000; Year 3 has premium 5,500 and losses 3,575. Calculate the combined ratio and assess the influence of Year 3.

Method: calculate each annual ratio and premium share, then combine. The annual ratios are $400/500=0.8000$, $1{,}000/2{,}000=0.5000$, and $3{,}575/5{,}500=0.6500$.

Total premium is $500+2{,}000+5{,}500=8{,}000$ thousand. The respective weights are $500/8{,}000=0.0625$, $2{,}000/8{,}000=0.2500$, and $5{,}500/8{,}000=0.6875$. They sum to $1.0000$.

The weighted calculation is $0.0625(0.8000)+0.2500(0.5000)+0.6875(0.6500)=0.05000+0.12500+0.446875=0.621875$.

Check by aggregation: losses total $400+1{,}000+3{,}575=4{,}975$ thousand, and $4{,}975/8{,}000=0.621875$. Rounded to two decimal places as a percentage, the combined ratio is 62.19%. Year 3 has 68.75% of the premium and therefore the greatest weight. Its ratio of 65.00% pulls the result more strongly toward itself than either smaller-premium year does. Common mistake avoided: treating all three annual ratios as equally influential would ignore that the years contribute very different amounts of premium.

Interpretation and Common Traps

Premium weighting is not a subjective adjustment or an optional preference. It follows from aggregating the dollars in the numerator and denominator. An annual ratio based on $5 million of premium represents more of the aggregate premium base than a ratio based on $500,000. The premium share expresses that difference exactly.

Do not average annual percentages without checking their weights. The simple arithmetic average is correct only when all included on-level premiums are equal.
Use the same premium basis in the ratios and weights. If the combined denominator uses on-level premium, the weights must be calculated from on-level premium, not from a different premium measure.
Do not confuse premium size with the direction of influence. A large-premium year above the current combined ratio raises the result; one below it lowers the result.
Check the range. With nonnegative premiums and a positive total, the combined ratio must be between the smallest and largest annual ratios. A result outside that range signals a calculation or data-basis error.

If an experience year has zero on-level premium, its annual ratio $L_j/P_j^{OL}$ is undefined, so it cannot be included as an ordinary annual ratio with a premium weight. The aggregate ratio may still be defined if total premium is positive, but a loss amount assigned to a zero-premium year requires careful review of the data and the intended aggregation basis. Do not silently assign that year a ratio or weight.

Key takeaway: Calculate the combined loss ratio as total adjusted losses divided by total on-level premium. Equivalently, take the premium-weighted average of the annual ratios, using each year’s share of total on-level premium as its weight.

Check Your Understanding

For each hypothetical scenario, calculate as directed and check that your result is consistent with the stated data.

Year A has $400 thousand of on-level premium and $220 thousand of adjusted losses. Year B has $600 thousand of on-level premium and $450 thousand of adjusted losses. Find each annual ratio and the combined ratio. (Answer: 55.00% and 75.00% annually; 67.00% combined.)
A hypothetical insurer has earned premiums of $900 thousand and $1,500 thousand, on-level factors of 1.10 and 0.96, and adjusted losses of $594 thousand and $1,008 thousand. Find the premium weight for the second year and the combined ratio. (Answer: second-year weight 59.26%; combined ratio 65.93%.)
Three years have on-level premiums of $200 thousand, $800 thousand, and $2,000 thousand, with annual loss ratios of 40%, 60%, and 80%. Find total losses and the combined ratio. (Answer: total losses $2,160 thousand; combined ratio 72.00%.)
Year A has $1,200 thousand of on-level premium and a 50% loss ratio. Year B has $800 thousand of on-level premium. If the combined ratio is 64%, find Year B’s loss amount and annual ratio. (Answer: $680 thousand of losses; 85.00% annual ratio.)
A current portfolio consists of $2,000 thousand of on-level premium at a 60% combined ratio. A new experience year contributes $3,000 thousand of on-level premium at a 72% ratio. Find the new combined ratio and state whether it rises or falls. (Answer: 67.20%; it rises.)

What You'll Learn