Why the Denominators Differ
A pure premium indication and a loss ratio indication both use projected losses, but they answer different questions. The pure premium method asks how much loss is expected per exposure unit. The loss ratio method asks how projected losses compare with the premium produced by the current rates. Keeping those questions separate prevents a common error: dividing losses by premium and treating the resulting ratio as a pure premium.
An exposure unit is the unit used to measure insured risk, such as one policy-year or one vehicle-year. Premium is the amount charged for insurance and depends on the rates in effect, as well as the exposures and coverage being considered. Thus, exposure measures the volume of risk; premium measures the price charged for that risk.
For the pure premium calculation, \(L\) and \(E\) must refer to the same coverage, population, and period. In an indication, losses and exposures are projected or adjusted as needed to represent the period being priced. Those selections are assumed already made in the calculations below. Premium is not needed to measure loss cost per exposure.
For a loss ratio calculation, premium has a different role: it is the denominator against which losses are compared. It must correspond to the same experience and coverage as the losses and must reflect the premium basis being evaluated. Dividing losses by premium at an unsuitable rate level or for a mismatched population can produce a misleading ratio.
From Loss Cost to an Indicated Average Rate
A loss cost alone is not necessarily the full premium rate. To connect it to a gross indicated average rate, suppose the target loss ratio is \(q\), where \(0<q\leq1\). The target loss ratio represents the share of premium intended to fund losses under the chosen ratemaking assumptions. The simplified relationship below assumes \(L\) is the projected loss amount to be funded, \(E\) is the corresponding exposure, and the same target \(q\) applies to the whole population. Other rate components must be treated consistently when setting \(q\).
The pure premium method starts with losses per exposure and uses the target loss ratio to convert that loss cost to an indicated average premium per exposure. By contrast, the loss ratio method starts with premium at the current rate level and calculates the premium change needed to reach the target ratio.
The rate factor is applied to the current premium basis. It is not itself a pure premium. If the current average premium per exposure is \(P/E\), the loss ratio method’s indicated average premium per exposure is \((P/E)(L/P)/q\), which simplifies to \(L/(qE)\). The two methods therefore agree under consistent data and assumptions, although their denominators and intermediate quantities have different meanings.
Why the Pure Premium Formula Works
The key is to express the target in terms of expected losses and premium. Under the simplified assumptions above, the indicated total premium is \(L/q\), because losses are to be \(q\) of that premium. Dividing that indicated premium by the number of exposure units gives the indicated average rate per exposure:
This also explains why premium is not the pure premium denominator. The quantity \(L/E\) has units of dollars per exposure unit. The quantity \(L/P\) has no units: dollars of loss divided by dollars of premium. Replacing exposure with premium changes the quantity being calculated; it does not produce another version of loss cost.
Use the projected exposure units corresponding to the projected losses. Calculate \(L/E\) to obtain loss dollars per exposure.
For a target loss ratio \(q\), divide the loss cost by \(q\) to obtain the indicated average premium per exposure.
Calculate \(L/P\) using corresponding premium, then divide by \(q\) to obtain a factor applied to the current premium basis.
Worked Examples
Worked Example: Converting a Loss Cost to an Indicated Average Rate
A hypothetical insurer projects losses of $1,260,000 for 4,200 policy-years. Premium at the current rate level for those exposures is $1,050,000. The target loss ratio is 0.75. Find the pure premium, indicated average rate per policy-year, and indicated factor relative to the current average premium. Assume the projected losses and exposures correspond, and use no discounting or separate mortality or probability model.
Method. Calculate projected losses per exposure, then use the target loss ratio to obtain the indicated average rate. Separately, calculate the loss ratio indication from current premium as a cross-check.
Pure premium.
The current average premium is \(\$1{,}050{,}000/4{,}200=\$250\) per policy-year. Therefore the indicated average rate is
As a cross-check, the current loss ratio is \(\$1{,}260{,}000/\$1{,}050{,}000=1.20\). The loss ratio indication is \(1.20/0.75=1.600\), the same factor. The indicated total premium is \(4{,}200\times\$400=\$1{,}680{,}000\), and \(\$1{,}260{,}000/\$1{,}680{,}000=0.75\), as required. The common mistake avoided here is reporting \(L/P=1.20\) as the pure premium; it is a loss ratio, not dollars per policy-year.
Worked Example: A Premium Denominator Gives a Different Quantity
For a hypothetical book, projected losses are $840,000 over 2,800 vehicle-years. Current premium for the same book is $980,000, and the target loss ratio is 0.70. Calculate the pure premium, the current loss ratio, and the indicated average premium per vehicle-year using both methods. Assume consistent coverage and period, with no discounting or separate mortality or probability model.
Method. Use exposure for the loss cost. Use premium for the loss ratio and its rate factor.
The pure premium is \(\$840{,}000/2{,}800=\$300\) per vehicle-year. The current average premium is \(\$980{,}000/2{,}800=\$350\) per vehicle-year. The pure premium indication gives
The current loss ratio is \(\$840{,}000/\$980{,}000=0.857142\ldots\). Its rate factor is \(0.857142\ldots/0.70=1.224489\ldots\). Applying that factor to the current average premium gives \(\$350\times1.224489\ldots=\$428.57\) per vehicle-year. The indicated total premium is \(2{,}800\times\$428.5714\ldots=\$1{,}200{,}000\); projected losses divided by that premium are 0.70. Dollar amounts are rounded to the nearest cent; the rate factor is shown to four decimal places as 1.2245. Dividing losses by premium gives 0.8571, a loss ratio, not the pure premium of $300.
Worked Example: Different Current Premiums, Same Indicated Rate
Two hypothetical territories each have 1,000 exposure units and projected losses of $200,000. Current premium is $250,000 in Territory A and $400,000 in Territory B. The target loss ratio is 0.75 in both. Find each territory’s pure premium, current loss ratio, indicated factor, and indicated average rate. Assume the stated losses, exposures, and premiums correspond; no discounting or separate mortality or probability model is used.
Method. Calculate loss cost by territory using exposure, and calculate each loss ratio indication using that territory’s current premium.
Each territory’s pure premium is \(\$200{,}000/1{,}000=\$200\) per exposure unit. The indicated average rate is \(\$200/0.75=\$266.67\) per exposure unit in each territory, rounded to cents.
In Territory A, the current loss ratio is \(\$200{,}000/\$250{,}000=0.80\). Its factor is \(0.80/0.75=1.0667\), and the current average premium is \(\$250\); thus the indicated average premium is \(\$250\times1.0667\ldots=\$266.67\).
In Territory B, the current loss ratio is \(\$200{,}000/\$400{,}000=0.50\). Its factor is \(0.50/0.75=0.6667\), and the current average premium is \(\$400\); thus the indicated average premium is \(\$400\times0.6667\ldots=\$266.67\). The factors differ because the current premium levels differ, while the loss cost per exposure is the same. The common mistake avoided is interpreting different loss ratio factors as evidence of different projected losses per exposure. Here both methods indicate the same average rate in both territories.
Common Denominator Traps
A quick units check often catches the error. If the desired result is a loss cost in dollars per exposure unit, the denominator must be an exposure count. If the denominator is premium in dollars, the result is a loss ratio. A ratio such as \(L/P\) can be useful in a rate indication, but only after it is interpreted as a comparison with the target loss ratio and applied to the relevant current premium basis.
- Using premium to calculate pure premium. \(L/P\) is a loss ratio; \(L/E\) is a pure premium.
- Applying a loss ratio factor to the wrong base. The factor adjusts the premium basis used in its calculation. It does not directly multiply exposure or loss cost.
- Mixing inconsistent data. Losses, premium, and exposures must correspond in coverage, population, and period. A mismatched premium can distort the loss ratio indication even if the loss cost calculation is sound.
- Comparing unlike intermediate results. A pure premium is a dollar amount per exposure; a loss ratio is a dimensionless proportion. Compare indicated rates or indicated premiums only after converting each method to the same basis.
Check Your Understanding
For each problem, distinguish the loss-cost calculation from the premium-based loss ratio calculation. Round dollar rates to the nearest cent and factors to four decimal places where needed.
- A hypothetical book has projected losses of $510,000, 1,700 policy-years, and current premium of $595,000. The target loss ratio is 0.75. Find the pure premium and the loss ratio indicated rate factor. Assume corresponding data and no discounting or separate mortality or probability model. (Answer: Pure premium $300 per policy-year; factor 1.1429.)
- A hypothetical portfolio has projected losses of $1,080,000 over 3,600 exposure units and current premium of $1,260,000. At a target loss ratio of 0.80, find the indicated average rate per exposure unit and check it using the loss ratio method. (Answer: $375.00 per exposure unit; loss ratio factor 1.0714.)
- Two hypothetical plans each have 3,200 exposure units and projected losses of $640,000. Current premium is $800,000 for Plan A and $1,000,000 for Plan B. With a target loss ratio of 0.80, give each plan’s pure premium, indicated factor, and indicated average rate. (Answer: Both pure premiums are $200; factors are 1.0000 for A and 0.8000 for B; both indicated average rates are $250.00.)
- A reviewer divides projected losses of $450,000 by premium of $600,000 and labels the result the pure premium. There are 1,500 matching exposure units and the target loss ratio is 0.60. Identify the reviewer’s ratio, then calculate the pure premium and indicated average rate. (Answer: The reviewer calculated a 0.7500 loss ratio; pure premium $300 per exposure unit; indicated average rate $500.00.)
- A hypothetical book has two groups: Group A has projected losses of $240,000, 800 exposures, and premium of $320,000; Group B has projected losses of $360,000, 1,200 exposures, and premium of $600,000. Using a 0.75 target loss ratio, calculate the combined pure premium, combined current loss ratio, and indicated average rate. Assume the data are consistent across groups. (Answer: Combined pure premium $300 per exposure; combined loss ratio 0.6522; indicated average rate $400.00 per exposure.)