Separating Rate-Level Changes From Loss-Cost Trends

Why the Two Adjustments Must Be Separated

A historical loss and premium experience period may not represent the conditions expected during the prospective rating period. Two distinct changes can explain part of that difference: the insurer may have changed its rate level, and the expected cost of claims may have changed over time. These effects are not interchangeable. A rate change affects the premium charged for an exposure; a loss-cost trend represents a change in expected claim cost.

Keeping the effects separate matters whenever historical losses and premiums are adjusted together. If the premium adjustment is mistaken for a loss trend, or vice versa, the resulting loss ratio or indicated rate level will not represent the intended prospective conditions.

Definition: A rate-level change is a change in the rates charged for coverage. An on-level adjustment restates historical premium to a specified rate level, commonly the current rate level. A loss-cost trend adjusts historical losses or loss costs to reflect an assumed change in expected claim costs between the historical and prospective cost levels. On-leveling applies to premium; loss-cost trending applies to losses or loss-cost components.

An on-level factor is treated here as an input already determined from the applicable rate history and premium data. A selected trend factor is likewise treated as an input based on the selected trend and its interval. The focus is on deciding which component each factor adjusts, rather than calculating either factor from rate-change dates or selecting a trend endpoint.

A Simple Model Explains the Distinction

Consider a portfolio with a fixed set of exposures. Let $E$ denote the number of exposure units, $r$ the average rate per unit, and $c$ the expected loss cost per unit. In this simplified model, premium is exposure multiplied by rate, and expected losses are exposure multiplied by expected loss cost.

P=E r,\qquad L=E c

Suppose rates change by a factor $k$, while the expected loss cost for the same risks and period is unchanged. Premium changes from $E r$ to $E(kr)$; expected losses remain $E c$. Thus, restating historical premium to the new rate level involves the rate factor, but it does not follow that historical losses should receive that factor.

Now suppose expected loss cost changes by a factor $t$, while the rate level is held fixed. Expected losses change from $E c$ to $E(tc)$, but the premium already charged does not automatically change. This is why trending losses does not mean trending premium.

Key result: For a historical loss-ratio analysis, adjust historical premium for rate-level differences and adjust historical losses for the selected loss-cost trend. Keep each adjustment attached to the quantity it describes. This interpretation assumes the rate-level change does not itself alter the expected loss cost for a fixed risk, and that the selected trend is intended to represent claim-cost change rather than a change in exposure volume or risk mix.

Worked Applications

Worked Example: On-Level Premium and Trended Losses

A hypothetical insurer has historical earned premium of $1,000,000 and historical losses of $600,000. The on-level factor to the selected rate level is 1.20. The selected aggregate loss-cost trend factor from the historical cost level to the prospective cost level is 1.08. Find the adjusted premium, trended losses, and resulting loss ratio. Treat the amounts and factors as deterministic inputs; no interest rate, mortality assumption, or probability distribution is used. Report money to the nearest cent and the loss ratio to two decimal places.

Method. On-level the premium and trend the losses separately. The common mistake this method avoids is applying the loss trend to premium or the rate-level factor to losses.

Step 1: On-level the premium. Multiply historical earned premium by the supplied on-level factor:

P_{\text{on-level}}=\$1{,}000{,}000(1.20)=\$1{,}200{,}000.00.

Step 2: Trend the losses. Multiply historical losses by the supplied loss-cost trend factor:

L_{\text{trended}}=\$600{,}000(1.08)=\$648{,}000.00.

Step 3: Calculate the adjusted loss ratio. Divide trended losses by on-level premium:

\frac{L_{\text{trended}}}{P_{\text{on-level}}} =\frac{\$648{,}000}{\$1{,}200{,}000} =0.54=54.00\%.

Check. The adjusted losses are greater than historical losses, consistent with an upward loss-cost trend. The on-level premium is greater than historical premium, consistent with an upward rate-level adjustment. The loss ratio is 54.00%.

A wrong calculation would apply the trend factor to premium and leave losses unchanged: $\$1{,}000{,}000(1.08)=\$1{,}080{,}000$, producing $\$600{,}000/\$1{,}080{,}000=55.56\%$. That calculation neither uses the supplied on-level factor nor adjusts losses to the prospective cost level, so it does not answer the stated question.

Worked Example: A Rate Decrease Does Not Reduce Historical Losses

A hypothetical book has historical earned premium of $2,000,000 and losses of $1,100,000. A rate decrease means that the supplied on-level factor to the selected rate level is 0.90. The selected loss-cost trend is 3% per year for two years. Find on-level premium, trended losses, and the adjusted loss ratio. Assume annual compounding of the selected trend, fixed exposure and risk mix, and no interest rate, mortality assumption, or probability distribution. Report money to the nearest cent and the loss ratio to two decimal places.

Method. Apply the rate factor to premium and compound the annual loss trend over two years before applying it to losses. The trap is to apply the rate decrease to losses, or to apply a one-year trend factor instead of a two-year factor.

Step 1: On-level premium.

P_{\text{on-level}}=\$2{,}000{,}000(0.90)=\$1{,}800{,}000.00.

Step 2: Calculate the two-year loss trend factor. With a 3% annual trend compounded over two years:

$$ (1.03)^2=1.0609. $$

Step 3: Trend losses.

L_{\text{trended}} =\$1{,}100{,}000(1.0609) =\$1{,}166{,}990.00.

Step 4: Calculate the adjusted loss ratio.

\frac{\$1{,}166{,}990}{\$1{,}800{,}000} =0.648327\ldots =64.83\%.

Check. A rate decrease reduces the restated premium from $2,000,000 to $1,800,000. The upward trend increases losses from $1,100,000 to $1,166,990.00. The loss ratio is therefore higher than the historical ratio of $\$1{,}100{,}000/\$2{,}000{,}000=55.00\%$, as these two specified changes imply.

It would be incorrect to multiply historical losses by 0.90 merely because rates decreased. The rate decrease changes the premium level, not the amount of claims expected for a fixed exposure and risk profile.

Worked Example: Applying Trend to Frequency and Severity

A hypothetical portfolio has 10,000 exposure units, historical claim frequency of 0.08 claims per exposure unit, and historical average severity of $2,500 per claim. The on-level factor for its historical premium is 1.20, and historical earned premium is $2,000,000. For the projection interval, selected annual frequency trend is 2% and selected annual severity trend is 5%; each applies for one year. Find projected frequency, projected severity, projected losses at unchanged exposure, on-level premium, and the resulting loss ratio. Assume fixed exposure and risk mix, annual compounding, and no interest rate, mortality assumption, or probability distribution. Report money to the nearest cent and the loss ratio to two decimal places.

Method. Trend the frequency and severity assumptions separately, then calculate expected losses using the fixed exposure count. Separately on-level premium. This avoids treating a change in premium rates as a change in claims or confusing exposure volume with loss-cost trend.

Step 1: Project frequency. Apply the one-year 2% trend:

0.08(1.02)=0.0816\text{ claims per exposure unit}.

At 10,000 exposure units, projected claim count is:

10{,}000(0.0816)=816\text{ claims}.

Step 2: Project severity.

\$2{,}500(1.05)=\$2{,}625.00\text{ per claim}.

Step 3: Calculate projected losses. Multiply projected claim count by projected severity:

L_{\text{projected}}=816(\$2{,}625)=\$2{,}142{,}000.00.

As a check, historical losses implied by the assumptions are $10{,}000(0.08)(\$2{,}500)=\$2{,}000{,}000$. The separate frequency and severity changes increase expected losses to $2,142,000.00 at the same exposure count.

Step 4: On-level premium and loss ratio.

P_{\text{on-level}}=\$2{,}000{,}000(1.20)=\$2{,}400{,}000.00, $$ $$ \frac{\$2{,}142{,}000}{\$2{,}400{,}000}=0.8925=89.25\%.

Check. The projected loss amount reflects frequency and severity assumptions; the premium amount reflects the rate-level adjustment. Neither calculation changes the stated 10,000 exposure units.

What Each Adjustment Does—and Does Not Do

Quantity	Rate-level adjustment	Loss-cost trend
Historical premium	Restates to a specified rate level	Not adjusted merely because claim costs trend
Historical losses	Not adjusted merely because rates changed	Restates expected claim costs to a specified cost level
Exposure count or volume	Not automatically changed by an on-level factor	Not automatically changed by a loss-cost trend
Frequency and severity	Not changed by a rate adjustment alone	May be trended separately when that matches the selected assumptions

The distinction depends on the quantities being compared. Premium can change because of rates, exposure volume, coverage, or business mix. Losses can change because of exposure volume, claim frequency, claim severity, coverage, or business mix. A rate-level adjustment is not a general-purpose correction for every premium difference, and a loss trend is not a correction for every loss difference.

In particular, hold the exposure basis and risk mix in view. If the number or type of exposures changes, that effect may require separate treatment. Likewise, a trend selection should represent the intended change in expected claim cost, not a change caused solely by a different portfolio composition. The simplified calculations above assume those other influences are controlled or addressed elsewhere.

Identify the quantity.
Determine whether the input is premium, losses, exposure, frequency, or severity. Do not choose a factor before identifying what it measures.

Identify the cause of change.
A change in charged rates points to premium on-leveling. A selected change in expected claim costs points to loss-cost trending.

Apply only the corresponding adjustment.
Use the rate-level factor on historical premium and the loss-cost trend on historical losses or the relevant loss-cost components.

Check the result in context.
Confirm that the adjusted premium reflects the stated rate level, the adjusted loss reflects the stated cost level, and exposure or mix changes have not been silently attributed to either factor.

Common Exam Traps

Trending premium. A loss-cost trend does not restate premium. If the question calls for prospective losses and on-level premium, the two components need different adjustments.
Applying a rate factor to losses. A rate decrease does not by itself reduce expected claims. A rate change can affect coverage choices or the portfolio over time, but those effects require separate assumptions and are not the direct meaning of on-leveling.
Changing exposure without noticing. A larger exposure count can increase aggregate losses even when loss cost per exposure is unchanged. That is not, by itself, loss-cost trend.
Using inconsistent intervals. A trend factor must correspond to the intended cost-level interval. Compounding a trend for the wrong duration changes losses even if the direction of the adjustment is correct.
Confusing an observed loss ratio with a projected one. The historical ratio uses historical amounts. A prospective analysis must put premium and losses on their intended levels before interpreting their ratio.

Key takeaway: Rate-level changes answer, “What premium would the historical exposure have earned at the selected rate level?” Loss-cost trends answer, “What claim cost would the historical losses represent at the selected cost level?” Keep the answers attached to premium and losses, respectively.

Check Your Understanding

For each problem, keep rate-level adjustments, loss-cost trends, and exposure assumptions distinct. Treat supplied factors as deterministic; no interest rate, mortality assumption, or probability distribution is involved. Round money to the nearest cent and percentages to two decimal places.

A hypothetical portfolio has historical losses of $500,000 and historical premium of $800,000. Its on-level factor is 1.10, and the selected loss trend factor is 1.06. Find adjusted losses, on-level premium, and the adjusted loss ratio. (Answer: $530,000.00; $880,000.00; 60.23%)
A book has $750,000 of historical losses and $1,200,000 of premium. The on-level factor is 0.95, and loss costs trend at 4% annually for two years. Find on-level premium, trended losses, and the adjusted loss ratio. (Answer: $1,140,000.00; $811,200.00; 71.16%)
A hypothetical risk group has $400,000 of losses and $600,000 of premium. The rate-level factor is 1.08, and the selected one-year loss trend is 3%. Calculate the adjusted loss ratio. Identify which factor applies to each amount. (Answer: Trend losses to $412,000.00; on-level premium to $648,000.00; adjusted loss ratio 63.58%)
A portfolio has 5,000 fixed exposure units, historical frequency of 0.12 claims per unit, and severity of $1,800. Frequency trends at 1% annually and severity at 4% annually for two years. Find projected claim count, severity, and losses. (Answer: 612.06 expected claims; $1,946.88 severity; $1,191,607.37 projected losses)
A hypothetical book has historical losses of $900,000 and premium of $1,500,000. Rates decreased, giving an on-level factor of 0.90; the selected loss-cost trend factor is 1.05. Calculate adjusted losses, on-level premium, and the adjusted loss ratio. (Answer: $945,000.00; $1,350,000.00; 70.00%)

What You'll Learn