Three Metrics, Three Questions
An excess-layer quote can be described by several ratios and dollar amounts, but they do not answer the same question. Rate on line compares premium with the layer’s stated limit. Expected loss ratio compares expected covered loss with premium. Technical premium is the premium indicated by expected loss, expenses, and any specified risk load. A layer can therefore have a high rate on line but a modest expected loss ratio, or the reverse.
Use a one-year pricing period and a hypothetical excess layer that pays losses above attachment point \(A\), up to limit \(L\). Its payment on a ground-up loss \(X\) is \(\min((X-A)_+,L)\), where \((z)_+=\max(z,0)\). Unless stated otherwise, examples treat the layer as having one annual premium, no reinstatements, and no other premium adjustments. The results are aligned with the CAS Exam 8 (Advanced Ratemaking) syllabus.
Calculate and Interpret Rate on Line
The denominator in rate on line is the limit of the layer, not the attachment point, the total amount of insurance above ground-up losses, or the expected loss. For a $5 million limit with a $1 million annual premium, the rate on line is \(1/5=20\%\). The attachment point affects the expected loss and may affect the premium, but it does not change the denominator in this definition.
Rate on line is a compact way to compare premiums relative to stated layer capacity. It is not an expected loss ratio and does not, by itself, reveal the probability or severity of losses. For example, a 20% rate on line means the annual premium is 20% of the stated limit. It does not mean the expected covered loss is 20% of the premium, nor does it mean that the limit will be exhausted once every five years.
For consistent comparisons, use the same premium basis and period for each quote. If a premium includes a particular expense or adjustment, compare it with another premium on the same basis. Also confirm that the limits are stated on a comparable basis; this tutorial uses the stated annual layer limit \(L\).
Expected Loss Ratio and Technical Premium
Let \(E_L\) denote expected covered layer loss for the pricing period. If the premium being assessed is \(P\), the expected loss ratio is \(E_L/P\). A higher premium lowers the expected loss ratio when expected loss is held constant; a higher expected loss raises it when premium is held constant. Always identify whether the denominator is a quoted premium or a technical indication.
To calculate a technical premium, specify how expenses and loads are represented. Suppose variable expenses are a proportion \(v\) of premium, fixed expenses for the period are \(F\), and a dollar risk load is \(R\). Under the assumptions that these components are additive and the variable expense rate applies to total premium, the indicated premium satisfies:
The premium appears on both sides because variable expenses are a percentage of premium. Solving for it requires \(v<1\):
This formula does not apply unchanged if the variable expense is defined as a percentage of losses rather than premium, if some expenses are calculated on another base, or if the components are not additive. In those cases, write the premium equation using the actual expense definitions before solving.
Once \(P\) has been calculated, use that same premium to calculate the technical rate on line \(P/L\) and technical expected loss ratio \(E_L/P\). When the premium equation includes only the components above, these components can also be expressed as shares of premium:
The identity is a useful reconciliation check. It is not a claim that actual losses or expenses in a future year will equal their expected or indicated values.
Worked Example: Pricing a Layer From Its Loss Outcomes
A hypothetical $4 million excess layer has these annual covered-loss outcomes: $0 with probability 0.60, $1 million with probability 0.20, $3 million with probability 0.15, and $4 million with probability 0.05. Variable expenses are 20% of premium, and the dollar risk load is $150,000. There are no fixed expenses. Calculate the technical premium, rate on line, and expected loss ratio.
Method and expected loss. The stated outcomes are already layer payments, so multiply each payment by its probability. In millions of dollars:
Thus, expected layer loss is $0.85 million. The premium equation is \(P=0.85+0.15+0.20P\), with all dollar amounts in millions. Therefore, \(0.80P=1.00\), and the technical premium is \(P=1.25\) million.
Rate on line. Divide the $1.25 million premium by the $4 million limit: \(1.25/4=0.3125\), or 31.25%.
Expected loss ratio. Divide expected loss by the same technical premium: \(0.85/1.25=0.68\), or 68.00%.
Check and common mistake avoided. Variable expenses are \(0.20(1.25)=0.25\) million. Expected loss, risk load, and variable expenses sum to \(0.85+0.15+0.25=1.25\) million, matching the premium. The premium shares are 68% expected loss, 12% risk load, and 20% variable expense. A common mistake is to add expected loss and risk load and call the $1.00 million subtotal the premium; that omits the $0.25 million variable expense.
Attachment and Limit Change the Loss Side
Changing a layer’s attachment point or limit changes which parts of ground-up losses the layer covers. It can therefore change expected layer loss and, in turn, a premium indication. The rate-on-line denominator is still the layer limit. A larger limit does not automatically imply a larger rate on line, because both the premium and the limit may change.
The following comparison uses one hypothetical annual ground-up loss \(X\) with the stated distribution. The layer payment in each outcome is calculated as \(\min((X-A)_+,L)\); its expected value is the probability-weighted average of those payments. Assume each layer has variable expenses of 20% of premium, no fixed expense, and a $250,000 risk load.
Worked Example: Comparing Two Layers on the Same Loss Distribution
In millions of dollars, annual ground-up loss is $0 with probability 0.50, $4 million with probability 0.25, $10 million with probability 0.15, and $20 million with probability 0.10. Compare a $5 million limit excess of $5 million with a $5 million limit excess of $10 million. For each, calculate expected layer loss, technical premium, rate on line, and expected loss ratio.
Method. Apply each layer’s attachment and limit to every ground-up loss before taking the probability-weighted average.
| Ground-up loss ($m) | Probability | $5m xs $5m payment ($m) | $5m xs $10m payment ($m) |
|---|---|---|---|
| 0 | 0.50 | 0 | 0 |
| 4 | 0.25 | 0 | 0 |
| 10 | 0.15 | 5 | 0 |
| 20 | 0.10 | 5 | 5 |
For the $5 million excess of $5 million layer, expected loss is \((0)(0.50)+(0)(0.25)+(5)(0.15)+(5)(0.10)=0.75+0.50=1.25\) million. Its technical premium is \((1.25+0.25)/(1-0.20)=1.50/0.80=1.875\) million. Its rate on line is \(1.875/5=37.50\%\), and its expected loss ratio is \(1.25/1.875=66.67\%\).
For the $5 million excess of $10 million layer, expected loss is \((0)(0.50)+(0)(0.25)+(0)(0.15)+(5)(0.10)=0.50\) million. Its technical premium is \((0.50+0.25)/0.80=0.9375\) million. Its rate on line is \(0.9375/5=18.75\%\), and its expected loss ratio is \(0.50/0.9375=53.33\%\).
Check and common mistake avoided. For the first layer, $1.25 million of expected loss plus $0.25 million of load plus 20% of the $1.875 million premium ($0.375 million) equals $1.875 million. For the second, $0.50 million plus $0.25 million plus 20% of $0.9375 million ($0.1875 million) equals $0.9375 million. A common mistake is to count the $4 million ground-up loss as a $4 million payment in the first layer. Because it does not reach the $5 million attachment, the layer pays zero.
Fixed Expenses, Risk Loads, and the Quoted Premium
Expenses and risk loads affect the metrics differently. Adding a dollar risk load or fixed expense increases the indicated premium, all else equal. That increase raises rate on line if the limit is unchanged and lowers the expected loss ratio if expected loss is unchanged. A variable expense provision also increases the premium through the gross-up. Do not count an expense twice: if a stated premium already includes it, do not add it again.
A quoted premium need not equal the technical indication. The quote’s rate on line is calculated from the quoted premium; its expected loss ratio is expected covered loss divided by that quote. Separately, the technical premium and technical metrics describe the indicated price under the stated assumptions. Keeping the quote and indication distinct makes it possible to discuss their difference without changing the definitions.
Worked Example: Gross-Up With Fixed Expenses
A hypothetical layer has a $4 million limit and expected covered loss of $900,000. Its annual fixed expenses are $120,000, its dollar risk load is $180,000, and variable expenses are 15% of premium. Calculate the technical premium, technical rate on line, and technical expected loss ratio.
Method. Put expected loss, fixed expense, and dollar risk load in the numerator, then gross up for variable expenses. All amounts in the premium equation are in thousands of dollars:
The technical premium is $1,411,764.71 before rounding, or $1,411,765 to the nearest dollar. The technical rate on line is \(1{,}411.7647/4{,}000=0.352941\), or 35.29%. The technical expected loss ratio is \(900/1{,}411.7647=0.6375\), or 63.75%.
Check and common mistake avoided. Variable expenses are \(0.15(1{,}411.7647)=211.7647\) thousand. Adding expected loss, fixed expense, risk load, and variable expense gives \(900+120+180+211.7647=1{,}411.7647\) thousand, which matches the premium. The component shares are 63.75% expected loss, 8.50% fixed expense, 12.75% risk load, and 15.00% variable expense; they total 100%. A common mistake is to calculate \(900+120+180=1{,}200\) thousand and stop. That subtotal does not fund variable expenses calculated as a percentage of premium.
Common Exam Traps and a Practical Sequence
The metrics are easiest to keep straight by writing down the numerator and denominator before calculating. Use \(P/L\) for rate on line and \(E_L/P\) for expected loss ratio. For a technical premium with variable expenses on premium, solve the premium equation rather than treating the pre-expense subtotal as the final premium.
- Using the wrong ROL denominator. Divide by the layer limit, not the attachment point or attachment plus limit.
- Confusing rate on line with expected loss ratio. The former uses premium divided by limit; the latter uses expected covered loss divided by premium.
- Applying a layer to the wrong amount. Calculate the layer payment from the ground-up loss and the stated attachment and limit before averaging.
- Grossing up the wrong expense basis. The \(1-v\) denominator is appropriate when variable expenses are \(v\) of premium. It is not a universal adjustment for every expense convention.
- Mixing quoted and indicated premiums. Identify which premium is used in the ELR denominator and report the basis explicitly.
A disciplined calculation proceeds from coverage to expected loss, then from premium assumptions to technical premium, and finally to ratios. This order helps separate what the layer is expected to pay from how the price is built.
Check Your Understanding
For each problem, state the premium basis used for the expected loss ratio. Dollar amounts are annual unless noted.
- A $6 million limit layer has expected covered loss of $720,000, a $180,000 risk load, no fixed expenses, and variable expenses equal to 25% of premium. Find technical premium, rate on line, and technical expected loss ratio. (Answer: technical premium $1,200,000; ROL 20.00%; ELR 60.00%.)
- A hypothetical quote is $1.6 million for a $8 million limit layer with expected covered loss of $960,000. Calculate the quote’s rate on line and expected loss ratio. (Answer: ROL 20.00%; expected loss ratio 60.00%.)
- A $3 million limit layer has annual payment outcomes of $0 with probability 0.50, $1 million with probability 0.30, and $3 million with probability 0.20. It has variable expenses of 10% of premium, no fixed expenses, and a $180,000 risk load. Calculate expected layer loss, technical premium, rate on line, and technical expected loss ratio. (Answer: expected loss $900,000; technical premium $1,200,000; ROL 40.00%; ELR 75.00%.)
- Ground-up annual loss is $0 with probability 0.50, $6 million with probability 0.30, and $14 million with probability 0.20. Compare a $4 million excess of $4 million layer with a $4 million excess of $10 million layer. Both have a $200,000 risk load, no fixed expense, and variable expenses of 20% of premium. Calculate each layer’s expected loss, technical premium, rate on line, and expected loss ratio. (Answer: excess of $4 million: expected loss $1,400,000, premium $2,000,000, ROL 50.00%, ELR 70.00%; excess of $10 million: expected loss $800,000, premium $1,250,000, ROL 31.25%, ELR 64.00%.)
- A $5 million limit layer has expected covered loss of $1,040,000, fixed expenses of $80,000, a $120,000 risk load, and variable expenses of 18% of premium. A proposed quote is $1,550,000. Calculate the technical premium and the quote’s rate on line and expected loss ratio. How does the quote compare with the technical premium? (Answer: technical premium approximately $1,512,195; quote ROL 31.00%; quote expected loss ratio approximately 67.10%; the quote is approximately $37,805 above the technical premium.)