Why Use Rate-on-Line and Payback Period?
A reinsurance quote has several numbers that answer different questions. The limit describes the stated amount of protection available under the contract. The premium is the price paid for that protection. Expected annual recoveries estimate the amount the cedent expects to collect, on average, under the covered loss model. Rate-on-line and expected payback period reduce these figures to ratios that are useful for describing and comparing layer quotes.
Neither measure is a complete pricing analysis. Rate-on-line relates premium to limit; it does not directly measure expected claims cost. Payback period relates limit to expected annual recoveries; it does not say how many years will actually pass before a loss occurs or the limit is recovered. Their interpretation depends on using consistent contract, period, and currency bases.
Use the premium and limit on the same currency basis. For example, if premium is quoted for a one-year contract, the rate-on-line is the one-year premium divided by the stated limit. For payback, \(E[R]\) must be the expected recovery over one year for the same covered contract and loss scope. If the contract has an occurrence limit, an annual expected recovery may include multiple covered occurrences, but it must reflect the contract’s actual terms and any applicable annual aggregate cap. This tutorial’s examples assume no reinstatements, no additional annual aggregate cap, and no expenses or discounting.
Calculating the Two Metrics
Rate-on-line is a price-per-unit-of-limit measure. A premium of $1.2 million for a $10 million limit gives a rate-on-line of 12%. A higher rate-on-line means more premium per dollar of stated limit; on its own, it does not establish that one layer is expensive or inexpensive relative to its expected losses or risk.
Expected payback period uses expected annual recovery rather than premium. If the expected annual recovery is $1 million and the limit is $10 million, the expected payback period is 10 years. This is a ratio of capacity to expected annual recovery, not a promise about the timing or amount of actual recoveries.
The arithmetic can be done in dollars, thousands, or millions, provided the numerator and denominator use the same units. Rate-on-line is dimensionless and is often shown as a percentage. Payback period has units of years when the recovery expectation is annual.
Worked Example: A Quote with a Discrete Recovery Distribution
A hypothetical one-year excess-of-loss layer has a $10 million limit and a quoted premium of $1.8 million. The modelled annual recovery \(R\), in millions of dollars, is 0 with probability 0.80, 5 with probability 0.15, and 10 with probability 0.05. Assume a 0% effective annual interest rate, no discounting, and no mortality assumption because this is a property loss model. The probabilities and recovery amounts specify the complete distribution for this calculation.
Step 1: Calculate expected annual recovery. Multiply each possible recovery by its probability and add:
Step 2: Calculate rate-on-line. Divide the annual premium by the limit:
Step 3: Calculate expected payback period. Divide the limit by expected annual recovery:
Check. The expected recovery is 12.5% of the limit per year, and \(1/0.125=8\) years. The rate-on-line is 18%, independently calculated from premium and limit. The method avoids substituting premium for expected recovery in the payback calculation.
Why Payback Is the Reciprocal of Expected Recovery as a Percentage of Limit
Define the expected annual recovery rate relative to limit as \(E[R]/L\). This is an annual expected amount divided by the layer’s capacity. The payback formula is its reciprocal:
The interpretation follows from linearity of expectation. If the expected recovery is \(E[R]\) each year, then over \(n\) years with the same expected annual recovery, the expected cumulative recovery is \(nE[R]\). The value of \(n\) that makes this expected cumulative amount equal to the limit solves \(nE[R]=L\). Rearranging gives \(n=L/E[R]\). This reasoning does not require recoveries to be identical in every year, but it does require the same expected annual recovery over the period and no change in the relevant limit or contract basis.
For instance, a 10-year expected payback does not mean a layer will pay one-tenth of its limit every year. Actual annual recoveries may be zero in many years and much larger in a loss year. The ratio summarizes an expectation; it does not describe the distribution of outcomes.
Worked Example: Comparing Two Layers
A hypothetical cedent compares two one-year property layers. Layer A has a $4 million limit and an annual premium of $880,000. Its annual recovery \(R_A\), in millions, is 0 with probability 0.70, 2 with probability 0.20, and 4 with probability 0.10. Layer B has an $8 million limit and an annual premium of $1.36 million. Its annual recovery \(R_B\), in millions, is 0 with probability 0.80, 4 with probability 0.15, and 8 with probability 0.05. Assume 0% effective annual interest, no discounting, and no mortality assumption because both are property layers. Each stated distribution is the complete model for that layer.
Layer A expected recovery:
Its rate-on-line is \(0.88/4=0.22\), or 22%. Its expected payback period is \(4/0.80=5\) years. Its expected annual recovery as a percentage of limit is \(0.80/4=20\%\), whose reciprocal is 5.
Layer B expected recovery:
Its rate-on-line is \(1.36/8=0.17\), or 17%. Its expected payback period is \(8/1.00=8\) years. Its expected annual recovery as a percentage of limit is \(1.00/8=12.5\%\), whose reciprocal is 8.
Comparison and check. Layer A has the higher rate-on-line and shorter expected payback. Those results say that A has more premium per dollar of limit and more expected recovery per dollar of limit under the stated models. They do not, by themselves, establish which quote is better: the layers have different limits and recovery distributions. The method avoids ranking the layers using just one ratio while ignoring what each ratio measures.
How to Read a Quote Comparison
The two metrics can move independently because they use different numerators. Holding limit constant, a higher premium increases rate-on-line but does not change payback if expected recoveries are unchanged. Holding premium and expected recoveries constant, a higher limit reduces rate-on-line and increases payback. A change in the loss model can change expected recovery and therefore payback even if premium and limit are unchanged.
| Metric | Calculation | What it summarizes | What it does not establish |
|---|---|---|---|
| Rate-on-line | Annual premium divided by limit | Premium per unit of stated limit | Expected claims cost, profitability, or adequacy of price |
| Expected payback period | Limit divided by expected annual recovery | Years for constant expected recoveries to equal the limit | Actual time until a loss, recovery volatility, or discounted value |
Both measures omit important features of the risk. They do not show how concentrated recoveries are in severe events, how uncertain the estimated expected recovery is, or how the contract’s wording affects which losses are recoverable. They also do not, on their own, account for expenses, brokerage, profit objectives, capital requirements, counterparty risk, or the time value of money. A quote comparison should therefore retain the underlying assumptions and contract details alongside the ratios.
Worked Example: A Premium Increase with Unchanged Expected Recovery
A hypothetical one-year layer has a $6 million limit. Under the unchanged loss model, annual recovery \(R\), in millions, is 0 with probability 0.75, 2 with probability 0.15, and 6 with probability 0.10. The original annual premium is $1.35 million; a renewal quote is $1.485 million. Assume a 0% effective annual interest rate, no discounting, and no mortality assumption because this is a property layer. The stated recovery distribution and probabilities apply to both quotes.
Step 1: Calculate expected annual recovery.
Step 2: Calculate the original metrics. The original rate-on-line is \(1.35/6=0.225\), or 22.5%. The expected payback period is \(6/0.90=6.666\ldots\) years, or 6.67 years to two decimals.
Step 3: Calculate the renewal metrics. The renewal rate-on-line is \(1.485/6=0.2475\), or 24.75%. The expected payback period remains \(6/0.90=6.666\ldots\) years, or 6.67 years to two decimals, because neither the limit nor the recovery expectation changed.
Check. The premium rose by \(1.485-1.35=\$0.135\) million, which is 10% of the original premium. Rate-on-line also rises by 10% relative to its original value: \(22.5\%\times1.10=24.75\%\). Payback is unchanged. This comparison isolates a premium change; it does not prove whether the renewal price is justified.
Common Mistakes and Exam Traps
A frequent error is to divide premium by expected recovery and call the result rate-on-line. That calculation is not rate-on-line: the denominator for rate-on-line is the limit. Another error is to divide premium by expected recovery and interpret the answer as expected payback. Payback uses the limit in the numerator, not premium. For the renewal example, the incorrect payback calculation \(1.485/0.90=1.65\) years uses premium and is not the expected payback period; the correct calculation is \(6/0.90=6.67\) years.
- Mixing units. Dividing a premium stated in dollars by a limit stated in millions without conversion gives a meaningless result. Convert both to dollars or both to millions first.
- Mixing time bases. A multi-year premium or a recovery estimate for several years cannot be inserted as though it were an annual figure without adjusting the basis.
- Using expected losses outside the layer. The recovery expectation must relate to recoveries under the layer’s actual terms, not total ground-up losses.
- Reading payback as an actual timeline. Expected recovery may be uneven across years; the ratio is not a prediction that the layer will be paid smoothly.
- Treating a lower ratio as proof of a better quote. A ratio cannot replace review of limits, coverage, loss assumptions, and other pricing considerations.
Check Your Understanding
For each problem, use the stated annual contract basis. Assume no discounting and no mortality where the exposure is property risk.
- A hypothetical layer has a $5 million limit, a $650,000 annual premium, and expected annual recovery of $250,000. Calculate rate-on-line and expected payback period. (Answer: 13%; 20 years.)
- A hypothetical $6 million property layer has an annual premium of $900,000. Annual recovery \(R\), in millions, is 0 with probability 0.80, 2 with probability 0.15, and 6 with probability 0.05. Calculate expected recovery, rate-on-line, and payback. (Answer: \(E[R]=\$0.60\) million; 15%; 10 years.)
- A hypothetical $12.5 million layer has an expected annual recovery of $1 million. What annual premium corresponds to a 16% rate-on-line, and what is the expected payback period? (Answer: $2 million; 12.5 years.)
- Compare two hypothetical layers. A has a $4 million limit, $720,000 premium, and $400,000 expected annual recovery. B has an $8 million limit, $1.2 million premium, and $640,000 expected annual recovery. Calculate both metrics for each and state which has the higher rate-on-line and which has the shorter payback. (Answer: A—18% and 10 years; B—15% and 12.5 years; A has both the higher rate-on-line and the shorter payback.)
- A hypothetical property layer has an unchanged $10 million limit and unchanged expected annual recovery of $800,000 at renewal. Its annual premium rises from $1.4 million to $1.54 million. Calculate both metrics before and after renewal, and explain which one changes. (Answer: Rate-on-line rises from 14% to 15.4%; payback remains 12.5 years.)