Why the Wording of a Limit Matters
A contract may limit the amount of ground-up loss eligible for coverage, or it may cap the insurer’s payment. These provisions can sound similar, especially when the same dollar amount appears in each. With a deductible, however, they can produce different payments: a deductible reduces the covered amount under a ground-up loss limit, while a payment cap is applied to the insurer’s amount after the deductible.
The key is to translate the contract wording into a payment function before calculating a probability or expectation. This tutorial uses nonnegative ground-up loss \(X\), an ordinary deductible \(d\), a ground-up loss limit \(m\), and a payment limit \(u\). The examples are hypothetical and use one loss at a time; they do not include expenses, coinsurance, or aggregate policy limits. The distinction is aligned with the SOA Exam ASTAM (Advanced Short-Term Actuarial Mathematics) syllabus.
Two Payment Functions
Under a ground-up loss limit \(m\), the amount of loss eligible for coverage is \(\min(X,m)\). Applying the deductible to that amount gives the insurer payment
Under a payment limit \(u\), first apply the deductible to the ground-up loss, giving \((X-d)^+\), and then cap the insurer’s payment:
The superscripts here distinguish the two payment functions: \(Y^{L}_{m,d}\) is the payment under a ground-up loss limit, and \(Y^{P}_{u,d}\) is the payment under a payment limit. The labels do not change the definition of the loss \(X\).
A useful equivalent expression for the loss-limit payment is
Thus, when \(m>d\), the maximum insurer payment under a ground-up loss limit is \(m-d\), not \(m\). Under a payment limit, the maximum is \(u\). The payment limit is reached when the ground-up loss is at least \(d+u\); the loss-limit payment reaches its maximum when the loss is at least \(m\).
Why the Formulas Hold
The formulas follow directly from the order of the contract terms. If the ground-up loss limit is applied first, the eligible loss cannot exceed \(m\). The deductible then removes the first \(d\) dollars of that eligible loss, with a floor of zero. If instead the deductible is applied to \(X\) first, the insurer’s preliminary payment is \((X-d)^+\); the payment limit then restricts that amount to at most \(u\).
The equivalent expression for \(Y^{L}_{m,d}\) can be checked by cases. If \(m\leq d\), the capped eligible loss never exceeds the deductible, so both \((\min(X,m)-d)^+\) and \(\min((X-d)^+,(m-d)^+)\) equal zero. If \(m>d\), consider three ranges of \(X\):
The deductible eliminates the eligible loss, so the payment is zero. Also, \((X-d)^+=0\), making the minimum in the equivalent expression zero.
The loss limit does not bind. The payment is \(X-d\), and \(\min((X-d)^+,m-d)=X-d\) because \(X-d<m-d\).
The eligible loss is \(m\), so the payment is \(m-d\). The equivalent expression also equals \(m-d\), since \((X-d)^+\geq m-d\).
These cases establish the identity for all nonnegative \(X\). They also show why the two limits generally differ. With the same numerical limit \(L\) and a positive deductible \(d<L\), the maximum payment under a ground-up loss limit \(m=L\) is \(L-d\), while a payment limit \(u=L\) allows payment up to \(L\).
Worked Example: Comparing Equal Numerical Limits
A hypothetical insurer covers a single ground-up loss \(X\), in thousands of dollars, with distribution \(P(X=2)=0.20\), \(P(X=6)=0.50\), and \(P(X=11)=0.30\). The probabilities sum to \(1\). The deductible is \(d=2\) thousand dollars. Compare a ground-up loss limit \(m=8\) thousand dollars with a payment limit \(u=8\) thousand dollars, and calculate each expected payment.
Method. Apply the deductible to the eligible loss under the ground-up loss limit; apply the payment cap after the deductible under the payment-limit alternative. The common mistake avoided here is treating both limits as if they cap the same quantity.
| Ground-up loss \(X\), thousands | Probability | \(Y^{L}_{8,2}=(\min(X,8)-2)^+\) | \(Y^{P}_{8,2}=\min((X-2)^+,8)\) |
|---|---|---|---|
| 2 | 0.20 | 0 | 0 |
| 6 | 0.50 | 4 | 4 |
| 11 | 0.30 | 6 | 8 |
For \(X=2\), both payments are zero because the loss does not exceed the deductible. For \(X=6\), both pay \(6-2=4\) thousand dollars because neither limit binds. For \(X=11\), the ground-up loss limit makes the eligible loss \(8\), so its payment is \(8-2=6\) thousand dollars. Under the payment limit, the preliminary payment is \(11-2=9\), capped at \(8\) thousand dollars.
The expected payment under the ground-up loss limit is
thousand dollars. The expected payment under the payment limit is
thousand dollars. The payment limit has the greater expected payment, \(4.4-3.8=0.6\) thousand dollars, because it allows an additional \(2\) thousand dollars at \(X=11\). As a check, that difference occurs with probability \(0.30\), so the expected difference is \(0.30(2)=0.6\) thousand dollars.
Interpreting Contract Language
The distinction is clearest when the contract wording identifies what is limited. “The first \(m\) dollars of ground-up loss are covered, subject to a deductible” describes a limit on eligible loss and leads to \(Y^{L}_{m,d}\). “The insurer pays the loss above the deductible, up to \(u\)” describes a cap on the insurer’s payment and leads to \(Y^{P}_{u,d}\). If the wording does not make the basis of the limit clear, the contract definition must be checked rather than inferred from the word “limit” alone.
The following comparison summarizes the effects of the two arrangements when \(m>d\) and \(u>0\).
| Feature | Ground-up loss limit \(m\) | Payment limit \(u\) |
|---|---|---|
| Amount limited | Eligible ground-up loss | Insurer payment after the deductible |
| Payment function | \((\min(X,m)-d)^+\) | \(\min((X-d)^+,u)\) |
| Maximum payment | \(m-d\) | \(u\) |
| Loss at which the maximum is reached | \(X\geq m\) | \(X\geq d+u\) |
If there is no deductible, the distinction disappears when the two limits have the same value: \(d=0\) gives \(Y^{L}_{m,0}=\min(X,m)\) and \(Y^{P}_{u,0}=\min(X,u)\), so choosing \(m=u\) makes the payment functions identical. With a positive deductible, equal numerical limits generally do not produce equal payments.
Worked Example: Reading a Ground-Up Loss Limit From a Scenario
A hypothetical equipment policy has a deductible of \(4\) thousand dollars. It covers no more than \(10\) thousand dollars of ground-up loss before the deductible. A single loss \(X\), in thousands of dollars, has distribution \(P(X=1)=0.30\), \(P(X=7)=0.40\), and \(P(X=13)=0.30\). Find the insurer’s expected payment. Then calculate what the expectation would be if the contract instead had a payment limit of \(10\) thousand dollars.
Method. Translate each contract description before averaging. In particular, “covers no more than \(10\) thousand dollars of ground-up loss” means \(m=10\), not \(u=10\).
Under the ground-up loss limit, \(Y^{L}_{10,4}=(\min(X,10)-4)^+\). The three payments are:
- At \(X=1\): \((\min(1,10)-4)^+=(1-4)^+=0\).
- At \(X=7\): \((\min(7,10)-4)^+=7-4=3\) thousand dollars.
- At \(X=13\): \((\min(13,10)-4)^+=10-4=6\) thousand dollars.
The expected payment is \(0.30(0)+0.40(3)+0.30(6)=0+1.2+1.8=3.0\) thousand dollars. The probabilities sum to \(0.30+0.40+0.30=1\).
For the alternative payment limit, \(Y^{P}_{10,4}=\min((X-4)^+,10)\). At \(X=1\), the payment is \(0\). At \(X=7\), it is \(\min(3,10)=3\) thousand dollars. At \(X=13\), it is \(\min(9,10)=9\) thousand dollars. Therefore,
thousand dollars. The alternative payment limit increases the expectation by \(3.9-3.0=0.9\) thousand dollars. This difference comes entirely from the loss of \(13\) thousand dollars: the ground-up loss limit pays at most \(10-4=6\), while the payment cap of \(10\) does not bind on the preliminary payment of \(13-4=9\).
When Can the Two Arrangements Match?
For a ground-up loss limit \(m>d\), its payment can be written as \(\min((X-d)^+,m-d)\). Therefore a payment limit \(u=m-d\) produces exactly the same payment for every \(X\):
This is a pointwise equivalence: for each possible loss amount, not just on average, the insurer payments agree. If \(m\leq d\), the ground-up loss limit gives zero payment for every loss, and a payment limit \(u=0\) also gives zero. In contrast, using \(u=m\) when \(m>d\) makes the payment cap larger than the maximum payment allowed by the ground-up limit.
Worked Example: Converting a Loss Limit to an Equivalent Payment Limit
A hypothetical policy has deductible \(d=1.5\) thousand dollars and a ground-up loss limit \(m=9.5\) thousand dollars. A single loss \(X\), in thousands of dollars, has distribution \(P(X=0)=0.10\), \(P(X=1.5)=0.20\), \(P(X=5)=0.25\), \(P(X=9.5)=0.25\), and \(P(X=14)=0.20\). Find the equivalent payment limit and the expected payment under either arrangement.
Method. Match the maximum payments by setting \(u=m-d\). This example avoids the mistake of using \(u=m\) and overlooking that the deductible reduces the ground-up limit’s maximum insurer payment.
The equivalent payment limit is \(u=9.5-1.5=8\) thousand dollars. Under the ground-up loss limit, the payment is \(Y^{L}_{9.5,1.5}=(\min(X,9.5)-1.5)^+\). Under the payment limit, it is \(Y^{P}_{8,1.5}=\min((X-1.5)^+,8)\).
| \(X\), thousands | Probability | Ground-up loss limit payment | Equivalent payment-limit payment |
|---|---|---|---|
| 0 | 0.10 | 0 | 0 |
| 1.5 | 0.20 | 0 | 0 |
| 5 | 0.25 | 3.5 | 3.5 |
| 9.5 | 0.25 | 8 | 8 |
| 14 | 0.20 | 8 | 8 |
For example, at \(X=5\), the loss-limit payment is \(\min(5,9.5)-1.5=3.5\), and the payment-limit payment is \(\min(5-1.5,8)=3.5\). At \(X=14\), the first arrangement pays \(9.5-1.5=8\), while the second pays \(\min(14-1.5,8)=8\). The probabilities sum to \(1\), and the matching entries confirm the equivalence at every listed loss.
The expected payment under either arrangement is \(0.10(0)+0.20(0)+0.25(3.5)+0.25(8)+0.20(8)=0+0+0.875+2+1.6=4.475\) thousand dollars. The largest payment is \(8\) thousand dollars, as required by both equivalent specifications.
Common Exam Trap
A frequent error is to read a ground-up loss limit as though it were a cap on the insurer’s payment. In the first worked example, that error would assign a payment of \(8\) thousand dollars to \(X=11\) under the ground-up loss limit. It would then produce the incorrect expectation
thousand dollars. The correction is to cap the eligible loss at \(8\) first and then subtract the \(2\)-thousand-dollar deductible. The payment at \(X=11\) is \(8-2=6\), not \(8\), so the correct expectation is \(0.20(0)+0.50(4)+0.30(6)=3.8\) thousand dollars.
A reliable check is to state the maximum insurer payment before doing any expected-value calculation. Under a ground-up loss limit \(m\), it is \((m-d)^+\); under a payment limit \(u\), it is \(u\). If a calculated payment exceeds that maximum, the contract has been translated incorrectly.
Check Your Understanding
For each problem, identify the limit basis before calculating. All losses and contract amounts are in the units stated in the problem.
- A hypothetical loss is \(X=9\) thousand dollars. The deductible is \(d=2\) thousand dollars and the ground-up loss limit is \(m=7\) thousand dollars. What is the insurer’s payment? (Answer: 5 thousand dollars)
- A hypothetical loss is \(X=12\) thousand dollars, with deductible \(d=3\) thousand dollars. Compare a ground-up loss limit \(m=10\) thousand dollars with a payment limit \(u=10\) thousand dollars. What are the two payments, in that order? (Answer: 7 thousand dollars and 9 thousand dollars)
- A hypothetical ground-up loss, in thousands of dollars, is \(X=2,5,9\) with probabilities \(0.20,0.50,0.30\), respectively. For deductible \(d=1\), find the expected payment under ground-up loss limit \(m=7\), then under payment limit \(u=7\). (Answer: 4.0 thousand dollars and 4.3 thousand dollars)
- A policy has deductible \(d=2.5\) thousand dollars and ground-up loss limit \(m=12\) thousand dollars. What payment limit gives the same insurer payment for every possible loss? (Answer: 9.5 thousand dollars)
- A hypothetical ground-up loss is \(X=6\) thousand dollars. The deductible is \(d=4\) thousand dollars, and the ground-up loss limit is \(m=3\) thousand dollars. What is the payment, and why? (Answer: 0; the limit-capped eligible loss does not exceed the deductible)