Why Combine Controls With Risk Transfer?
Controls and risk transfer address different parts of an exposure. A control can prevent an event, reduce its severity, or improve detection and response. A contract or insurance policy can shift some of the resulting financial loss to another party, subject to its terms. A sound design considers both: transfer does not prevent the event, and a control does not necessarily eliminate the organization’s financial responsibility.
The calculation must follow the terms of the arrangement. First determine the loss remaining after controls. Then identify which portion is covered, apply deductibles and limits, and calculate the amount the organization retains. Premiums and control costs are additional costs of the design; they are not themselves claim payments or reductions in the loss allocated among parties.
The payment formula assumes that the deductible applies before the insurer’s limit, that the limit caps the insurer’s payment, and that there is no separate aggregate limit, coinsurance, sublimit, or counterparty default affecting recovery. If the contract has different terms, the calculation must reflect those terms instead. In the examples below, amounts are hypothetical and all policy payments are assumed collectible.
How the Allocation Works
For a simple covered loss, the policy pays nothing when the loss is at or below the deductible. Above the deductible, it pays up to the limit. The organization therefore retains the deductible portion and any covered amount above the insurer’s limit. If a loss is excluded, it remains with the organization unless another arrangement transfers it.
Here \(R(z)\) is the retained covered loss. To see why the payment formula holds, consider the three possible ranges. If \(z\le d\), there is no amount above the deductible, so the payment is zero. If \(d<z\le d+u\), the amount above the deductible is \(z-d\), which is below the limit. If \(z>d+u\), the amount above the deductible exceeds the limit, so payment stops at \(u\). In each range the payment cannot be negative or exceed the policy limit.
A control changes the loss to which this allocation is applied. For a claim, calculate the loss after the control, not the pre-control loss. For an expected-cost comparison, model how the control changes event probability, severity, or both, and then apply the transfer terms to the controlled loss. The control’s effect need not be independent of event severity: if control performance and loss size are related, their joint behavior must be reflected in the model.
Worked Example: Control Reduction Followed by a Policy Payment
A hypothetical equipment failure would cause a covered loss of $96,000. An installed shutoff control limits the damage by $18,000 when the failure occurs. The organization has a policy with a $10,000 per-occurrence deductible and a $50,000 per-occurrence payment limit. Assume the control works as specified, the loss is covered, and the insurer pays in full. Allocate the loss among the reduction due to the control, the insurer, and the organization.
Method. Apply the control first, then apply the policy terms to the remaining covered loss.
Step 1: Calculate the post-control loss. The control reduces the loss by $18,000:
Step 2: Calculate the amount above the deductible.
Step 3: Apply the payment limit. The insurer pays the lesser of $68,000 and the $50,000 limit:
Step 4: Calculate the retained amount and check.
The allocation reconciles: $18,000 is avoided through the control, $50,000 is paid by the insurer, and $28,000 remains with the organization. The retained amount consists of the $10,000 deductible plus $18,000 above the covered amount reached by the policy limit. The check is $18,000 + $50,000 + $28,000 = $96,000, the original loss.
Common mistake avoided: Applying the deductible and limit to the original $96,000 would ignore the control’s reduction. The policy responds to the remaining $78,000 loss, not the pre-control amount.
Comparing Expected Costs of a Combined Design
For an annual expected-cost comparison, include expected retained claims, insurance premiums, and control costs on a consistent time basis. The following calculation uses a simple model with at most one event per year. It is not a general formula for every frequency or severity model; when multiple events are possible, each event and any aggregate policy terms must be modeled.
Worked Example: Comparing a Control-and-Insurance Package
A hypothetical organization faces one possible equipment-loss event per year. Without a control, the event occurs with probability 0.10 and causes a covered loss of $200,000; otherwise, the loss is zero. A proposed control costs $4,000 per year, reduces the event probability to 0.06, and limits the loss to $150,000 if the event occurs. An insurance policy costs $1,000 per year, has a $25,000 per-occurrence deductible, and pays at most $75,000 per occurrence. Assume all payments are collectible and there are no other costs. Compare expected annual costs for no mitigation, insurance without the control, the control without insurance, and both together.
Method. For each alternative, calculate the event loss remaining after any control and insurance payment, weight it by the event probability, and add applicable annual costs.
No mitigation. Expected annual loss is:
Insurance without the control. The amount above the deductible is $200,000 - $25,000 = $175,000. The insurer’s payment is capped at $75,000, so the organization retains $200,000 - $75,000 = $125,000 if the event occurs. Expected retained loss is \(0.10(\$125{,}000)=\$12{,}500\). Including the $1,000 premium, expected annual cost is $13,500.
Control without insurance. The controlled event loss is $150,000 and its probability is 0.06. Expected annual claim loss is \(0.06(\$150{,}000)=\$9{,}000\). Adding the $4,000 control cost gives $13,000.
Control and insurance together. For a $150,000 controlled loss, the amount above the deductible is $150,000 - $25,000 = $125,000. The $75,000 limit applies, leaving $150,000 - $75,000 = $75,000 with the organization if the event occurs. Expected retained loss is \(0.06(\$75{,}000)=\$4{,}500\). Add the annual control cost and premium:
The combined alternative has the lowest expected annual cost among these four alternatives: $9,500 compared with $13,000, $13,500, and $20,000. This comparison is limited to the stated model and cost assumptions. It does not establish that the alternatives have equal volatility or that the lowest expected cost is automatically the preferred design.
Common mistake avoided: For the controlled loss, subtracting the deductible and treating the result of $125,000 as the insurer’s payment ignores the $75,000 limit. The correct payment is $75,000, leaving $75,000 per event with the organization. The expected retained claim is therefore $4,500, not $1,500.
Allocating Multiple Layers and Excluded Loss
With layered coverage, determine each layer’s attachment point and limit before calculating recoveries. For a layer attaching at amount \(a\) from the ground up, with payment limit \(u\), the payment on a covered loss \(z\) is \(\min(\max(z-a,0),u)\). This form assumes that the attachment point is stated on a ground-up basis and that applicable underlying coverage responds as intended. An exclusion is handled separately: it is not made covered merely because a policy layer is available.
Worked Example: Control, Primary Coverage, Excess Coverage, and an Exclusion
A hypothetical incident would cause a $310,000 loss before a containment control. The control prevents $40,000 of covered physical damage. After the control, the loss consists of $230,000 of covered damage and $40,000 of excluded business-interruption loss. A primary policy has a $20,000 deductible and a $100,000 per-occurrence payment limit. An excess policy attaches at $120,000 of covered loss from the ground up and pays at most $100,000. Assume the layers respond as written, all payments are collectible, and the limits are not otherwise exhausted. Calculate the control reduction, each insurer’s payment, and the organization’s retained loss.
Method. Separate covered and excluded amounts after the control; calculate the primary and excess payments on the covered amount; then add the excluded loss to the organization’s retention.
Step 1: Confirm the controlled loss. The control prevents $40,000, leaving $310,000 - $40,000 = $270,000. This agrees with the stated components: $230,000 covered + $40,000 excluded = $270,000.
Step 2: Calculate the primary payment. The amount of covered loss above the $20,000 deductible is $230,000 - $20,000 = $210,000. The primary limit caps payment at $100,000.
Step 3: Calculate the excess payment. The excess layer attaches at $120,000 of covered loss. The covered amount above attachment is $230,000 - $120,000 = $110,000. The excess limit caps payment at $100,000:
Step 4: Calculate retained loss. The organization retains $230,000 - $100,000 - $100,000 = $30,000 of covered loss, plus the $40,000 excluded loss. Total retention is $70,000.
Check. The original $310,000 loss is allocated as $40,000 prevented by the control, $200,000 paid by insurers, and $70,000 retained: $40,000 + $200,000 + $70,000 = $310,000. The retained amount also reconciles by component: $30,000 covered retention + $40,000 excluded loss = $70,000.
Common mistake avoided: Treating the excess policy’s $100,000 limit as a further payment on top of an incorrectly calculated primary layer can create overlapping or inconsistent coverage. Calculate each payment using its stated attachment and limit, then reconcile total payments against the covered loss.
Design Considerations and Exam Traps
A combined design should make clear which risk each part addresses. A preventive control may reduce event probability; a containment control may reduce severity; a policy may transfer only covered amounts within specified attachment and limit terms. The organization should identify the residual exposure after all three effects rather than describing the risk as simply “insured” or “controlled.”
Expected cost is only one comparison measure. Two designs with equal expected cost can leave very different retained losses in severe scenarios. The organization should also consider control reliability, policy conditions, exclusions, limits, and whether the proposed transfer responds to the specific residual exposure. These considerations do not change the arithmetic, but they affect whether the modeled arrangement is a credible design.
- Do not apply transfer to the pre-control loss. Unless the contract explicitly provides otherwise, apply the terms to the loss remaining after controls.
- Do not confuse amount above the deductible with the insurer’s payment. The payment may be capped by a per-occurrence limit.
- Keep excluded loss separate. A covered-loss layer does not pay an excluded amount just because the total incident loss exceeds the attachment point.
- Keep claims and program costs distinct. Retained claims, premiums, and control costs belong in a total-cost comparison, but they are different components.
- Check the policy basis. Per-occurrence limits, aggregate limits, sublimits, attachment definitions, and counterparty collectibility can change the result.
This treatment is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.
Check Your Understanding
For each hypothetical case, apply the control first, respect coverage terms, and check that the allocation reconciles.
- A $84,000 covered loss is reduced by $12,000 through a control. The policy has a $8,000 deductible and a $45,000 per-occurrence limit. Find the insurer’s payment and the organization’s retained loss. (Answer: insurer pays $45,000; organization retains $27,000.)
- A single event can occur at most once per year. Its probability is 0.08 and its covered loss is $120,000. A control reduces event probability to 0.05 without changing severity. The policy has a $10,000 deductible, a $50,000 per-occurrence payment limit, and a $1,200 annual premium; the control costs $2,000 annually. Find expected annual cost with both control and insurance. (Answer: $6,700.)
- After a control, an incident consists of $190,000 of covered loss and $25,000 of excluded loss. Primary coverage has a $15,000 deductible and an $80,000 limit. Excess coverage attaches at $95,000 from the ground up and has a $60,000 limit. Find total insurer payments and total organization retention. (Answer: insurers pay $140,000 in total; the organization retains $75,000.)
- A control reduces a covered event loss from $50,000 to $35,000 but does not change its annual probability of 0.10. Insurance has a $5,000 deductible, a $20,000 per-occurrence limit, and a $500 annual premium. The control costs $800 annually. Find expected annual cost with both measures. (Answer: $2,800.)
- A control changes the annual probability of a covered loss from 0.12 to 0.08 and reduces its severity from $100,000 to $80,000. It costs $2,200 annually. Insurance costs $1,500 annually, has a $10,000 deductible, and pays at most $40,000 per occurrence. Assume at most one event per year and full collectibility. Find expected annual cost with both measures. (Answer: $6,900.)