Why Capital Enters Product Pricing
A product’s expected claims and operating expenses are not the only considerations in a pricing decision. The insurer also commits financial resources to support the risk. Those resources have an opportunity cost: capital committed to one product cannot be used elsewhere, and the organization may require a return on that capital. A pricing indication can therefore include a capital charge alongside expected losses and expenses.
The key distinction is between the amount of capital allocated to a product and the cost of that capital over the pricing period. If a product is assigned $5,000 of capital per policy and the annual required return is 8%, its one-year capital charge is $400 per policy, not $5,000. Adding the full capital amount to the price confuses a balance-sheet resource with the annual cost of providing that resource.
This framework is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus. It is a cost decomposition, not a universal rule that every premium must equal exactly these three components. Taxes, reinsurance costs, investment income, contingencies, and profit objectives may also matter, but their treatment must be stated and used consistently. The examples below isolate expected loss, expenses, and the capital charge.
Build the Cost Stack on a Consistent Basis
A pricing calculation is meaningful only if its components use compatible units. For a one-year policy, expected losses, expenses, allocated capital, and the capital return rate should all correspond to that policy-year. If a result is calculated per policy, every component should be per policy; if it is calculated for a block, all components should be for that block.
Expected loss is the mean of the loss variable for the pricing unit. For example, if the number of claims \(N\) is independent of claim severities \(Z_j\), and each claim has the same finite mean severity, then for aggregate loss \(X=\sum_{j=1}^{N}Z_j\),
This identity does not require a Poisson claim count; it requires that the claim count and severities satisfy the stated independence and identical-mean assumptions. A different loss model may require a different calculation of \(E[X]\). In every case, the capital charge is calculated from the capital assigned to the relevant pricing unit, not from expected loss and not from the total premium.
Decide whether the calculation is per policy, per exposure, or for a portfolio, and make the time horizon explicit.
Use assumptions that match the unit and period. Do not substitute a risk measure or capital amount for expected loss.
Use the capital allocated to the product or pricing unit under the chosen allocation basis. Do not silently replace it with standalone capital.
Multiply allocated capital by the annual capital return rate for a one-year period, then add this charge to the other stated costs.
Any difference is the residual after the included costs; it should not be mislabeled as capital or as an already included expense.
Why the Capital Charge Is a Rate Times Capital
Suppose the organization commits \(K\) dollars of capital to a product for one year and requires an annual return of \(k\) dollars per dollar of capital. The required earnings attributable to that capital are \(k\) for each dollar committed. Summing over \(K\) dollars gives the annual capital cost:
More precisely, the rate interpretation means that each dollar of capital carries a required annual earnings amount of \(k\) dollars, so \(K\) dollars carry \(Kk\) dollars of required annual earnings. The charge is an opportunity-cost allowance, not a repayment of the capital principal. The principal remains capital supporting the business; it is not ordinarily consumed as an annual claim or expense.
If a quoted premium is \(P\), a useful reconciliation for this simplified one-year basis is:
where \(r\) is the residual after the three specified components. It may represent a targeted profit amount or other omitted pricing elements, depending on the model. This identity is an accounting decomposition, not a claim that \(r\) is necessarily positive or that the premium is adequate under every other consideration. When expenses depend on premium, or the capital rate and holding period differ from the one-year assumptions, the calculation needs a consistent treatment of those features rather than a mechanical application of this simple form.
Worked Example: A One-Year Policy Cost Indication
A hypothetical insurer prices a one-year equipment policy. Claim count \(N\) is Poisson with mean \(0.12\), and each claim has a fixed severity of $25,000. Claim severities are independent of claim count. Expenses are $420 per policy-year. The policy is assigned $4,000 of capital, and the required annual return is 8%. A proposed premium is $4,250. Calculate the cost indication from the stated components and the residual at the proposed premium.
Method. Calculate expected aggregate loss, calculate the annual capital charge, and add both to expenses. This keeps the capital amount separate from its cost.
For a Poisson count with mean \(0.12\), \(E[N]=0.12\). The severity mean is \(E[Z_1]=\$25{,}000\). By the aggregate-loss expectation formula,
The capital charge is 8% of the assigned $4,000:
The cost indication is therefore
At the proposed premium, the residual after these components is
Check. The components plus residual total $3,000 + $420 + $320 + $510 = $4,250, the quoted premium. The common mistake is to add the full $4,000 of capital instead of its $320 annual cost. That incorrect calculation would give a cost indication of $7,420, treating a capital resource as if it were an annual expense.
Use Allocated Capital for the Pricing Decision
Product pricing often uses a capital allocation: a share of portfolio capital assigned to a product under a specified method. The relevant capital charge for that pricing decision is the required return applied to the assigned amount. A product’s standalone capital may differ from its allocated capital because the product is part of a diversified portfolio. Replacing the allocated amount with standalone capital changes the basis of the decision.
Earlier material on capital allocation covers methods for assigning portfolio capital and the importance of reconciling assigned amounts. For pricing, a practical check is that charges across products add to the charge on the allocated portfolio total when the same rate and period apply:
where \(K_j\) is the amount allocated to product \(j\), and the allocations fully sum to the portfolio capital \(K_{\text{portfolio}}\). If different capital rates are applied to different products, the product charges need not equal a single common rate times total capital; the rate assumptions must be made explicit.
Worked Example: Pricing From a Portfolio Allocation
A hypothetical insurer has $1,200,000 of portfolio capital allocated across two one-year products. Product A receives $450,000 across 300 policies. Product B receives the remaining $750,000 across 500 policies. The same 10% annual required return applies to both allocations. For a Product A policy, claim count is Bernoulli with probability 0.04 of one claim, and the claim severity is fixed at $20,000. Expenses are $125 per policy-year. If the proposed premium is $1,150, calculate the per-policy cost indication and residual. Also check the portfolio capital charge.
Method. Use allocated capital per policy for Product A, calculate expected loss using the stated one-claim model, and then reconcile aggregate product charges to the portfolio charge.
Product A capital per policy is
For a Bernoulli claim count, the expected claim count is the claim probability, \(0.04\). Thus the expected loss per policy is
The per-policy capital charge is \(0.10(\$1{,}500)=\$150\). Adding the stated components gives
The residual at the proposed premium is \(\$1{,}150-\$1{,}075=\$75\) per policy. For the portfolio check, Product A’s annual charge is \(300(\$150)=\$45{,}000\). Product B has \(750{,}000/500=\$1{,}500\) allocated capital per policy, so its charge is \(500(0.10)(\$1{,}500)=\$75{,}000\). Together they produce \(\$45{,}000+\$75{,}000=\$120{,}000\), which equals \(0.10(\$1{,}200{,}000)=\$120{,}000\).
Check. The per-policy amount uses allocated capital, not Product A’s total allocation of $450,000. The aggregate charges also reconcile exactly to the common-rate charge on the full portfolio allocation.
Comparing Capital Charges Across Product Tiers
The allocated capital per policy can vary by product tier, even when the same capital return rate applies. That means the capital charge can differ between tiers independently of their expected loss and expenses. Pricing comparisons should therefore show the components separately. Otherwise, a larger capital charge might be mistaken for a larger expected claim cost, or a product with lower expected loss might appear cheaper without accounting for its capital requirement.
In the next example, capital is allocated by tier and the pricing calculation is performed for each policy. The portfolio totals provide a useful arithmetic check.
Worked Example: Two Tiers With Different Allocations
A hypothetical one-year property product has 200 high-risk policies and 300 low-risk policies. The high-risk tier receives $600,000 of allocated capital, and the low-risk tier receives $300,000. The required annual return is 9% for both tiers. For a high-risk policy, the probability of one claim is 0.06 and its severity is fixed at $30,000; expenses are $250. For a low-risk policy, the probability of one claim is 0.025 and severity is fixed at $28,000; expenses are $160. Proposed premiums are $2,500 and $1,100, respectively. Calculate each tier’s cost indication and residual, then check the aggregate capital charge.
Method. Divide each tier’s allocated capital by its policy count, calculate expected loss from the Bernoulli claim model, and apply the same annual return rate to each per-policy capital amount.
High-risk allocated capital per policy is \(\$600{,}000/200=\$3{,}000\). Expected loss is \(0.06(\$30{,}000)=\$1{,}800\), and the capital charge is \(0.09(\$3{,}000)=\$270\). Its cost indication is
The high-risk residual is \(\$2{,}500-\$2{,}320=\$180\) per policy.
Low-risk allocated capital per policy is \(\$300{,}000/300=\$1{,}000\). Expected loss is \(0.025(\$28{,}000)=\$700\), and the capital charge is \(0.09(\$1{,}000)=\$90\). Its cost indication is
The low-risk residual is \(\$1{,}100-\$950=\$150\) per policy. The high-risk tier’s total capital charge is \(200(\$270)=\$54{,}000\). The low-risk tier’s is \(300(\$90)=\$27{,}000\). The total is $81,000, equal to \(0.09(\$600{,}000+\$300{,}000)=\$81{,}000\).
Check. Each per-policy premium reconciles to its expected loss, expenses, capital charge, and residual. The portfolio charge reconciles to the required return on the full $900,000 allocation.
Common Mistakes and Practical Checks
Capital charging is straightforward arithmetic, but several modeling choices can create material errors. Before relying on a price indication, check the capital basis, rate, period, and treatment of other components.
- Adding capital instead of its cost. If allocated capital is \(K\), a one-year charge at rate \(k\) is \(kK\), not \(K\). Capital is the resource supporting risk; the charge is the required return on that resource.
- Using the wrong capital amount. Use the allocated amount when the pricing basis calls for allocated capital. Standalone capital and allocated capital answer different questions and should not be substituted without changing the stated basis.
- Mixing time periods. A one-year required return should not be applied without adjustment or explanation to capital held for a different period. Match the rate convention and the pricing horizon.
- Mixing units. Do not add a portfolio-level capital charge to per-policy expected losses. Convert all components to a common basis before summing.
- Counting an item twice. If a stated expense or other loading already includes a capital-related amount, adding the same amount again overstates the cost stack. State which components are included.
- Treating the residual as automatically adequate profit. The residual is simply premium less the specified components. Its interpretation depends on which other costs, taxes, and objectives have or have not been included.
A useful final review is to multiply the per-unit capital charge by the number of units, then compare the result with the rate applied to total allocated capital. When the allocation fully sums to portfolio capital and the same rate applies, those totals should agree. This check can reveal a missing exposure count, a misplaced decimal, or use of the wrong capital allocation.
Check Your Understanding
For each hypothetical one-year pricing problem, calculate the requested quantity using only the stated components.
- A product has a Poisson claim count with mean 0.10 and fixed claim severity of $18,000. Expenses are $210 per policy, allocated capital is $2,500 per policy, and the required annual return is 8%. Find the cost indication. (Answer: $2,210)
- A policy has probability 0.03 of one claim, with fixed severity $40,000. Expenses are $275, allocated capital is $5,000, and the annual required return is 6%. Find the cost indication. (Answer: $1,775)
- A product receives $250,000 of allocated capital across 50 policies. Each policy has expected loss $900 and expenses of $100. The required annual return is 7%, and the quoted premium is $1,500 per policy. Find the capital charge and residual per policy. (Answer: $350 capital charge; $150 residual)
- A portfolio allocates $2,000 of capital to each of 120 policies in one tier and $5,000 to each of 80 policies in another. At an 8.5% annual required return, find each tier’s total capital charge and the portfolio total. (Answer: $20,400 and $34,000; total $54,400)
- A one-year policy has expected loss $2,600, expenses $340, allocated capital $12,000, and a 9% required annual return. Its quoted premium is $4,000. Find the cost indication and state whether the quote exceeds it. (Answer: $4,020; the quote is $20 below the indication)