Maximizing Expected Underwriting Profit

From Expected Contribution to Underwriting Profit

A price that produces positive expected contribution does not necessarily produce positive underwriting profit. Contribution measures premium revenue less expected claims and variable expenses. Underwriting profit also deducts fixed expenses attributable to the business being evaluated. The distinction matters when assessing whether a portfolio or product covers its full costs, and when comparing pricing choices.

The calculations below use a single evaluation period, such as one policy year. A prospect either buys or does not buy, and the expected claims and expenses for a buyer are measured over that same period. When expected costs differ among buyers, use the expected cost conditional on purchase; the prospects who accept a price need not have the same cost profile as those who decline it.

Definition: Expected contribution is expected premium revenue minus expected claims and other variable expenses, before fixed expenses. Expected underwriting profit is expected contribution minus the fixed expenses included in the evaluation. A fixed expense is not incurred in proportion to the number of policies written during the period being modeled; a variable expense changes with premium, policy count, claims, or another specified activity measure.

Suppose there are $n$ prospects. At price $p$, each prospect has purchase probability $q(p)$. For a policy that is purchased, let $C(p)$ be expected claims, $V(p)$ be other variable expenses per policy, and $r(p)$ be a variable expense rate applied to premium. Let $F$ be fixed expenses for the portfolio over the evaluation period. If these quantities are the same for all prospects, then:

\begin{aligned} E[\text{premium revenue}] &= n q(p)p,\\ E[\text{claims}] &= n q(p)C(p),\\ E[\text{other variable expenses}] &= n q(p)V(p),\\ E[\text{premium-based expenses}] &= n q(p)r(p)p,\\ E[\text{contribution}] &= n q(p)\left[p-C(p)-V(p)-r(p)p\right],\\ E[\text{underwriting profit}] &= E[\text{contribution}]-F. \end{aligned}

If expenses include no premium-based component, set $r(p)=0$. Costs must be classified consistently: for example, a per-policy acquisition expense belongs in $V(p)$, while a commission expressed as a percentage of premium belongs in $r(p)p$. Do not count the same expense in both places.

Why the Expected-Profit Formula Holds

The calculation follows from adding each prospect’s expected result. Let $I_i$ equal 1 if prospect $i$ buys and 0 otherwise. For a common premium $p$, the random contribution from that prospect is $I_i$ times the premium less the costs incurred if the policy is written. Since $E[I_i]=q(p)$, the expected contribution from one prospect is the purchase probability multiplied by the expected per-policy contribution among buyers.

E\left[I_i\left(p-C_i-V_i-r(p)p\right)\right] = q(p)\left[p-E[C_i\mid I_i=1]-E[V_i\mid I_i=1]-r(p)p\right].

Summing over $n$ prospects gives the portfolio contribution when their purchase probabilities and conditional expected costs are the same. The calculation does not require purchase decisions to be independent: expectation is additive. If risks have different probabilities or costs, calculate each prospect’s expected contribution separately and sum. Finally, subtract $F$ once if that fixed expense applies to the portfolio for the period.

This conditional formulation is important. If price changes who buys, then the expected claims cost among buyers may also change. The relevant quantity is not automatically the cost of the original prospect pool; it is the expected cost for the risks that purchase at the price being evaluated.

Worked Pricing Comparisons

Worked Example: Comparing Prices After Fixed Expenses

A hypothetical insurer evaluates one-year prices for 1,000 prospects. At a price of $500, the purchase probability is 0.80; at $550 it is 0.70; and at $600 it is 0.55. For every policy written, expected claims are $300 and other variable expenses are $40. A further expense equals 10% of premium. The portfolio has $9,000 of fixed expenses, incurred regardless of the number of policies sold. Calculate expected contribution and underwriting profit at each price, then identify the best of these three choices by expected profit.

Method. For each price, first calculate the expected number of policies, then the contribution per policy, total expected contribution, and finally profit after subtracting the fixed expense once.

At $500, expected policies are $1{,}000(0.80)=800$. The premium-based expense per policy is $0.10(500)=\$50$. Contribution per policy is $500-300-40-50=\$110$, so total expected contribution is $800(110)=\$88{,}000$. Expected underwriting profit is $88{,}000-9{,}000=\$79{,}000$.

At $550, expected policies are $1{,}000(0.70)=700$. The premium-based expense is $0.10(550)=\$55$, so contribution per policy is $550-300-40-55=\$155$. Total expected contribution is $700(155)=\$108{,}500$, and expected underwriting profit is $108{,}500-9{,}000=\$99{,}500$.

At $600, expected policies are $1{,}000(0.55)=550$. The premium-based expense is $0.10(600)=\$60$, so contribution per policy is $600-300-40-60=\$200$. Total expected contribution is $550(200)=\$110{,}000$, and expected underwriting profit is $110{,}000-9{,}000=\$101{,}000$.

Among the three stated prices, $600 gives the greatest expected underwriting profit: $101,000. The fixed expense is the same at all three prices, so it changes each profit level but not their ranking. Common mistake avoided: multiplying the $9,000 fixed expense by the purchase probability would treat it as incurred only when a prospect buys, contrary to the problem statement.

Worked Example: Combining Segments and Allocating Fixed Expense

A hypothetical insurer considers one uniform annual price of $700 for two segments. Segment A has 600 prospects, a purchase probability of 0.60, and expected claims of $380 per policy written. Segment B has 400 prospects, a purchase probability of 0.30, and expected claims of $250 per policy written. Both segments have $30 of other variable expense per policy and a premium-based expense rate of 8%. The portfolio has $20,000 in fixed expenses. Find total contribution and total underwriting profit. Then allocate fixed expenses to the segments in proportion to their expected policy counts and calculate the resulting fully allocated segment profits.

Method. Calculate each segment’s expected policy count and contribution using its own expected claims cost. Add segment contributions, then subtract the portfolio fixed expense. The allocation is a separate reporting step.

The premium-based expense per policy is $0.08(700)=\$56$. Segment A expects $600(0.60)=360$ policies. Its contribution per policy is $700-380-30-56=\$234$, giving total contribution of $360(234)=\$84{,}240$.

Segment B expects $400(0.30)=120$ policies. Its contribution per policy is $700-250-30-56=\$364$, giving total contribution of $120(364)=\$43{,}680$. Total expected policies are $360+120=480$, and total contribution is $84{,}240+43{,}680=\$127{,}920$. Total underwriting profit is $127{,}920-20{,}000=\$107{,}920$.

Segment A accounts for $360/480=0.75$ of expected policies, so its allocated fixed expense is $0.75(20{,}000)=\$15{,}000$. Its fully allocated profit is $84{,}240-15{,}000=\$69{,}240$. Segment B accounts for $120/480=0.25$ of expected policies, so its allocated fixed expense is $0.25(20{,}000)=\$5{,}000$. Its fully allocated profit is $43{,}680-5{,}000=\$38{,}680$. The allocated profits sum to $69{,}240+38{,}680=\$107{,}920$, which checks against total portfolio profit. Common mistake avoided: deducting the $20,000 fixed expense in full from each segment would subtract it twice.

Fixed Expenses and Product Decisions

A fixed expense may be fixed relative to policy count but still depend on a decision. For example, a product may require a one-time setup cost that is incurred only if the product launches. If the launch is optional and the setup cost can be avoided by not launching, it belongs in the total profit comparison between launch and no launch. If the cost has already been incurred and cannot be recovered, it is sunk for the decision; it remains part of historical or reported profit, but it does not change the incremental profit from choosing one price over another.

Worked Example: A Positive Contribution but a Loss-Making Launch

A hypothetical insurer considers launching a one-year specialty product. There are 900 prospects, each with a purchase probability of 0.50 at an annual premium of $1,200. For a policy written, expected claim frequency is 0.10 claims per year and expected severity per claim is $7,000. Other variable expenses are $80 per policy, and commission is 8% of premium. Launching incurs an avoidable fixed setup expense of $160,000 for the year. Calculate expected contribution and underwriting profit, and determine whether launching has positive expected profit.

Method. Calculate expected claims from frequency and severity, then calculate contribution per written policy and multiply by expected policy count. Subtract the avoidable setup expense once.

Expected claims per policy are $0.10(7{,}000)=\$700$. The commission per policy is $0.08(1{,}200)=\$96$. Contribution per policy is $1{,}200-700-80-96=\$324$. Expected policies are $900(0.50)=450$, so expected contribution is $450(324)=\$145{,}800$.

Expected underwriting profit if the insurer launches is $145{,}800-160{,}000=-\$14{,}200$. If not launching avoids the setup cost and produces no contribution, its modeled profit is $0. Thus, under the stated assumptions, launching has negative expected profit relative to not launching. Common mistake avoided: concluding the launch is profitable merely because contribution is positive. Contribution of $145,800 does not cover the $160,000 fixed setup expense.

Accounting Choices and Common Traps

The profit measure is only as sound as its cost definitions. Identify the evaluation period and the relevant portfolio, then decide which expenses are variable and which are fixed for that evaluation. A fixed expense can be allocated to segments for reporting, but an allocation does not change the total expense or total portfolio profit. A different allocation basis may change reported segment profits without changing the portfolio result.

Expected underwriting profit is also not the same as premium revenue, nor is it necessarily the same as a broader economic-profit measure. The calculation above deducts expected claims and the specified expenses. If an analysis separately includes reinsurance costs, taxes, a catastrophe provision, or another explicit cost, it must be included according to the chosen accounting definition and only once. An assumed expense or risk load should not be silently counted both as a premium deduction and as an additional cost.

Multiplying a portfolio fixed expense by take-up. If the expense is incurred for the portfolio regardless of sales, subtract it once, not once per buyer or in proportion to purchase probability.
Subtracting fixed expenses per policy before aggregating. This may be useful for a stated allocation, but it should not replace the direct portfolio calculation unless the allocation sums to the full fixed expense.
Using the wrong expected claims cost. Use the cost conditional on purchase at the price under evaluation when buyer selection changes the risk mix.
Counting a premium-based expense twice. A rate $r$ applied to premium contributes a per-policy cost of $rp$; do not also include that same commission in other variable expenses.
Treating a sunk expense as avoidable. A sunk cost belongs in reported profit but does not affect a decision between alternatives if it is already unavoidable.

Key takeaway: Expected underwriting profit equals expected contribution minus the fixed expenses included in the evaluation. Calculate revenue, expected claims, and variable expenses for the policies expected to be written; subtract portfolio fixed expenses at the appropriate level and only once.

Check Your Understanding

For each problem, calculate expected contribution before fixed expenses and then expected underwriting profit where requested. Use the stated cost assumptions and evaluation period.

A hypothetical program has 800 prospects. At an annual price of $640, purchase probability is 0.45. Expected claims are $390 per written policy, other variable expenses are $25 per policy, premium-based expenses are 5% of premium, and fixed expenses are $9,000. Find expected policies, contribution, and profit. (Answer: 360 policies; contribution $69,480; profit $60,480.)
A hypothetical portfolio charges $500. Segment A has 300 prospects, purchase probability 0.50, and expected claims of $260 per policy. Segment B has 500 prospects, purchase probability 0.20, and expected claims of $340 per policy. Both segments have $20 of variable expense per policy and a 6% premium expense rate. Fixed expenses are $12,000. Find total contribution and profit. (Answer: contribution $39,500; profit $27,500.)
A hypothetical insurer has 2,000 prospects. At a price of $1,500, purchase probability is 0.30. For each buyer, expected claim frequency is 0.06 and expected severity is $10,000. Other variable expenses are $100 per policy, premium-based expenses are 4%, and fixed expenses are $50,000. Find expected claims per policy, contribution, and profit. (Answer: expected claims $600 per policy; contribution $444,000; profit $394,000.)
A hypothetical book has 600 expected policies and total expected contribution of $84,000. Fixed expenses are $30,000 and are allocated to segments in proportion to expected policy counts. What is total underwriting profit, the fixed expense allocation per expected policy, and the fully allocated contribution per expected policy? (Answer: total profit $54,000; fixed allocation $50 per policy; fully allocated contribution $90 per policy.)
A hypothetical insurer has 1,200 prospects and compares two annual prices. At $700, purchase probability is 0.70; at $800, it is 0.50. Expected claims are $350 per written policy at either price, other variable expenses are $40 per policy, premium-based expenses are 10%, and fixed expenses are $25,000. Calculate profit at each price and select the better of the two choices. (Answer: profit at $700 is $176,600; profit at $800 is $173,000; choose $700 among the two stated choices.)

What You'll Learn