Fair Allocation Across Policy Classes

Why Class-Level Fairness Needs a Benchmark

An allocation can be fair in total and still treat policy classes inconsistently. A proposal may distribute the entire surplus pool while directing too much to a class that contributed relatively little to the surplus available for allocation. Alternatively, a class may have a contractual minimum that a purely contribution-based formula would fail to respect. A useful assessment therefore starts with both guarantees and contribution, then checks whether the proposed amounts are consistent with those features.

There is no single formula that defines fairness in every participating arrangement. Contracts, applicable requirements, the insurer’s stated allocation principles, and the nature of the surplus all matter. The calculation below is an analytical benchmark: it helps identify differences that need explanation, but it does not by itself determine a legal entitlement or replace the terms of a contract.

Definition: Let $A$ be the total amount available for allocation among $n$ policy classes. For class $j$, let $g_j$ be its minimum allocation arising from guarantees or other binding commitments, and let $c_j$ be its nonnegative contribution measure for the surplus being allocated. The residual surplus is $R=A-\sum_{j=1}^{n}g_j$. If $R\geq 0$ and $\sum_j c_j>0$, a contribution-weighted benchmark assigns class $j$ the amount $b_j=g_j+R c_j/\sum_k c_k$.

The contribution measure should be defined consistently with the surplus pool. It may reflect class-attributable experience for the period, such as investment, mortality or morbidity, expenses, and lapse experience, with favorable and unfavorable effects included using the same conventions. A contribution measure is not automatically the same as premium volume, policy count, or reserve size. Those quantities may help explain contribution, but using them as a proxy requires justification.

Building and Interpreting the Benchmark

The method has two distinct stages. First, account for binding minimums. Second, divide only the residual surplus in proportion to the selected contribution measures. Separating these stages prevents a guarantee from being mistaken for a contribution to this period’s surplus.

Define the pool and period.
Specify the amount available, the classes included, and the period to which the surplus and experience relate. The pool and contribution measures must use compatible definitions.

Identify class minimums.
Set out each class’s guaranteed or otherwise binding minimum allocation. Check that their sum does not exceed the pool.

Measure class contributions.
Attribute relevant favorable and unfavorable experience to classes using a consistent method. Explain any material exclusions or adjustments.

Allocate the residual and compare.
Calculate the benchmark, then compare each proposed allocation with its benchmark amount and with the class’s minimum.

The benchmark has two useful checks. First, it must exhaust the pool: $\sum_j b_j=A$. Second, after removing minimums, the benchmark residual allocations must be proportional to the contribution measures: $ (b_j-g_j)/c_j=R/\sum_k c_k$ whenever $c_j>0$. A proposal that differs from the benchmark is not automatically unfair, but the size and direction of the difference should have a reason grounded in the arrangement’s principles.

\sum_{j=1}^{n} b_j = \sum_{j=1}^{n} g_j+ R\frac{\sum_{j=1}^{n}c_j}{\sum_{j=1}^{n}c_j} = \sum_{j=1}^{n}g_j+R =A.

This identity also gives a short justification for the formula: each class receives its minimum plus its share of the residual, and the contribution shares sum to one. Thus the calculation preserves every minimum and allocates exactly the available amount, provided the stated conditions hold.

If all contribution measures are zero, proportional allocation is undefined; another stated basis is needed. If a measure is negative, applying the formula mechanically can produce an allocation below the class minimum. Negative experience should be handled through an explicit risk-sharing or attribution rule rather than hidden by an unsuitable weight. If the sum of the minimums exceeds the pool, the proposed pool cannot meet all minimums; there is no nonnegative residual to allocate by this method.

Worked Example: Comparing a Proposal With a Two-Class Benchmark

A hypothetical insurer has $500,000 available for two policy classes. Binding minimum allocations are $80,000 for Class A and $120,000 for Class B. Their measured contributions to the surplus being distributed are $300,000 and $100,000, respectively. A proposed allocation is $300,000 to A and $200,000 to B. Assess it against the contribution-weighted benchmark.

Method. Deduct the minimums first, then allocate the remaining pool in the ratio of contributions. This avoids applying contribution weights to amounts already required by the minimums.

The minimums total $80{,}000+120{,}000=200{,}000$. Therefore the residual is $500{,}000-200{,}000=300{,}000$. Total contributions are $300{,}000+100{,}000=400{,}000$, so the contribution shares are $300{,}000/400{,}000=0.75$ for A and $100{,}000/400{,}000=0.25$ for B.

A’s residual share is $300{,}000(0.75)=225{,}000$, giving a benchmark of $80{,}000+225{,}000=\$305{,}000$. B’s residual share is $300{,}000(0.25)=75{,}000$, giving a benchmark of $120{,}000+75{,}000=\$195{,}000$. The benchmark totals $305{,}000+195{,}000=\$500{,}000$.

The proposal also totals $500,000. Its deviations from the benchmark are $300{,}000-305{,}000=-\$5{,}000$ for A and $200{,}000-195{,}000=+\$5{,}000$ for B. The deviations sum to zero, as they must when both allocations exhaust the same pool. Relative to this benchmark, the proposal transfers $5,000 from A to B; the proposal would need a separate rationale for that departure.

Common mistake avoided: Do not describe A’s deviation as positive $5,000: proposal minus benchmark is $300{,}000-305{,}000=-5{,}000$, not $+5{,}000$. The sign identifies which class receives less or more than the benchmark.

Experience and Duration in a Class Comparison

A class’s contribution measure should reflect the experience relevant to the surplus being distributed, not just a single favorable component. If investment experience is favorable but morbidity experience is unfavorable, both effects should be considered under the chosen attribution method. Otherwise, a class may appear to have generated surplus simply because an adverse component was omitted.

Duration also requires care. A class with longer-standing policies may have different guarantees, reserve patterns, exposure, or experience from a newer class. Those differences can matter when measuring contribution or identifying minimum commitments. But elapsed time by itself does not establish that a class should receive a larger share of current surplus. The analyst should identify the specific duration-related feature and explain how it affects the pool, guarantees, or contribution measure.

Worked Example: Combining Experience With Different Class Durations

A hypothetical insurer allocates a $400,000 pool among three classes. Their minimum allocations are $50,000, $70,000, and $30,000. For this pool, the attribution method produces the following class contributions: Class A has $50,000 favorable investment experience, $10,000 unfavorable morbidity experience, and $20,000 favorable lapse experience; Class B has $30,000 favorable investment, $15,000 favorable morbidity, and $5,000 unfavorable lapse experience; Class C has $20,000 favorable investment, $5,000 unfavorable morbidity, and $10,000 favorable lapse experience. Their average policy durations are 12, 5, and 2 years. The proposed allocations are $160,000, $160,000, and $80,000. Assess the proposal.

Method. Net the stated experience components within each class, preserve the class minimums, and allocate the residual by the resulting nonnegative contributions. Treat duration as relevant context, not as an additional weight unless a stated allocation principle supports doing so.

Class A’s contribution is $50{,}000-10{,}000+20{,}000=\$60{,}000$. Class B’s is $30{,}000+15{,}000-5{,}000=\$40{,}000$. Class C’s is $20{,}000-5{,}000+10{,}000=\$25{,}000$. Total contributions are $60{,}000+40{,}000+25{,}000=\$125{,}000$.

The minimums total $50{,}000+70{,}000+30{,}000=\$150{,}000$, leaving $400{,}000-150{,}000=\$250{,}000$ residual. The contribution shares are $60/125=0.48$, $40/125=0.32$, and $25/125=0.20$. The residual allocations are therefore $120,000, $80,000, and $50,000. Adding the minimums gives benchmarks of $170,000 for A, $150,000 for B, and $80,000 for C. These total $400,000.

The proposal’s deviations, stated as proposal minus benchmark, are $-\$10{,}000$ for A, $+\$10{,}000$ for B, and $0 for C. The deviations sum to zero. A has the longest average duration, but that fact alone does not establish that its benchmark should be increased or that B’s allocation is justified. The proposal’s shift from A to B calls for an explanation, such as a relevant guarantee or an allocation principle not represented in the stated contribution measures.

Common mistake avoided: Do not add a separate duration multiplier just because classes have different average policy durations. First determine whether duration changes a guarantee, exposure, or contribution to the pool; otherwise the multiplier would introduce an unsupported second weighting basis.

A Frequent Error: Weighting the Whole Pool

Contribution weights apply to the residual only in this benchmark. Applying them to the whole pool and then adding minimums double-counts part of the amount. Ignoring minimums altogether can also give a result that fails to meet a binding allocation. The next example isolates this error.

Worked Example: Correcting a Whole-Pool Weighting Error

A hypothetical $240,000 pool is allocated between two classes. Each class has a minimum of $60,000. Their contributions are $90,000 for Class A and $30,000 for Class B. A proposed calculation allocates the entire pool using contribution shares, producing $180,000 for A and $60,000 for B. Assess the calculation against the benchmark.

Method. Apply the contribution shares to the pool remaining after the minimums, not to the full pool. Then compare the proposal with the resulting class amounts.

The minimums total $60{,}000+60{,}000=\$120{,}000$, so the residual is $240{,}000-120{,}000=\$120{,}000$. Contributions total $90{,}000+30{,}000=\$120{,}000$, giving shares $90/120=0.75$ and $30/120=0.25$. The residual allocations are $120{,}000(0.75)=\$90{,}000$ to A and $120{,}000(0.25)=\$30{,}000$ to B. The benchmark totals are therefore $150,000 for A and $90,000 for B.

The proposed amounts, $180,000 and $60,000, do total $240,000 and satisfy both minimums. However, compared with the benchmark, A receives $180{,}000-150{,}000=+\$30{,}000$ and B receives $60{,}000-90{,}000=-\$30{,}000$. Thus the proposal shifts $30,000 from B to A relative to the benchmark.

Common mistake avoided: Weighting the whole $240,000 gives $180,000 and $60,000, but this applies contribution weights to amounts that include the two minimums. The corrected benchmark first reserves $120,000 for those minimums and weights only the remaining $120,000.

How to Report a Fairness Assessment

A useful conclusion separates arithmetic from judgment. State whether the pool is fully allocated, whether each minimum is met, and how the proposed amount for each class compares with the benchmark. Then explain which features could justify a departure. Depending on the arrangement, relevant considerations may include contractual participation rights, experience attribution, class exposure, or a duration-related guarantee. A difference without an identified reason is a point for investigation, not proof by itself of unfair treatment.

Keep the comparison at the class level consistent with the underlying allocation principles. In particular, do not compare one class’s proposed total with another class’s residual share, or mix contribution measures from different periods. The benchmark is meaningful only when the pool, minimums, and contributions refer to compatible classes and periods.

Key takeaway: A transparent class-level comparison preserves binding minimums, allocates the remaining pool using a clearly defined contribution basis, and reports each proposal-minus-benchmark difference. Experience and duration can inform the analysis, but every adjustment needs a stated connection to the guarantees or surplus being allocated.

Check Your Understanding

For each problem, calculate the benchmark where the stated conditions permit it. In comparison problems, define deviations as proposal minus benchmark.

A $260,000 pool has minimums of $40,000 and $50,000. The classes’ contributions are $51,000 and $34,000. Find the two benchmark allocations. (Answer: $142,000 and $118,000.)
A $500,000 pool has minimums of $100,000, $50,000, and $50,000. Contributions are $80,000, $40,000, and $30,000. Find the benchmark for each class and verify the total. (Answer: $260,000, $130,000, and $110,000; total $500,000.)
A $300,000 pool has minimums of $30,000 and $60,000, with contributions of $70,000 and $30,000. A proposal assigns $180,000 and $120,000. Find the benchmark and the two deviations. (Answer: Benchmark $177,000 and $123,000; deviations +$3,000 and −$3,000.)
A $420,000 pool has minimums of $60,000, $90,000, and $30,000. Net experience contributions are $60,000, $40,000, and $20,000. The proposed allocations are $170,000, $180,000, and $70,000, and average durations are 10, 6, and 2 years. Find the benchmark and explain whether duration alone changes it. (Answer: Benchmark $180,000, $170,000, and $70,000; deviations −$10,000, +$10,000, and $0; duration alone does not change the benchmark.)
A $180,000 pool has binding minimums of $80,000, $70,000, and $50,000. Can a nonnegative residual be allocated using contribution weights? (Answer: No. Minimums total $200,000, exceeding the pool by $20,000; the stated pool cannot meet all minimums.)

What You'll Learn