Contribution-Based Allocation Among Policies

Purpose of Contribution-Based Allocation

Once an insurer has determined the surplus available for allocation, a further question is how to divide it among groups of participating policies. A contribution-based method links each group’s share to its measured contribution to the surplus experience. The method is straightforward to calculate, but only when the contribution basis is defined consistently and the underlying data support the attribution.

The contribution base and the distributable pool are not necessarily the same amount. The contribution base measures the relative weights of the groups; the pool is the amount to allocate. For example, some surplus may be retained, or the amount available to policyholders may have been set separately. The contribution weights determine proportions, not the size of the pool.

Definition: Suppose a surplus pool $D$ is to be allocated among $n$ policy groups. Let $C_j$ be the measured contribution of group $j$ on the selected attribution basis, and let $C=\sum_{j=1}^{n}C_j$. If each $C_j$ is nonnegative and $C>0$, the contribution weight of group $j$ is $w_j=C_j/C$, and its allocation is $D_j=Dw_j$. Thus each group receives the same proportion of the pool as its proportion of the total contribution base.

Contributions can be built from components such as mortality or morbidity experience, expenses, investment experience, lapse experience, and changes in reserves, provided those components are defined consistently. A favorable contribution may be positive and an unfavorable contribution negative. The relevant group total is the signed sum of its included components, not the sum of their absolute values.

The basic proportional formula assumes a nonnegative contribution total for each group and a positive overall contribution base. If a group total is negative, the formula can produce a negative allocation; if the total base is zero, it is undefined. Those cases require an explicit allocation rule rather than an automatic use of the ordinary proportional formula.

Why the Proportional Formula Works

The contribution principle gives each group a common allocation rate per unit of contribution. Let that rate be $k$, so group $j$ receives $D_j=kC_j$. Requiring the allocations to exhaust the pool means their sum must equal $D$. Summing the group allocations gives $k\sum_j C_j=D$, which determines the common rate and therefore the allocation formula.

\sum_{j=1}^{n}D_j=D \quad\Longrightarrow\quad k\sum_{j=1}^{n}C_j=D \quad\Longrightarrow\quad k=\frac{D}{C} \quad\Longrightarrow\quad D_j=\frac{D}{C}C_j=D\frac{C_j}{C}.

This justification shows both the appeal and the limitation of the method. It automatically reconciles to the pool when the contributions sum to a positive total, and it gives a simple interpretation to the allocation rate. But it does not establish that the chosen contribution measure is appropriate: that depends on the data and the attribution decisions made before the calculation.

Worked Applications

Worked Example: Allocating a Pool from Signed Experience Components

A hypothetical participating fund has determined that $ \$22{,}000 $ is available for allocation to two policy groups. On a consistent basis, the groups’ contribution components, in thousands of dollars, are as follows: group G has claim contribution 8, expense contribution 3, investment contribution 4, and reserve-change contribution $-1$; group H has corresponding contributions 5, 2, 3, and 0. Allocate the pool by the contribution principle.

Method. First sum the signed components for each group. Then calculate the groups’ shares of the total contribution base and apply those shares to the $ \$22{,}000 $ pool. The negative reserve-change contribution for G reduces its net contribution; it is not discarded.

Group G contribution. In thousands, $C_G=8+3+4-1=14$.

Group H contribution. In thousands, $C_H=5+2+3+0=10$.

The total contribution base is $C=14+10=24$ thousand. The weights are $w_G=14/24=7/12$ and $w_H=10/24=5/12$. They sum to $7/12+5/12=1$.

Allocations. Group G receives $ \$22{,}000(14/24)=\$12{,}833.33$, and group H receives $ \$22{,}000(10/24)=\$9{,}166.67$, rounded to the nearest cent.

Check. The allocations total $ \$12{,}833.33+\$9{,}166.67=\$22{,}000.00$. The method avoids the common mistake of assigning weights of 0.60 and 0.40: those weights do not follow from contributions of 14 and 10, whose correct shares are $14/24$ and $10/24$.

Worked Example: Combining Experience Components Across Three Groups

A hypothetical insurer allocates a $ \$50{,}000 $ pool among groups A, B, and C. Each group’s contributions, in thousands of dollars, are measured on the same basis. A has investment contribution 18, lapse contribution $-2$, and expense contribution 3. B has corresponding contributions 8, 4, and 1. C has corresponding contributions 6, $-1$, and 3. Find the allocation.

Method. Net the signed components within each group before calculating any proportions. In particular, negative experience reduces the relevant group’s contribution; it is not a separate negative weight if the group’s net total remains positive.

For A, $C_A=18-2+3=19$ thousand. For B, $C_B=8+4+1=13$ thousand. For C, $C_C=6-1+3=8$ thousand. The total contribution base is $C=19+13+8=40$ thousand.

The respective weights are $19/40=0.475$, $13/40=0.325$, and $8/40=0.200$. They sum to $0.475+0.325+0.200=1.000$.

The allocations are:

D_A=\$50{,}000\left(\frac{19}{40}\right)=\$23{,}750,\qquad D_B=\$50{,}000\left(\frac{13}{40}\right)=\$16{,}250,\qquad D_C=\$50{,}000\left(\frac{8}{40}\right)=\$10{,}000.

Check. The amounts sum to $ \$23{,}750+\$16{,}250+\$10{,}000=\$50{,}000$. All net contributions are positive and the total contribution base is $ \$40{,}000$, even though two component entries are negative. The common error to avoid is taking absolute values of negative components, which would change the agreed contribution measure.

Worked Example: Converting Per-Policy Experience into Contribution Weights

A hypothetical fund has a $ \$36{,}000 $ pool to allocate among three policy groups. Its experience analysis assigns a positive surplus contribution of $ \$20 $ per policy to group X, $ \$15 $ per policy to group Y, and $ \$10 $ per policy to group Z. The respective numbers of policies in the measured groups are 900, 600, and 300. Calculate each group’s allocation.

Method. The contribution base is the per-policy contribution multiplied by the matching group count. Do not use the per-policy amounts alone as the group weights, because the groups contain different numbers of policies.

Group X contributes $900(\$20)=\$18{,}000$. Group Y contributes $600(\$15)=\$9{,}000$. Group Z contributes $300(\$10)=\$3{,}000$. The total contribution base is $ \$18{,}000+\$9{,}000+\$3{,}000=\$30{,}000$.

The weights are $18/30=0.60$, $9/30=0.30$, and $3/30=0.10$. Applying them to the $ \$36{,}000 $ pool gives X $ \$36{,}000(0.60)=\$21{,}600$, Y $ \$36{,}000(0.30)=\$10{,}800$, and Z $ \$36{,}000(0.10)=\$3{,}600$.

Check. The allocations total $ \$21{,}600+\$10{,}800+\$3{,}600=\$36{,}000$. The pool is larger than the contribution base, but that does not prevent a proportional allocation: the base supplies the relative weights, while the specified pool supplies the dollars allocated.

Data Needed to Support an Allocation

A calculation is only as credible as its contribution basis. Before calculating the weights, an actuary needs enough information to determine which policies belong to each group, what experience items are attributed to those policies, and how each item is measured. The basis should also match across groups: comparing contributions measured over different periods or with different expense definitions would make the resulting weights difficult to interpret.

Specify the groups and measurement period.
Identify the policies included in each group and the period over which the surplus experience is measured. Define how policies entering, leaving, or changing groups are treated.

Define the contribution measure.
List the included sources, such as claims, expenses, investment experience, lapses, or reserve changes, and state the sign convention and valuation basis for each.

Assemble and reconcile the data.
Use policy-level or group-level data that can be traced to the specified experience totals. Confirm that common items are not omitted or counted twice.

Calculate weights and allocations.
Sum each group’s signed contributions, verify the total base is positive, apply the weights to the specified pool, and check that the allocations sum to that pool.

Depending on the chosen basis, useful information may include policy counts or exposure measures, premiums, claims, expenses, investment amounts, reserve movements, and rules for attributing shared costs or experience. The relevant items are determined by the contribution definition, not by a fixed universal list. If a source cannot be attributed reliably to groups, it should not be silently assigned in a way that implies unsupported precision.

Conditions, Interpretation, and Common Traps

The ordinary formula is most direct when every group’s net contribution is nonnegative and the total is positive. A negative component is not itself a problem; as the examples show, it can be included in a positive net group total. A negative net group contribution is different: proportional allocation would give that group a negative amount, which may not be meaningful as a distribution. The actuary must then refer to an explicit governing allocation rule rather than force the ordinary method to produce nonnegative shares.

Likewise, if all contributions sum to zero, contribution weights cannot be calculated. If the contribution totals are incomplete or measured on inconsistent bases, arithmetic precision does not make the allocation well supported. These are data and method limitations, not reasons to replace missing information with policy counts or another convenient proxy without justification.

Using the pool as the denominator. The denominator for the contribution weights is the total contribution base $C$, not the distributable pool $D$.
Forgetting signed components. A loss or unfavorable item reduces a group’s net contribution; it should not be omitted or converted to a positive value unless the defined basis requires that treatment.
Confusing per-policy and group contributions. A per-policy amount must be combined with the relevant policy count or exposure when the intended base is the group total.
Ignoring reconciliation. The weights should sum to 1, and the allocations should sum to the pool, subject only to stated rounding.

Key takeaway: Contribution-based allocation assigns each group the same share of the distributable pool as its share of a defined, positive total contribution base. The formula is simple; the essential actuarial work is defining, measuring, and reconciling the contributions consistently.

Check Your Understanding

For each problem, assume the stated contribution basis is appropriate and allocate the entire specified pool by the contribution principle unless the question asks otherwise.

Groups R and S have net contributions of $ \$12{,}000 $ and $ \$8{,}000 $. Allocate a $ \$25{,}000 $ pool. (Answer: R receives $ \$15{,}000 $; S receives $ \$10{,}000 $.)
Group A’s components, in thousands, are 9, 2, and $-1$; group B’s are 7, 1, and 2. Allocate a $ \$18{,}000 $ pool. (Answer: each group receives $ \$9{,}000 $.)
Three groups have net contributions of $ \$15{,}000 $, $ \$9{,}000 $, and $ \$6{,}000 $. Allocate a $ \$24{,}000 $ pool. (Answer: $ \$12{,}000 $, $ \$7{,}200 $, and $ \$4{,}800 $, respectively.)
Groups P, Q, and R have contributions of $ \$14{,}000 $, $ \$8{,}000 $, and $ \$3{,}000 $. Allocate a $ \$50{,}000 $ pool. (Answer: P receives $ \$28{,}000 $; Q receives $ \$16{,}000 $; R receives $ \$6{,}000 $.)
Groups X, Y, and Z have respective policy counts of 600, 500, and 400, and contributions per policy of $ \$20 $, $ \$12 $, and $ \$5 $. Allocate a $ \$31{,}000 $ pool using total group contributions. (Answer: X receives $ \$18{,}600 $; Y receives $ \$9{,}300 $; Z receives $ \$3{,}100 $.)

What You'll Learn