Tutorials › CAS Exam 9 - Risk Management for Actuaries › Worked Proportional Allocations Across Business Units

CAS Exam 9 · Capital Allocation Methods · Tutorial 354 of 995

Worked Proportional Allocations Across Business Units

Apply proportional weights to allocate aggregate capital across business units, then check and reconcile the resulting amounts when reported figures are rounded.

Advanced 12 min read

What You'll Learn

  • Define proportional weights from each unit’s standalone capital and the portfolio total.
  • Calculate allocations from an aggregate capital target and verify full reconciliation.
  • Handle zero standalone capital and a zero aggregate target without invalid ratios.
  • Reconcile rounded allocations to a rounded aggregate using fractional remainders.
  • Identify arithmetic and rounding traps that can misstate unit allocations.

Why Proportional Allocation Requires Care

An organization may set an aggregate capital target for a portfolio and need to assign that target across business units. A proportional allocation uses each unit’s standalone capital as its weight. The calculation is direct, but a sound result still depends on using consistent units, checking the denominator, and reconciling both exact and rounded amounts.

The rule is an allocation convention: it distributes a specified target according to standalone capital. It does not assert that the target equals the sum of standalone capital, nor does it explain how the target was determined. The calculations below focus on applying and checking the rule.

Definition: For \(n\) business units, let \(K_i\) be the nonnegative standalone capital for unit \(i\), and let \(C\) be the nonnegative aggregate capital target to allocate. If \(K=\sum_{j=1}^{n}K_j>0\), the proportional weight and allocated capital for unit \(i\) are $$ w_i=\frac{K_i}{K}, \qquad C_i=Cw_i=C\frac{K_i}{\sum_{j=1}^{n}K_j}. $$ The weights sum to one, and the allocations sum to the target. If \(K=0\), the proportional weights are undefined; another allocation basis or an explicit rule is needed.

Deriving the Allocation Rule

The formula follows from two requirements. First, each allocation is proportional to its unit’s standalone capital, so all units use the same nonnegative scaling factor \(\lambda\): \(C_i=\lambda K_i\). Second, the allocations must sum to the target. Summing the first relationship gives \(C=\lambda\sum_i K_i\). When \(\sum_iK_i>0\), solving for \(\lambda\) gives the proportional allocation formula.

$$ \sum_{i=1}^{n} C_i =\sum_{i=1}^{n}\lambda K_i =\lambda\sum_{i=1}^{n}K_i =C \quad\Longrightarrow\quad \lambda=\frac{C}{\sum_{i=1}^{n}K_i}, \qquad C_i=C\frac{K_i}{\sum_{j=1}^{n}K_j}. $$

This derivation also shows why the rule reconciles before rounding: the common factor is chosen precisely so that the allocated amounts sum to \(C\). A unit with \(K_i=0\) receives \(C_i=0\). If \(C=0\) while total standalone capital is positive, all allocations are zero. In either case, division by a unit’s allocated amount is not needed to express proportionality: \(C_iK_j=C_jK_i\) holds for every pair of units.

Worked Example: Allocating a Target Across Four Units

A hypothetical insurer has standalone capital of $6.4 million, $4.8 million, $2.8 million, and $2.0 million for four business units. Allocate an aggregate target of $12.8 million in proportion to those figures. Show the weights and verify the total.

Method. First total the standalone capital, in millions of dollars: \(6.4+4.8+2.8+2.0=16.0\). The target-to-total factor is \(12.8/16.0=0.8\).

The weights are \(6.4/16.0=0.40\), \(4.8/16.0=0.30\), \(2.8/16.0=0.175\), and \(2.0/16.0=0.125\). They sum to \(1.00\).

Multiplying each standalone amount by the common factor gives:

$$ \begin{aligned} C_1&=6.4(0.8)=5.12\text{ million},\\ C_2&=4.8(0.8)=3.84\text{ million},\\ C_3&=2.8(0.8)=2.24\text{ million},\\ C_4&=2.0(0.8)=1.60\text{ million}. \end{aligned} $$

Check. \(5.12+3.84+2.24+1.60=12.80\) million, the stated target. The method avoids the mistake of allocating the target by adding or subtracting the same dollar amount for every unit; proportional allocation scales each unit according to its own standalone amount.

Exact Allocations and Rounded Reports

An exact proportional calculation can produce amounts that are not convenient to report, such as dollars with cents when management reports in thousands. Rounding every unit independently may make the reported allocations fail to sum to the reported aggregate target. The remedy is to separate the calculation into two stages: calculate exact allocations first, then reconcile the rounded amounts.

Suppose the reporting increment is \(q\), such as \(q=\$1{,}000\). Round the aggregate target to the required reporting precision. For each exact allocation, take the number of whole \(q\)-increments below it. Then compare the sum of those floors with the rounded target. Give the required additional increments to the units with the largest fractional remainders.

Rounding procedure: Express each exact allocation in units of the reporting increment \(q\). Take the floor of each amount, then calculate how many increments must be added for the unit amounts to sum to the rounded aggregate target. Add those increments to the units with the largest fractional remainders. This preserves the rounded total while keeping the adjustments as small as possible under this procedure.

Worked Example: Reconcile Four Allocations to the Rounded Target

A hypothetical portfolio has standalone capital of $1.3 million, $2.7 million, $3.1 million, and $2.9 million. Its aggregate target is $8,765,432. Allocate the target proportionally, then report each unit to the nearest $1,000 so the reported amounts sum to the target rounded to the nearest $1,000.

Method. Total standalone capital is \(1.3+2.7+3.1+2.9=10.0\) million. The weights are \(0.13, 0.27, 0.31,\) and \(0.29\); their sum is \(1.00\). Multiplying the target by each weight gives the exact allocations:

$$ \begin{aligned} C_1&=8{,}765{,}432(0.13)=\$1{,}139{,}506.16,\\ C_2&=8{,}765{,}432(0.27)=\$2{,}366{,}666.64,\\ C_3&=8{,}765{,}432(0.31)=\$2{,}717{,}283.92,\\ C_4&=8{,}765{,}432(0.29)=\$2{,}541{,}975.28. \end{aligned} $$

Check the exact total. The first two amounts sum to $3,506,172.80. Adding the third gives $6,223,456.72; adding the fourth gives $8,765,432.00.

The aggregate target rounded to the nearest $1,000 is $8,765,000. In thousands of dollars, the exact unit amounts are 1,139.50616, 2,366.66664, 2,717.28392, and 2,541.97528. Their floors are 1,139, 2,366, 2,717, and 2,541, which sum to 8,763. Therefore, two additional increments of $1,000 are needed. The fractional remainders, in thousands, are 0.50616, 0.66664, 0.28392, and 0.97528. The largest are for Units 4 and 2, so those units receive the two increments.

The reported allocations are therefore $1,139,000, $2,367,000, $2,717,000, and $2,542,000. Their sum is $8,765,000.

Common mistake and correction. A calculation error can be hidden when two incorrect amounts happen to offset in the final sum. For example, writing \(8{,}765{,}432(0.31)=\$2{,}717{,}783.92\) is incorrect; the correct product is $2,717,283.92. The incorrect Unit 3 amount would also change its remainder and could misdirect the rounding increments. Check each multiplication and the exact total before applying the rounding procedure.

Zero Weights and a Second Rounding Check

A zero standalone amount is not a problem if the denominator is positive: the unit’s weight and allocation are both zero. A zero denominator is different. If every standalone amount is zero, the formula cannot produce proportional weights, even if an aggregate target has been supplied. Do not divide by zero or treat an undefined weight as zero without an explicit alternative rule.

Worked Example: Include a Zero-Capital Unit and Round in Thousands

A hypothetical portfolio has standalone capital of $1.75 million, $2.50 million, $0.75 million, $3.00 million, and $0 for five units. The aggregate target is $7.25 million. Calculate the exact allocations and report them to the nearest $1,000 so their sum equals the target.

Method. The standalone total is \(1.75+2.50+0.75+3.00+0=8.00\) million. The weights are \(0.21875,\ 0.3125,\ 0.09375,\ 0.375,\) and \(0\), which sum to \(1\). The exact allocations are:

$$ \begin{aligned} C_1&=7{,}250{,}000(0.21875)=\$1{,}585{,}937.50,\\ C_2&=7{,}250{,}000(0.3125)=\$2{,}265{,}625.00,\\ C_3&=7{,}250{,}000(0.09375)=\$679{,}687.50,\\ C_4&=7{,}250{,}000(0.375)=\$2{,}718{,}750.00,\\ C_5&=7{,}250{,}000(0)=\$0. \end{aligned} $$

The exact amounts sum to $7,250,000. In thousands, their floors are 1,585, 2,265, 679, 2,718, and 0, summing to 7,247. Three increments are needed to reach the target of 7,250 thousand. The nonzero fractional remainders are 0.9375 for Unit 1, 0.625 for Unit 2, 0.6875 for Unit 3, and 0.75 for Unit 4. The three largest belong to Units 1, 4, and 3.

The reported amounts are $1,586,000, $2,265,000, $680,000, $2,719,000, and $0. They sum to $7,250,000. The zero-capital unit remains at zero; its presence does not alter the other units’ relative weights.

Checks That Prevent Allocation Errors

A proportional allocation should pass several checks before it is reported. These checks are especially useful when figures are transferred between a model, a spreadsheet, and a management report.

1
Check the basis.
Confirm that all standalone amounts use the same currency and scale, and that the aggregate target is in matching units.
2
Check the denominator.
Calculate the total standalone capital and confirm it is positive before calculating weights.
3
Check the weights.
Verify that each weight is \(K_i/\sum_jK_j\) and that the weights sum to one, allowing only for display rounding.
4
Check exact allocation.
Sum the unrounded allocations and confirm they equal the aggregate target.
5
Check reported allocation.
After rounding, sum the reported amounts and compare that sum with the target rounded to the same increment.

A frequent trap is to round the weights first and then treat those rounded values as exact. For instance, displayed weights of 13%, 27%, 31%, and 29% are exact in the preceding example, but displayed percentages in other cases may not sum to 100% because of rounding. Use the underlying standalone amounts to calculate allocations; use displayed weights for explanation only unless they are specified as the actual allocation basis.

Key takeaway: Calculate each proportional allocation from the same aggregate target and the same total standalone basis. Reconcile exact amounts first; if reports use a coarser increment, use the rounded target and fractional remainders to reconcile the reported amounts.

Check Your Understanding

For each problem, calculate the proportional allocations and check the stated total. Treat amounts as dollars unless a unit is specified.

  1. A hypothetical group has standalone capital of $9 million, $6 million, and $5 million. Allocate a $16 million target proportionally. (Answer: $7.2 million, $4.8 million, and $4.0 million.)
  2. Three hypothetical units have standalone capital of $3.6 million, $2.4 million, and $1.2 million. Allocate a $9.9 million target. (Answer: $4.95 million, $3.30 million, and $1.65 million.)
  3. Standalone capital is $2 million, $3 million, and $5 million, and the target is $4,321,000. Find each exact proportional allocation. (Answer: $864,200, $1,296,300, and $2,160,500.)
  4. Standalone capital is $1 million, $2 million, $3 million, and $4 million; the target is $6,543,210. Allocate proportionally, then report to the nearest $1,000 so the rounded allocations sum to the target rounded to the nearest $1,000. (Answer: $654,000, $1,309,000, $1,963,000, and $2,617,000; total $6,543,000.)
  5. A hypothetical portfolio has three units, each with zero standalone capital, and an aggregate target of $5 million. Can the proportional weights be calculated from these data? (Answer: No. The denominator is zero, so proportional weights are undefined; an alternative allocation basis or rule is required.)
Clinical Biostats