Tutorials › CAS Exam 9 - Risk Management for Actuaries › Capital Charges For Growth And Business Mix Changes

CAS Exam 9 · Regulatory Capital in Practice · Tutorial 889 of 995

Capital Charges For Growth And Business Mix Changes

Estimate how exposure growth and changes in business mix or asset allocation affect category charges, then trace those changes through a covariance-based capital aggregation.

Advanced 10 min read

What You'll Learn

  • Calculate premium and reserve charges from exposure bases and prescribed factors
  • Separate the effect of volume growth from the effect of business-line mix
  • Estimate how an asset reallocation changes its category charge
  • Aggregate changed category charges using a stated covariance formula
  • Identify when linear scaling and fixed-factor estimates are not appropriate

Why Growth and Mix Affect Capital

Growth does not translate into regulatory capital through volume alone. The result depends on which exposures grow, the factors applied to those exposures, and how the resulting category charges aggregate. A company adding premium in a higher-charge line may increase its premium charge more than a company adding the same amount in a lower-charge line. Moving assets between classes can also change the asset charge even when total assets remain constant.

The calculations below use hypothetical factors and a hypothetical covariance aggregation. They illustrate the mechanics, not the requirements of any particular jurisdiction. In practice, the applicable formula may prescribe its own exposure bases, factors, adjustments, and aggregation rules. Apply those rules consistently rather than assuming that a factor or basis is universal.

Definition: A category charge is a formula-based capital amount assigned to a specified risk category using an exposure measure and the applicable factor or calculation rule. A growth effect is the change in a charge caused by a change in exposure. A mix effect is the change caused by redistributing exposure among categories or business lines with different charge characteristics.

Calculating Category Charges as Exposures Change

When a formula applies a constant factor \(k_j\) to exposure \(E_j\) in line or class \(j\), the charge for that exposure is \(C_j=k_jE_j\). If the exposure changes by \(\Delta E_j\) and the factor remains unchanged, the charge changes by \(\Delta C_j=k_j\Delta E_j\). This scaling is valid only when the calculation is linear over the range considered and the exposure definition and factor have not changed.

For example, a premium charge might use a prescribed premium basis, while a reserve charge might use a prescribed reserve basis. Whether the basis is gross or net, written or earned, and which adjustments apply must come from the applicable formula. Growth estimates are meaningful only if the same basis is used before and after the change.

Worked Example: Premium and Reserve Growth

A hypothetical insurer has two business lines. The premium factors are 12% for Line A and 20% for Line B. Its current premium bases are $40 million and $25 million, respectively. New business adds $10 million to Line A and $5 million to Line B. Separately, a reserve exposure subject to an 8% factor grows from $60 million to $75 million. Calculate the charge changes. Treat the premium and reserve charges as separate categories.

Method. Apply each factor to its own exposure before and after growth. The calculation assumes the factors and exposure definitions remain unchanged.

Current premium charge. Line A contributes \(0.12(40)=4.80\) million. Line B contributes \(0.20(25)=5.00\) million. The current premium charge is \(4.80+5.00=9.80\) million.

Premium charge after growth. Line A's new exposure is \(40+10=50\) million, giving \(0.12(50)=6.00\) million. Line B's new exposure is \(25+5=30\) million, giving \(0.20(30)=6.00\) million. The new premium charge is \(6.00+6.00=12.00\) million, so the increase is \(12.00-9.80=2.20\) million. Equivalently, the increments are \(0.12(10)=1.20\) million and \(0.20(5)=1.00\) million, totaling \(2.20\) million.

Reserve charge. Before growth, the charge is \(0.08(60)=4.80\) million. After growth, it is \(0.08(75)=6.00\) million. The reserve charge increases by \(6.00-4.80=1.20\) million, which also equals \(0.08(75-60)=1.20\) million.

Check and common mistake. Each new charge equals its old charge plus the calculated increment. The premium increment is not found by applying one factor to total new premium: the two lines have different factors. Also, the premium and reserve charges are separate category amounts; adding them does not, by itself, produce diversified aggregate capital.

Separating Volume Growth from Business Mix

For a set of lines with the same total premium, changing the share in each line can change the charge. If line \(j\) has factor \(k_j\) and premium exposure \(P_j\), the premium charge under a linear factor model is:

$$ C_P=\sum_j k_jP_j $$

At fixed total premium \(P=\sum_jP_j\), let \(w_j=P_j/P\) be the share in line \(j\). Then \(C_P=P\sum_j k_jw_j\). The quantity \(\sum_j k_jw_j\) is the exposure-weighted average factor. It changes when the shares change, even if total premium does not.

Worked Example: A Shift Toward a Higher-Charge Line

A hypothetical portfolio has $100 million of premium in two lines. Line A has an 8% factor and Line B has an 18% factor. Initially, 70% of premium is in A and 30% in B. The portfolio shifts to an even split, with total premium unchanged. Find the old and new premium charges and the mix effect.

Method. Calculate each line's exposure from its share of total premium, then apply that line's factor.

Initial mix. Line A has \(0.70(100)=70\) million of premium and contributes \(0.08(70)=5.60\) million. Line B has \(0.30(100)=30\) million and contributes \(0.18(30)=5.40\) million. The initial charge is \(5.60+5.40=11.00\) million.

New mix. Each line now has \(0.50(100)=50\) million. Line A contributes \(0.08(50)=4.00\) million; Line B contributes \(0.18(50)=9.00\) million. The new charge is \(4.00+9.00=13.00\) million. The mix change therefore increases the charge by \(13.00-11.00=2.00\) million, with no increase in total premium.

Check and common mistake. The initial weighted factor is \(0.70(0.08)+0.30(0.18)=0.11\), giving \(100(0.11)=11.00\) million. The new weighted factor is \(0.50(0.08)+0.50(0.18)=0.13\), giving \(100(0.13)=13.00\) million. A mistaken calculation would carry forward the old 11% weighted factor and report no mix effect. That is wrong because the line shares changed.

Asset Allocation and Aggregate Capital

An asset charge can also change through mix rather than growth. Under a simplified class-factor method, if asset class \(j\) has amount \(A_j\) and factor \(q_j\), its charge is \(D_j=q_jA_j\). Moving a fixed amount from one class to another changes the charge by the difference between the factors applied to that amount. This calculation assumes that the prescribed asset classification, valuation basis, and factors are unchanged.

Category charges are not necessarily additive when the formula applies a covariance aggregation. For this tutorial, suppose category charges are collected in a vector \(C\), and a specified correlation matrix \(R\) is used. The illustrative aggregate is \(K=\sqrt{C^\mathsf{T}RC}\). The diagonal elements of \(R\) are 1, and its off-diagonal elements are the prescribed or assumed correlations. The formula is appropriate here because it is explicitly part of the hypothetical model; a different regulatory formula may use a different aggregation rule.

Aggregation formula: For category charges \(C_1,\ldots,C_n\) and a specified correlation matrix \(R\), the covariance-form aggregate is \(K=\sqrt{C^\mathsf{T}RC}\). For three categories with correlations \(\rho_{12},\rho_{13},\rho_{23}\), this becomes \(K=\sqrt{C_1^2+C_2^2+C_3^2+2\rho_{12}C_1C_2+2\rho_{13}C_1C_3+2\rho_{23}C_2C_3}\).

The cross terms explain why a change in a category charge does not generally equal the change in aggregate capital. To see this, let the category vector change from \(C\) to \(C+\Delta\), with the same symmetric matrix \(R\). Squaring the aggregate before and after gives:

$$ K_{\mathrm{new}}^2-K_{\mathrm{old}}^2 =(C+\Delta)^\mathsf{T}R(C+\Delta)-C^\mathsf{T}RC =2C^\mathsf{T}R\Delta+\Delta^\mathsf{T}R\Delta. $$

The first term captures the interaction of the change with existing category charges; the second captures the change's own squared contribution and interactions within the change vector. If both aggregate amounts are positive, factoring the left-hand side also gives \(K_{\mathrm{new}}-K_{\mathrm{old}}=(K_{\mathrm{new}}^2-K_{\mathrm{old}}^2)/(K_{\mathrm{new}}+K_{\mathrm{old}})\). This justifies comparing aggregate capital by recalculating the full formula, rather than treating a category-charge increase as an equal increase in aggregate capital.

Worked Example: Asset Reallocation Through the Aggregation Formula

A hypothetical insurer has $50 million in a lower-charge asset class with a 2% factor and $30 million in a higher-charge class with an 8% factor. It transfers $10 million from the lower-charge class to the higher-charge class; total assets remain $80 million. Its other two category charges are \(C_P=10\) million and \(C_R=8\) million. Use the illustrative aggregation correlations \(\rho_{PR}=0.25\), \(\rho_{PA}=0.10\), and \(\rho_{RA}=0.20\), where \(A\) denotes the asset category. Calculate the asset charge before and after the transfer and the resulting aggregate capital.

Method. First calculate the asset charge from class amounts and factors. Then insert that charge into the full covariance formula. All charge amounts are in millions of dollars.

Asset charge before the transfer. The lower-charge class contributes \(0.02(50)=1.00\) million. The higher-charge class contributes \(0.08(30)=2.40\) million. Thus \(C_A=1.00+2.40=3.40\) million.

Asset charge after the transfer. The class amounts are \(50-10=40\) million and \(30+10=40\) million. Their charges are \(0.02(40)=0.80\) million and \(0.08(40)=3.20\) million. Thus \(C_A=0.80+3.20=4.00\) million, an increase of \(4.00-3.40=0.60\) million.

Aggregate before the transfer. With charges \((10,8,3.4)\), the squared aggregate is \(10^2+8^2+3.4^2+2(0.25)(10)(8)+2(0.10)(10)(3.4)+2(0.20)(8)(3.4)\). The terms are \(100+64+11.56+40+6.80+10.88=233.24\). Therefore \(K_{\mathrm{old}}=\sqrt{233.24}=15.272\) million, or $15.27 million rounded to the nearest cent.

Aggregate after the transfer. With charges \((10,8,4)\), the squared aggregate is \(10^2+8^2+4^2+2(0.25)(10)(8)+2(0.10)(10)(4)+2(0.20)(8)(4)\). The terms are \(100+64+16+40+8+12.8=240.8\). Therefore \(K_{\mathrm{new}}=\sqrt{240.8}=15.518\) million, or $15.52 million rounded to the nearest cent. The aggregate increase is \(15.518-15.272=0.246\) million, or $0.25 million to the nearest cent.

Check and common mistake. The asset charge rises by $0.60 million, while aggregate capital rises by about $0.25 million because the aggregation formula includes the other categories and their correlations. Reporting a $0.60 million aggregate increase would confuse a category-charge change with an aggregate-capital change. The total asset amount remains $80 million, which confirms that this is a mix effect, not an asset-growth effect.

When a Simple Growth Estimate Is Not Enough

Linear scaling is useful for a quick estimate, but it is not a substitute for the applicable calculation. A factor may change at thresholds, an exposure may be subject to a cap or floor, or the formula may combine several bases before applying a charge. A change in business mix may also change more than premium shares: it can affect reserves, catastrophe exposure, reinsurance recoverables, or other categories. Each relevant category must be recalculated using its own prescribed basis.

Keep the following distinctions clear when estimating capital effects:

  • Exposure growth versus mix change. Growth changes total exposure; mix change redistributes exposure. Both can occur at the same time.
  • Category charge versus aggregate capital. Recalculate the aggregation after updating category charges. Do not assume the aggregate moves dollar-for-dollar with one charge.
  • Consistent measurement basis. Use the same applicable exposure definition, valuation date, and factor basis for the before-and-after comparison.
  • Formula estimate versus full recalculation. Treat a fixed-factor estimate as conditional on the factor and formula remaining applicable. Recalculate when thresholds, adjustments, or model rules change.

This approach is useful for assessing the direction and approximate size of capital effects when planning growth or evaluating a portfolio shift. It is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus. The central discipline is to identify what changed, calculate each affected category on its own terms, and then apply the stated aggregation method to the revised charge vector.

Check Your Understanding

For each hypothetical problem, use only the factors and aggregation rules stated in the question. Report dollar amounts in millions unless indicated otherwise.

  1. A portfolio has $30 million of premium at a 10% factor and $20 million at a 16% factor. It adds $5 million to the first line and $4 million to the second, with factors unchanged. What is the new premium charge and its increase? (Answer: new charge $7.34 million; increase $1.14 million)
  2. Total premium is $80 million. Line A has a 7% factor and Line B a 15% factor. The mix changes from 75% A and 25% B to 50% in each line. Find the old and new charges and the mix effect. (Answer: old $7.20 million; new $8.80 million; increase $1.60 million)
  3. A reserve exposure subject to a 9% factor grows from $120 million to $135 million. Calculate the reserve charge before and after growth and the charge increase. (Answer: $10.80 million before; $12.15 million after; increase $1.35 million)
  4. An insurer transfers $10 million from an asset class with a 2% factor to one with an 8% factor. The total asset amount is unchanged. By how much does the asset category charge increase? (Answer: $0.60 million)
  5. Two category charges are initially $6 million and $4 million, with correlation \(0.25\). They become $8 million and $5 million, with the same correlation. Using \(K=\sqrt{C_1^2+C_2^2+2(0.25)C_1C_2}\), calculate aggregate capital before and after the change, rounding to the nearest cent in millions. (Answer: before \(\sqrt{64}=8.00\) million; after \(\sqrt{109}=10.44\) million)
Clinical Biostats