Integrated Rating Capital Calculation For A Composite Insurer

From Balance-Sheet Inputs to a Capital Assessment

A composite insurer’s capital assessment brings together two different questions: how much capital is available after relevant balance-sheet adjustments, and how much capital is required for the risks being assessed. A defensible calculation keeps those quantities distinct, states how risk charges are aggregated, and interprets the resulting ratio against explicitly defined thresholds.

The case below uses a hypothetical insurer and an illustrative rating model. It is not a formula for any particular rating agency. In particular, the model treats each stand-alone risk charge as a standard-deviation scale and uses a correlation-based aggregation. Actual models may prescribe different charge definitions, calibrations, aggregation rules, or adjustments. There is no interest-rate or mortality assumption in this balance-sheet capital calculation; its inputs are monetary capital amounts and assumed risk correlations.

Definition: Let $A$ be available capital after the stated balance-sheet adjustments, and let $C_{\mathrm{agg}}$ be aggregate required capital under the stated risk model. The capital ratio is $A/C_{\mathrm{agg}}$. A larger ratio indicates more available capital relative to modeled requirements, but its interpretation depends on the model and the thresholds being applied.

Set Out the Model Before Calculating

For this tutorial, the insurer has four risk categories: underwriting, invested assets, catastrophe, and operational risk. Their stand-alone charges are denoted $C_1,\ldots,C_4$, and the correlation between categories $i$ and $j$ is $\rho_{ij}$. The illustrative model assumes centered, jointly normally distributed risk proxies, with standard deviations equal to the stated stand-alone charges. Thus the charges and correlations are sufficient to calculate the standard deviation of the aggregate proxy.

C_{\mathrm{agg}} = \sqrt{\sum_{i=1}^{4} C_i^2 +2\sum_{i<j}\rho_{ij}C_iC_j}

This expression is appropriate here because the charges are explicitly defined as standard-deviation scales for the risk proxies. If a model instead supplies capital charges defined as quantiles, or uses a different aggregation method, they cannot automatically be substituted into this formula as though they were standard deviations. Follow the model’s own definitions and rules.

Derivation: Write the four centered risk proxies as $Z_i$, with $\operatorname{SD}(Z_i)=C_i$ and $\operatorname{Corr}(Z_i,Z_j)=\rho_{ij}$. The aggregate proxy is $Z=\sum_i Z_i$. By the variance-of-a-sum identity, $\operatorname{Var}(Z)=\sum_i\operatorname{Var}(Z_i)+2\sum_{i<j}\operatorname{Cov}(Z_i,Z_j)$. Since $\operatorname{Var}(Z_i)=C_i^2$ and $\operatorname{Cov}(Z_i,Z_j)=\rho_{ij}C_iC_j$, taking the square root gives the stated aggregate standard deviation.

Worked Composite Insurer Case

The hypothetical insurer reports equity of $500 million. For the simplified available-capital calculation, a $12 million adjustment is admitted, while $17 million of nonadmitted assets and $10 million of intangible assets are deducted. Its stand-alone charges are based on the following assumed inputs: net premiums of $1,100 million at a 15% underwriting factor; invested assets of $1,200 million at a 5% asset factor; a $77 million catastrophe scenario charge; and an operational exposure base of $500 million at a 5% factor. These factors and amounts are invented solely for this example.

Worked Example: Build the Base Capital Assessment

Method. First calculate available capital and each stand-alone charge. Then aggregate the charges using the stated base correlations and divide available capital by aggregate required capital. The model uses the standard-deviation interpretation described above; there is no separate interest or mortality assumption.

Available capital. The stated adjustment is added and the two stated deductions are subtracted:

A=500+12-17-10=485\text{ million dollars}.

Stand-alone charges. Underwriting capital is $1{,}100(0.15)=165$ million dollars. The invested-asset charge is $1{,}200(0.05)=60$ million dollars. The catastrophe charge is given as 77 million dollars, and the operational charge is $500(0.05)=25$ million dollars. The charge vector is therefore $(165,60,77,25)$, in millions.

For the base case, assume the underwriting–asset correlation is 0.10, underwriting–catastrophe is 0.05, underwriting–operational is 0.05, asset–catastrophe is 0, asset–operational is 0, and catastrophe–operational is 0.10. The sum of the four squared charges is:

165^2+60^2+77^2+25^2 =27{,}225+3{,}600+5{,}929+625 =37{,}379.

The six correlation-weighted pair products are $0.10(165)(60)=990$, $0.05(165)(77)=635.25$, $0.05(165)(25)=206.25$, $0(60)(77)=0$, $0(60)(25)=0$, and $0.10(77)(25)=192.5$. Their sum is $2{,}024$. The aggregate charge is:

C_{\mathrm{agg,base}} =\sqrt{37{,}379+2(2{,}024)} =\sqrt{41{,}427} \approx203.536\text{ million dollars}.

The available-to-required capital ratio is $485/203.536\approx2.383$, rounded to three decimals. As a check, the aggregated charge exceeds the largest stand-alone charge of 165 million dollars, but is below the sum of all four charges, $327$ million dollars. Both are reasonable for this positively correlated, diversified aggregation.

Common mistake avoided: Do not add the stand-alone charges and call their sum the square-root aggregate. That would give $327$ million dollars and ignore the model’s diversification calculation. Conversely, diversification does not mean the aggregate must be below every individual charge; here it is above 165 million dollars.

Stress the Dependence Assumptions

A sensitivity assessment can change correlations while holding available capital and stand-alone charges fixed. This isolates the effect of dependence: higher positive correlations generally reduce diversification credit under the square-root formula. The stressed correlations below are illustrative assumptions, not observed relationships.

Worked Example: Recalculate Under Stressed Correlations

Method. Keep the four charges at $(165,60,77,25)$ million dollars and use stressed correlations: underwriting–asset 0.25, underwriting–catastrophe 0.15, underwriting–operational 0.10, asset–catastrophe 0.05, asset–operational 0.05, and catastrophe–operational 0.20. The centered multivariate-normal risk-proxy assumption and charge parameters are unchanged; only the correlations change.

The correlation-weighted pair products are $0.25(165)(60)=2{,}475$, $0.15(165)(77)=1{,}905.75$, $0.10(165)(25)=412.5$, $0.05(60)(77)=231$, $0.05(60)(25)=75$, and $0.20(77)(25)=385$. Their sum is $5{,}484.25$. Therefore:

C_{\mathrm{agg,stress}} =\sqrt{37{,}379+2(5{,}484.25)} =\sqrt{48{,}347.5} \approx219.88065\text{ million dollars}.

Using the same available capital of $485 million, the stressed ratio is $485/219.88065\approx2.20574$, or 2.206 to three decimals. The stress increases required capital by about $219.88065-203.53624=16.34441$ million dollars relative to the base case. The ratio falls because the numerator is unchanged while the denominator increases.

Common mistake and correction. A tempting but incorrect multiplication is $0.15(165)(77)=1{,}732.5$. First calculate $165(77)=12{,}705$, then multiply by 0.15: $12{,}705(0.15)=1{,}905.75$. Using the incorrect product would understate the sum of pair products and the aggregate charge. The corrected pair-product total is $5{,}484.25$, giving stressed required capital of approximately $219.88 million and a ratio of 2.206.

Interpret Thresholds and Headroom

A ratio has no rating interpretation by itself. Thresholds must be stated, and a numerical result should not be presented as an actual agency rating unless the relevant agency methodology and all of its inputs have been applied. For illustration only, suppose this simplified model labels ratios below 2.0 as the lower band, ratios from 2.0 up to but not including 2.5 as the middle band, and ratios of 2.5 or more as the upper band. On these invented thresholds, both the base ratio of 2.383 and stressed ratio of 2.206 are in the middle band.

Threshold headroom can be expressed in capital dollars. At a fixed required-capital amount, the available capital needed for a threshold ratio $r_0$ is $r_0C_{\mathrm{agg}}$. The difference between current available capital and that amount is the available-capital headroom above the threshold. A ratio near a boundary merits attention to the size and direction of plausible changes; it is not enough to report only the current band.

Worked Example: Measure Stressed Threshold Headroom

Method. Use the stressed aggregate charge of $219.88065 million and available capital of $485 million. Assume the same hypothetical bands: below 2.0, from 2.0 to less than 2.5, and at least 2.5. No distributional recalibration is made; this is a ratio and threshold calculation using the stressed result already derived.

The available capital corresponding to the lower boundary of the middle band is $2.0(219.88065)=439.76130$ million dollars. Current headroom above that boundary is $485-439.76130=45.23870$ million dollars. If available capital falls by $50 million, it becomes $435 million, and the resulting ratio is $435/219.88065\approx1.978$, which is below 2.0 and therefore in the lower band under the stated illustration.

At the upper boundary, the required available capital is $2.5(219.88065)=549.701625$ million dollars. The insurer is short of that boundary by $549.701625-485=64.701625$ million dollars. This does not change the current middle-band classification; it quantifies the gap to the next boundary.

Check. Multiplying the stressed denominator by the current ratio recovers available capital to rounding: $219.88065(2.20574)\approx485$ million dollars. The threshold calculation uses available capital divided by required capital, not the inverse.

Practical Checks Before Reporting a Result

A concise assessment should make its construction auditable. Separate adjustments to equity from risk charges; show the charge basis and factor for each category; list the correlations used; and calculate the ratio only after aggregation. If a stress changes both charges and correlations, identify both changes so the result cannot be mistaken for a dependence-only sensitivity.

Check	Question to answer	Why it matters
Available capital	Which adjustments are admitted, and which items are deducted?	Reported equity and usable capital need not be the same quantity.
Stand-alone charges	What exposure base and factor, scenario, or model output produced each charge?	Charges are not comparable unless their definitions and units are clear.
Aggregation	Are the correlations paired with the correct risk categories and used once in the pair sum?	A wrong pair or arithmetic error directly distorts required capital.
Threshold interpretation	Are the boundaries explicit, and is the ratio numerator divided by denominator in the intended order?	A correct number can still be assigned to the wrong band if the rule is misread.

The point of an integrated calculation is not merely to obtain one ratio. It is to show how balance-sheet adjustments, risk charges, dependence assumptions, and thresholds work together. In this hypothetical case, the stressed dependence assumptions raise required capital and reduce the ratio, while the illustrative threshold remains unchanged. A professional assessment should state those assumptions and avoid treating a simplified model result as a complete rating decision.

Check Your Understanding

For each problem, use the model or threshold assumptions stated in that problem. Dollar amounts are hypothetical; report capital amounts in millions and ratios to three decimals where applicable.

A hypothetical insurer reports equity of $360 million. It receives a $8 million admissible adjustment and deducts $13 million of intangible assets and $5 million of nonadmitted assets. What is its adjusted available capital? (Answer: $350 million.)
Two centered jointly normal risk proxies have standard deviations of $80 million and $45 million, with correlation 0.30. Under the square-root standard-deviation aggregation, calculate aggregate required capital. (Answer: $\sqrt{10{,}585}\approx\$102.883$ million.)
Four standard-deviation risk charges are 100, 40, 30, and 20 million dollars. All correlations are zero except the correlation between the first and second risks, which is 0.25, and between the third and fourth risks, which is 0.50. Calculate aggregate required capital. (Answer: $\sqrt{15{,}500}\approx\$124.499$ million.)
An insurer has available capital of $300 million and required capital of $125 million. Under hypothetical bands of below 2.0, from 2.0 to less than 2.4, and at least 2.4, calculate its ratio and band. (Answer: $300/125=2.400$; upper band.)
A company has $400 million of available capital and $180 million of required capital. Under a hypothetical threshold ratio of 2.1, what is the maximum decrease in available capital it can sustain before falling below the threshold, assuming required capital is unchanged? (Answer: $400-2.1(180)=\$22$ million.)

What You'll Learn