From Statutory Exposures to a Final Requirement
A regulatory capital calculation is a chain of linked steps. Statutory exposure amounts are converted into risk charges using prescribed factors or rules; the resulting charges are combined using the prescribed aggregation method; and formula adjustments are then applied in the specified order. An error early in the chain can carry through to the final requirement, so each intermediate result should be visible and independently checked.
The calculations here use hypothetical schedules and inputs for teaching. A regulator’s actual factors, risk categories, correlations, caps, and adjustment rules depend on the applicable regime and reporting instructions. The examples therefore demonstrate a calculation method, not a universal regulatory formula. The approach is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.
The Calculation Sequence
Keep the sequence explicit. First identify the statutory exposure bases and their units. Next apply the factors or rules given in the problem to calculate the individual charges. Combine charges within a category when the prescribed method requires it, and then aggregate the category charges using the supplied correlation inputs. Finally apply any offsets and additions. Do not assume an adjustment belongs inside the covariance formula unless the stated rule says so.
Record each exposure, its unit, the applicable prescribed factor, and any charge already supplied. Separate statutory amounts from calculated charges.
Apply each factor to its stated exposure base. Keep intermediate results in a consistent unit, such as thousands of dollars.
Use the specified correlations and formula. If subcategory charges are aggregated first, do that before using the resulting category charge in the next level.
Calculate the offset, enforce any cap or floor, and then apply any required addition in the specified order.
Check units, squared terms, cross terms, and adjustment signs. Report intermediate values as well as the final requirement.
Covariance Aggregation and Its Justification
For category charges \(C_1,\ldots,C_n\), a prescribed covariance-style aggregation uses the following form, where \(\rho_{ij}\) is the supplied correlation parameter for categories \(i\) and \(j\):
The formula contains each category’s squared charge and a cross term for each distinct pair. A pair contributes \(2\rho_{ij}C_iC_j\), not just \(\rho_{ij}C_iC_j\). A zero correlation removes that pair’s cross term; it does not remove either category’s squared charge.
One justification comes from the variance of a sum. If random quantities \(Z_i\) have standard deviations \(C_i\) and pairwise correlations \(\rho_{ij}\), then the variance of their sum is the sum of their variances plus twice the sum of their pairwise covariances. Since \(\operatorname{Cov}(Z_i,Z_j)=\rho_{ij}C_iC_j\), taking the square root gives the displayed expression. A regulatory formula may prescribe this structure as an aggregation approximation; the calculation does not, by itself, establish that real losses are normally distributed or that a charge is a standard deviation.
The correlations must be stated or otherwise determined by the applicable instructions. They should also be compatible with a valid correlation structure. If the problem supplies only some pairwise correlations and leaves another required pair unspecified, the aggregate is generally not determined. Do not silently fill the gap with zero or with a value from another setting.
Worked Example: Exposure Inputs Through the Final Requirement
A hypothetical schedule reports dollar exposures in thousands. It prescribes asset factors of 1% for a bond exposure of 8,000 and 20% for an equity exposure of 500. Two premium exposures, 1,200 and 700, have factors of 8% and 12%. Two reserve exposures, 1,000 and 600, have factors of 5% and 10%. The catastrophe charge is 45. Premium, reserve, and catastrophe charges are aggregated with correlations 0.25, 0.10, and 0.20, respectively, for premium-reserve, premium-catastrophe, and reserve-catastrophe. The resulting underwriting charge is combined with the asset charge, a credit charge of 35, and an operational charge of 20. The correlations are asset-underwriting 0.20, asset-credit 0.10, underwriting-credit 0.15; operational risk has zero correlation with each of the other modules. A tax offset is 10% of the preliminary aggregate, capped at 30, and a required addition is 12. Find the final requirement, in thousands.
Method and unit check. All exposures, charges, the cap, and the addition are expressed in thousands of dollars. Multiply each exposure by its factor to obtain its charge.
The asset charge is \(8{,}000(0.01)+500(0.20)=80+100=180\).
The premium charge is \(1{,}200(0.08)+700(0.12)=96+84=180\). The reserve charge is \(1{,}000(0.05)+600(0.10)=50+60=110\).
Aggregate the underwriting components first:
Therefore \(C_{\mathrm{UW}}=\sqrt{60{,}025}=245.000\), since \(245^2=60{,}025\).
Now combine asset, underwriting, credit, and operational charges. The operational charge contributes its square, \(20^2\), but its zero correlations mean it contributes no cross terms:
Thus \(C_{\mathrm{pre}}=\sqrt{115{,}522.5}=339.886\), to three decimal places. The uncapped offset is \(0.10(339.886)=33.989\). The cap is 30, so the allowed offset is \(\min(33.989,30)=30\). The final requirement is \(339.886-30+12=321.886\), or 321.89 thousand to the nearest cent in thousands.
Check and common mistake avoided. The squared aggregate is positive and the final requirement is below the preliminary aggregate by 18, which equals the net adjustment \(30-12\). A common error is to subtract the uncapped offset of 33.989; the cap limits the subtraction to 30.
Worked Calculations from Given Category Charges
A problem may provide charges already calculated from statutory exposures. In that case, do not apply exposure factors a second time. Use the supplied charges as the inputs to the aggregation formula, while still checking whether the problem requires one or more aggregation levels.
Worked Example: Three Correlated Modules and an Offset Cap
A hypothetical calculation provides asset, underwriting, and credit charges of 100, 150, and 40 thousand. The pairwise correlations are 0.20 for asset-underwriting, 0.10 for asset-credit, and 0.30 for underwriting-credit. Operational risk has a charge of zero. The offset is 12% of the preliminary aggregate, capped at 24 thousand, and the required addition is 7 thousand. Calculate the final requirement.
Method. Calculate the squared aggregate with one cross term for each specified pair:
Therefore \(C_{\mathrm{pre}}=\sqrt{44{,}500}=210.950\) thousand, to three decimal places. The uncapped offset is \(0.12(210.950)=25.314\) thousand. Applying the cap gives an offset of \(\min(25.314,24)=24\) thousand. The final requirement is \(210.950-24+7=193.950\), or 193.95 thousand.
Check and common mistake avoided. The operational charge is zero, so it adds neither a squared term nor cross terms. The nonzero charges still contribute their own squares and all three specified pairwise cross terms. The cap applies to the offset, not to the preliminary aggregate or final requirement.
Worked Example: Distinguishing Cross Terms from Formula Adjustments
A hypothetical formula provides charges of 80 thousand for assets, 120 thousand for underwriting, 30 thousand for credit, and 25 thousand for operational risk. The only nonzero correlations are asset-underwriting 0.25, asset-credit 0.10, and underwriting-credit 0.20. The offset equals 8% of the preliminary aggregate, capped at 18 thousand, and the required addition is 4.5 thousand. Find the final requirement.
Method. Operational risk is uncorrelated with the other modules, but its squared charge remains in the aggregate:
Thus \(C_{\mathrm{pre}}=\sqrt{29{,}045}=170.426\) thousand, to three decimal places. The uncapped offset is \(0.08(170.426)=13.634\) thousand, which is below the cap of 18 thousand. The offset is therefore 13.634 thousand. The final requirement is \(170.426-13.634+4.5=161.292\), or 161.29 thousand.
Check and common mistake avoided. Since the offset is below its cap, the cap does not change it. A common error is to omit the operational squared charge because operational risk is independent; independence removes cross terms, not the charge’s own contribution.
Exam Checks and Common Traps
A reliable calculation makes the model and the order of operations apparent. Before taking a square root, confirm that every required squared charge and every stated pairwise cross term is present. After taking the square root, apply offsets and additions according to their own definitions. Rounding intermediate values too early can also move a capped offset or alter the reported final amount, so retain sufficient precision until the final step.
- Missing correlation input: If the formula needs a pairwise correlation and the problem does not specify it, do not invent a value. State that the calculation is underdetermined unless another stated rule supplies the input.
- Missing the factor of two: Each distinct pair contributes \(2\rho_{ij}C_iC_j\) in the displayed aggregation formula.
- Confusing zero correlation with zero charge: A zero correlation removes cross terms for that pair; it does not remove the risk charge’s square.
- Mixing units: Convert exposures and charges to a consistent unit before combining them. An amount in dollars cannot be combined directly with an amount in thousands of dollars.
- Applying adjustments in the wrong place: Calculate the preliminary aggregate first when the offset is defined as a percentage of that aggregate. Then enforce the cap and apply any addition in the specified order.
A useful final check is to verify that the squared aggregate equals the sum of the displayed squared charges and cross terms. Then check the adjustment algebra separately: if the offset is \(O\) and the addition is \(A\), the final amount must equal \(C_{\mathrm{pre}}-O+A\). This simple separation makes sign and cap errors easier to detect.
Check Your Understanding
For each problem, use the stated hypothetical formula and keep all charges in thousands. Round final amounts to two decimal places.
- Asset and underwriting charges are 95 and 70, and the credit charge is 25. Their pairwise correlations are asset-underwriting 0.20, asset-credit 0.10, and underwriting-credit 0.30. Operational risk has a zero charge. The offset is 10% of the preliminary aggregate, capped at 12, and the addition is 3. Find the final requirement. (Answer: 127.88 thousand)
- Asset, underwriting, and credit charges are 120, 100, and 40. Their pairwise correlations are 0.15, 0.10, and 0.25, respectively, in that order. Operational risk has a zero charge. The offset is 5% of the preliminary aggregate, capped at 10, and the addition is 2. Find the final requirement. (Answer: 173.42 thousand)
- Asset, underwriting, credit, and operational charges are 60, 90, 20, and 30. The nonzero correlations are asset-underwriting 0.20, asset-credit 0.10, and underwriting-credit 0.30; all correlations involving operational risk are zero. The offset is 15% of the preliminary aggregate, capped at 15, and the addition is 6. Find the final requirement. (Answer: 119.37 thousand)
- Premium, reserve, and catastrophe charges are 80, 60, and 40. Their correlations are premium-reserve 0.25, premium-catastrophe 0.10, and reserve-catastrophe 0.20. Calculate the underwriting aggregate before any adjustments. (Answer: 124.90 thousand)
- Asset, underwriting, credit, and operational charges are 100, 120, 30, and 10. The nonzero correlations are asset-underwriting 0.20, asset-credit 0.10, and underwriting-credit 0.25; operational risk is uncorrelated with the other modules. The offset is 9% of the preliminary aggregate, capped at 20, and the addition is 5. Find the final requirement. (Answer: 169.30 thousand)