Why ORSA Scope Depends on the Insurer
An insurer’s risk assessment is useful only if it reflects how that insurer actually earns premiums, pays claims, invests assets, and transfers risk. A life insurer with long-term guarantees faces a different pattern of exposure from a property-casualty insurer exposed to concentrated catastrophe losses. A health insurer may be especially sensitive to medical-cost trends and the timing of premium changes. Applying the same scenarios and emphasis to all three can miss the risks that matter to each business.
The purpose of mapping ORSA to the risk profile is to connect the insurer’s business model to its assessment: identify how exposures arise, consider how they could affect solvency over time, and choose analysis that captures their important features. This is not a requirement to use a unique model for every risk. It is a reasoned explanation of why the assessment’s coverage, assumptions, and scenarios suit the insurer being assessed.
A practical mapping starts with the insurer’s activities rather than with a generic risk taxonomy. For each major business feature, ask what can change, how that change reaches earnings or financial resources, how quickly it can emerge, and whether the insurer can respond in time. A risk may appear in several parts of the assessment: for example, a catastrophe can create an ultimate underwriting loss and an immediate need for cash to pay claims.
Record the products, customers, geographic concentrations, distribution arrangements, liability features, assets, and risk-transfer arrangements relevant to the insurer.
Explain how a change in a driver could affect claims, expenses, asset values, cash flows, capital, or the ability to meet obligations when due.
Select scenarios, projections, sensitivities, or qualitative analysis that represent the exposure and its time pattern. State important assumptions and limitations.
Consider whether the analysis misses concentrations, dependencies, delayed effects, or constraints on management responses.
What Business Characteristics Change the Assessment?
The same broad risk label can describe materially different exposures. Interest-rate risk for a life insurer with long-term guarantees is not necessarily the same problem as interest-rate risk for a short-tail property-casualty insurer holding liquid assets. The mapping should explain the pathway from the underlying exposure to the insurer’s own financial position.
| Business characteristic | Questions for ORSA mapping | Possible analytical emphasis |
|---|---|---|
| Products and guarantees | What must be paid, when, and under what contract conditions? | Changes in claim frequency, severity, policyholder behavior, or guarantee costs |
| Liability duration and cash-flow pattern | How long do obligations remain exposed to changing conditions? | Multi-period projections, asset-liability stresses, and reinvestment effects |
| Assets and investment strategy | What market movements or asset features could affect value or liquidity? | Market shocks, spread changes, concentration, and asset liquidity |
| Reinsurance and other risk transfer | What is retained, what can be recovered, and when might recovery arrive? | Net-loss analysis, counterparty assumptions, coverage limits, and cash timing |
| Operations and external dependencies | Which systems, providers, locations, or processes are critical? | Service disruption, data or model limitations, and operational continuity |
The time horizon is part of the mapping. A short-tail claim shock can affect current-year results and liquidity, while a change to long-duration guarantees may compound through future cash flows. A one-year view may be useful for a near-term stress, but it should not be treated as a complete account of exposures that develop over longer periods. The assessment should make clear what each projection period captures and what it does not.
Worked Example: Long-Term Guarantees at a Life Insurer
Worked Example: Long-Term Guarantees at a Life Insurer
A hypothetical life insurer has assets of $800 million with modified duration 6 and liabilities of $760 million with modified duration 9. Its starting surplus is $40 million. Assess the approximate effect on surplus of a parallel 1.00 percentage-point decline in yields. Treat the calculation as a first-order sensitivity: assume cash flows are unchanged, ignore convexity, and make no tax, policyholder-behavior, or management-action adjustment.
Method. Use the duration approximation for each balance-sheet value. For a small parallel yield change \(\Delta y\), the approximate percentage change in value \(P\) is \(-D_{\text{mod}}\Delta y\), where \(D_{\text{mod}}\) is modified duration. A 1.00 percentage-point decline means \(\Delta y=-0.0100\).
The asset value increases by approximately $48 million, from $800 million to $848 million. Apply the same calculation to liabilities:
The liability value increases by approximately $68.4 million, from $760 million to $828.4 million. Surplus is assets minus liabilities, so its change is approximately \(48-68.4=-20.4\) million. The resulting surplus is \(40-20.4=\$19.6\) million.
Check. The liability duration is greater than the asset duration, so this rate decline increases the liability value more than the asset value in this simplified calculation. The surplus therefore falls. The answer is consistent with that direction. The common mistake is to report the asset gain of $48 million as the surplus effect while omitting the liability revaluation.
ORSA mapping. This insurer’s long-term guarantees and duration mismatch make interest-rate stresses and multi-period asset-liability projections relevant. The calculation is only a local, first-order sensitivity: it does not represent a complete ORSA projection, and its assumptions may be unsuitable for large yield changes or when cash flows respond to rates.
Why the Duration Approximation Works
The sensitivity above follows from the definition of modified duration. Let \(P(y)\) be a value as a function of yield \(y\). Modified duration is \(D_{\text{mod}}=-(1/P)(dP/dy)\). For a sufficiently small yield change, a first-order Taylor approximation gives the change in value as the local slope multiplied by the yield change:
This is an approximation, not an exact repricing formula. It assumes the selected duration adequately represents the local response and does not include curvature, or convexity. It can be misleading for large changes, non-parallel yield-curve movements, or cash flows that change with rates. Those limitations matter when translating a sensitivity into ORSA conclusions: the method must fit the exposure and the question being tested.
Worked Example: Catastrophe Exposure and Reinsurance
Worked Example: Catastrophe Exposure and Reinsurance
A hypothetical property-casualty insurer has $120 million of available financial resources before a stress. Two severe events in one year produce gross claims of $95 million and $70 million. The insurer has a per-occurrence reinsurance layer of $60 million excess of $20 million, with an annual aggregate cap of $90 million on recoveries. Assume the reinsurer pays valid recoveries in full, ignore expenses, premiums, taxes, and all other changes, and measure the stress on a consistent dollar basis. Calculate the net claims and remaining resources.
Method. For each occurrence, recovery before the aggregate cap is the lesser of the layer limit and the loss above the retention:
For the first event, \(R_1=\min(\max(95-20,0),60)=\min(75,60)=\$60\) million. For the second event, \(R_2=\min(\max(70-20,0),60)=\min(50,60)=\$50\) million. Their total before the aggregate cap is \(60+50=\$110\) million. Applying the $90 million aggregate cap gives total recovery of \(\min(110,90)=\$90\) million.
Total gross claims are \(95+70=\$165\) million. Net claims after reinsurance are therefore \(165-90=\$75\) million. Remaining resources under the stated assumptions are \(120-75=\$45\) million.
Check. Recovery cannot exceed the annual aggregate cap, so a $110 million recovery is impossible. The corrected $90 million recovery leaves net claims of $75 million. A common mistake is to calculate the per-occurrence recoveries correctly but forget to apply the aggregate cap.
ORSA mapping. The event losses test catastrophe exposure and the contractual effect of risk transfer. The $75 million net claim amount alone does not describe every relevant consequence: the insurer may have to pay gross claims before receiving recoveries. The assessment should therefore consider the timing of claim payments and recoveries as well as the ultimate net loss. It should also explain how the assumed coverage terms and counterparty payment affect the result.
Worked Example: Medical-Cost Pressure at a Health Insurer
Worked Example: Medical-Cost Pressure at a Health Insurer
A hypothetical health insurer covers 50,000 members. Its expected annual claims are $4,000 per member, and annual premiums are fixed at $220 million for the period considered. Test a combined stress in which medical cost per service is 8% higher and utilization is 5% higher. Assume the two effects multiply, all 50,000 members remain covered, and premiums, expenses, investment income, taxes, and all other items are unchanged. Starting surplus is $30 million. Calculate stressed claims, underwriting result, and approximate remaining surplus.
Method. First calculate the base claims: \(50{,}000\times \$4{,}000=\$200\) million. Because the two stress factors act multiplicatively, the stressed claim amount is:
The increase in claims is \(226.8-200=\$26.8\) million. With premiums fixed at $220 million, the stressed underwriting result before other expenses is \(220-226.8=-\$6.8\) million. Under the stated simplifying assumption that the incremental claim cost reduces surplus dollar for dollar, remaining surplus is \(30-26.8=\$3.2\) million.
Check. The separate increases are not simply added: multiplying 1.08 by 1.05 gives 1.134, or a 13.4% combined increase. The result is below the 13% that would follow from incorrectly adding the percentages, and the calculations are internally consistent. The common mistake is to add the two stress percentages without accounting for the interaction.
ORSA mapping. Medical trend and utilization are distinct drivers that can affect claims together. The fixed-premium assumption makes the timing of repricing important: a longer projection could show whether and when premiums can respond, subject to the insurer’s contracts and operating constraints. A useful assessment should state those timing assumptions rather than imply that a one-period stress captures the full effect.
Use a Tailored View Without Losing Coverage
These examples illustrate why the same broad scenario set need not receive the same emphasis for every insurer. Long-term guarantees point toward careful analysis of liability cash flows and asset-liability sensitivity. Catastrophe concentration points toward event severity, reinsurance terms, and the timing of recoveries. Health claims point toward medical-cost drivers, utilization, and premium adjustment timing. Each profile can also involve other exposures; emphasis on a particular driver is not a reason to ignore the rest of the business.
A tailored assessment should distinguish what is quantified from what is described qualitatively. Quantitative projections can show the effect of specified assumptions, but uncertain data, operational dependencies, or possible changes in behavior may require explicit discussion of limitations. Where several drivers could occur together, the assessment should make clear whether the scenario combines them, treats them separately, or does not capture their interaction.
- Do not map by label alone. Explain how a risk reaches the insurer’s claims, assets, cash flows, or financial resources.
- Do not assume risk transfer removes risk. Apply contract features and consider timing and counterparty performance.
- Do not force a single horizon onto every exposure. Match the analysis period to how the risk develops and when the insurer can respond.
- Do not treat a simple sensitivity as a full projection. State its assumptions and limitations, including omitted interactions or actions.
The central discipline is traceability: a reader should be able to connect an insurer’s business characteristics to the exposures assessed, the analysis selected, and the conclusions drawn. That connection is what makes an ORSA an insurer-specific view of risk and solvency, aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.
Check Your Understanding
For each problem, identify the relevant mapping point and complete the requested calculation where applicable.
- A hypothetical life insurer has assets of $500 million with modified duration 4 and liabilities of $480 million with modified duration 7. Using a first-order approximation and ignoring convexity, estimate the surplus change for a parallel 0.50 percentage-point rise in yields. (Answer: Assets decrease by $10 million, liabilities decrease by $16.8 million, and surplus increases by $6.8 million.)
- A hypothetical property-casualty insurer has two event losses of $45 million and $75 million. Each event is covered by a $40 million layer excess of $10 million, and total annual recovery is capped at $65 million. What is total recovery and net claims? (Answer: Per-event recoveries are $35 million and $40 million; the $65 million cap applies, so net claims are $120 million − $65 million = $55 million.)
- A health insurer expects base claims of $150 million. A stress increases medical cost per service by 6% and utilization by 4%, multiplicatively. What are stressed claims and the increase over base? (Answer: Stressed claims are \(150(1.06)(1.04)=\$165.36\) million; the increase is $15.36 million.)
- A hypothetical insurer expects $30 million of catastrophe recoveries, but the claims payment date precedes expected reinsurance collection by two months. What additional feature should its ORSA analysis consider, beyond the ultimate net loss? (Answer: The interim cash-flow or liquidity need created by paying claims before collecting recoveries.)
- A health insurer can adjust premiums only at annual renewal, while claims costs can change during the year. Why should its ORSA projection state the renewal timing? (Answer: Because the timing determines how long increased claims may be incurred before premiums can respond, affecting projected earnings and solvency.)