From ORSA Assessment to Business Choice
An ORSA assessment becomes useful when it informs choices the insurer must make: whether to expand a line, distribute capital, transfer risk, change investments, or raise capital. The relevant question is not simply whether a plan is profitable or acceptable on a current balance sheet. It is whether the plan remains consistent with the insurer’s stated risk appetite across the projected scenarios and time horizons that matter to the decision.
The projections used below are hypothetical and are for internal decision analysis. They do not represent a prescribed regulatory capital calculation. In particular, an insurer’s internal capital need and risk-appetite threshold must be defined by its own governance and assessment; the ratios here are illustrative management measures.
A coverage ratio and headroom provide related but different views. The ratio compares resources with needs in relative terms; headroom gives the dollar difference between them. If a decision threshold is 125%, resources of 500 and capital need of 400 exactly meet it: \(500/400=1.25\). The headroom is 100, but that 100 is not all distributable if management requires the 125% threshold to be maintained.
Use comparable quantities. Capital resources and capital need must refer to the same date, scenario, business perimeter, and basis. If a scenario includes management actions, include them only when their feasibility, timing, and modeled effects are supported. A projected ratio should not be treated as a complete decision rule: earnings, liquidity, execution risk, counterparty exposure, and strategic considerations may also matter.
A Practical Decision Sequence
A disciplined use of ORSA findings links a proposal to its modeled effects, checks those effects against risk appetite, and makes any conditions on approval explicit. The following sequence helps keep the analysis decision-focused.
Specify the proposed action, when it would begin, and the projection date or dates relevant to the decision.
Use the base projection and the adverse scenarios that expose vulnerabilities relevant to the proposal. State whether the figures are before or after modeled management actions.
Reflect both sides of the decision where appropriate: an action can change capital resources, internal capital need, or both.
Check each relevant scenario against the stated risk-appetite measure, not just against a favorable base projection.
Record assumptions, limits, and triggers that could change the decision. A proposal that passes only under an uncertain action or narrow margin may need redesign or conditional approval.
Worked Example: Testing a Growth Proposal
Worked Example: Testing a Growth Proposal
A hypothetical insurer is considering expanding a specialty coverage product. At the end of year 3, its current plan projects base-scenario resources of 560 million dollars and internal capital need of 440 million dollars. Under a relevant adverse scenario, it projects resources of 500 million dollars and need of 420 million dollars. The proposed expansion changes the year-3 base projection to resources of 550 million dollars and need of 450 million dollars; in the adverse scenario it changes resources to 465 million dollars and need to 435 million dollars. The insurer’s internal minimum coverage target is 120%. Should the proposal proceed unchanged?
Method. Calculate coverage in each scenario before and after expansion, then compare each result with 120%. All figures are in millions of dollars and refer to the same year-3 date.
Current plan, base scenario. Coverage is \(560/440=1.2727\), or 127.27%. Its headroom is \(560-440=120\).
Current plan, adverse scenario. Coverage is \(500/420=1.1905\), or 119.05%. The adverse projection is already below the 120% target; its headroom is \(500-420=80\), whereas the resources needed to meet the target are \(1.20(420)=504\). It is therefore short by 4 million dollars relative to the target.
Expansion, base scenario. Coverage is \(550/450=1.2222\), or 122.22%. The proposal meets the target in the base scenario, with headroom \(550-450=100\). Its target resources are \(1.20(450)=540\), so it has 10 million dollars above the target amount.
Expansion, adverse scenario. Coverage is \(465/435=1.0690\), or 106.90%. Required resources at the target are \(1.20(435)=522\), so the proposal is short by \(522-465=57\) million dollars.
Decision. The expansion meets the target in the base projection but fails it materially in the adverse projection. On these assumptions, it should not proceed unchanged. Management could assess a smaller expansion, additional risk mitigation, or a capital plan, then rerun the relevant projections. Approval would require evidence that the revised choice meets the threshold under the scenarios the insurer has identified as decision-relevant.
Common mistake avoided. Approving the proposal because its base coverage is 122.22% would ignore the adverse result of 106.90%. A base-case pass does not establish that the decision is consistent with a risk appetite intended to hold across adverse conditions.
Dividends and the Cost of Distributing Capital
A dividend generally reduces capital resources. If it does not also change the modeled capital need, its direct effect on coverage is straightforward: replace \(C\) with \(C-D\), where \(D\) is the dividend. The relevant amount is not simply the current surplus. It is the amount that can be paid while preserving the required coverage in the scenarios management has chosen to test.
This follows directly from the threshold definition. After a dividend, coverage must satisfy \((C-D)/K \ge r\). Multiplying by \(K>0\) and rearranging gives \(D \le C-rK\). Thus, \(C-rK\) is the maximum dividend under this simplified constraint. It is not automatically a recommended dividend: uncertainty, liquidity, restrictions on distributions, and other business needs can justify a smaller amount.
Worked Example: Setting a Dividend Limit From a Stress Projection
A hypothetical insurer projects stressed resources of 510 million dollars and internal capital need of 430 million dollars at a one-year horizon. Its internal minimum coverage target is 115%. The board is considering a 30 million dollar dividend. Assume for this calculation that the dividend reduces resources dollar-for-dollar and does not change capital need.
Method. Apply the coverage test to the stressed projection and calculate the maximum distribution consistent with the target.
Before the dividend. Coverage is \(510/430=1.1860\), or 118.60%. Target resources are \(1.15(430)=494.5\) million dollars, leaving \(510-494.5=15.5\) million dollars above the target.
Test the proposed dividend. Post-dividend resources would be \(510-30=480\) million dollars. Coverage would be \(480/430=1.1163\), or 111.63%, which is below 115%. The dividend therefore fails the stated stressed-scenario constraint.
Find the maximum under the simplified test. The largest dividend satisfying the target is \(D_{\max}=510-1.15(430)=15.5\) million dollars. If the board instead considers a 10 million dollar dividend, resources would be \(510-10=500\) million dollars and coverage would be \(500/430=1.1628\), or 116.28%. That alternative passes the numerical threshold, with \(500-494.5=5.5\) million dollars above target resources.
Decision. The proposed 30 million dollar dividend should not be approved on this projection. A 10 million dollar distribution passes this one test, but the board should consider whether a 5.5 million dollar margin over target is adequate and whether other scenarios or constraints are binding.
Common mistake avoided. The pre-dividend headroom \(510-430=80\) million dollars is not the amount available for distribution while maintaining 115% coverage. The correct limit is \(C-rK=15.5\) million dollars.
Reinsurance and Investment Choices
Risk-transfer and investment decisions may affect both projected resources and capital need. A reinsurance arrangement can reduce modeled losses or internal need, while its premium, recoverable amounts, counterparty exposure, and contract terms can affect the net result. An investment change can improve projected income in one scenario while increasing losses or capital need in another. The decision should reflect the complete modeled effect, rather than counting only the favorable component.
The following example treats the hypothetical reinsurer’s modeled recovery as collectible in the stated scenario. In practice, an ORSA decision should consider whether recovery timing, contractual terms, and counterparty risk are adequately reflected. A modeled improvement is not a substitute for checking that the risk transfer is effective and operationally feasible.
Worked Example: Comparing a Reinsurance Arrangement
A hypothetical insurer’s adverse projection before reinsurance has resources of 450 million dollars and internal capital need of 440 million dollars. A proposed treaty costs 8 million dollars, is projected to produce 24 million dollars of recoveries in that adverse scenario, and reduces internal capital need by 20 million dollars after the insurer’s modeled allowance for remaining risk. The target coverage ratio is 108%. Assume the recovery is included in resources and the premium is deducted from resources.
Method. Compare adverse-scenario coverage before and after the treaty, adjusting resources for both the premium and recovery and adjusting capital need for the modeled reduction.
Without the treaty. Coverage is \(450/440=1.0227\), or 102.27%. This is below the 108% target.
With the treaty. Resources are \(450+24-8=466\) million dollars. Capital need becomes \(440-20=420\) million dollars. Coverage is \(466/420=1.1095\), or 110.95%, which exceeds the 108% target. Required resources are \(1.08(420)=453.6\) million dollars, leaving \(466-453.6=12.4\) million dollars above target resources.
Decision. The treaty improves the modeled adverse coverage from 102.27% to 110.95% and passes the numerical target under the stated assumptions. That supports further consideration, not automatic acceptance. The insurer should confirm that the recovery assumptions, contract wording, price, and counterparty treatment are suitable and that the result is not dependent on an untested operational assumption.
Common mistake avoided. Counting the 24 million dollar recovery but ignoring the 8 million dollar premium would overstate resources as 474 million dollars. The correct amount is \(450+24-8=466\) million dollars. The capital-need reduction must also be supported by the modeled residual risk rather than assumed to equal the recovery.
Investment Decisions and Capital Plans
Investment decisions should be tested as joint changes to projected income, stressed asset losses, and internal capital need. A higher expected return does not by itself demonstrate that an allocation is suitable: the adverse projection may worsen, and the resulting coverage may breach risk appetite. If hedging or another risk-reduction action is included, its cost and effect should both appear in the projection.
Worked Example: Testing a Higher-Yield Investment Strategy
A hypothetical insurer’s current plan projects base-scenario resources of 510 million dollars and capital need of 400 million dollars. In an adverse market scenario it projects resources of 470 million dollars and capital need of 420 million dollars. A proposed higher-yield strategy adds 6 million dollars to base resources, but without a hedge would produce adverse resources of 454 million dollars and need of 440 million dollars. A hedge costs 3 million dollars, adds 18 million dollars to adverse resources relative to the unhedged strategy, and reduces adverse capital need to 425 million dollars. The internal target is 110%.
Method. Evaluate the unhedged and hedged strategy separately in both scenarios, including the hedge cost wherever it applies.
Unhedged base projection. Resources are \(510+6=516\) million dollars; need remains 400 million dollars. Coverage is \(516/400=129.00\%\).
Unhedged adverse projection. Coverage is \(454/440=1.0318\), or 103.18%. This fails the 110% target.
Hedged base projection. After the 3 million dollar hedge cost, resources are \(516-3=513\) million dollars. Coverage is \(513/400=128.25\%\).
Hedged adverse projection. Resources are \(454+18-3=469\) million dollars, and capital need is 425 million dollars. Coverage is \(469/425=1.1035\), or 110.35%. Target resources are \(1.10(425)=467.5\) million dollars, so the margin above target resources is only \(469-467.5=1.5\) million dollars.
Decision. The hedge changes the adverse result from a 103.18% failure to a 110.35% pass, but the margin is narrow. Management could approve only with appropriate limits and confidence in the hedge assumptions, or seek a larger margin before adopting the strategy.
Common mistake avoided. Looking only at the 129.00% unhedged base coverage would overlook the adverse shortfall. Also, adding the hedge benefit without deducting its cost would give incorrect resources: the correct adverse amount is \(454+18-3=469\), not 472 million dollars.
When a projection falls short of a threshold, the ORSA can inform a capital plan rather than force a binary choice between proceeding and abandoning the proposal. Potential responses include slowing growth, retaining earnings, changing risk transfer, revising an investment limit, or arranging a capital injection. Each response needs a quantified effect, a credible implementation path, and a test of residual solvency after the action. For a simple capital injection \(Q\) that increases resources without changing need, the target condition becomes \(C+Q \ge rK\), so the minimum injection under that assumption is \(Q_{\min}=\max(0,rK-C)\). If the action also changes capital need, both effects must be modeled.
Check Your Understanding
For each hypothetical case, use the stated internal target and assumptions. Amounts are in millions of dollars unless noted.
- A growth proposal projects base resources of 620 and capital need of 480, but adverse resources of 598 and need of 500. The target is 125%. Calculate both coverage ratios and the adverse resource shortfall, if any, relative to target. (Answer: Base 129.17%; adverse 119.60%; adverse target resources are 625, so the shortfall is 27.)
- Stressed resources are 455 and capital need is 400. The target is 110%, and a proposed dividend of 22 reduces resources dollar-for-dollar without changing need. Does it pass, and what is the maximum dividend under this single constraint? (Answer: Post-dividend coverage is 108.25%, so it fails; maximum dividend is \(455-1.10(400)=15\).)
- Before a treaty, stressed resources are 390 and capital need is 410. A treaty costs 6, produces modeled stressed recoveries of 22, and reduces capital need by 18. The target is 102%. Calculate post-treaty coverage and determine whether it passes. (Answer: Resources \(390+22-6=406\); need \(410-18=392\); coverage \(406/392=103.57\%\), so it passes.)
- A proposed portfolio shift has stressed resources of 632 before hedging and capital need of 612. A hedge costs 2, adds 14 to stressed resources, and reduces capital need by 20. The target is 109%. Calculate the hedged coverage and the resource shortfall or margin relative to target. (Answer: Resources \(632+14-2=644\); need \(612-20=592\); coverage \(108.78\%\); target resources are 645.28, so the shortfall is 1.28.)
- A capital plan projects stressed resources of 340 and capital need of 300 against a 115% target. A proposed injection of 18 increases resources without changing need. Calculate post-injection coverage and the amount by which resources exceed target resources. (Answer: Coverage \(358/300=119.33\%\); target resources are 345, so resources exceed target by 13.)