Making a Decision When No Single Measure Is Enough
A market-risk action can reduce a stressed loss and still be a poor choice. A hedge may consume scarce collateral; an asset sale may realize a loss; an option may be effective but costly. Choosing among actions therefore requires testing the same candidate plans against several constraints, rather than ranking them on risk reduction alone.
This tutorial uses a hypothetical insurer case to show how to compare estimated stress losses, capital usage, liquidity needs, and cost. The numerical results are planning estimates, not observations about an actual insurer. This decision discipline is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.
A Compact Insurer Case
Suppose a hypothetical insurer has an estimated market stress loss of $18.0 million before taking further action. Its hard stress-loss limit is $12.0 million. Management sets a more cautious planning target of $10.0 million, so a plan can pass the hard limit yet still fall short of the desired buffer. Available capital is $33.0 million and the relevant internal capital floor is $26.0 million. The resulting capital headroom is $7.0 million. Of $11.0 million in identified liquidity resources, $5.0 million must remain available for operations, leaving $6.0 million deployable.
The insurer is considering three hypothetical actions. Amounts are in millions of dollars. For this case, capital consumption includes the stated premium, transaction cost, or realized loss; no tax effects, recoveries, or other offsets are assumed. Liquidity use is measured relative to the $6.0 million deployable amount: a negative number means the action releases liquidity. The estimates below are stipulated planning inputs. For the combinations considered later, the insurer's planning model treats the stated stress-loss reductions as additive; that assumption must be checked rather than presumed in an actual decision.
| Action | Estimated stress-loss reduction | Capital consumption | Net liquidity use | Description |
|---|---|---|---|---|
| H: futures hedge | $4.5 | $0.8 | $4.0 | Cash expense and collateral encumbrance |
| P: put options | $6.5 | $2.2 | $1.0 | Premium and associated liquidity use |
| R: asset rotation | $7.5 | $3.5 | −$2.0 | Realized loss and transaction cost; releases liquidity |
A compact test makes the trade-offs explicit. If \(L_0\) is the starting stress loss and \(\Delta L(a)\) is the estimated reduction from action plan \(a\), then residual stress loss is \(L(a)=L_0-\Delta L(a)\). If \(C(a)\) is capital consumption and \(Q(a)\) is net liquidity use, feasibility requires \(L(a)\leq 12.0\), \(C(a)\leq 7.0\), and \(Q(a)\leq 6.0\). Meeting the internal target additionally requires \(L(a)\leq 10.0\). All quantities in these inequalities are in millions of dollars.
Screening Individual Actions
Start by calculating each action's residual loss and remaining resources. This first screen distinguishes an action that is infeasible from one that is feasible but does not achieve the preferred target.
Worked Example: Screening Three Standalone Actions
Using the hypothetical insurer's assumptions, assess H, P, and R separately against the $12.0 million hard stress-loss limit, the $10.0 million internal target, $7.0 million of capital headroom, and $6.0 million of deployable liquidity. Identify the method and the main mistake to avoid.
Method. For each action, subtract its estimated reduction from the $18.0 million starting loss. Separately subtract capital consumption from capital headroom, and net liquidity use against deployable liquidity.
Action H. Residual stress loss is \(18.0-4.5=13.5\). It exceeds the hard limit by \(13.5-12.0=1.5\), so H fails the risk constraint. Capital remaining is \(7.0-0.8=6.2\), and deployable liquidity remaining is \(6.0-4.0=2.0\). The action fits the capital and liquidity constraints, but that does not cure its risk-limit failure.
Action P. Residual stress loss is \(18.0-6.5=11.5\). It passes the hard limit with \(12.0-11.5=0.5\) of margin, but misses the internal target by \(11.5-10.0=1.5\). Capital remaining is \(7.0-2.2=4.8\), and deployable liquidity remaining is \(6.0-1.0=5.0\). P is feasible against the hard constraints, but does not meet the more cautious target.
Action R. Residual stress loss is \(18.0-7.5=10.5\). It passes the hard limit with \(12.0-10.5=1.5\) of margin but misses the internal target by \(0.5\). Capital remaining is \(7.0-3.5=3.5\). Since R releases $2.0 million, its net liquidity use is \(-2.0\), leaving \(6.0-(-2.0)=8.0\) of deployable liquidity. This is an increase of $2.0 million, not a use of liquidity.
Check. H fails the hard risk limit; P and R pass it but fail the internal target. Each action remains within the stated capital and liquidity constraints. The common mistake is to call P or R a complete solution because it passes the hard limit. The calculation shows why the distinction between a hard constraint and an internal target matters.
Comparing Combinations and Justifying a Choice
When no standalone action meets the internal target, combinations can be assessed using the same calculations. In this hypothetical case, the planning model provides additive estimates for the combinations shown below. This is a case-specific assumption, not a general property of hedges or portfolio changes. Real positions can overlap in their stress response, so combined effects should be recalculated or otherwise validated before implementation.
Worked Example: Finding the Least-Cost Pair That Meets the Target
Consider H+P, H+R, and P+R. Assume their stress-loss reductions, capital consumption, and net liquidity use are the sums of the corresponding standalone estimates. Which pair has the lowest stated capital consumption while meeting the $10.0 million internal stress-loss target and both resource constraints?
H+P. The estimated reduction is \(4.5+6.5=11.0\), giving residual loss \(18.0-11.0=7.0\). Capital consumption is \(0.8+2.2=3.0\), leaving \(7.0-3.0=4.0\) of headroom. Net liquidity use is \(4.0+1.0=5.0\), leaving \(6.0-5.0=1.0\) deployable.
H+R. The estimated reduction is \(4.5+7.5=12.0\), giving residual loss \(18.0-12.0=6.0\). Capital consumption is \(0.8+3.5=4.3\), leaving \(7.0-4.3=2.7\) of headroom. Net liquidity use is \(4.0+(-2.0)=2.0\), leaving \(6.0-2.0=4.0\) deployable.
P+R. The estimated reduction is \(6.5+7.5=14.0\), giving residual loss \(18.0-14.0=4.0\). Capital consumption is \(2.2+3.5=5.7\), leaving \(7.0-5.7=1.3\) of headroom. Net liquidity use is \(1.0+(-2.0)=-1.0\), leaving \(6.0-(-1.0)=7.0\) deployable.
All three pairs meet the $10.0 million target and satisfy the stated capital and liquidity constraints. Their capital consumption is $3.0 million for H+P, $4.3 million for H+R, and $5.7 million for P+R. Therefore H+P has the lowest stated capital consumption among the qualifying pairs. This result alone does not make it the best operational choice: it leaves only $1.0 million of deployable liquidity, compared with $7.0 million for P+R.
The logic behind the residual-loss calculation is direct. Under the case's stated accounting convention, the starting stressed loss is \(L_0\), and the action is assumed to offset \(\Delta L(a)\) of that loss under the same stress. The portion not offset is therefore the starting loss less the offset, \(L(a)=L_0-\Delta L(a)\). This relationship requires the reduction estimate to refer to that same stress and to include the effects relevant to the chosen action. If a stress response is nonlinear, the reduction is not known, or action effects interact, the subtraction remains an accounting identity only after \(\Delta L(a)\) has been estimated consistently; standalone reductions should not simply be added without support.
Separate the hard risk limit from the internal target, and calculate capital and deployable liquidity headroom before ranking plans.
Use the same stress, units, valuation date, and treatment of expenses and liquidity for every candidate.
A plan that breaches any binding constraint is not made acceptable by favorable results on another measure.
Among feasible choices, weigh cost and remaining buffers, then challenge assumptions that could change feasibility.
Testing the Recommendation Under Uncertainty
A plan that passes using central estimates may fail when collateral needs rise or risk mitigation proves less effective than expected. The next example applies two explicit adverse assumptions to illustrate how such a challenge can change the preferred action.
Worked Example: Choosing a Plan That Survives Adverse Assumptions
Suppose the insurer requires its chosen pair to meet the $10.0 million internal target after a 20% haircut to estimated stress-loss reduction. It also tests collateral needs 25% higher than estimated. For this test, assume R still releases $1.0 million of liquidity after a conservative haircut from its estimated $2.0 million release. Which of H+P and P+R meets all constraints under these assumptions?
H+P stress test. Its estimated reduction is $11.0 million. After a 20% haircut, the usable reduction is \(11.0\times(1-0.20)=8.8\). Residual stress loss is \(18.0-8.8=9.2\), which meets the $10.0 million target. Capital consumption remains $3.0 million, within the $7.0 million headroom.
The original H+P liquidity use is \(4.0+1.0=5.0\). With collateral needs 25% higher, it becomes \(5.0\times1.25=6.25\). This exceeds the $6.0 million deployable limit by \(6.25-6.0=0.25\). Thus H+P fails the stressed liquidity constraint, even though it passes the stressed risk and capital tests.
P+R stress test. Its estimated reduction is $14.0 million. After the same 20% haircut, usable reduction is \(14.0\times0.80=11.2\). Residual stress loss is \(18.0-11.2=6.8\), meeting the target. Capital consumption is \(2.2+3.5=5.7\), leaving \(7.0-5.7=1.3\) of headroom.
For liquidity, apply the 25% increase to P's $1.0 million use, giving \(1.0\times1.25=1.25\). Use the stipulated conservative $1.0 million release from R, so net use is \(1.25-1.0=0.25\). Deployable liquidity remaining is \(6.0-0.25=5.75\), which is within the constraint.
Recommendation and check. Under these specified adverse assumptions, recommend P+R: residual stress loss is $6.8 million, capital consumption is $5.7 million, and net liquidity use is $0.25 million. It passes all three tests. H+P would be less costly in capital under central estimates, but its stressed liquidity use of $6.25 million breaches the limit. The recommendation depends on the stated assumptions and the decision-maker's requirement to withstand these particular adverse tests; different constraints or estimates could justify a different choice.
Common Traps and Decision Discipline
A feasibility calculation is only as useful as its definitions and assumptions. In particular, do not mix a risk measure with a resource constraint, or silently treat uncertainty as if it were a guaranteed hedge benefit. A concise recommendation should identify the selected plan, demonstrate that each binding constraint is met, and state the assumption most likely to reverse the decision.
- Passing one test is not passing all tests. H+P meets the stressed target but fails the stressed liquidity cap. Its risk result cannot offset that breach.
- Do not add unlike quantities. A $1 million risk reduction does not create $1 million of capital headroom unless the capital calculation explicitly recognizes that effect.
- Keep signs consistent. An action that releases liquidity has negative net liquidity use. Subtracting a negative amount increases the deployable amount remaining.
- Do not confuse cost with benefit. Capital consumption reduces headroom: for H+P, the correct calculation is \(7.0-3.0=4.0\), not \(7.0+3.0=10.0\).
- Challenge combinations. Additive action effects are an assumption in the hypothetical case, not a universal rule. Interactions, basis risk, model uncertainty, execution, and changing collateral terms can alter results.
Check Your Understanding
Use the hypothetical insurer case and its stated assumptions. Amounts are in millions of dollars.
- For P alone, calculate residual stress loss, capital headroom remaining, and deployable liquidity remaining. Does it meet the hard limit and the internal target? (Answer: Residual loss $11.5; capital remaining $4.8; deployable liquidity remaining $5.0; passes the $12.0 hard limit but misses the $10.0 target.)
- For H+R under the additive planning assumption, calculate residual stress loss, capital remaining, and deployable liquidity remaining. (Answer: Residual loss $6.0; capital remaining $2.7; deployable liquidity remaining $4.0.)
- For P+R, apply a 20% haircut to the estimated combined stress-loss reduction. Calculate the residual loss and state whether it meets the internal target. (Answer: Haircut reduction is \(14.0\times0.80=11.2\); residual loss is \(18.0-11.2=6.8\), so it meets the target.)
- For H+P, calculate net liquidity use if both actions' liquidity requirements rise by 25%. Does the plan meet the $6.0 liquidity constraint, and by how much does it breach or remain within it? (Answer: \(5.0\times1.25=6.25\); it breaches the constraint by $0.25.)
- Suppose a committee imposes capital consumption of at most $5.0 and net liquidity use of at most $4.0, while requiring residual stress loss no greater than $10.0. Among H+P, H+R, and P+R, which is the lowest-capital-consumption qualifying pair? (Answer: H+R; it has residual loss $6.0, capital consumption $4.3, and liquidity use $2.0. H+P exceeds the liquidity limit, and P+R exceeds the capital limit.)