Tutorials › CAS Exam 9 - Risk Management for Actuaries › Common Errors in Rating-Driven Capital Planning

CAS Exam 9 · Ratings, Capital, and Management · Tutorial 940 of 995

Common Errors in Rating-Driven Capital Planning

Use agency-specific measures, realistic timing, and conditional action credit to test whether a capital plan remains adequate beyond its headline threshold.

Advanced 13 min read

What You'll Learn

  • Distinguish a planning threshold from a target that includes a buffer.
  • Keep rating agency, regulatory, and internal capital measures separate.
  • Calculate how capital changes and requirement changes affect a capital ratio.
  • Allow for implementation delays and the date an action becomes effective.
  • Test proposed actions using only the credit that is supported and recognized.
  • Identify common calculation and communication errors in rating-driven plans.

Why Capital Plans Can Look Stronger Than They Are

A capital plan can appear adequate in a presentation and still fail to support the intended rating outcome. A common reason is that a headline ratio is treated as the plan itself: the ratio meets a threshold under one set of assumptions, so management assumes the plan is secure. That conclusion may ignore uncertainty, the exact measure being assessed, when actions take effect, or whether an agency will recognize their full modeled benefit.

This tutorial focuses on those planning errors. The calculations use hypothetical, simplified capital measures to illustrate how to test a plan. They are not formulas for any particular rating agency: agency methodologies, definitions, and rating judgments differ. A real plan must use the relevant agency’s current measure and documented assumptions.

Definition: For a simplified capital-planning measure, let \(A\) be the capital amount recognized for the measure and \(C\) its positive modeled capital requirement. The capital ratio is \(R=A/C\). A threshold is a specified minimum ratio used in a decision; a planning target is the ratio management chooses to seek, potentially above that threshold to allow for adverse movement and uncertainty. The numerator and denominator must be defined consistently with the measure being assessed.

The ratio is a useful planning summary, but it is not a rating by itself. A rating outcome can depend on factors beyond a single capital ratio, and the ratio itself may not be the agency’s complete or published measure. Therefore, a ratio comparison is meaningful only when the measure, valuation basis, timing, and assumptions are stated.

Threshold Compliance Is Not the Same as Resilience

Managing only to a threshold asks whether a plan meets a minimum under a selected projection. Resilient planning asks whether it can continue to meet the relevant objective after plausible adverse changes. Those changes might include lower available capital, higher required capital, delayed earnings, or less favorable recognition of a management action. The appropriate buffer depends on the organization’s circumstances; it should not be chosen by treating a threshold as a guarantee.

Worked Example: A Plan That Meets the Threshold Only Before Stress

A hypothetical insurer has adjusted capital of $144 million and a modeled capital requirement of $120 million. Its planning threshold is a ratio of 115%. A proposed business expansion is expected to reduce adjusted capital by $6 million and increase the requirement by $8 million. Determine the projected ratio and the additional capital needed to reach the threshold after expansion. Then calculate the capital needed for a 125% planning target.

Method. Recalculate both the numerator and denominator after the stated changes, then compare projected capital with the amount implied by each target ratio. This avoids the common mistake of applying a capital change to the numerator while leaving the requirement unchanged.

Before expansion, the ratio is \(144/120=1.20\), or 120%. After expansion, adjusted capital is \(144-6=138\) million, and the requirement is \(120+8=128\) million. Thus:

$$ R=\frac{138}{128}=1.078125=107.81\%. $$

The projected ratio is below the 115% threshold. At a requirement of $128 million, capital of \(1.15(128)=147.20\) million is needed to meet that threshold. The shortfall is \(147.20-138=9.20\) million. For a 125% planning target, required capital is \(1.25(128)=160.00\) million, a shortfall of \(160.00-138=22.00\) million.

Check. The amounts needed are greater than current projected capital, and the 125% target requires more capital than the 115% threshold: $22.00 million versus $9.20 million. The pre-expansion ratio of 120% did not establish that the expansion was affordable.

A useful check is to translate a target ratio into a dollar requirement. If the target is \(k\) and the applicable requirement is \(C\), the corresponding capital amount is \(kC\). This does not make the target a universal agency standard; it makes the planning implication of a chosen target explicit.

Keep Capital Measures Separate

A rating agency measure, a regulatory solvency ratio, and an internal economic capital measure may use different capital definitions, risk charges, aggregation methods, valuation bases, and treatment of diversification. A strong result under one measure does not automatically establish strength under another. Even when two measures are expressed as ratios, their numerators and denominators may not be comparable.

A disciplined analysis identifies the measure used for each conclusion and calculates it from the appropriate inputs. Converting one ratio into another by assuming a constant relationship is not justified unless a specific, validated mapping is available and relevant to the decision.

Worked Example: A Strong Solvency Ratio Does Not Settle a Rating Measure

A hypothetical insurer reports rating-measure adjusted capital of $154 million against a rating-measure requirement of $125 million. Separately, it reports regulatory eligible own funds of $180 million against a regulatory capital requirement of $120 million. A proposed asset revaluation reduces rating-measure capital by $12 million, while a change in the rating measure increases its requirement by $5 million. Calculate both initial ratios and the revised rating-measure ratio.

Method. Calculate each ratio only from its own numerator and denominator. Then apply the stated changes only to the rating measure. This avoids treating the regulatory ratio as a substitute for the rating measure.

The initial rating-measure ratio is \(154/125=1.232\), or 123.20%. The regulatory ratio is \(180/120=1.50\), or 150.00%. They describe different measures.

After the revaluation and requirement change, rating-measure capital is \(154-12=142\) million, and the rating-measure requirement is \(125+5=130\) million. The revised ratio is:

$$ \frac{142}{130}=1.092307\ldots=109.23\%. $$

The regulatory ratio remains 150.00% under the information given, but that does not reverse or offset the rating-measure result. No regulatory changes were specified, and no validated mapping between the measures was provided.

Check. The revised rating ratio is below its initial 123.20%, while the separate regulatory ratio is still 150.00%. The appropriate conclusion is that the two measures give different information, not that one cancels the other.

Allow for Timing and Recognition

A management action contributes to a capital plan only when it can be executed and when its effects are recognized in the relevant measure. Signing, approval, settlement, operational implementation, and agency recognition may occur on different dates. A plan that uses a future benefit to support an earlier decision can therefore overstate near-term capital strength.

Recognition is also not automatically equal to a modeled estimate. The amount credited may depend on the terms of the action, supporting evidence, the measure’s rules, and the agency’s assessment. A careful plan distinguishes the full modeled benefit from the amount supported for the decision date.

Worked Example: Delayed Contribution and Partial Reinsurance Credit

A hypothetical insurer has adjusted capital of $126 million and a capital requirement of $105 million. A growth plan increases the requirement by $10 million. Management proposes a $10 million contribution that will arrive two quarters later and reinsurance modeled to reduce the requirement by $8 million once effective. At the interim decision date, the contribution has not arrived and only $5 million of reinsurance reduction is recognized. Calculate the ratio at the interim date and after both actions are fully effective.

Method. Calculate the interim position using only changes effective and recognized at that date; calculate the later position separately. This avoids counting future or unrecognized benefits as if they were already available.

After growth, before either action is recognized, capital remains $126 million and the requirement is \(105+10=115\) million. At the interim date, the contribution has not arrived, so it adds nothing to capital. The recognized reinsurance reduction is $5 million, so the requirement is \(115-5=110\) million. The interim ratio is:

$$ \frac{126}{110}=1.145454\ldots=114.55\%. $$

After the contribution arrives, capital is \(126+10=136\) million. Once the full modeled reinsurance reduction is effective and recognized, the requirement is \(115-8=107\) million. The later ratio is:

$$ \frac{136}{107}=1.271028\ldots=127.10\%. $$

Check. The later ratio exceeds the interim ratio because capital increases by $10 million and the requirement falls by a further $3 million, from $110 million to $107 million. The denominator calculation uses the full $8 million reinsurance reduction from the post-growth requirement: \(105+10-8=107\).

Test the Credit for Proposed Actions

A proposal is not the same as an executed, recognized action. For planning purposes, distinguish at least three amounts: the benefit in management’s model, the benefit supported by available evidence and terms, and the benefit recognized in the measure at the relevant date. If these differ, the plan should show the difference rather than silently assuming the most favorable amount.

This discipline is especially important when a plan appears to pass a threshold only if every proposed action receives full credit. The analysis should state what happens under the supported case and what additional evidence or execution would be needed before relying on a larger benefit.

Worked Example: Full Modeled Credit Changes the Conclusion

A hypothetical insurer has adjusted capital of $160 million and a modeled requirement of $130 million. A portfolio risk-transfer action is modeled to reduce the requirement by $6 million. Based on current documentation, only $2 million is supported for recognition. A 120% threshold applies. Calculate the ratio using full modeled credit and using supported credit, and state whether each meets the threshold.

Method. Hold capital constant and reduce the requirement by the applicable credit in each case. This avoids the common mistake of assuming that a modeled reduction is automatically fully recognized.

With full modeled credit, the requirement is \(130-6=124\) million, so the ratio is:

$$ \frac{160}{124}=1.290322\ldots=129.03\%. $$

With supported credit of $2 million, the requirement is \(130-2=128\) million, so the ratio is:

$$ \frac{160}{128}=1.25=125.00\%. $$

Both calculations exceed the 120% threshold, but the supported-credit calculation is the appropriate basis under the stated evidence. The difference between the two ratios is \(129.03\%-125.00\%=4.03\) percentage points.

Check. A smaller reduction in the requirement leaves a larger denominator and therefore a lower ratio. The supported case still passes, but it has less headroom than the full-credit case.

A Short Ratio Derivation for Plan Review

The effect of changing capital and the requirement can be stated exactly. Suppose the starting amounts are \(A\) and \(C>0\), and a plan changes them by \(\Delta A\) and \(\Delta C\), with \(C+\Delta C>0\). The starting ratio is \(A/C\), and the revised ratio is \((A+\Delta A)/(C+\Delta C)\). Subtracting the starting ratio gives:

$$ \begin{aligned} \frac{A+\Delta A}{C+\Delta C}-\frac{A}{C} &=\frac{C(A+\Delta A)-A(C+\Delta C)}{C(C+\Delta C)}\\ &=\frac{C\Delta A-A\Delta C}{C(C+\Delta C)}. \end{aligned} $$

The denominator of this difference is positive under the stated conditions. Thus, an increase in capital (\(\Delta A>0\)) raises the ratio if the requirement is unchanged; a reduction in the requirement (\(\Delta C<0\)) also raises it if capital is unchanged. If both change, the net effect depends on \(C\Delta A-A\Delta C\). This is why a plan must calculate both sides of the ratio rather than describe an action simply as “capital positive.”

Common Errors in Rating-Driven Planning

The following checks help expose optimistic or internally inconsistent plans before they are used to support a decision:

  • Stopping at the threshold. A point estimate that barely meets a threshold may not withstand adverse experience or changes in assumptions. State the chosen planning target and show the sensitivity or buffer behind it.
  • Substituting one measure for another. Do not use a regulatory ratio, internal model result, or another agency’s measure as proof of a result under the measure actually being assessed.
  • Mixing dates. Do not combine current capital with future contribution proceeds, or use a post-implementation benefit to justify a pre-implementation position. Show each decision date separately.
  • Crediting a proposal as though it were complete. Separate modeled, supported, and recognized benefits. Apply only the amount appropriate to the decision date and evidence.
  • Changing only one side of the calculation. Growth, risk transfer, or an asset change can affect both capital and the requirement. Recalculate both where the assumptions say both change.
  • Using an inconsistent denominator. For example, if a post-growth requirement is $115 million and a recognized action reduces it by $5 million, the revised requirement is $110 million—not $105 million. State the starting point for every reduction.

A reliable review makes the conclusion traceable: name the measure, define its numerator and denominator, show the effective date and recognition basis for each action, and recalculate under adverse or less favorable assumptions. This process is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.

Key takeaway: A capital plan should not rely on a threshold-only result, a different capital measure, a future benefit, or unconfirmed full credit. Its conclusion should follow from the correct measure, the correct date, and the amount of each action that is actually supported.

Check Your Understanding

For each hypothetical situation, calculate the requested result using only the stated measure and timing.

  1. A company has adjusted capital of $172 million and a requirement of $140 million. Expansion reduces capital by $4 million and increases the requirement by $6 million. What is the revised ratio, and how much additional capital is needed to reach a 120% target? (Answer: \(168/146=115.07\%\); target capital is \(1.20(146)=\$175.20\) million, so the shortfall is $7.20 million.)
  2. A rating measure has capital of $138 million and a requirement of $120 million. A separate regulatory measure has eligible funds of $168 million and a requirement of $112 million. A rating-measure haircut reduces its capital by $6 million and its requirement rises by $4 million. What is the revised rating ratio, and may the regulatory ratio be substituted for it? (Answer: Rating ratio \(=(138-6)/(120+4)=132/124=106.45\%\); regulatory ratio is \(168/112=150.00\%\), but it is a separate measure and cannot be substituted without a validated mapping.)
  3. A plan starts with capital of $90 million and a requirement of $75 million. Expansion reduces capital by $2 million and adds $5 million to the requirement. A proposed action would reduce the requirement by $3 million, but only $1 million is recognized before the decision date. What is the ratio at that date? (Answer: \(88/(75+5-1)=88/79=111.39\%\).)
  4. A company has capital of $210 million and a requirement of $180 million. Growth adds $12 million to the requirement. A proposed contribution adds $15 million to capital, and a risk-transfer action reduces the requirement by $10 million. At the decision date, only $8 million of the contribution has arrived and only $6 million of the requirement reduction is recognized. Calculate the ratio using full proposed credit and the ratio at the decision date. (Answer: Full proposed case \(225/(180+12-10)=225/182=123.63\%\); decision-date case \((210+8)/(180+12-6)=218/186=117.20\%\).)
  5. A plan has capital of $250 million and a requirement of $200 million. A stress reduces capital by $9 million and increases the requirement by $16 million. For a 125% target, calculate the stressed ratio, the additional capital needed if the requirement stays fixed, and the requirement reduction needed if capital stays fixed. (Answer: Stressed ratio \(241/216=111.57\%\); target capital is \(1.25(216)=\$270\) million, so $29 million additional capital is needed; with capital fixed, the requirement must be at most \(241/1.25=\$192.80\) million, a reduction of $23.20 million.)
Clinical Biostats