Growth Plan Rating Assessment Workshop

Introduction

A growth plan can improve expected earnings while reducing the capital cushion that supports a target financial strength rating. It can also make earnings less predictable, even when the new business has a positive expected return. A useful assessment therefore traces several effects together: expected earnings, capital consumption, volatility, and the headroom remaining after growth.

This is a decision framework, not a universal rating-agency formula. Agencies use their own methodologies and may adjust reported figures or apply qualitative judgment. For an internal assessment aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus, management should make its assumptions explicit and test whether the plan remains within a stated capital buffer.

Definition: Available capital is the capital resource recognized for the assessment. Required capital is the selected rating or management capital benchmark. Capital headroom is available capital less required capital. A headroom buffer is a management-selected minimum amount of headroom to retain; it is not necessarily an agency-published threshold.

$$ H=A-K $$

Here $A$ denotes available capital, $K$ denotes required capital, and $H$ denotes headroom. These measures must use a consistent valuation basis, perimeter, and date. In particular, a reported capital measure should not be compared with a required-capital figure calculated for a different group of entities or on a different basis without a reconciliation.

Trace the Plan Through Earnings and Capital

Begin with a defined starting position. Estimate the plan’s incremental earnings on a consistent basis, including material expenses and the cost of any reinsurance or financing. Then estimate its incremental required capital, net of any reinsurance recognition that is actually allowed in the chosen assessment. Finally, determine how much of the earnings is expected to remain in available capital.

For a one-period projection, if incremental earnings are retained in full and there are no other changes to available capital, then $A_1=A_0+E$, where $E$ is incremental retained earnings. If required capital changes by $\Delta K$, then $K_1=K_0+\Delta K$. Subtracting the second equation from the first gives the headroom bridge:

H_1=(A_0+E)-(K_0+\Delta K) =H_0+E-\Delta K

This identity explains why positive earnings alone do not establish that growth is affordable. A plan increases headroom only when retained earnings exceed the associated increase in required capital. If earnings are partly distributed, taxed differently from the stated measure, or not recognized as available capital, replace $E$ with the amount actually retained and recognized.

Worked Example: Test a Full Growth Proposal Against a Buffer

A hypothetical insurer has available capital of $620 million and required capital of $500 million. Management requires at least $80 million of headroom. A proposed annual underwriting expansion is expected to generate $12 million of after-tax earnings, all retained. Its gross increase in required capital is estimated at $90 million; recognized reinsurance relief reduces the net increase by $20 million. Calculate projected headroom and determine whether the full plan meets the buffer.

Method. Calculate starting headroom, calculate the net capital increase, and apply the one-period headroom bridge. Assume no dividends, capital issuance, other business changes, or additional changes in required capital.

Starting headroom is:

H_0=620-500=120\text{ million}

The net required-capital increase is the gross increase less recognized reinsurance relief:

\Delta K=90-20=70\text{ million}

Available capital after retaining the projected earnings is $620+12=632$ million. Required capital becomes $500+70=570$ million. Therefore:

H_1=632-570=62\text{ million}

The result is $18 million below the $80 million buffer. Equivalently, the bridge gives $120+12-70=62$ million. The full plan does not meet the stated condition. The method avoids the common mistake of treating the $12 million earnings contribution as sufficient evidence of capacity while ignoring capital consumption.

Assess Earnings Volatility Alongside Expected Earnings

Expected earnings do not show how widely actual results may vary around the expectation. A simple assessment can compare the standard deviation of existing annual earnings with that of the proposed growth portfolio. If $X$ is existing earnings and $Y$ is growth-plan earnings, their combined earnings are $X+Y$. For finite variances:

\operatorname{Var}(X+Y) =\operatorname{Var}(X)+\operatorname{Var}(Y) +2\operatorname{Cov}(X,Y)

If the standard deviations are $\sigma_X$ and $\sigma_Y$, and their correlation is $\rho$, then $\operatorname{Cov}(X,Y)=\rho\sigma_X\sigma_Y$. The formula therefore requires estimates for both standalone volatility and the dependence between the two earnings streams. Adding standard deviations directly is generally not the correct calculation; assuming zero correlation without support can also misstate the combined volatility.

This is a measure of variability, not a full probability model for losses or a rating-agency capital calculation. It does not capture every tail feature or qualitative consideration. Use it as one comparison in the decision, and state the period, units, and dependence assumption.

Worked Example: Combine Existing and Growth Earnings Volatility

A hypothetical insurer estimates annual existing earnings to have a mean of $50 million and a standard deviation of $30 million. A proposed growth portfolio has mean earnings of $12 million and a standard deviation of $16 million. The estimated correlation between the two earnings amounts is 0.25. Calculate the combined mean and standard deviation, and compare the latter with existing volatility.

Method. Add the means, then use the variance aggregation formula with the stated correlation. Assume both standard deviations refer to the same annual period and are measured in millions of dollars.

The combined expected earnings are:

E[X+Y]=50+12=62\text{ million}

The covariance is:

\operatorname{Cov}(X,Y) =0.25(30)(16)=120

The combined variance, in squared millions of dollars, is:

\operatorname{Var}(X+Y) =30^2+16^2+2(120) =900+256+240 =1{,}396

Thus the combined standard deviation is $\sqrt{1{,}396}=37.363$ million, or $37.36 million to the nearest cent. This is $7.36 million above the existing standard deviation of $30 million. The growth adds expected earnings, but it also increases the measured variability of total earnings.

A common error is to add the standard deviations and report $30+16=46$ million. That calculation ignores diversification from the positive but less-than-one correlation. The covariance calculation gives the correct standard deviation under the stated assumptions.

Translate the Assessment Into Conditions

A recommendation should distinguish between the economics of the plan and the conditions under which it fits the capital position. A useful test is whether projected headroom remains above the selected buffer after expected retained earnings and incremental capital requirements are included. If the plan is too large, management can assess a smaller scale, additional reinsurance, a different mix of business, or an explicit capital action. Any proposed remedy should be recalculated for both its capital effect and its effect on earnings.

A scale calculation is useful when the plan’s effects can reasonably be approximated as proportional to its size. Let $x$ be the fraction of the full plan, where $x=1$ represents the full proposal. If the full plan produces incremental retained earnings $E$ and an incremental required-capital amount $\Delta K$, then:

H_1(x)=H_0+xE-x\Delta K =H_0+x(E-\Delta K)

For a required buffer $B$, the condition is $H_1(x)\geq B$. This linear approach is only appropriate when earnings and capital effects scale proportionally over the range being considered. If pricing, expenses, diversification, reinsurance terms, or capital requirements change with volume, the plan must be recalculated at each relevant scale rather than extrapolated mechanically.

Worked Example: Find a Feasible Scale and Compare Reinsurance

Use the hypothetical insurer’s starting headroom of $120 million and buffer of $80 million. At full scale, the plan adds $12 million of retained earnings and $70 million of required capital after $20 million of reinsurance relief. Assume both effects scale linearly. Find the maximum scale that meets the buffer. Then assess an alternative reinsurance arrangement that increases total relief to $35 million but reduces full-scale retained earnings to $9 million because of its cost.

Method. Express headroom as a function of the plan fraction and solve the buffer inequality. Assume no other changes to capital, and assume the stated earnings and capital effects are proportional to $x$.

Under the original arrangement:

$$ H_1(x)=120+x(12-70)=120-58x $$

Require $120-58x\geq80$. Thus $58x\leq40$, so $x\leq40/58=20/29=0.689655$. The maximum plan scale is approximately 68.97%. At this scale, retained earnings are $12(20/29)=8.2759$ million and incremental required capital is $70(20/29)=48.2759$ million. The resulting headroom is $120+8.2759-48.2759=80.0000$ million, subject to rounding.

Under the alternative arrangement, net full-scale required capital is $90-35=55$ million, and retained earnings are $9 million. Therefore:

$$ H_1(x)=120+x(9-55)=120-46x $$

The buffer condition gives $46x\leq40$, or $x\leq40/46=20/23=0.869565$. The maximum scale is approximately 86.96%. At that scale, earnings are $9(20/23)=7.8261$ million and required capital increases by $55(20/23)=47.8261$ million, leaving $120+7.8261-47.8261=80.0000$ million. The alternative permits greater scale but produces lower retained earnings at full scale. It is not automatically preferable: management should consider the value and terms of the reinsurance as well as its capital effect. The calculation avoids the mistake of counting extra reinsurance relief without also recognizing its stated earnings cost.

Common Decision Traps

Comparing incompatible measures. Available capital and required capital must refer to a consistent date, entity perimeter, and basis.
Counting gross relief twice. Apply only the reinsurance relief recognized in the assessment, and do not subtract it again if the reported capital increase is already net of relief.
Equating expected earnings with capital capacity. Expected earnings affect available capital only to the extent they are retained and recognized; required capital may rise by more than those earnings.
Ignoring dependence. Growth earnings may move with existing results. Volatility aggregation needs a covariance or correlation assumption, not just two standalone standard deviations.
Relying on a single projected endpoint. State the buffer and identify the assumptions that could change the result. A recommendation can be conditional on verified pricing, reinsurance, and capital effects.

A disciplined recommendation states what the plan is expected to earn, how much required capital it consumes, how total earnings variability changes, and how much headroom remains against a chosen buffer. It should also specify conditions for proceeding, such as a maximum scale, confirmed reinsurance recognition, retained earnings, or a defined capital action. These are decision controls, not claims that any single calculation determines an external rating outcome.

Check Your Understanding

For each hypothetical case, use the stated assumptions and show the relevant headroom or volatility calculation before checking the final answer.

A company has available capital of $450 million and required capital of $370 million. A plan adds $6 million of retained earnings and $25 million of required capital. What is projected headroom, and how far is it above or below a $70 million buffer? (Answer: $61 million of headroom, $9 million below the buffer.)
Existing annual earnings have a standard deviation of $20 million. A growth portfolio has a standard deviation of $15 million, and their correlation is 0.40. If their means are $30 million and $8 million, calculate combined mean and standard deviation. (Answer: Mean $38 million; standard deviation $\sqrt{865}=$$29.41 million.)
Starting headroom is $100 million and the required buffer is $65 million. A full plan adds $7 million of retained earnings and $30 million of required capital. Under linear scaling, what is full-scale projected headroom and does it meet the buffer? (Answer: $77 million; yes, it exceeds the buffer by $12 million.)
Starting headroom is $90 million and the buffer is $60 million. Each full-plan unit adds $8 million of retained earnings and $38 million of required capital, with both effects proportional to scale $x$. Find the maximum feasible $x$. (Answer: $x=1.00$; the full plan leaves exactly $60 million of headroom.)
Starting headroom is $120 million and the buffer is $75 million. Proposal A adds $10 million of retained earnings and $55 million of required capital. Proposal B adds $14 million of retained earnings and $70 million of required capital. Assuming all earnings are retained, compare projected headroom and identify which proposal meets the buffer. (Answer: Proposal A leaves $75 million and meets the buffer; Proposal B leaves $64 million and falls $11 million short.)

What You'll Learn