CAS Exam 9 · ERM in Strategic Decisions · Tutorial 179 of 995

Building Risk-Informed Business Cases

Build a business case that connects expected value to the assumptions, resources, uncertainties, downside consequences, and responses that decision-makers need to assess.

Intermediate 12 min read

What You'll Learn

  • Define a business case that brings benefits, costs, resources, and risk assumptions into one decision record
  • Organize scenario cash flows and calculate expected net present value
  • Show how a planned response changes expected costs and downside outcomes
  • Make resource requirements and their costs explicit rather than treating staff capacity as free
  • Identify common calculation errors, including omitting probability or discounting

Why a Business Case Needs a Risk View

A business case is more than a forecast of benefits and an initial budget request. It is a structured argument for a decision: what the organization expects to gain, what it must commit, which assumptions support the estimate, how uncertain the outcome is, and what could happen if important assumptions fail. A positive expected value is useful evidence, but it does not by itself show whether the downside is acceptable or whether the organization can deliver the plan.

For an actuarial decision-maker, the business case should make the chain from assumptions to consequences visible. If expected savings depend on a new process reducing handling time, state the assumed reduction and identify what happens if it is smaller. If a response is proposed, record its cost and the exposure it is expected to address. This makes the analysis reviewable and helps decision-makers distinguish an attractive result from a result that depends on unsupported assumptions.

Definition: A risk-informed business case is a decision record that presents expected benefits and resource requirements together with the assumptions, uncertainties, adverse consequences, and proposed responses that may affect whether those benefits are achieved. Its financial analysis is one input to the decision, not a substitute for assessing strategic fit, risk appetite, operational feasibility, or the quality of the evidence.

Build the Case From Explicit Assumptions

A useful case identifies its decision and evaluation period first. It then separates initial investment, ongoing operating costs, expected benefits, and resource use. Resources can include staff time, systems capacity, expertise, and management attention. If internal staff time has a meaningful opportunity cost, include that cost in the financial analysis as well as reporting the resource quantity; otherwise, a project can appear inexpensive simply because its capacity demand is hidden.

Assumptions should be specific enough to test. “Efficiency improves” is not a usable assumption. “The process saves 1,000 staff hours annually after implementation, at a loaded cost of $50 per hour” is testable. The case should also state the evidence and uncertainty behind the estimate. Estimates may be informed by historical experience, external evidence, expert judgment, or a pilot; those sources do not have equal certainty, so the basis should be clear.

  • Benefits: State the benefit, its timing, how it is measured, and the assumptions required to realize it.
  • Requirements: Show initial and recurring expenditure, internal resource demand, dependencies, and any capacity constraints.
  • Uncertainty: Describe important sources of variation and use scenarios or sensitivity tests where appropriate.
  • Downside: State plausible adverse outcomes, their timing and consequences, and who or what would be affected.
  • Response: Identify proposed actions, their costs, and the assumptions about how they change the exposure.
  • Limitations: Record material omissions, data weaknesses, and factors not captured in the quantified analysis.

For a scenario-based financial analysis, suppose there are mutually exclusive and collectively exhaustive scenarios \(s\), with probabilities \(p_s\). Let \(C_0\) be the initial cost at time 0, and let \(CF_{t,s}\) be the net cash flow at the end of year \(t\) in scenario \(s\), after including the costs and benefits specified for that scenario. If all cash flows are annual, measured consistently, and discounted at effective annual rate \(i\), scenario net present value is

$$ NPV_s=-C_0+\sum_{t=1}^{n}\frac{CF_{t,s}}{(1+i)^t}, \qquad E[NPV]=\sum_s p_s NPV_s. $$

This calculation assumes the probabilities are appropriate for the decision being analyzed and sum to 1, the cash-flow timing is correctly represented, and the discount rate is suitable for the valuation purpose. It does not mean that a decision-maker should treat the expected value as a likely outcome. A scenario table or a separate downside measure is needed to show how widely outcomes may vary.

Worked Example: Comparing Scenarios for a Claims Process Upgrade

A hypothetical insurer is considering a claims process upgrade. All amounts are in thousands of dollars. The initial implementation cost is $300 at time 0. After the implementation and planned operating arrangements, the annual net cash flow is estimated to be $140 in the base scenario, $90 in the downside scenario, and $180 in the upside scenario, paid at each year-end for three years. The scenario probabilities are 0.50, 0.30, and 0.20, respectively. Assume an effective annual discount rate of 5%, no taxes, no inflation, no salvage value, and no other cash flows. Calculate each scenario NPV and expected NPV.

Method. Discount each scenario’s three annual net cash flows using the same annuity factor, then weight the resulting scenario NPVs by their probabilities. This keeps the scenario consequences visible instead of replacing them with a single average forecast.

Step 1: Calculate the three-year annuity factor.

$$ \frac{1}{1.05}+\frac{1}{1.05^2}+\frac{1}{1.05^3} =0.952381+0.907029+0.863838 =2.723248. $$

Step 2: Calculate scenario NPVs. In the base scenario, the present value of the annual net cash flows is \(140(2.723248)=381.2547\). Therefore, \(NPV_{\text{base}}=-300+381.2547=81.2547\). In the downside scenario, the present value is \(90(2.723248)=245.0923\), so \(NPV_{\text{downside}}=-300+245.0923=-54.9077\). In the upside scenario, the present value is \(180(2.723248)=490.1846\), so \(NPV_{\text{upside}}=-300+490.1846=190.1846\).

Step 3: Weight the outcomes. The expected NPV is \(0.50(81.2547)+0.30(-54.9077)+0.20(190.1846)=40.6274-16.4723+38.0369=62.1920\) thousand dollars. Rounded to the nearest dollar, the expected NPV is $62,192.

Check. The probability-weighted annual net cash flow is \(0.50(140)+0.30(90)+0.20(180)=133\). Its present value is \(133(2.723248)=362.1920\); subtracting the $300 initial cost also gives $62.1920 thousand. The downside NPV is negative even though the expected NPV is positive, so the expected result alone does not describe the range of outcomes.

Why Probability-Weighted NPV Works

The probability-weighted NPV follows from linearity of expectation. This result is useful because it explains when a scenario calculation can be checked by first averaging cash flows and then discounting them. It does not eliminate the need to show the scenarios: the two methods give the same expected value, but only the scenario presentation directly displays the possible adverse outcome.

1
Write the scenario cash flows.
For each scenario, \(NPV_s=-C_0+\sum_{t=1}^{n}CF_{t,s}/(1+i)^t\), where the initial cost is the same in every scenario.
2
Take the probability-weighted average.
Multiply each scenario NPV by \(p_s\) and sum. Because the probabilities sum to 1, the initial-cost term remains \(-C_0\).
3
Rearrange the finite sums.
\(\sum_s p_sNPV_s=-C_0+\sum_{t=1}^{n}\sum_s p_sCF_{t,s}/(1+i)^t=-C_0+\sum_{t=1}^{n}E[CF_t]/(1+i)^t\). Thus discounting expected cash flows gives the expected NPV under the stated assumptions.

The rearrangement is valid because there are finitely many scenarios and years, the cash flows have finite expectations, the discount rate is fixed across scenarios, and the same evaluation horizon and timing conventions are used. If scenario-specific discount rates are used, or if cash-flow timing differs, the formula must be adapted rather than applied mechanically.

Make Responses and Downside Consequences Visible

A proposed response belongs in the business case as both an action and an assumption. State its resource or cash cost, the exposure it is intended to change, and the expected residual consequence. Do not silently reduce a loss estimate because a control or other response is planned. Instead, identify the before-and-after assumptions and make clear whether the response cost is included in the cash flows.

A common way to represent uncertainty is to show a central scenario alongside adverse and favorable cases. A sensitivity test changes one material assumption at a time, such as adoption rate or annual savings, to show how the result responds. A scenario can instead combine several related conditions. Whichever approach is used, avoid giving a precise-looking probability where the evidence does not support it; explain the basis and uncertainty of the estimate.

Worked Example: Including a Risk Response in a Technology Case

A hypothetical insurer evaluates a technology investment. The initial project cost is $400 thousand, and expected annual net operating cash flow before a specified incident loss is $180 thousand at each year-end for three years. A planned safeguard costs $25 thousand at time 0 and is assumed to reduce the probability of a $200 thousand incident loss in year 2 from 20% to 5%. Assume the incident causes no other cash-flow effects, the loss is paid at the end of year 2, the effective annual discount rate is 6%, and there are no taxes, inflation, or other cash flows. Compare expected NPV with and without the safeguard, and show the downside if the incident occurs despite it.

Method. Discount the regular annual cash flows and the year-2 loss separately. Include the safeguard cost with certainty, but include the incident loss in expected NPV only after multiplying it by the relevant probability.

Step 1: Discount the regular cash flows. The three-year annuity factor is \(1/1.06+1/1.06^2+1/1.06^3=0.943396+0.889996+0.839619=2.673012\). The present value is \(180(2.673012)=481.1422\) thousand dollars.

Step 2: Discount the incident loss. Its year-2 present value if it occurs is \(200/1.06^2=200/1.1236=177.9993\) thousand dollars. With the safeguard, expected incident cost in present-value terms is \(0.05(177.9993)=8.9000\) thousand dollars. Expected NPV with the safeguard is \(-400-25+481.1422-8.9000=47.2422\) thousand dollars.

Step 3: Compare with no safeguard. Without the safeguard, expected incident cost is \(0.20(177.9993)=35.5999\) thousand dollars. Expected NPV without the safeguard is \(-400+481.1422-35.5999=45.5423\) thousand dollars. Under these assumptions, the safeguard increases expected NPV by \(47.2422-45.5423=1.6999\) thousand dollars.

Step 4: Show the adverse outcome. If the incident occurs despite the safeguard, NPV is \(-400-25+481.1422-177.9993=-121.8571\) thousand dollars. If it does not occur, NPV is \(-400-25+481.1422=56.1422\) thousand dollars. Weighting these outcomes gives \(0.05(-121.8571)+0.95(56.1422)=47.2422\) thousand dollars, consistent with the expected-NPV calculation.

Check and common mistake. The response improves expected NPV by about $1,700, but the incident case remains materially negative. A common error is to subtract the full $200 thousand incident loss as if it were certain, or to subtract only \(0.05(200)\) without discounting it. The correct expected present-value loss is \(0.05(200/1.06^2)=8.9000\) thousand dollars. The conclusion depends on the assumed probability, loss amount, and effectiveness of the safeguard; those assumptions should be recorded, not treated as established facts.

Include Resource Requirements in the Economics

A business case should state both the quantity of resources needed and their financial treatment. If an initiative requires staff time, report the expected capacity demand even when those employees are already on payroll. If the organization assigns a cost to that time, include it once in the cash-flow analysis; do not also subtract it a second time as a separate “resource charge.” If a resource is genuinely constrained, its availability can affect feasibility even where its accounting cost is small.

Worked Example: Testing the Cost of Internal Capacity

A hypothetical insurer proposes a three-year process improvement. It requires an initial $240 thousand and 0.5 full-time-equivalent year of specialist capacity in each of the three years. For valuation, the capacity is assigned a cost of $80 thousand per full-time-equivalent year. Other annual operating costs are $60 thousand. Gross annual benefits are estimated at $120 thousand in the downside scenario, $180 thousand in the base scenario, and $240 thousand in the upside scenario, each at year-end. Their probabilities are 0.25, 0.50, and 0.25. Assume an effective annual discount rate of 4%, no taxes, inflation, or salvage value, and that the capacity cost is included in annual operating costs. Calculate expected NPV and the break-even gross annual benefit.

Step 1: Record the resource cost once. Annual specialist capacity costs \(0.5(80)=40\) thousand dollars. Total annual cost is therefore \(40+60=100\) thousand dollars. The resource requirement is also reported separately as 0.5 full-time-equivalent year annually, or 1.5 full-time-equivalent years over the evaluation period.

Step 2: Calculate annual net cash flows. Downside net cash flow is \(120-100=20\) thousand dollars. Base net cash flow is \(180-100=80\) thousand dollars. Upside net cash flow is \(240-100=140\) thousand dollars.

Step 3: Calculate the annuity factor and scenario NPVs. The three-year factor at 4% is \(1/1.04+1/1.04^2+1/1.04^3=0.961538+0.924556+0.888996=2.775091\). The downside NPV is \(-240+20(2.775091)=-184.4982\) thousand dollars. The base NPV is \(-240+80(2.775091)=-17.9927\) thousand dollars. The upside NPV is \(-240+140(2.775091)=148.5127\) thousand dollars.

Step 4: Calculate expected NPV. The probability-weighted annual net cash flow is \(0.25(20)+0.50(80)+0.25(140)=5+40+35=80\) thousand dollars. Thus expected NPV is \(-240+80(2.775091)=-17.9927\) thousand dollars, or -$17,993 to the nearest dollar.

Step 5: Find the break-even gross annual benefit. To make NPV zero, required annual net cash flow is \(240/2.775091=86.484\) thousand dollars. Adding the $100 thousand annual cost gives a break-even gross annual benefit of approximately $186.48 thousand per year. The probability-weighted gross benefit is \(0.25(120)+0.50(180)+0.25(240)=180\) thousand dollars, below that threshold.

Check and common mistake. The expected gross benefit of $180 thousand less annual costs of $100 thousand is the expected net cash flow of $80 thousand used in the NPV calculation. A common mistake is to omit the $40 thousand capacity cost because it is internal, or to count it both in annual costs and again as a separate deduction. With the assumptions given, the expected financial case is negative and the capacity demand is explicit; decision-makers can now consider whether the benefits, assumptions, or feasible design warrant reconsideration.

Common Business-Case Traps

  • Reporting a point estimate without its assumptions. A forecast is not self-explanatory. Record the drivers that must hold for benefits to occur and identify the evidence supporting them.
  • Confusing expected value with a guaranteed or typical outcome. A positive expected NPV can coexist with a negative downside scenario. Present both, with the scenario assumptions.
  • Leaving resources out of the case. Show staff capacity and other scarce inputs, and include their economic cost consistently when appropriate.
  • Counting a response without its cost or uncertainty. A proposed response is not automatically effective. State its cost and the assumed change in exposure, and show residual downside.
  • Mixing timing or units. Use a consistent evaluation horizon, annual cash-flow timing, currency scale, and discount rate. Discount a year-2 loss for two years, not one.
  • Presenting precision beyond the evidence. Numerical detail can aid comparison, but it does not make uncertain assumptions certain. Explain material data limitations and avoid unsupported probability precision.

A complete business case supports a decision by making the benefits and commitments comparable, the uncertain assumptions inspectable, and the consequences of being wrong understandable. Its recommendation should follow from that evidence and the organization’s objectives and risk appetite, rather than from the expected NPV alone. This approach is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.

Check Your Understanding

Assume year-end cash flows, an effective annual discount rate, and no taxes, inflation, or salvage value unless a question states otherwise. Amounts are in thousands of dollars.

  1. A proposal costs $250 at time 0 and produces net cash flow of $100 at each year-end for three years. At 4%, calculate its NPV. (Answer: $27.51 thousand.)
  2. A project requires $500 initially and produces net cash flow of $220 at each of the next two year-ends. At 10%, calculate NPV and state whether the financial NPV is positive or negative. (Answer: -$118.18 thousand; negative.)
  3. A project has a 10% probability of a $300 loss at the end of year 2. At 5%, what is the expected present value of this loss? (Answer: $27.2109 thousand.)
  4. Three mutually exclusive scenarios have probabilities 0.20, 0.50, and 0.30, and NPVs of -$100, $40, and $120, respectively. Calculate expected NPV. (Answer: $36 thousand.)
  5. An investment costs $420 initially and returns the same net cash flow at each of four year-ends. At 5%, what annual net cash flow gives an NPV of zero? (Answer: $118.44 thousand per year.)
Clinical Biostats