Study the CBCF syllabus by mapping each tool to the question it answers: NPV measures value added at the cost of capital, IRR states a return rate, payback measures exposure time, and the profitability index measures value per unit of scarce budget. Drill incremental cash flow treatment and rationing math with worked examples until the choice of rule is automatic.
Keeping projected cash flows truly incremental
Project cash flows must capture only what changes if the project is accepted: exclude sunk costs, include opportunity costs and side effects on existing products, and trace working capital from investment to recovery.
Three adjustments decide most estimation questions. Sunk spending, such as market research already paid for, never enters the analysis because no decision can recover it. Opportunity costs enter even though no cash moves: if the project uses a warehouse the firm could rent out, the forgone rent is a real cost of accepting.
Side effects and working capital complete the picture. Cannibalization of an existing product line is a cost of the new project; a boost to complementary sales is a benefit. Working capital tied up at launch is an outflow at time zero and an inflow when recovered at the end. Interest and financing flows stay out of project cash flows because the discount rate already prices financing.
Worked scenario: a firm contemplates a machine costing 400,000 with 30,000 of working capital invested at launch, producing 190,000 of annual pre-tax cash flow for three years. It already spent 80,000 on research, and the launch will cut 25,000 per year from an existing product's cash flow. The discount rate is 10%. The common mistake subtracts the sunk research, ignores the erosion, and ignores the working capital recovery: 190,000 x 2.4869 = 472,511, minus 510,000, gives roughly -37,500, so the analyst rejects. The better decision counts only incremental items: 165,000 per year (190,000 less 25,000 erosion) x 2.4869 = 410,339, plus 30,000 recovered at year 3 x 0.7513 = 22,539, minus 430,000, gives roughly +2,900, so the project is marginally acceptable. The decision flips entirely on classification, which is why this topic deserves its own drill.
- Sunk costs: always excluded, regardless of size
- Opportunity costs: included at the value of the next best use
- Erosion and synergy: included as incremental side effects
- Working capital: outflow at start, inflow at recovery, both discounted
- Financing charges: excluded from cash flows, captured in the discount rate
NPV versus IRR when mutually exclusive projects differ in timing or scale
For mutually exclusive projects, rank by NPV at the cost of capital. IRR can disagree with NPV when projects differ in scale or in when cash arrives, and NPV is the rule that ties directly to value created.
IRR carries a reinvestment assumption that inflows are reinvested at the IRR itself, while NPV implicitly reinvests at the discount rate. When two projects conflict, the timing of the difference is the tell: the project returning cash earlier shows the higher IRR, while the one with larger later cash flows can show the higher NPV. Non-conventional cash flows, with a sign change after year zero, can produce multiple IRRs or none, leaving the rule unusable.
Worked scenario: Project S costs 1,000 and returns 600 in each of years 1 and 2; Project L costs 1,200 and returns 300 per year for six years; the discount rate is 10%. Project S has an IRR near 13.1% and an NPV of about 41; Project L has an IRR near 13.0% and an NPV of about 107. The mistake is choosing S because its IRR edges out L's. The better decision is L: at a 10% cost of capital it adds far more value because its cash flows compound longer and its scale is larger. The lesson is to compute both measures, then, for mutually exclusive choices, let NPV decide and use the timing of the cash flow gap to explain the conflict.
Note also that a project with an IRR above the discount rate always has a positive NPV when its conventional cash flow pattern holds, so IRR works acceptably as an accept-reject screen; it breaks down specifically in ranking problems and with non-conventional flows.
| Rule | Decision benchmark | Strength | Blind spot |
|---|---|---|---|
| NPV | NPV > 0; highest NPV among mutually exclusive projects | Directly measures value added | Needs a valid discount rate |
| IRR | IRR > cost of capital | Intuitive rate; no rate needed to compute | Conflicts with NPV on ranking; multiple IRRs possible |
| Profitability index | PI > 1; rank under budget constraints | Values each unit of scarce capital | Misleads when projects are indivisible |
| Payback | Payback within a policy cutoff | Simple exposure measure | Ignores time value and post-cutoff flows |
| Discounted payback | Cutoff on discounted flows | Respects time value | Still silent on flows after the cutoff |
Using payback and discounted payback as screens without letting them reject value
Payback measures how quickly the initial outlay is recovered and works as a liquidity or exposure screen; neither payback variant measures value creation, so a project failing a payback cutoff can still have strongly positive NPV.
Plain payback sums undiscounted cash flows until the initial investment is recovered; discounted payback does the same using present values, so it is always longer. Discounted payback has one property neither IRR nor plain payback shares: if a project reaches discounted payback within its life, its NPV must be positive, because the remaining discounted flows can only add value. That makes it a conservative screen.
The weakness is symmetric to the strength. Both measures ignore everything after the cutoff, so a long-horizon project with enormous late flows looks worthless, and neither rule says how good a project is, only how fast capital returns. In practice, treat a short payback as one supporting argument under capital scarcity or rapid obsolescence, and always pair it with NPV. When a practice question gives a payback cutoff and an NPV, check whether the two rules conflict and state which one governs the accept-reject decision: value, which means NPV, with payback as a secondary consideration.
Sensitivity, scenarios, and the three break-even levels
Risk analysis in capital budgeting changes one input at a time in sensitivity analysis, moves several inputs together in scenario analysis, and locates break-even points: accounting break-even, cash break-even, and the higher NPV break-even.
Sensitivity analysis recomputes NPV while varying a single input, such as unit sales or price, holding the rest fixed; it shows which variable the NPV estimate leans on most. Scenario analysis moves a consistent set of inputs together, such as a recession case with lower volume, lower price, and higher costs, and reports NPV for each named case. Simulation extends this by drawing inputs from distributions, but sensitivity and scenarios carry most of the analytical weight in coursework.
Break-even analysis gives the scenario analysis real teeth. Accounting break-even is (fixed costs plus depreciation) divided by the contribution margin per unit; it covers book costs only. Cash break-even drops depreciation from the numerator, so it sits lower. NPV break-even, sometimes called financial break-even, is higher still: it is the volume at which NPV equals zero, so it must also cover the opportunity cost of the invested capital. A project that clears accounting break-even can still destroy value, which is exactly the trap the three-level distinction exposes. When you see a break-even question, first identify which level is requested, then build the numerator accordingly.
- Sensitivity analysis: one variable moves, others held constant
- Scenario analysis: named cases move several variables together
- Accounting break-even: (fixed costs + depreciation) / contribution per unit
- Cash break-even: fixed costs / contribution per unit
- NPV break-even: volume where NPV = 0; the highest of the three
Allocating a fixed budget with the profitability index
Under capital rationing with divisible projects, rank projects by profitability index, defined as present value of inflows divided by initial investment, and fund down the ranking until the budget is exhausted.
The profitability index measures value created per unit of capital consumed, which is the right objective when the budget, not project quality, is the binding constraint. PI equals the present value of future cash flows divided by the initial investment, and its NPV equivalent is PI = 1 + (NPV / investment). Divisible projects make the ranking logic exact: you can take a fraction of the next project on the list. Indivisible projects break the simple ranking, and the correct bundle may skip a high-PI project to fit a combination with more total NPV.
Worked scenario: the budget is 500,000. Project A costs 300,000 with PV of inflows 450,000, so PI 1.50 and NPV 150,000; Project B costs 250,000 with PV 350,000, PI 1.40, NPV 100,000; Project C costs 200,000 with PV 250,000, PI 1.25, NPV 50,000. The mistake is ranking by raw NPV, which picks A plus C for 200,000 of NPV. The better decision, assuming projects are divisible, is A plus 80% of B: 150,000 plus 80,000, or 230,000 of NPV, using the full budget along the PI ranking. The difference of 30,000 is the direct payoff of matching the rule to the constraint. If projects were indivisible, you would enumerate feasible bundles instead and verify A + C against other fits.
A two-week practice sequence with a scored self-check rubric
Build one full model per week: week one covers a single conventional project through every decision rule, week two covers a mutually exclusive pair and a rationing problem, scored against a rubric of named checkpoints.
Week one exercise: invent a project with an initial outlay, three years of cash flows, working capital, and a discount rate of your choosing, then compute NPV, IRR, both paybacks, PI, and all three break-evens. Expected observations: discounted payback exceeds plain payback; PI equals 1 plus NPV divided by investment; IRR above the discount rate coincides with positive NPV; NPV break-even volume exceeds accounting break-even volume. Any contradiction points to a calculation error, not to theory.
Week two exercise: replicate the mutually exclusive pair from this guide, then the rationing set, and change one number, such as raising the discount rate, and watch which rules flip. Self-check rubric, scored 0-2 per item for a maximum of 14: (1) sunk costs excluded correctly; (2) erosion and opportunity costs included; (3) working capital timed at start and recovery; (4) NPV computed with the annuity factor correctly; (5) IRR interpreted with its reinvestment caveat stated; (6) mutually exclusive ranking resolved by NPV with the conflict explained; (7) rationing answer ranked by PI and budget fully allocated. Scores of 12 or above on a fresh, self-written problem are a reasonable milestone that you can apply the rules independently; the rubric measures practice mastery, not exam outcomes.
Then condense each rule to one card: definition, decision benchmark, and the one situation where it misleads. Review the cards against your written models rather than in the abstract, so each card is anchored to a number you computed.
Readiness checks before you sit the CBCF exam
You are ready when you can classify any listed cash flow as incremental or not within seconds, resolve an NPV-IRR conflict in writing, and allocate a rationing budget by PI without hesitating over the formula.
Run three timed checks. First, take a fresh cash flow list containing a sunk cost, an opportunity cost, an erosion effect, and working capital, and produce the correct incremental series in under three minutes. Second, present two mutually exclusive projects and state, in two sentences, which rule wins and why the other disagrees. Third, allocate a three-project, fixed-budget problem using PI and state the total NPV achieved.
Finish with a concept sweep of the syllabus topics: the capital budgeting process steps, the reinvestment assumptions behind NPV and IRR, the relationship between payback variants, the three break-even levels, and the divisibility condition that makes PI ranking exact. Any topic where you must pause to reconstruct the logic goes back into the week-two exercise rather than into passive rereading. Administrative details such as scheduling and registration belong with the issuer; see the note below the FAQ.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.