A factory expansion, a software replacement, an environmental retrofit, and a new product line all compete for the same pool of capital. Capital allocation decides which wins, and the math matters more than the slide deck.
Capital investments are long-lived outlays whose cash returns spread across many years. Analysts describe four standard types of capital investment, ordered here from the most predictable to the least:
- Going concern (maintenance) projects. Replace worn equipment, upgrade IT, sustain current operations. Lowest risk because the cash baseline is well known.
- Regulatory, safety, and environmental. Mandated by law or regulator. Often produce no measurable revenue. Failure to invest means losing the license to operate.
- Expansion of existing business. Scale current products or geographies. Risk is moderate. The cash-flow model rests on familiar products.
- New lines of business and other. Greenfield products, research and development (R&D), pet projects. Highest uncertainty. This is where bias and politics show up.
KEY: Forecast uncertainty climbs from going-concern (lowest) to new business (highest). Required returns and scrutiny should rise with the category.
Common mistakes
- Picking IRR over NPV for mutually exclusive projects. In Example 4 the compact line has the higher IRR (18.9%) but only $271,906.27 of NPV. The full line has a 15.3% IRR and $568,848.27 of NPV. Trap: choosing the compact line because the percentage looks better.
- Using net income in ROIC. ROIC uses NOPAT, EBIT × (1 − t), because invested capital includes debt. Trap: $18M ÷ $190M = 9.47% instead of the correct 11.84% in Example 3. A second trap divides by ending capital ($22.5M ÷ $200M = 11.25%) instead of the $190M average.
- Including sunk costs. R&D already spent on a product is irrelevant to the go/no-go decision. Trap: subtracting $2,000,000 of prior research from the project's incremental cash flows.
Bottom line
- Four investment types: going concern (maintenance), regulatory/compliance, expansion of existing business, and new lines of business, climbing in risk and required return
- NPV rule: accept if NPV > 0. IRR rule (conventional cash flows): accept if IRR > required return. IRR is found by trial: bracket the rate where NPV changes sign, then interpolate
- ROIC = NOPAT ÷ average invested capital, where NOPAT = EBIT × (1 − t) and invested capital = interest-bearing debt + equity. Value is created when ROIC > WACC
- When NPV and IRR conflict on mutually exclusive projects, follow NPV. For conventional incremental flows, an incremental IRR above the required return confirms the larger project
Exam shortcut
When NPV and IRR conflict on mutually exclusive projects, always pick NPV. The IRR's reinvestment assumption and scale blindness are the trap. For an IRR without a calculator, bracket two rates with opposite-sign NPVs and interpolate. For ROIC, check the numerator first: a choice built on net income is the trap, and the right one uses EBIT × (1 − t) over average debt plus equity.
The full lesson (about 4,189 words, 28 min read) adds 10 worked examples, all 6 common mistakes, a self-check, free in the app.
Learning objectives
- capital investments and capital allocation
Browse all free CFA Level I lessons or jump into free CFA Level I practice questions.