CI 5 Capital Investments and Capital Allocation
Capital investments, also called capital projects, are outlays with a useful life of one year or more, so they are carried among a firm long-term assets rather than passing straight through the income statement. Like most assets they are booked at cost. The income statement then absorbs the cost gradually as a non-cash depreciation or amortization charge across the asset useful life, which smooths the spending and lines it up with the benefits the asset produces. The statement of cash flows, by contrast, records the cash outlay when it happens.
Once in service, a capital asset is carried at cost less accumulated depreciation or amortization. As that accumulated charge grows, the net carrying value falls toward zero or toward a salvage value. The table below traces a single piece of equipment costing 120,000 that is depreciated 40,000 per year.
| End 20X0 | End 20X1 | End 20X2 | |
|---|---|---|---|
| Equipment (cost) | 120,000 | 120,000 | 120,000 |
| Less: accumulated depreciation | 0 | 40,000 | 80,000 |
| Equipment, net | 120,000 | 80,000 | 40,000 |
These assets are not only tangible property and equipment. A growing share of capital investment goes into digital capabilities and other intangible assets. The tools in this lesson apply either way, because the focus throughout is on cash flows, not on the physical form of the asset.
The four categories
Capital projects can be sorted by whether they maintain the business or grow it, which also lines up with their risk. Projects that keep operations running tend to be lower risk and easier to appraise; projects that push into new scale, scope, or entirely new lines carry more uncertainty, longer horizons, and larger commitments of capital.
| Category | Purpose | Risk |
|---|---|---|
| Going concern (maintenance) | Continue current operations, improve efficiency, manage risk | Low |
| Regulatory or compliance | Meet safety, compliance, or supervisory standards imposed by third parties | Variable, often unavoidable |
| Expansion of existing business | Grow scale or scope, research and development and acquisitions within the core | Low to moderate |
| New lines of business and other | Investments and acquisitions outside the current business | Often high |
Going concern projects
Going concern projects, or maintenance capital expenditures, keep the current business at its existing size: replacing assets that are wearing out, refreshing IT hardware and software, and steadily improving existing facilities, such as swapping data center cooling units for more efficient units. Because they usually replicate what the firm already does, they are relatively easy to evaluate; efficiency projects in particular come down to weighing the upfront cost against the periodic savings expected over time.
Managers and their lenders generally try to match the term of the financing to the life of the new asset. A utility that installs replacement generation equipment expected to last 30 years might fund it with a 30-year bond. This match funding lowers financing risk. Funding a long-lived asset with shorter debt introduces rollover risk, meaning uncertain cost or availability of refinancing before the asset reaches the end of its life, while borrowing for longer than needed can mean paying a higher long-term rate or buying back debt that is no longer required.
Issuers do not have to disclose how much of their capital spending is maintenance. Analysts often approximate annual maintenance capital expenditure with reported depreciation and amortization expense. That approximation is closer to the truth when the expected asset life is near the actual life and when historical cost is near replacement cost, both of which hold better for shorter-lived assets.
Regulatory compliance projects
Regulatory compliance projects are not discretionary; they are mandated to satisfy specific standards, for instance a new pollution limit or a rule obliging banks to monitor and report transactions. Such projects often raise expenses without adding revenue, yet they must be done to avoid fines or to keep operating. They can also cut both ways competitively: incumbents may find that new standards raise a barrier to entry that protects their margins, and firms that help shape the standards, or that adopt early, can reduce uncertainty and gain an edge. As standards tighten, a firm has to judge whether a business still clears its minimum return once the extra costs are counted; sometimes those costs can be passed to customers as higher prices, and sometimes the right answer is to wind the affected operation down.
Expansion and new lines of business
Expansion projects increase the scale of existing operations or extend the firm into adjacent products, regions, or markets. They introduce execution risks such as sourcing more inputs, clearing production and distribution bottlenecks, or underbudgeting the cost of winning new customers. These risks are greatest for early-stage firms without established operations, whose expansion is therefore usually equity financed; more established firms with a track record of successful expansion can more often use debt. Spending here can be heavy: pharmaceutical and energy exploration companies frequently invest more than 10 percent of annual revenue in new drugs and new reserves.
New lines of business and other projects push into activities largely unrelated to the current business, either as a startup-style bet on a new technology or model or as an acquisition in a new industry. They are usually the riskiest category, exposed to the hazards of an unfamiliar business and to the risk of overpaying. Analysts gauge growth prospects and management priorities by tracking the level and trend of expansion spending, often estimated as total capital expenditure minus the maintenance portion.
Because issuers rarely break out maintenance versus growth capital expenditure, a common analyst shortcut is to treat depreciation and amortization expense as the maintenance figure and to read the remainder of total capital expenditure as expansion. The estimate is rough: it holds up best for short-lived assets, where book life and replacement cost stay close to reality, and it weakens for long-lived assets bought many years ago at prices well below today cost.
Classify each project into maintenance (going concern) or growth (expansion or new business), and identify the specific type.
Capital allocation is the process a firm management and board use to make investment and capital-return decisions, aiming to earn risk-adjusted returns above what investors could get on similarly risky opportunities elsewhere. It resembles the way investors build portfolios, but it works at a far more granular level and draws on proprietary, non-public information: issuers invest in projects, not just in whole companies. For an outside analyst, the judgement that matters is whether management will steward capital well over the long run, which means examining the process, its adherence to sound principles, and above all the historical track record.
Steps in the process
- Idea generation. Ideas can come from anywhere, but they work best when management understands the competitive setting and the firm own capabilities. Many ideas come from operating managers and involve extending existing activities or adjacent businesses.
- Investment analysis. Managers project the size, timing, duration, and variability of the cash flows an investment is expected to produce, then test whether it is a wise use of capital. This is where NPV and IRR are applied.
- Planning and prioritization. Management ranks profitable opportunities that together add the most value on a risk-adjusted basis. A project that looks good alone can be less attractive once existing operations, competing projects, ESG factors, or financing limits are considered. When value-creating opportunities run out, remaining capital should be returned to shareholders so they can redeploy it.
- Monitoring and post-investment review. Actual results are tracked against projections, and investment levels may be raised or cut. This validates assumptions, exposes systematic errors such as chronic optimism, enforces operating discipline, and often seeds the next round of ideas.
The next three tools appear in order of how much project detail they need. NPV and IRR value individual projects from management own cash flow forecasts. ROIC works from the consolidated financial statements that an outside analyst can actually see.
Net present value
The net present value of an investment is the present value of its expected after-tax cash inflows minus its costs. In the simple case a single outlay at inception is followed by inflows.
The decision rule follows directly: invest when NPV is greater than or equal to zero, and decline when NPV is below zero. A positive NPV adds to shareholder wealth; a negative NPV destroys it. A zero-NPV project just meets the required return and leaves no margin for forecasting error. In practice NPV is one input among several, so a non-negative NPV is best read as necessary but not sufficient for going ahead.
Gerhardt Corporation weighs an outlay of €50 million today that returns after-tax cash flows of €16 million per year for four years plus €20 million in Year 5. The required rate of return is 10 percent.
Many real investments have unconventional patterns, with outflows not only at inception but also on later dates, for example when capacity is added mid-project. Here the timing must be specified explicitly, using a spreadsheet or the XNPV function, which does not assume evenly spaced periods.
Gerhardt weighs the same €50 million outlay but with an uneven schedule (amounts in millions of euros), including an outflow at t = 1.5.
| t | 0 | 1 | 1.5 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|---|
| Cash flow | −50 | 10 | −5 | 13 | 16 | 19 | 23 |
Internal rate of return
The internal rate of return is the rate of discount that forces the NPV to zero. It shares the same structure as the NPV equation, with IRR in the denominator in place of r.
The rule is to invest when IRR is at least the required return r, which in this role is called the hurdle rate, and to decline when IRR is below it. One caution: IRR equals an investor realized (geometric) return only if interim cash flows are reinvested at the IRR itself. If they are reinvested at a lower rate, the realized return is lower, and vice versa. NPV instead assumes reinvestment at r, which is often the more realistic assumption.
Return to the conventional Gerhardt project: €50 million out, €16 million for four years, then €20 million in Year 5.
An investment has cash flows of −1,000 at t = 0, +5,000 at t = 1, and −6,000 at t = 2. The signs change twice.
NPV and IRR usually agree, but when two projects are mutually exclusive one may have the larger NPV while the other has the higher IRR. Choose the higher NPV. NPV states the wealth increase in currency terms, whereas IRR reports only a rate and ignores the size of the project and the length of time the rate is earned. Many practitioners still prefer IRR because a percentage above the hurdle is easy to read, while a currency amount needs to be judged against firm size; once the data exist for NPV, the IRR is trivial to add. The theoretically sound criterion, though, is NPV.
Return on invested capital
Outside analysts cannot audit management project-level NPVs or IRRs; they see consolidated, highly aggregated statements. Return on invested capital (also called return on capital employed) fills that gap. It measures the after-tax profit earned on all the capital management has put to work, so it is an annual, company-wide return rather than a project return.
An issuer reports after-tax operating profit of 24,395 in Year 2 and the balance sheet below.
| End Year 1 | End Year 2 | |
|---|---|---|
| Cash | 4,364 | 6,802 |
| Short-term assets | 40,529 | 52,352 |
| Long-term assets | 287,857 | 279,769 |
| Total assets | 332,750 | 338,923 |
| Accounts payable | 35,221 | 50,766 |
| Short-term debt | 21,142 | 5,877 |
| Long-term debt | 112,257 | 106,597 |
| Share capital | 15,688 | 15,688 |
| Retained earnings | 148,442 | 159,995 |
| Total liabilities and equity | 332,750 | 338,923 |
ROIC can be split into a profit margin and a turnover term, which shows two routes to a strong return.
The left factor is the after-tax operating margin; the right factor is capital turnover, the sales generated per unit of invested capital. A high-margin firm can still post a low ROIC if turnover is weak, and a low-margin firm can post an attractive ROIC if turnover is high. ROIC has real strengths: it uses data an outside analyst can obtain, it accounts for the capital needed to earn the profit (unlike a bare margin), it is an aggregate view of value creation across all investments (which matters because investors buy the whole company, not single projects), and it can be set against the required return on both debt and equity to see whether value is being created. Because it blends debt and equity capital, it should be compared to a blended required return; using the equity return alone would overstate the bar.
Its limits are equally important. ROIC is an accounting measure, not a cash measure, so operating profit and cash flow can diverge. It is backward looking and can swing year to year with investment activity and conditions, so trends matter more than any single reading, and profitable investments can take years to earn competitive returns. As a highly aggregated figure it can hide strong and weak areas. There is also less agreement on how to measure it, especially the denominator, where practitioners may strip out some intangibles, excess cash, or certain long-term liabilities such as pension and deferred tax balances.
Larissa Soroka analyzes two mutually exclusive projects for ABC company. Investors require 8 percent. Management asks her to include the loss of revenue from existing products but to treat recently completed market research as a sunk cost. Cash flows (currency millions):
| Time | 0 | 1 | 2 | 3 | 4 |
|---|---|---|---|---|---|
| Project A | −18.5 | 4.5 | 6.0 | 6.0 | 5.5 |
| Project B | −33.5 | −2.5 | −1.0 | 24.0 | 25.5 |
The tools are simple to compute, but they leave wide room for error in how the inputs are built. A few principles keep the modeling honest, and a set of recurring pitfalls, some analytical, some behavioral, explain many poor decisions.
Principles
- Use after-tax cash flows. Decisions should rest on after-tax cash flows, not on accounting profit. That means reflecting taxes on the project cash flows, including the tax savings from non-cash charges like depreciation and amortization where these are deductible.
- Include only incremental cash flows, but look broadly. Ignore sunk costs, meaning outlays already incurred. Count only the cash flows that exist because of the investment. Crucially, a project often affects the rest of the firm: a positive spillover such as cost savings, or a negative one such as cannibalized sales of a similar product. Both are incremental and both belong in the analysis.
- Respect the timing of cash flows. The forecast timing, duration, volatility, and possible change in direction of the cash flows all matter. Shifting cash flows from one period to another changes both NPV and IRR.
Bradshaw forecasts pre-tax (gross) cash flows and related depreciation for a project, at a 6 percent required return and an 18 percent tax rate. In this jurisdiction, depreciation cannot be deducted for tax.
| Time | 0 | 1 | 2 | 3 |
|---|---|---|---|---|
| Gross cash flow | −7.50 | 4.50 | 4.50 | 6.00 |
| Depreciation | 0.00 | −1.00 | −1.00 | −1.00 |
Cognitive errors
Internal forecasting errors. Managers can get the inputs wrong, and outsiders may be unable to spot it until weak results reveal it. Common cases include mis-estimated costs or required returns; overhead items such as management time, IT support, and financial systems are especially hard to pin down, and firms often fail to model how competitors will respond.
Ignoring the cost of internal financing. The main funding source for large issuers is operating cash flow, and many teams treat that money as free but scarce, budgeting it against prior-year amounts while treating external debt or equity as expensive and reserving it for big items such as acquisitions. That logic is flawed. Internally generated cash is equity financing: it could have been paid out to shareholders, so withholding it still carries their opportunity cost. Whatever the source, the same risk-adjusted required return should apply; internally funded projects do not deserve a lower r. The bias is hard to isolate, but a reluctance both to raise outside capital and to hand capital back is a warning sign.
Inconsistent treatment of inflation. Analysis can be done in nominal terms (cash flows that include inflation) or in real terms (cash flows adjusted to strip inflation out), but the cash flows and the discount rate must match: nominal flows with a nominal rate, real flows with a real rate. Inflation rarely moves every price and cost by the same amount. Rising oil prices help a producer selling oil, yet over longer periods they lift production and capital costs, and they squeeze the refining and chemical operations many oil firms also own, for which crude is an input. And if actual inflation differs from what was assumed, after-tax cash flows come in better or worse than forecast, regardless of how carefully the real-versus-nominal choice was made.
Behavioral biases
Inertia. McKinsey researchers who examined over 1,600 listed US companies found a 0.92 correlation between a segment capital investment in one year and the next, a sign that budgets are anchored to prior-year figures. If spending in a segment stays flat or rises while its returns fall, the analyst should question the justification and whether the capital would do better elsewhere.
Leaning on accounting metrics like EPS. Managers are often rewarded for boosting return on equity, net income, or per-share earnings. Many value-adding investments actually depress these figures in the near term, while buybacks and cost cuts flatter them, so an over-focus on short-run accounting can steer a firm away from its shareholders long-run interest. Analysts look first at how compensation is structured, then compare capital spending to history and to peers; that said, low spending can simply signal a lack of good opportunities, in which case returning capital is the right call.
Pet project bias. Favored projects may skip rigorous analysis, or get analyzed with inflated projections. They are hard to see from outside because statements are aggregated, so the tell is in governance: concentrated ownership or control, weak board oversight, and executive pay that is not aligned with stakeholders.
Failure to consider alternatives or scenarios. Many firms never generate enough alternatives, and many skip breakeven, scenario, and simulation analysis of different outcomes. This can come from thin investment experience, for instance never having made a divestiture or lived through a failed investment. Failure is unwelcome, but a total absence of it over time may mean management is not taking enough risk.
The process so far assumed a firm decides everything at inception and holds one course for the life of a project. In reality many investments are a sequence of decisions: some taken now, others deferred until future events or information arrive. These deferrable choices are real options. Like a financial option, a real option gives the firm the right, not the obligation, to act later, and a firm should exercise one only when it adds value. Because that flexibility is contingent on how the future unfolds, it can lift the NPV of an investment well above what a single fixed plan would show.
Types of real option
- Timing option. Rather than invest now, the firm can wait, giving up near-term returns in exchange for better information. Sequencing investments means an initial project can create the option to make later ones.
- Sizing option. An abandonment option lets the firm exit after starting if results disappoint: if the cash from abandoning exceeds the present value of continuing, it should abandon. A growth (or expansion) option is the mirror image, letting the firm invest more when results are strong.
- Flexibility option. Once an investment is running, the firm may adjust operations. A price-setting option lets it raise prices when demand exceeds capacity, capturing value it could not get by producing more. A production-flexibility option lets it change output, for example by adding overtime or shifts, when demand differs from the forecast.
- Fundamental option. The whole value can hinge on an outside factor such as a commodity price: an oil well or refinery depends on the price of oil, a mine on the price of gold. A firm would drill only when prices are high. Many research and development projects behave the same way.
Evaluating projects with real options
Three approaches are common. First, if a project has a positive NPV even before counting its real options, and those options can only add value, the NPV is a floor and the firm should invest. Second, the firm can value the project without options, then subtract the option cost and add the option value.
Third, for investments that involve a series of future decisions and alternative outcomes, decision trees or option pricing models assign probabilities and timing to future events and fold them into the NPV. More sophistication does not guarantee more accuracy, because these models depend on unobservable inputs.
Gerhardt considers a €500 million facility for a new product, using a 10 percent required return. It assigns 60 percent to a successful launch worth €750 million in one year. If the product fails (40 percent), then at t = 2 the facility can be repurposed for an alternative product (30 percent) worth €600 million, sold to another firm (30 percent) for €400 million, or abandoned (40 percent) for €0.
ScolarCorp spends $30 million on market research at t = 0, with a 7.5 percent required return. If research shows high interest (70 percent), it upgrades the plant for −$55 million and then, at t = 2, a launch succeeds (50 percent) for +$200 million, fails (40 percent) for −$80 million, or the plant is sold (10 percent) for +$100 million. If interest is low (30 percent), it sells the plant at t = 1 for +$90 million.