Capital allocation in an uncertain world: how CFOs are rethinking the portfolio

In 2022, Meta committed roughly $87 billion in capital expenditurecapital expenditureCapital Expenditure (CapEx) is money spent to acquire, upgrade, or extend long-lived assets like equipment, property, or software that deliver value over multiple years.View full definition → across two years, betting heavily on the metaverse. By mid-2023, the company had reversed course, slashed headcount, and redirected capital toward AI infrastructure. The financial markets rewarded the pivot aggressively: the stock more than tripled between late 2022 and early 2024. The episode is not a story about the metaverse being wrong. It is a story about the speed at which large capital commitments can become liabilities when the underlying thesis shifts, and about the organizational capacity required to reallocate without being paralyzed by sunk-cost logic.

That capacity, more than any single investment decision, is what separates CFOs who create durable value from those who simply manage a budget cycle.

The structural forces reshaping capital allocation

Several forces are converging in 2026 to make allocation decisions harder and more consequential at the same time.

Interest rates, while no longer at their 2023 peaks in most advanced economies, remain structurally higher than the 2010s baseline. The cost of capital for a mid-sized industrial company sitting at 8 to 10 percent WACC changes the math on almost every long-duration investment, from factory builds to software platform bets. Projects that looked attractive at a 4 percent hurdle rate simply do not clear the bar anymore.

Geopolitical fragmentation is forcing supply-chain and manufacturing footprint decisions that are, in effect, capital allocation decisions in disguise. The CHIPS and Science Act in the United States, the European Chips Act, and a wave of industrial policy in Southeast Asia have created a landscape where governments are actively competing for corporate investment. For CFOs at companies like TSMC, Samsung, or Infineon, capital programs once driven purely by demand forecasts now carry a political risk dimension that must be priced explicitly.

AI infrastructure spending has introduced a new asymmetry into portfolio construction. Hyperscalers (Microsoft, Google, Amazon, Meta) are committing hundreds of billions collectively to data centers and custom silicon through 2028. For the vast majority of companies that are not hyperscalers, the question is not how much to spend on AI infrastructure but where in the value chain to participate as a user, an integrator, or a provider of specialized services. Getting that positioningpositioningThe mental space you want your brand to occupy in your target customer's mind relative to alternatives.View full definition → wrong has capital consequences that will compound for years.

Finally, activist pressure has shortened the patience horizon for underperforming asset portfolios. Elliott Management's interventions at BP, Southwest Airlines, and Honeywell between 2023 and 2025 all centered, in different ways, on the argument that management was misallocating capital or maintaining a portfolio structure that destroyed rather than created value. The activist thesis is rarely purely financial. It is a commentary on the CFO's allocation discipline.

What this means for the CFO

The practical implication is that capital allocation can no longer be treated as a once-a-year event embedded in the budgeting cycle. The companies that executed well through the rate shock of 2022 to 2024 shared a common characteristic: they had built mechanisms for real-time portfolio review, not just annual strategic planning. That means establishing a dedicated capital reallocation process with its own governance, separate from the operational budget, with explicit criteria for redeployment triggers.

McKinsey research published over several years consistently shows that companies in the top quartile of capital reallocation (defined as willingness to shift more than 30 percent of capital expenditure across business units over a decade) generate roughly 50 percent more total shareholder return than low-reallocators. The drag from inertia is that large. Most companies reallocate far less than they think they do, because business unit leaders resist, and because the CFO lacks the political standing or the analytical infrastructure to force the conversation.

Building that analytical infrastructure matters more than most finance leaders acknowledge. Granular business unit return on invested capital, tracked at a frequency that allows quarterly decision-making, is table stakes. Less common but more valuable: a portfolio heat mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → that pairs ROIC with growth trajectory and competitive position, updated continuously rather than in the annual strategy review. Alphabet's internal capital committee model, which requires business units to compete for incremental capital against explicit IRRIRRThe Internal Rate of Return is the discount rate that makes a project's net present value equal zero. It expresses an investment's expected annualized return.View full definition → thresholds and strategic alignment scores, is one operational example worth studying, even if not every company can replicate it at scale.

The cost-of-capital discipline deserves particular attention. With WACC now a live variable rather than a stable assumption, sensitivity analysis on discount rates is not optional. CFOs at manufacturing companies, in particular, need to stress-test capital programs under multiple rate scenarios before committing. The difference between a project that returns 9 percent IRR and a 10.5 percent WACC is a value-destroying investment approved by a team that anchored to last year's hurdle rate.

Portfolio exits matter as much as entries. One of the most persistent failure modes in large corporations is the reluctance to divest businesses that are absorbing capital without generating returns above the cost of capital. General Electric's prolonged struggle to right-size its portfolio before the eventual breakup into GE Vernova, GE Aerospace, and GE HealthCare is the canonical case. The CFO role in forcing divestiture discipline, with specific timelines and hard financial thresholds, is underappreciated.

Translating this into practice

• Run a portfolio ROIC audit now, not at year-end. Identify which business units are generating returns above WACC and which are structurally below it, and be honest about the trend direction, not just the current snapshot.
• Separate the capital reallocation conversation from the budget defense conversation. When the two are conflated in a single annual process, inertia wins almost every time.
• Price geopolitical risk explicitly in investment cases. A new manufacturing facility in a jurisdiction affected by tariff or subsidy policy should carry a scenario analysis that models two or three distinct political environments over the asset's life, not a single base case.
• Establish a divestiture shortlist with named assets and financial exit criteria. The shortlist does not need to become public, but it should exist, be reviewed quarterly, and carry a named executive owner.
• When evaluating AI-related investments, distinguish between operational spending (faster payback, lower risk) and platform bets (longer horizon, binary outcomes). Blending them into a single "digital transformation" budget obscures the risk profile and makes accountability impossible.

The CFO who controls capital allocation controls strategy, regardless of what the organizational chart says. That is not a theoretical point. It is what separates the finance function that influences long-term value from the one that reports on it after the fact.