Capital reallocation is the real portfolio decision

What the paper develops

In the portfolios I have governed under a fixed funding ceiling, the annual plan was usually treated as the decision. Leaders argued for months. They agreed on a split of money and people across initiatives. Then they defended that split until the next planning round. The plan felt like governance. It was mostly administration.

The decision that actually governs a portfolio is made continuously, not once a year. Every time the evidence shifts, someone chooses whether to move funding and capacity toward the work that is now paying off, or to leave them where last year's argument put them. An organization that can move resources on evidence governs its portfolio; one that can only wait for the next budget cycle administers a plan it has already outgrown.

The gap is measurable. McKinsey's State of Organizations 2026 drew on more than 10,000 senior executives across 15 countries and 16 industries. It found that 47 percent of leaders review budget and talent only annually or less often, while only 30 percent of organizations reallocate resources enterprise-wide. The figures do not set a universal cadence, but they show how rarely the organization makes an enterprise-level move. Moving resources is how a strategy stops being a statement and becomes an allocation.

What an annual budget freezes

An annual plan is a snapshot of conviction, formed on the evidence available at one moment. The trouble is not that the snapshot is wrong but that the organization keeps acting on it long after the evidence behind it has changed. Research on major projects finds that front-end estimates of cost and benefit are poor predictors of what happens. So the numbers that justified last year's allocation were never accurate enough to freeze for twelve months.

When the plan becomes the decision, three things lock in place. Funding sticks to initiatives rather than to evidence, so a program missing every milestone keeps its capacity. Scarce specialists stay assigned to work that has become less valuable, because reassigning them feels like admitting the plan was wrong. And the portfolio review shifts into a conformance posture. It asks whether each initiative is on track against its plan, not whether it is still the best use of the capacity it holds. A project can be perfectly on plan and still be the wrong place for the next dollar and the next scarce specialist.

Reallocation is the portfolio decision

Portfolio management, done seriously, is the continuous act of selecting, prioritizing, accelerating, de-prioritizing, killing, and reallocating work under limited resources, not selection followed by a year of monitoring. The gates exist to accelerate winners and fail fast on losers, not to wave everything through.

There is a longer-run performance case for treating reallocation as the real decision. It belongs here as background evidence for the discipline, not as investment advice. McKinsey's study of resource reallocation tracked companies over 15 years. The top third of reallocators shifted an average of 56 percent of capital across their businesses. That top third earned roughly 30 percent higher total shareholder returns each year than the bottom third, and were about 13 percent more likely to avoid acquisition or bankruptcy. The mechanism is not financial engineering but discipline: companies that keep moving capital toward their better opportunities compound an advantage, while others slowly fund their own decline.

That evidence carries a caveat the discipline depends on. The same research is explicit that over spans under about three years, high reallocators underperformed. Moving on every quarterly number is churn, and churn destroys value. The advantage comes from consistent, evidence-led reallocation held over the medium term: moving when the evidence is real and durable, and holding when it is only noise.

The short-cycle review

The mechanism that makes this real is a short-cycle reallocation review. It is a standing routine, run monthly or quarterly, whose only job is to decide what should move — not a status meeting with reallocation bolted on. It rests on four elements: a trigger, a defined condition that puts an initiative's capacity back on the table; an evidence standard of baseline, current signals, and an honest outside-view read, not advocacy; a decision owner, a single accountable sponsor with the standing to move resources across initiative boundaries; and a move, an actual change to the allocation, or a recorded decision to hold and why.

The forces that keep money in place are ordinary and strong. There is the sunk-cost reflex. There is the political economy of holding a budget. There is weak evidence at the point of decision. And there is the simple absence of anyone with the authority to move resources across boundaries. That is why reallocation has to be engineered as a routine rather than left to good intentions. Left to good intentions, the money does not move.

The portfolio view that makes a move legitimate

The unit of analysis is the contested resource and the credible alternatives that could use it, not the one project that happens to be in trouble. A project-status deck is organized around activity. A reallocation brief must be organized around a choice: the resource at stake, the time window, the original benefit and baseline, the observed evidence, dependencies, alternatives, and the consequence of holding. Inaction also allocates capacity.

Government investment-management guidance makes the control concrete: use actual investment data, predefined performance thresholds, and portfolio-level review to adjust resources among investments when necessary. A threshold does not replace judgment; it makes the condition for reconsidering an allocation known before a sponsor is under pressure to defend it. The decision still has to consider dependencies, capacity, risk, and the effects on other work.

The trade-off is the point of portfolio management. Project sponsors can explain local evidence; the portfolio owner decides whether that evidence still warrants the capacity it consumes when compared with the enterprise alternatives. The decision record should preserve the trigger, evidence, alternatives, accountable owner, and expected next observation. That makes a move—or a deliberate hold—reversible on evidence rather than politics.

The honest test of portfolio governance is not the quality of the annual plan; any competent organization can produce one. The test is what happens in month four, when the evidence has moved and the plan has not. The plan tells you what you believed a year ago. Reallocation is what you do about what you know now. Only one of them is governance.