The largest capital market in the economy is internal — roughly $640 billion moved inside US multibusiness companies each year, more than equity and corporate debt issuance combined — and most of it does not move at all. This paper takes the measured premium for breaking allocation inertia and rebuilds capital allocation as what it should be: a standing, evidence-fed process rather than an annual negotiation.
This paper argues one thing: capital-allocation agility carries a measured shareholder premium, allocation inertia is a measured disease, and the binding constraint is neither analysis nor conviction — it is cadence. Three arguments carry the conclusion.
First, the evidence. Across more than 1,600 US companies over fifteen years, McKinsey found a third of businesses received almost exactly the prior year’s capital — a 0.99 correlation — while the most active third of reallocators earned on average 30% higher annual shareholder returns than the least active. Extending the sample twenty years widened the annual gap to 3.9 percentage points — enough, compounded, to make the dynamic reallocator worth roughly twice its sluggish peer.
Second, inertia is structural, not cyclical. The same research found the financial crisis changed reallocation behavior not at all — and there is a rational core to the cowardice: over horizons under three years, active reallocators underperform. The premium pays only to institutions with the information and the nerve to hold the position.
Third, the fix is cadence built on live evidence. When agents keep unit economics current continuously — return on capital by cell, not by annual template — allocation stops being an event and becomes a standing review with pre-agreed triggers. Machines keep the map current; people argue the strategy and make the call.
Executives speak of “going to the market” for capital, but the market that matters most sits inside the building. The roughly $640 billion allocated and reallocated annually within US multibusiness companies exceeds what equity and corporate debt markets supply together — which makes the CFO’s allocation table a larger capital market than Wall Street, run without a price mechanism. And its revealed behavior is remarkable: for a third of business units, next year’s capital is this year’s, at a correlation of 0.99. The single best predictor of allocation is not strategy, returns, or opportunity. It is last year’s spreadsheet.
The cost of that inertia is not rhetorical. Top-third reallocators — shifting an average of 56% of capital across business units over fifteen years — delivered roughly 30% higher annual total shareholder returns than bottom-third peers, a result McKinsey found consistent across sectors from mining to consumer goods. Over twenty years the annual TRS gap between high and low reallocators reached 3.9 percentage points; compounding does the rest, with the dynamic reallocator ending the period worth about twice as much. The nuance that separates this from a slogan: under three-year horizons, the relationship inverts. Reallocation depresses near-term results before it pays — the market loves it in the long run and punishes it in the short run. The premium is real, and it is gated behind patience.
Three forces hold the pattern. Anchoring: the annual process opens from last year’s base, and every deviation must be argued uphill against an incumbent number. Silo defense: the executives who control information about a business’s prospects are the ones whose capital is at stake, so the map of opportunity arrives pre-negotiated. And latency: allocation decisions are made annually on snapshots assembled quarterly — by the time the case for moving capital is visible in the pack, it has been true for months. The short-horizon penalty completes the trap: a leadership team measured on next year has a rational reason to leave the money where it sleeps.
The remedy is not a braver annual meeting; it is removing the annual meeting’s monopoly. In the target state, agents maintain the allocation map continuously — return on capital by cell, computed from live actuals rather than templates; options surfaced as conditions cross pre-agreed thresholds; the equivalent of a zero-based review available on demand because the evidence never goes stale (The Self-Steering Plan is the instrument; this paper is the discipline built on it). Allocation reviews move to a quarterly cadence with triggers, not templates. People do what only people can: set the strategy the map serves, argue the bets, and hold the multi-year position through the J-curve — with a board narrative built in advance for the years the premium costs before it pays.
The same inertia governs the resource that capital is supposed to follow. McKinsey’s research on talent flows finds companies that rapidly reallocate their people are 2.2 times more likely to outperform peers on total shareholder returns than slow movers — and every operator knows why the two stick together: a business unit’s best argument for keeping its capital is the team already sitting on it, and a team’s best protection is the budget line it occupies. Reallocation programs that move money without moving people ship capital to units that cannot absorb it; programs that move neither are strategy documents. The standing process this paper describes therefore keeps two maps current, not one — where the capital works and where the capability sits — and treats a divestment or an exit as what the arithmetic says it is: the highest-yield reallocation available, and the one the annual process is structurally least able to propose, because no unit ever nominates itself.
This is also where the machine’s neutrality earns its keep. An agent-maintained allocation map has no silo to defend and no career staked on an incumbent budget line; it will surface the exit candidate, the starved winner, and the over-fed legacy with the same indifference. The politics do not disappear — they move to where they belong, in the open, around a map nobody authored to protect themselves. That, more than any model, is what a standing allocation process buys: a negotiation that finally starts from the truth.
First, publish your reallocation rate. One number — the share of capital that actually moved — in the pack beside ROIC. What the board can see, the organization cannot quietly anchor away.
Second, move allocation to a quarterly cadence with pre-agreed triggers, fed by a continuously current map. The annual process can remain as ceremony; the decisions should stop waiting for it.
Third, pre-fund the courage. Agree with the board, in advance, the multi-year story for the years reallocation costs before it pays — because the evidence says it will cost first, and the institutions that harvest the premium are the ones that already said so out loud.
The Lights-Out Maturity Index: six questions, two minutes, no scales to interpret. Your anonymous result joins the inaugural Lights-Out Finance Survey — the benchmark this publication reports on.
Take the Maturity IndexBrowse all papersFounder & Editor of Lights-Out Finance. Big 4 partner in CFO Advisory & Finance Transformation with two decades across the Americas, EMEA, and APAC; DEng in AI (George Washington), MBA in Finance (Cornell), Master in Financial Engineering (Queen’s Smith), MEng/MBA (École des Ponts ParisTech); US CPA, CGMA, FRM, CQF, CTP, CDAA.