The Asset That Became a Flow

23 min read

There is a number in Heidrick & Struggles' 2026 report that ought to unsettle more than it celebrates: requests for interim COOs grew 250% in a single year [1]. This is not an isolated figure. Requests for CEOs and chairs rose 38% year over year; demand for interim C-suite leadership has accumulated 151% growth since 2021; and interim CFOs now represent 51% of all requests for temporary leadership [1]. Read slowly what this means: the role that exists precisely to know a company's financial innards — its commitments, its skeletons, its accounting memory — has become the most rented position at the top of the hierarchy. The chief financial officer, historic guardian of continuity, is now the most fungible piece on the board.

Sunny Ackerman, who authors the report, frames it as a triumph: independent talent, she writes, "are moving to the center of how critical work gets done" [1]. And from the standpoint of the market Heidrick operates in, it is a triumph — liquidity has risen, matching has accelerated, finished specialization finds a buyer in days rather than quarters. But read through the lens of brain capital, the phrase describes something else: the conversion of executive discernment — the slowest asset in the entire economy to form — into a spot commodity. What is being rented at 250% annual growth is not working hours. It is accumulated judgment: the capacity to decide well in contexts that never repeat themselves, distilled over decades inside organisations that paid for that distillation.

This is the thesis this essay will sustain, with the sources arbitrating one another: the economy that celebrates cognition "on tap" — human leadership rented at record pace, or artificial intelligence agentically deployed — is consuming an inheritance whose replenishment it has stopped financing. The liquidity of brain capital is not a new model of value creation. It is the liquidation of a stock formed by the previous regime — long, broad and slow employment — which the very existence of the rental market is corroding. Every interim leader today was formed by a permanent job of yesterday. And no one is forming tomorrow's.

For more than half a century, human capital theory — from Gary Becker to the organisational economics descended from it — rested on a premise so obvious it was rarely stated: companies invested in the training of their staff because they expected to keep them. The long career was the financing vehicle for discernment. The large corporation functioned as an implicit school: internal rotations, mistakes absorbed by the organisation, gradual exposure to decisions of increasing complexity, unbilled mentorship. The permanent employment contract was, in practice, a capitalisation contract — the company advanced decades of learning against the promise of harvesting mature judgment at the end. The 151% growth in interim leadership since 2021 [1] measures, above all, the quiet dissolution of that contract: mature judgment came to be bought ready-made, and the question of who produces it was left orphaned.

Microsoft's Work Trend Index gives this mutation its sharpest vocabulary: "intelligence on tap," human-agent teams, a human-agent ratio proposed as a new management metric [2]. Intelligence as a utility — turn on the tap, pay for the flow, turn off the tap. The temptation is to attribute the change to the tool: AI agents made cognition liquid, therefore liquidity is AI's offspring. But Heidrick's data dismantles that causality with unwitting elegance: the same rental logic appears exactly where agents cannot reach — in the chief executive's office, in operations leadership, in the finance function, with increases of 38%, 250% and 14% respectively [1]. When the same pattern appears where the tool is absent, the tool is not the cause. AI is an instance — the most visible, not the deepest — of an earlier mutation: cognition changing status, from asset held to flow rented. The convergence between the two reports, reached by independent paths, confirms that the phenomenon is structural, not sectoral. And neither asks the next question: where will what they rent continue to come from.

There is, amid Heidrick's numbers, one that deserves a contrarian reading: demand for human-capital specialists — workforce planning, organisational change — grew 129% in a year [1]. The report presents this as proof of market sophistication. But Darrell Rigby and Zach First, in the Harvard Business Review of January-February 2026, offer the inverse key: "Too much change can traumatize your organization" — and the remedy they propose is not to manage constant transformation better, but to minimise the need for it [3]. Read against HBR, the 129% stops looking like a solution and starts looking like a symptom: organisations outsourcing the management of their own perpetual mutation instead of stopping it. It is human energy rented to compensate for architectural failures — a pattern this publication has already described in relation to organisational exhaustion, and which reappears here with a price tag. The previous edition of the same magazine proposed Jana Werner and Phil Le-Brun's "Octopus Organization" as a model of adaptability [4]; the market, meanwhile, chose the opposite path to internal adaptation — buying externally, piecemeal, what the organisation no longer generates internally.

The arbiter of this dispute arrives from where it would least be expected: a study on exceptional achievement cited by The Economist of 17 January 2026. Its conclusions are uncomfortable for any model of talent acceleration: child prodigies rarely become adult prodigies, and the mechanism that best predicts mature excellence is what researchers describe as "search and match" — broad interests, prolonged exploration, late specialisation [5]. Now, Heidrick's spot market pays precisely for the finished product of that process: finance, FP&A, PMO leadership, transformation — finished specialisation, immediately available, with interim CFOs absorbing 51% of requests [1]. The collision is head-on and no one flags it: the market buys the outcome of a slow, broad, protected journey — and its own logic makes that journey impossible to finance. One does not do "search and match" over twenty years in an economy that hires by the day. Rental presupposes the stock; the stock presupposes the regime that rental is destroying.

The lens of brain capital reveals here what conventional lenses — the labour market's, technology's, operational efficiency's — let slip: the difference between price and reproduction. The independent talent market is setting prices with increasing precision; it is not forming talent. And the distinction between what becomes commoditised and what escapes commoditisation is already visible at the top of the distribution. Michael Steinberger, in his portrait of Alex Karp, documents the limiting case: Palantir's cofounder received a compensation package worth $1.1 billion in 2020, becoming the highest-paid CEO of a listed company that year [6]. That sum, note, was paid to a mind defined by absolute non-fungibility — a man who never learned to drive, who communicated almost exclusively in German with Austrian and Swiss assistants, and whose personal obsessions shaped one of Silicon Valley's most controversial companies [6]. The contrast with the 51% of interim CFOs is not anecdotal: it is the snapshot of brain capital's bifurcation. The middle of the cognitive distribution — competent, certified, interchangeable — becomes a commodity rentable at efficient-market prices; the untradeable idiosyncratic, that which no platform can standardise, retreats to the extreme and captures monopoly rents. Liquidity does not distribute value: it concentrates it in whatever refuses to be liquid.

This publication wrote, in April, about the extraction of knowledge — intellectual labour converted into AI infrastructure, the patrimony of those who think transformed into the raw material of those who deploy agents. What the 2026 figures show is the next step in the same logic: it is no longer just accumulated knowledge that is extracted; it is the top human itself that becomes rentable infrastructure, and the victim has shifted from being the stock to being the process that produced it. Graham Weaver, who spent roughly twelve years studying how to win without crushing the soul, describes thirty years spent "running" — from high school through university and career, always expecting the next stage would finally bring the feeling of being enough, which never arrived [7]. His testimony and the study cited by The Economist converge, via paths that share nothing in common, on the same verdict: the acceleration model misjudges what actually creates value. Weaver's individual race is the intimate version of the liquidation that Heidrick's numbers measure at market scale — and it is through that door, the human cost of a balance sheet no one consolidates, that the argument must proceed.

The human cost has an accounting counterpart, and it is this that the lens of brain capital forces us to see: when an organisation replaces a permanent employee with an interim leader, it is not simply swapping one contract for another — it is moving discernment from the asset column to the operating-cost column. Heidrick & Struggles' figures, read this way, cease to describe a labour market and start to describe a balance sheet: requests for CEOs and chairs growing 38% in a year, interim COOs up 250%, CMOs 100%, and interim C-suite leadership demand that has accumulated 151% growth since 2021 [1]. Sectoral concentration — health and life sciences, consumer, technology and services — confirms this is not a niche: these are precisely the sectors where competitive advantage depends on specialised judgment, and they have been fastest to convert it into rented flow [1]. And the most requested skills — finance, FP&A, project management and PMO leadership — are exactly those requiring the most years of maturation inside real organisations, with real mistakes and real consequences [1]. The spot market buys maturation; it does not produce it.

The official thesis attributes this mutation to artificial intelligence, and it is here that arbitration becomes necessary. Microsoft's Work Trend Index describes a world of "intelligence on tap": human-agent teams, a new human-agent ratio as a management metric, cognition available at the tap like electricity [2]. But if AI were the cause, the rental pattern should be weaker precisely where agents cannot reach — and it is there that it is most pronounced. No agent replaces an interim CFO before a board of directors, and yet interim CFOs represent 51% of all interim leadership requests recorded by Heidrick [1]. When the same phenomenon appears with and without the tool, the tool is not the cause: it is an instance. Cognition has changed status — from asset held to flow contracted — and AI simply extends to the middle of the pyramid a logic the top had already adopted on its own account. The convergence between Heidrick's "independent talent moving to the center of how critical work gets done" and Microsoft's intelligence on tap is not a coincidence of language: it is the same mutation seen from two different floors of the building [1][2].

The report's most revealing figure is also the most uncomfortable when confronted with contemporary management literature: demand for human-capital specialists grew 129% in a year [1]. Heidrick presents this as proof of sophistication — companies taking workforce planning seriously. Rigby and First, in the January-February Harvard Business Review, offer the inverse reading: "Too much change can traumatize your organization," they write, arguing that the remedy is not to manage constant transformation better but to minimise the need for it [3]. Read against HBR, the 129% stops looking like a solution and starts looking like a symptom — organisations outsourcing the management of their own perpetual mutation instead of stopping it, renting firefighters instead of ceasing to set fires. The previous edition of the same magazine points the same way by another route: Christopher Marquis proposes "strategic hibernation" as a resilience model, and Werner and Le-Brun "the Octopus Organization" as adaptive architecture [4] — both structural alternatives to what the interim market monetises as a service. And there is a parallel data point that reinforces the suspicion: Işık and Goswami, in MIT Sloan Management Review, show that AI risk management lags in organisations due to internal cultural and structural problems [5] — precisely the kind of problem no specialist rented at 129% annual growth can solve from outside, because culture cannot be hired by the hour.

It is here that The Economist enters as arbiter of the dispute — not over whether the rental market works, but over where what it rents comes from. The study on exceptional achievement cited in the 17 January edition is categorical on one point: child prodigies rarely become adult prodigies [6]. The mechanism that best predicts adult excellence is the inverse of acceleration — broad interests, prolonged exploration, late specialisation: a "search and match" process that requires protected time, tolerance for deviation and, above all, structures that finance unproductive years before talent finds its fit [6]. Now, those structures used to be called careers. Permanent employment, for all its supposed rigidity, was the long-term financing vehicle for discernment: it paid for the years of broad search, absorbed the cost of formative mistakes, and delivered, after two decades, the CFO who today represents 51% of interim requests [1][6]. The incompatibility is structural, not cyclical: the spot market can only buy the finished product of a process whose own logic — slow, broad, late — the spot market itself has made impossible to finance.

What this lens makes visible, and which none of the sources sees in isolation, is the depreciation no one accounts for. Every interim leader hired in 2026 was formed by a permanent job of yesterday — by an organisation that paid, unknowingly, the positive externality that Heidrick now lives off. The report celebrates that independent talent has moved "to the center of how critical work gets done" [1]; Microsoft celebrates intelligence on tap [2]; neither asks who pays for the next tap. A market that rents discernment without financing its formation is not creating value: it is consuming a vintage. And vintages, by definition, run out — not suddenly, but cohort by cohort, as the generation formed in stable careers ages and the next generation, raised in flow, reaches the top without the 20 or 30 years of broad search that The Economist identifies as the raw material of adult excellence [6].

For a board of directors, the implication is not nostalgic — it is allocative. If the middle of the cognitive distribution becomes commoditised (interim CFOs at 51%, COOs at +250%) and the idiosyncratic captures monopoly rents (Karp's $1.1 billion in 2020 [6][1]), then the board's question stops being "what talent do we rent?" and becomes "what cognition do we insist on retaining?" Cook and Nohria, in the November-December HBR, offer the criterion: the best CEOs do not accumulate talent — they build systems and cultures that help everyone execute better [4]. That is precisely the internal formation engine that rental does not replace. The company that rents everything rentable and cultivates nothing ends up with a cognitive balance sheet composed exclusively of short-term liabilities — efficient until the day the market, having exhausted the stock it inherited, has nothing rare left to rent it.

BOX — The Cognitive Vintage: stock, flow and invisible depreciation

Call it the cognitive vintage: the stock of discernment accumulated by a generation of executives formed in long, broad and slow careers, which the on-demand talent market consumes without replenishing. The fundamental distinction is accounting-based. A flow is rented and paid for by the day; a stock is built over decades and depreciates in silence. The systemic error of 2026 is treating the latter as if it were the former.

The stock forms through a mechanism The Economist documents: broad search, late specialisation, "search and match" [6]. There is no known shortcut — accelerated prodigies rarely turn into adult prodigies, and Graham Weaver's testimony, thirty years spent "running" without any achievement ever producing the feeling of being enough, is the intimate version of the same verdict [6][7]. What creates top-tier cognitive value is slow; what the market buys is immediate.

The flow, in turn, is measured in Heidrick's numbers: +38% in CEO requests, +151% in interim leadership since 2021, 51% of all requests concentrated in interim CFOs [1]. Every one of these transactions is rational for buyer and seller alike. The irrationality is aggregate: no one, at any point in the chain, finances the formation of the stock that every transaction presupposes.

The depreciation is invisible because it appears on no individual balance sheet. The company that rents an interim COO does not record the erosion of the vintage; nor does the platform that places them; nor does the executive themselves, in monetising decades of training paid for by previous employers, least of all. It is the classic problem of the commons, transposed onto cognition: the pasture is the discernment accumulated by the previous employment regime, and every shepherd is innocent.

The extreme of the distribution confirms the rule by inversion: the $1.1 billion paid to Alex Karp in 2020 [6] is the market price of what no tap supplies — a mind that is non-negotiably idiosyncratic. When the middle liquefies, rents take refuge in what refuses to be liquid. For the board, the lesson is direct: identify which cognition, within the house, is vintage and not flow — and protect it, because the market no longer produces it.

Implications for decision-makers

For the CEO, the first consequence is accounting before it is strategic: the company's balance sheet does not distinguish between rented cognition and owned cognition, and this blindness carries a price. Heidrick & Struggles' report documents that requests for CEOs and chairs grew 38% year over year and that interim C-suite leadership has accumulated +151% since 2021 [1] — numbers any board reads as flexibility gained. Read through the lens of brain capital, they say something else: a growing share of the company's decision-making capacity has moved outside its perimeter and into the market, subject to spot pricing, competing demand, and unavailability at the critical moment. The question the board should ask is not "how much does renting cost?" but "which decisions have we stopped knowing how to make alone?" — because the second bill never appears on the first.

For the investor, the bifurcation the data reveals redefines where to look for durable rents. The same market that treats the CFO as a fungible unit — 51% of all interim leadership requests are concentrated in that function [1] — paid Alex Karp $1.1 billion in 2020, the highest package of any listed CEO that year [6]. The implication is direct: when the middle of the cognitive distribution becomes commoditised, the premium migrates to the untradeable — to cognition with no spot-market substitute. In valuing a company, the relevant question stops being "what talent does it have?" and becomes "what talent does it have that cannot be rented?" That is the fraction of human capital that sustains multiples; the rest is operating cost under a different name.

For those managing transformation, the cross-reading of HBR and Heidrick imposes a diagnostic inversion. Rigby and First argue that "too much change can traumatize your organization" and that the remedy is to minimise the need for constant transformation [4]; Heidrick, meanwhile, records 129% growth in a year in demand for human-capital specialists [1]. Read together, the two figures suggest that a significant share of demand for transformation talent is not leverage — it is painkiller. Organisations that need to rent, year after year, someone to manage their own mutation are not solving a skills problem; they are masking an architecture problem. The lucid decision-maker treats every renewed interim contract as a diagnostic signal, not as a renewed solution: if the same type of gap reappears for three straight quarters, the gap is not one of people — it is one of design.

For the head of talent, The Economist's evidence on exceptional achievement — early prodigies who rarely become adult prodigies, and the "search and match" mechanism as the best predictor of excellence [6] — has an uncomfortable operational translation: high-potential acceleration programmes, as usually designed, optimise for the wrong indicator. Early specialisation produces visible performance early and scarce discernment late; broad search produces the reverse. In an economy where finished discernment can be rented by the day [1] but no one finances its formation anymore, the company that maintains long, broad and protected career paths will be building, almost alone, the asset all its competitors will have to rent. It is a structural arbitrage position — expensive in the short run, monopolistic in the long run.

For the board as a whole, the final implication is one of governance: the ratio between owned cognition and rented cognition should become a regular reporting metric, with the same seriousness as financial leverage is reported. Microsoft's report on the frontier of work already proposes the human-agent ratio as a management metric for "intelligence on tap" [2]; the same discipline should apply to the human tap. A company whose critical decisions depend more than half on external cognition — interim, advisory or agentic — is leveraged in an asset it does not control, in a market whose stock, as the data suggests, is running out without replenishment [1][6].

The liquidity paradox

The central tension of this analysis is that talent liquidity appears to solve exactly the problem it worsens. The company that rents an interim COO — a segment whose demand grew 250% in a year [1] — obtains immediate execution and avoids the cost of building that capability internally. But every rational rental reduces, across the economy, the incentive for the long-term investment that produced that COO in the first place. The spot market is efficient at allocation and destructive at formation: it allocates perfectly a stock it prevents from replenishing itself. It is the cognitive version of unregulated fishing — every boat is rational, the fish stock is finite, and the efficiency of each catch accelerates the collapse of the whole.

The paradox deepens when one observes who benefits from the liquidity. Today's interim executive monetises, at spot-market prices, decades of training that previous employers paid for under the permanent employment regime [1]. It is the perfect deal for their generation — and impossible for the next, because the employers who would pay for that training are, precisely, replacing it with rental. Brain capital's liquidity is, in this exact sense, an intergenerational transfer disguised as market innovation: the generation formed by the old regime sells to the new regime what the new regime no longer produces.

There is a second, more subtle tension between discourse and mechanism. Heidrick states that independent talent is moving "to the center of how critical work gets done" [1] — and it is right on the descriptive plane. But critical work, by definition, is that whose context matters: top-level decision depends on knowing the company's history, the board's balances, the scars of past crises. Rented cognition arrives without that context and departs before accumulating it. The more critical the work, the greater the context discount — and yet it is precisely in the most critical work, the C-suite, that rental grows most [1]. The market is applying commodity logic to the segment where commodity logic least applies.

The third face of the paradox is temporal. Weaver describes thirty years spent "running," in which every early achievement never produced the feeling of being enough [7]; The Economist shows that excellence forced early rarely survives into adulthood [6]. The liquid-cognition economy institutionalises exactly that race: careers fragmented into short mandates, evaluated by immediate deliverables, without the broad time in which "search and match" operates. The system that most depends on mature discernment is the one that least tolerates the conditions under which discernment matures.

What comes next

The most likely scenario for the 2027-2030 horizon is the continuation of the bifurcation already visible in the data. The middle of the cognitive distribution — specialised, replicable, well-documented functions — will keep commoditising, pressured simultaneously by the human interim market [1] and by agents' "intelligence on tap" [2]. Prices in that layer will fall as agent-based supply matures; the interim CFOs who today represent 51% of requests [1] will face, within a few years, competition from systems that execute FP&A without fatigue or notice period. The idiosyncratic extreme, by contrast, will see its premiums grow — the Karp precedent [6] is not an anomaly, it is an early signal of where rents take refuge once everything else turns liquid.

The second scenario, less visible but more consequential, is the exhaustion of the vintage. If the top-tier discernment stock was mostly formed by the stable employment regime, and that regime is contracting at an accelerating pace, then somewhere in the next decade the interim market will face a supply problem neither of the current reports anticipates [1][2]. The first symptoms will be subtle: lengthening placement times, fee inflation without quality improvement, recycling of the same names on ever-faster rotation. When those signs appear, companies that maintained internal formation capacity will be in the position of holding the only mine in an economy of traders.

The third scenario is the intelligent organisational response: the emergence of companies that treat the formation of discernment as a deliberate competitive advantage, not as an inherited cost. The lesson from private equity that HBR documents — "successful PE firms have found a more straightforward way to create surplus value" [3] — applies here by analogy: the surplus value of the next decade will lie in owning the formation process, not just the product. Companies that offer long, broad and protected paths — the opposite of the interim mandate — will attract precisely the young talent who realise the spot market only pays well for those already formed by someone else. Scarcity will reverse bargaining power: whoever forms, retains; whoever rents, competes at auction.

There remains the bleak scenario, one the data makes impossible to ignore: that no one corrects anything, because no individual actor has an incentive to. Every quarter of interim-market growth [1] confirms, for each participant, that the model works. The depreciation of the stock will only become visible once aggregated — and by then, it will be late. Tragedies of the commons are rarely resolved by the shepherds' spontaneous lucidity; they are resolved when someone fences off a pasture and proves the fence pays off.

Conclusion

What this lens reveals, in the end, is a silent inversion in the nature of human capital: for a century, companies owned cognition and rented capital; in 2026, they rent cognition and own capital. Heidrick's numbers [1], Microsoft's tap [2], and Karp's premium [6] are three snapshots of the same movement — intelligence changing its accounting status, from an asset one cultivates to a flow one contracts. The change looks neutral because every transaction is voluntary and every price is fair. But spot prices only measure the value of what exists; they have never measured the cost of ceasing to produce it.

The board that understands this first will have a decade's head start over those still celebrating flexibility. Because discernment — the only raw material this market trades — cannot be manufactured at the speed at which it is rented, and the stock currently in circulation was paid for by a regime that circulation itself is extinguishing. Renting intelligence is easy; the hard part, and the decisive one, is being the last house that still knows how to make it.