For the past four years, the US stock market has done something strange: it has climbed steadily despite high interest rates, trade wars, geopolitical shocks, and valuations that look stretched by almost every historical measure. Most commentators credit the AI boom or aggressive investors. Robert Buckland, the former chief global equity strategist at Citigroup, offers a different explanation — one that has less to do with who's buying and more to do with what's available to buy.
Buckland calls the phenomenon 'de-equitisation': the gradual shrinking of the total supply of publicly traded shares. Companies have been buying back their own stock and cancelling it. Private equity firms have been taking listed companies private. Meanwhile, new listings — IPOs — have slowed to a trickle. The result is a market where the pool of available shares actually contracts year after year. Basic supply and demand does the rest: with fewer shares chasing the same investor dollars, prices climb. Buckland calls the resulting price-support mechanism a 'de-equitisation put,' borrowing market slang for an implicit floor beneath prices.
The scale is striking. The 'Magnificent 7' — Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia and Tesla — bought back roughly $230bn of their own shares in 2025. That cash came not from new debt or new equity issuance, but from the companies' own operating profits, making this tech bull market unusually self-financed. Previous tech-driven booms looked very different. The canal mania of the 1800s, the railway boom of the 1840s, and the dotcom bubble of the 1990s were all capital-intensive: companies issued huge volumes of new shares to fund their expansion, eventually flooding the market and helping precipitate the crash. Big Tech's 'capex-lite' business model — software margins, minimal infrastructure — broke that pattern.
But here's the catch: AI is changing the equation. Alphabet, Amazon, Meta and Microsoft are expected to invest a combined $725bn in AI infrastructure this year. That money has to come from somewhere, and increasingly it's coming from the buyback budget. Meta and Alphabet have already paused their buyback programs — a sharp reversal from the $92bn they collectively returned to shareholders in 2024. Apple has announced another $100bn in buybacks for 2026, but at a $4.3tn market capitalisation, $100bn buys a lot less Apple stock than it used to.
Meanwhile, the IPO pipeline is filling up. SpaceX, OpenAI and Anthropic are exploring public listings at a combined valuation of up to $4tn — which would represent roughly a 6% expansion of the US public equity market, a figure last seen during the late-1990s dotcom bubble. OpenAI and Anthropic alone are expected to privately raise a combined $152bn in 2026 before any IPO. Buckland's point is not that public markets will suddenly choke on these new shares — they won't have to absorb everything at once. But early-stage private investors and employees will eventually want to cash out, creating an 'overhang' of unsold stock that may weigh on prices for years.
The bears, Buckland argues, have spent the bull market waiting for valuations to crack and missing the real mechanism at work — that shrinking supply and rising demand mathematically push prices up. They may finally be vindicated, but not for the reasons they expected. As one veteran banker once told Buckland: if the ducks are quacking, feed them. The ducks have been quacking loudly for US big-cap tech. They are, he suspects, about to be fed.
For four years, US stocks have defied gravity not because investors kept buying — but because the supply of shares kept shrinking. That trick may finally be running out.
Robert Buckland, a veteran equity strategist, argues that today's bull market has been powered less by frenzied buying and more by a quiet shrinkage of the share supply. Big US tech companies have spent enormous sums buying back their own stock, while private equity has pulled other companies off public exchanges entirely. Fewer shares chasing the same money equals higher prices.
But Buckland sees that dynamic about to flip. SpaceX, OpenAI and Anthropic are eyeing IPOs at a combined valuation of up to $4 trillion, and the Magnificent 7 are pouring an estimated $725bn into AI infrastructure this year — money that used to fund buybacks. The 'de-equitisation put' that has propped up valuations may be about to fade.
Markets are governed by supply and demand just like anything else. If you want to understand why prices rise, don't only look at how much people are buying — look at how much there is to buy.
If you have a 401(k) in your future, an index fund in a custodial account, or just a passing interest in why headlines keep saying 'stocks at record highs,' this matters. The mechanics Buckland describes determine whether the market you'll invest in for the next 40 years is overpriced, fairly priced, or about to correct. It also helps explain why companies like OpenAI and SpaceX — household names you've never been able to buy stock in — may finally hit public markets soon.
A $4tn collective valuation for SpaceX, OpenAI and Anthropic would represent roughly a 6% expansion of the US public equity market — a scale last seen during the late-1990s dotcom bubble. If their early investors and employees rush to cash out, that 'overhang' of unsold stock could weigh on prices for years. Watch for two signals: whether Big Tech resumes aggressive buybacks once AI capex peaks, and whether the IPO window actually opens or stays jammed. Either way, the simple story that 'stocks always go up because of AI' is about to get more complicated.