Last week’s violent market reaction to Microsoft’s otherwise solid quarterly results exposed a growing fissure beneath the AI euphoria: investors are no longer willing to treat vast, multi‑year capital spending as an article of faith.
Microsoft reported results that many corporate watchers would call healthy, yet attention immediately shifted to two numbers investors dislike — a pause in Azure revenue acceleration and guidance that the company will spend more than $100 billion in capital expenditures this year. Stocks punished that uncertainty. Microsoft plunged almost 10% in one session and shed roughly $381 billion of market value across two trading days, its worst weekly performance since March 2020.
That sell‑off was not an isolated bout of profit‑taking. Meta, which had briefly rallied after signalling resumed revenue momentum, fell back after disclosing plans to increase capital spending by up to 87% through 2026. The market’s message was blunt: investors will tolerate heavy spending only while growth keeps pace. When growth wobbles, valuations are vulnerable.
The nervousness has a clear arithmetic. Big tech’s three‑year AI rally has been predicated on the idea that these firms’ scale and willingness to pour cash into compute, data centres and specialised chips would secure them the leading edge in the next platform shift. That calculus requires a second step — converting that investment into durable, above‑market returns. Several market managers quoted in recent days say that proof is not yet evident.
The scale of the wager is staggering. Alongside Microsoft and Meta, Alphabet and Amazon are expected to push the industry’s collective capital spending well into the hundreds of billions this year; estimates compiled by market watchers suggest the four firms’ capex could top $500 billion, with the bulk funnelled into AI compute infrastructure. Google’s booming share price — up more than 70% over six months on the back of Gemini and custom AI chips — encapsulates the upside investors see when execution appears to match ambition.
But rising expectations cut both ways. Indexes that track the seven largest US tech names are off their recent highs even as the broader S&P 500 has edged higher, and active fund managers have been trimming technology allocations for months. Hedge funds and retail investors have also rotated into cyclicals and commodity sectors. The question is no longer whether these companies can spend money; it is whether that spending will generate returns that justify current valuations.
There are also rising concerns about the downstream effects of universal capex growth. Startups such as OpenAI and other AI‑focused ventures — many not yet profitable — are demanding large and rising pools of compute. That pushes up the market price of scarce data‑centre capacity and specialised chips, increasing the industry’s breakeven points. If the largest cloud providers expand capacity faster than demand materialises, margins may compress and price competition could intensify.
Investors will get fresh data points this week: Alphabet and Amazon are due to report after the bell, and markets will scrutinise not just revenue and margins but explicit capex plans and the near‑term trajectory of cloud and AI monetisation. The balance between growth and investment has become the central valuation battleground for big tech.
For the companies themselves, the policy is straightforward but stark: demonstrate that AI capex converts into revenue growth and margin expansion, or face further market re‑rating. For investors, the near term is likely to be volatile as the market reassesses how much of the AI story is hype, and how much will be realised in cash flows over the coming years.