The Smartest Money Just Walked Out of Tech. Goldman Says It’s the Biggest Exit in a Decade.
Goldman Sachs runs one of the most closely watched prime brokerage operations on Wall Street. When it tells you that hedge funds just executed the largest selloff of technology stocks in a decade, that’s not a headline designed to get clicks. It’s a data point from the people who process the actual trades.
The selloff happened. The scale is real. The question worth asking isn’t why — the answer is more specific than tariffs — but what it signals about where institutional conviction in the technology sector actually stands in 2026.
Tariff exposure is a real operational headwind for tech hardware supply chains. But hedge funds navigated tariff cycles in 2018 and again in 2025 without triggering the largest tech selloff in a decade. Something changed this time. Institutional money is doing what it always does ahead of a structural disruption: it’s getting out before the story changes permanently.
What Goldman’s Data Actually Shows
Goldman Sachs’s prime brokerage data covers the flow of hedge fund trading across its client base — one of the largest such datasets in the world. According to Fortune’s coverage of the Goldman prime brokerage note, technology, telecoms, and media-focused hedge funds were down as much as 2.78% in a single day during the selloff — among the worst single-day losses for that strategy cluster in nearly a year.
The broader picture is more striking. Goldman’s analysis, as reported by Yahoo Finance, describes the global technology sector as having one of its weakest periods of relative returns in the last 50 years compared to global equities broadly. That’s not a bad quarter. That’s a generational rotation signal — the kind of underperformance that historically marks a structural, not cyclical, shift in investor conviction.
The rotation landed in defensives: consumer staples, healthcare, utilities. These are the sectors you buy when you’re not sure what you’re holding in tech is still worth what you paid for it, and you want to park capital somewhere it’s unlikely to be disrupted by a technology trend you don’t fully understand yet.
Goldman itself, somewhat paradoxically, simultaneously released a “buy tech” note on valuation grounds. That’s not a contradiction — it’s the difference between what the data shows long-term and what the trading desk is actually doing short-term. Institutions often hold two views at once: they believe in the sector over a five-year horizon while exiting the position over a three-month one.
The Disruption These Funds Actually Fear
Tariffs and macro uncertainty are real, but they’re not new. Hedge funds navigated tariff cycles in 2018 and 2025 without triggering the largest tech selloff in a decade. Something is different this time, and the most coherent explanation for the scale and speed of this exit is AI disruption risk.
Not AI as an opportunity — as a threat.
The technology sector that hedge funds have loaded up on for the past decade is dominated by software companies: enterprise SaaS, cloud infrastructure, productivity tools, workflow automation, data management. These businesses were built on the premise that the complexity of operating modern organizations creates recurring demand for specialized software at scale.
AI is now systematically attacking every layer of that premise.
When a single AI model can draft legal documents, write and debug code, analyze financial reports, manage customer communications, summarize research, and route support tickets — all at a fraction of the cost of the specialized software that previously did each of those things — the addressable market for those software companies doesn’t just grow more competitive. It contracts. The pricing power that justified the multiples disappears. The switching costs that locked in annual contracts erode.
This isn’t hypothetical. Enterprise software companies that built their valuations on workflow automation are already seeing it in their pipeline conversations. Customers are asking questions they weren’t asking eighteen months ago: why do we need this subscription if the AI handles it? Why are we paying per seat for a tool that a general-purpose model replicates for a fraction of the cost?
Hedge funds that have studied this thesis are making a logical trade: exit the companies most exposed to AI disruption before the earnings reports start confirming what the AI demos already suggest.
The Beneficiaries and the Casualties Are Not the Same Companies
The important nuance in this selloff is that “tech” is not a monolith. The companies building the AI infrastructure — the hyperscalers, the chip manufacturers, the model providers — are not equally exposed to disruption. In many cases, they are the disruption.
What’s being sold is the layer above the infrastructure: the specialized software that sits between general-purpose computing and specific business problems. That layer — the one that generated enormous returns on small switching costs — is now being compressed by models that can substitute across many verticals simultaneously.
This means the selloff is a rotation within tech as much as it’s a rotation out of tech. Funds moving out of enterprise SaaS and workflow software while maintaining or building positions in AI infrastructure companies aren’t pessimistic about technology broadly. They’re pessimistic about the specific segment that happened to dominate the “tech” portfolio for the past decade.
For retail investors who hold tech-heavy index funds or ETFs, this distinction matters enormously. An index that was dominated by enterprise software valuations five years ago is being repriced not because tech is failing but because the specific bet the index was making — on software subscription moats — is being tested against a disruption the index construction never accounted for.
What This Means for Crypto and Web3
Crypto markets move in correlation with institutional risk-off events, and this selloff is no exception. When hedge funds reduce growth asset exposure broadly, crypto absorbs selling pressure in the first wave.
The more specific signal sits at the infrastructure layer. The Graph — which provides decentralised indexing for blockchain data queries — and Chainlink — which supplies off-chain data to smart contracts — are both positioned as essential Web3 middleware. Their moat argument is structurally identical to the enterprise SaaS moat argument hedge funds just repriced: that the complexity of the underlying problem creates durable demand for a specialised intermediary. If general-purpose AI agents can query blockchain data and interface with off-chain sources without a dedicated protocol layer sitting in between, that moat erodes on the same timeline as enterprise workflow software.
Neither project is in immediate danger. Both have substantial installed bases, active developer ecosystems, and protocol-level integrations that aren’t trivially replaced. But the hedge fund thesis — that AI substitutes for specialised middleware rather than complementing it — applies to on-chain infrastructure as directly as it applies to SaaS. The question isn’t whether The Graph or Chainlink survive. It’s whether their current valuations already assume a world where AI doesn’t route around them.
The largest hedge fund exit from tech in a decade is not a “sell everything” signal. It’s a “reprice the bet on specialised middleware moats” signal. That bet runs through Web3 infrastructure as much as it runs through enterprise software.
The Goldman Contradiction Worth Watching
Goldman’s simultaneous “buy tech” valuation call and “largest selloff in a decade” flow data are not contradictory in isolation, but they are worth holding in tension. When the institution managing the trades says the smart money is selling while its research desk says valuations are attractive, the question is which Goldman is right.
Historically, in structural inflection points, flow data leads price recovery. The valuation call is right eventually — but the timing requires waiting through a repricing that could be extended, and institutions with short holding periods can’t afford to wait through it. They’re not wrong to sell. They’re just operating on a different clock than the analysts who write the buy notes.
For long-term investors, the Goldman buy call may ultimately be correct: technology will remain the dominant economic sector, and the companies that survive and adapt to AI disruption will trade at compressed multiples only temporarily. But “temporarily” can mean years — and it requires betting that specific companies will be among the survivors rather than the casualties.
That’s a much harder bet to make confidently right now, which is exactly why the smartest money is currently reducing its exposure to avoid having to make it at all.
Implications for Q2 and Q3 2026
Goldman’s data covers a specific period, but the structural conditions that drove the selloff haven’t resolved. Enterprise software companies are heading into an earnings season where forward guidance will be watched closely for signs of AI-driven customer churn, pricing pressure, and elongated sales cycles. Any meaningful weakness in those metrics will give hedge funds additional confirmation for the thesis driving the exit.
Conversely, if leading AI infrastructure companies — the picks-and-shovels layer — post strong numbers showing that AI adoption is accelerating enterprise demand rather than replacing it, the rotation thesis breaks down and some of this capital flows back.
The single most important data point over the next 90 days is whether AI is showing up in enterprise earnings as an accelerator or a headwind. The hedge funds that executed this trade are watching the same data. If they’re right, the selloff continues. If they’re wrong, the reversion will be fast — and everyone who sold will look foolish in hindsight.
That’s the uncomfortable situation in a structural inflection point: the thesis is compelling, the timing is unknowable, and the cost of being right too early looks identical to the cost of being simply wrong.
Frequently Asked Questions
What did Goldman Sachs report about hedge fund tech selloffs?
Goldman Sachs’s prime brokerage data showed hedge funds executing the largest selloff of technology stocks in a decade. Technology, telecom, and media-focused hedge funds were down as much as 2.78% in a single session. Goldman simultaneously released a “buy tech” research note on valuation grounds, reflecting different time horizons across the institution.
Why are hedge funds selling tech stocks now?
The primary driver appears to be AI disruption risk to enterprise software business models rather than tariff or macro concerns alone. Hedge funds are repricing the valuation case for specialized software companies whose moats — switching costs, workflow lock-in, per-seat pricing — are being tested by general-purpose AI that substitutes across multiple verticals simultaneously.
Does the hedge fund tech selloff affect crypto markets?
Directly, crypto correlates with institutional risk-off events and typically faces early selling pressure. Indirectly, the thesis driving the selloff — that AI compresses the value of specialized middleware — has implications for Web3 infrastructure projects built on the assumption that specialized tooling creates durable competitive advantages.
Should retail investors follow hedge funds out of tech?
Goldman’s own research argues that tech valuations are attractive on a longer time horizon. Retail investors with a multi-year holding period are in a different position than hedge funds operating on quarterly performance cycles. The risk is concentration in the wrong segment of tech — specifically, enterprise software exposed to AI substitution — rather than technology broadly.
Sources
- Fortune: Hedge funds selling global tech stocks at fastest pace in 6 months
- Yahoo Finance: Goldman Sachs Says Buy Tech as Valuations Turn Attractive
- Motley Fool: Goldman Sachs Says the AI Software Sell-Off Was Overdone
- DefiCryptoNews: Microsoft — Trouble Ahead
- VaaSBlock — Web3 risk and governance analysis
