The Great Rotation Out of Tech — Season 9, Episode 47
And they get it wrong again
CNBC this week: “The much-hyped great rotation out of tech for 2026 may be over already.” Six weeks. That’s how long this one lasted before the obituary got written.
If you’ve been trading this bull market since the October 2022 low, you’ve seen this movie roughly twice a year. Same script, different macro villain, same ending. At some point, it stops being a coincidence and becomes a business model — for media, for strategists, for tech haters and value investors — anyone who needs something new to say every Monday morning.
Here’s the honest chronology.
Nine consolidations, nine rotation calls
February 2023 — SVB. Regional banking crisis, credit crunch fears. Tech sold off into the panic. Narrative: “Duration assets get hit first when liquidity breaks.” Resolution: Nvidia’s May guidance beat, triggering +53% in a day. Semis went vertical for the rest of the year.
August–October 2023 — The rate scare. 10-year yield ran from 3.8% to 5.0%. QQQ down 12%. Every strategist on television: “Higher for longer kills tech multiples, rotate into value.” Resolution: Powell pivot in December. NDX +25% in ten weeks.
April 2024 — Sticky CPI. Hot inflation print blew up rate-cut hopes. Nasdaq –7% inside a week. Narrative: “No landing, multiples have to compress.” Six weeks later: new all-time highs.
July–August 2024 — The Yen carry unwind. IWM +11% in a single week, Mag-7 –15%. Every macro desk called the regime change. “Breadth is broadening. The mega-cap era is over. Equal-weight finally wins.” Small caps rolled over within four weeks. Nasdaq made new highs through December.
September 2024 — AI capex bubble, wave one. Goldman’s “Too Much Spend, Too Little Benefit” paper went viral. Narrative: “Hyperscaler capex has no ROI, semis crack.” Hyperscalers raised capex guidance the next quarter. SMH made new highs.
January 2025 — DeepSeek. NVIDIA –17% in a single session on R1’s release — roughly 600 billion in market cap gone in an afternoon. “Compute thesis is dead. Efficiency kills infrastructure demand.” Jevons paradox played out in two weeks. Back to all-time highs by spring.
March–April 2025 — Liberation Day. Tariff shock, broad de-risking. “Supply chains unwind, semis are uninvestable.” Exemptions came. A pause came. Full recovery in weeks.
Summer 2025 — AI capex bubble, wave two. Stronger hyperscaler reports, and yet: “Circular financing, OpenAI deals aren’t real, this whole thing is a Ponzi.” Guidance kept climbing.
Q1 2026 — the current one. Energy leads. Staples hold. Russell 2000 up 9.5% in three weeks. Every strategist note: “The equal-weight trade was right all along.” Six weeks later, the CNBC obituary you just read.
Nine windows. Nine rotation narratives. Same ending every time.
Why the script never works
There’s a specific reason the “rotation out of tech” call keeps failing, and once you see it, you can’t unsee it:
“Tech” isn’t a sector. It’s the label markets attach to whichever segment of the economy is compounding innovation fastest at any given moment.
The theme rotates constantly — mainframes, PCs, enterprise software, internet, mobile, cloud, social, SaaS, semis, AI infrastructure, memory, advanced packaging. Within this bull market alone, the baton has changed hands four or five times. Software led into 2023. Semis took over mid-2023. Hyperscalers dominated 2024. In 2025, the weight moved down the stack to compute infrastructure. Right now in 2026, it’s rotating again — memory +61% YTD, advanced packaging +43%, Rubin in early ramp.
From 30,000 feet, this looks like “tech is leading.” From closer range, it’s a relentless reshuffle within the category. What strategists consistently misread as “rotation out of tech” is almost always rotation within Tech — from last cycle’s winners into next cycle’s. The exit door they keep pointing at isn’t actually a door. It’s a mirror.
The two conditions under which the call is actually right
I’m not saying the rotation-out-of-tech thesis is always wrong. I’m saying it’s right under exactly two conditions, and neither is in play right now.
Condition one: a hard rate shock. Duration assets compress first and most violently when the discount rate moves abruptly. 2022 was the textbook case — Fed Funds from 0% to 4.5% in 12 months, NDX down 33% — the rotation call was correct for roughly 13 months. Since the October 2022 low, no macro print has been large enough to repeat that. The 2023 yield spike came closest and still resolved in Tech’s favor within a quarter.
Condition two: the innovation layer is private. This one gets underappreciated, and for long-term allocators, it matters more than the first. Between roughly 2004 and 2012, public tech indices genuinely underperformed in stretches — because the compounding was happening in Facebook, Twitter, Uber, Airbnb, Palantir, SpaceX, all private. The public-tech investor of that era was looking at a shell of the real opportunity set.
The same structural risk exists today, arguably larger. OpenAI, Anthropic, xAI, SpaceX, Stripe, Databricks — collectively among the highest-growth entities in the global economy, none listed. If your exposure is purely public-market Tech, a meaningful slice of the actual innovation is not in your portfolio, full stop. That’s not “tech is weak.” That’s “the object of desire isn’t on the ticker tape yet.” The correct response is access, not rotation.
Everything between those two conditions — the stuff that fills 95% of the rotation commentary — is noise dressed up as a regime change.
The 2026 version is no exception
Q1 handed the rotation camp a convincing story. Small caps ripped. Energy led through mid-February. Materials outperformed. Strategists who’d been calling for equal-weight for three years finally got their victory lap. Then six weeks after the March 30 low, information technology was up 20% and communication services up 16.5%, versus 11.4% for the S&P. Energy, the Q1 hero, is now down 10.3%. Staples flat. Materials trailing.
The rotation was real. It just went the direction rotations in this market almost always go — from the previous cohort of tech winners into the next. From last generation’s Hopper-era exposure into the Rubin early-ramp names. From generalist semiconductor exposure into the specific links of the stack where the capex is landing this quarter.
Next time the headline comes back around — and it will, probably by late August and last through September/October (seasonal pattern) — save yourself the anxiety and pull up this list.
Nine in a row. You don’t need a tenth data point to see a pattern.