Big Tech is having its worst stretch in years.

Everyone is worried that Amazon (AMZN), Microsoft (MSFT), Meta (META), and Alphabet (GOOGL) are spending too much on AI. 

As much as Jensen Huang, Mark Zuckerberg and others keep pushing the “AI changes everything” narratives – and as much as Sam Altman keeps signing multi-billion-dollar deals – these once-loved tech titans can’t seem to catch a break from Mr. Market. 

The bear thesis – that these companies overinvested in AI and will pay the price – sounds convincing enough. After all, $700 billion-plus in yearly AI-related capex is a staggering amount, the likes of which we’ve never seen before. 

But that still doesn’t mean the bears are right. 

In fact, I’d argue that the recent weakness in Big Tech stocks may be the best buying opportunity in the group that we’ve seen in almost three years… 

Why Investors Are Worried About Magnificent 7 Stocks

The bears aren’t wrong about the facts, just the conclusion.

Start with free cash flow. Amazon, Microsoft, Alphabet, and Meta are projected to post combined free cash flow of roughly $94 billion this year – down from $205 billion in 2025 and $230 billion in 2024. That is a $136 billion deterioration in a single year. 

When a group of companies burns through more than a hundred billion dollars in incremental cash, the question of “where is all that money going?” is entirely legitimate.

Then there’s the Iran War overhang. Operation Epic Fury has injected genuine macro uncertainty into markets, with oil prices spiking, risk premiums rising, and investors rotating hard into international equities and defensive sectors. From October 2025 through February 2026, the Bloomberg “Magnificent 7” index fell 7.3% while the S&P 500 Equal Weight index climbed 8.9%, led by energy and materials. That kind of divergence doesn’t happen unless the market is making a real call about the near-term risk/reward of owning high-capex tech.

And then there is Nvidia (NVDA), the poster child for the entire AI saga. At the company’s recent GTC event, Jensen Huang forecasted $1 trillion in data center sales through 2027 and announced Chinese government approval to resume AI chip sales. Yet, the stock still ended the week down 4.1%. 

When the most bullish possible catalysts still produce a down week, the market is sending a clear message: it doesn’t trust the numbers.

AI Capex Is Building Long-Term Advantage

But this is where the narrative breaks down.

It’s not like that $136 billion in free cash flow just evaporated. It was invested in GPU clusters, hyperscale data centers, fiber networks, and cooling infrastructure – the AI economy’s physical backbone. And those assets are already generating returns. 

Google is already seeing AI Overviews drive more than 10% additional queries in the searches where they appear. Meta’s AI-powered ad machine is working, with price per ad up 6% and impressions up 18%. AWS just posted 24% growth – its fastest in 13 quarters – while Bedrock has become a multibillion-dollar run-rate business. And Microsoft 365 Copilot has already crossed 15 million paid seats, embedding itself into enterprise workflows across the Fortune 500.

The fundamental error in the bear case is that it treats AI capex as an expense rather than an investment. When your AI infrastructure is making your core businesses more efficient, accelerating cloud revenue growth, and cementing competitive moats that no startup can realistically breach in the next decade – that’s not waste. That’s the most consequential capital allocation in corporate history.

Also, as we explored in yesterday’s issue, the U.S.-Iran War seems likely to be a fading headwind rather than a permanent condition. 

And here’s the irony the bears may be missing: the flight to international equities and defensive sectors that punished Big Tech over the past several months? That trade is now becoming dangerously crowded – and crowded trades have a habit of unwinding violently. 

As geopolitical risk subsides, the most obvious beneficiary of capital returning to U.S. equities is exactly what everyone has been selling: Big Tech.

Big Tech Earnings Support the Bull Case

The numbers that cut through all the noise is the earnings growth. 

The Magnificent 7 is expected to grow profits 19% in 2026, compared with 14% for the other 493 companies in the S&P 500. That 500-basis-point gap reflects something structural and durable: these companies have monopolistic positions in the highest-growth industries on the planet, and they are compounding into those positions with every dollar they spend on AI.

Over long enough time horizons, as go earnings, so go stocks. You can have a narrative war about capex efficiency, ROI timelines, and geopolitical risk premiums. But at the end of the day, the company growing earnings the fastest wins. And right now, that’s still Big Tech – by a wide margin.

These aren’t companies that need to prove themselves; they’ve over-delivered quarter after quarter for a decade. Their businesses are entrenched, their balance sheets are fortress-grade, and their management teams have compounded capital through a pandemic, a rate spike, a bear market, and now a war. 

The idea that heavy AI spending is somehow going to break Alphabet or Meta or Microsoft might be amusing if it weren’t so fashionable.

We’ve Seen This Setup Before – and It Was Bullish

Now for the timing aspect.

The correlation between the Magnificent 7 and the equal-weighted S&P 500 turned negative on February 23 and has continued falling since. In other words, Big Tech and the rest of the market are moving in opposite directions. That’s unusual. In fact, it’s only happened at this extreme level once before in the past decade – the first quarter of 2023.

At that time, ChatGPT’s debut ignited AI excitement, and Big Tech surged while the rest of the market languished in bear-market mud. The correlation broke negative for the same structural reason it is today: Big Tech was marching to a different drummer than the broader economy. 

What happened next? From the start of 2023 through February 2026, the Mag 7 index soared more than 300% while the equal-weight S&P 500 rose just 42%.

Now, it’s worth acknowledging that the mechanism is different this time. In 2023, the decoupling reflected Big Tech surging while the rest lagged. Today, the decoupling reflects Big Tech falling less than the broader market in a risk-off environment. The signal is relative defensiveness, not pure euphoria. But the underlying structural story is identical: Big Tech is operating on a different fundamental plane than the rest of the market, and the correlation breakdown is the market’s early signal that it’s starting to figure that out.

A Valuation Reset Could Be an Opportunity

Six months ago, you could make a reasonable argument that Big Tech was priced for perfection. The Magnificent 7 was trading at roughly 33 times estimated earnings last October. Today it trades at under 25x – below its own 10-year average of 29x and at the lowest level since the “Liberation Day” tariff tantrum last April.

Think about that for a moment. Investors are being offered the best earnings compounders in the world – companies with dominant market positions, fortress balance sheets, and superior earnings growth – at a below-average multiple. 

The bull case from here doesn’t require Nvidia to print another 1,100% return or for AI to achieve AGI this year. It just requires that these companies keep doing what they’ve been doing for a decade – growing earnings faster than everyone else – and that the market, eventually, notices the discount.

The Bottom Line

AI capex is eye-watering. Markets are noisy. Wars are unpredictable. The bears have real, compelling data on their side.

But zoom out. In 2026, the Magnificent 7 is growing earnings 19% compared to 14% for the rest of the S&P 500. 

Their capex is converting into deeper moats and higher long-run terminal values, not evaporating into thin air. 

Their valuations have been reset to below-average levels after months of underperformance. 

The correlation between Big Tech and the broader market has broken down to a level we’ve only seen once in the past decade – and the last time it happened, it marked the beginning of a 300%-plus run.

The doom and gloom around Big Tech right now is real. But it may also be the setup for the next great buying opportunity.

Prepare for a Major Capital Rotation

When everyone is scared of the best business models in the world, at below-average valuations, with superior earnings growth and AI compounding into their monopolies – that’s an invitation.

Big Tech is starting to look interesting again.

If de-escalation holds, capital could rotate back into the Magnificent 7 — just as it has in past risk-off → relief cycles.

But whether money is flowing into Big Tech… or getting pulled out of it during periods of volatility… the hundreds of billions being spent on AI still have to go somewhere.

That “somewhere” is the physical backbone of the AI economy – exactly where my colleague Eric Fry is focused right now.

In his FutureProof 2026 briefing, he breaks down what he calls AI’s “Golden Rivets” – the critical materials, energy systems, and components that the entire AI buildout depends on.

And here’s the key: those companies don’t need perfect sentiment around Big Tech to win. They just need the spending to continue – and it already is.

That means they could benefit not only from the AI boom itself…

But from the massive reallocations of capital that tend to happen as cycles like this evolve.

You can watch the full presentation here – but hurry. It’s only available until midnight tonight.