Tech Sector’s $650 Billion AI Binge: Are Investors Right to Doubt the Hyperscaler Frenzy?

If you want to see what happens when Silicon Valley’s imagination collides with Wall Street’s skepticism, look no further than the current AI arms race. The so-called 'Big Four' hyperscalers, Alphabet, Amazon, Microsoft, and Meta, are on a $650 billion spending spree, building data centers as if Moore’s Law is about to be repealed by congressional fiat. The market, however, is not impressed. Tech stocks, especially those with the most AI exposure, have stumbled out of the gate in 2026. The narrative has shifted from 'AI will save us all' to 'AI will bankrupt us all' in less than a quarter.

Let’s not pretend this is just another sector rotation. The numbers are staggering. According to MarketWatch, hyperscalers are barreling ahead with capital expenditures that would make even SoftBank blush. Yet, the market reaction has been a collective shrug, if not outright derision. The Nasdaq 100 is flatlining, and the Technology Select Sector SPDR Fund ($XLK) is stuck at $141.06, refusing to budge. That’s a rounding error for a sector that used to move two percent before the opening bell.

There’s a whiff of 1999 here, but without the euphoria. Instead, the mood is clinical, almost surgical. The market is dissecting every earnings report, every guidance update, looking for signs that this AI capex binge will actually produce profits, not just headlines. The latest round of Big Tech earnings gave us little to work with. Microsoft and Alphabet talked up AI, but their margins are compressing. Meta is spending like a drunken sailor on both AI and the metaverse, and Amazon’s AWS growth is slowing. The only thing that’s up is the skepticism.

The context is even more damning when you zoom out. The S&P 500 Equal Weight Index just hit an all-time high, while the cap-weighted index is treading water. That’s not a sign of broad-based tech strength. It’s a sign that money is flowing out of the AI darlings and into the old-economy stocks, banks, industrials, even utilities. The 'flight to safety' is now a 'flight to boredom.' And yet, the hyperscalers keep spending.

Why? Because they have no choice. The AI narrative has become existential. If you’re not building, you’re dying. But the market is calling their bluff. Investors want to see actual returns on this $650 billion bet, not just another slide deck about machine learning. The divergence between capital spent and capital returned has never been wider.

Strykr Watch

The technicals are as uninspiring as the price action. $XLK is rangebound between $140 and $143, with RSI stuck in the mid-40s. There’s no momentum, no volume, just a slow grind sideways. The 50-day moving average is converging with the 200-day, threatening a dreaded 'death cross.' If $XLK breaks below $140, there’s air down to $135. On the upside, a close above $143 could trigger a short squeeze, but the odds are fading by the day. Option flows are skewed to the downside, with put-call ratios at multi-month highs. The market is hedged, not hopeful.

The risk is that this malaise turns into a rout. If hyperscalers miss another quarter, or if AI capex guidance gets ratcheted up again without a corresponding revenue boost, the bottom could fall out. The market is already pricing in disappointment. The only thing keeping tech afloat is the sheer size of the buybacks. But even that has its limits.

The opportunity, if you can call it that, is to fade the hype. Shorting $XLK on rallies has been a profitable trade, and there’s little reason to change that stance until the fundamentals improve. If you’re looking for a contrarian play, wait for a capitulation wick below $140 and then nibble for a bounce. But don’t expect miracles. This is a market that punishes hope.

Strykr Take

The AI spending spree has become a liability, not a catalyst. Until the hyperscalers can prove that $650 billion buys more than headlines, tech will remain a trade, not an investment. The smart money is already rotating out. Don’t be the last one holding the bag.

datePublished: 2026-02-08 00:15 UTC