I beforehand wrote how the inventory market was at a crossroads, with a ticking clock working towards its quest for 7,000. Time ran out, provided that the S&P 500 and technology-laden Nasdaq have retreated, together with a 1.2% and 1.6% drop on Feb 5.
On the floor, there are three headline catalysts derailing shares’ rally:
- The nomination of traditionally hawkish Kevin Warsh to interchange Fed Chairman Jerome Powell
- Ballooning massive tech spending that is shifting the narrative from AI alternative to dangers
- Jobs knowledge exhibiting the employment market stays weak, elevating recessionary fears
These are legitimate considerations, however my thirty-year investing profession has taught me that when the market will get overbought and largely one-sided, resembling in January, nearly any catalyst will get the blame pretty much as good old school profit-taking snowballs.
Think about this: Warsh has been largely dovish in his feedback this previous 12 months, believing inflation would in the end be cured by AI. Skyrocketing AI capex at hyperscalers Meta Platform, Amazon, and Alphabet are removed from something new, and whereas job market knowledge is not nice, that is not a brand new phenomenon, both.
Rising unemployment was behind the Fed’s pivot again to rate of interest cuts in September, and, if something, ongoing weak point will increase the chance of extra cuts in 2026.
Nonetheless, energetic buyers are utilizing the most recent information to unwind outsized bets, significantly in tech shares, and the ache has been substantial, particularly in software program, which has turn out to be the goal du jour of agentic AI disruption.
B
Jobs knowledge comes up quick
We have seen a slate of regarding unemployment knowledge this week that provides to the recessionary narrative.
The Job Openings and Labor Turnover Survey,JOLTS, confirmed {that a} development towards fewer open unfilled jobs stays intact. There have been 6.5 million open jobs in December, down practically a million 12 months over 12 months (966,000).
We additionally noticed a surge in unemployment claimsfinal week to 231,000, up
22,000 from the prior week and well-above the 4-week shifting common of 212,250, which additionally elevated by 6,000 from the earlier week.
If that wasn’t sufficient, ADP’s Employment Report additionally upset. In keeping with ADP’s survey confirmed solely 22,000 jobs had been created by the U.S. financial system in December. Wall Road hoped for 45,000.
The true kicker, although, was the layoffs knowledge launched by Challenger, Grey & Christmas. The agency had already famous final month that U.S. employers laid off 1.2 million American employees in 2025, the Seventh-worst 12 months since 1989.
In January, there wasn’t a lot reduction, as 108,435 individuals had been laid off in the course of the month, up 118% 12 months over 12 months, and the worst exhibiting for the month of January since 2009, when the Nice Recession was hitting labor markets laborious.
“This is a high total for January. It means most of these plans were set at the end of 2025, signaling employers are less-than-optimistic about the outlook for 2026,” said Andy Challenger.
AI shifts from unlimited opportunity to prove-it reality
The AI spending frenzy has yet to peak, given guidance from major hyperscalers Amazon, Alphabet, and Meta Platforms.
The biggest hyperscalers, including Microsoft and Oracle, totaled over $390 billion in 2025, according to Goldman Sachs. In 2026, Goldman Sachs was targeting spending to surge past $500 billion — a figure that may shockingly be too low.
Hyperscaler 2026 capex (estimated or annualized):
- Microsoft: $100 billion, annualized run rate estimate based on $35 billion in Q1.
- Alphabet: $175B – $185B, with focus on infrastructure and Gemini development
- Amazon: $200 billion, AWS data center build out, robotics, and custom trainium chips
- Meta Platforms: $115- $135 billion, Llama training and reality applications
- Oracle: $50 billion, cloud data center scaling to deliver on $523 billion RPO backlog.
For perspective, Alphabet’s spending alone is double 2025.
Overall, hyperscaler spending on liquid-cooled server racks packed with next-gen Nvidia (and sometimes AMD) AI chips, connected by super-fast interconnects and switches, isn’t bad for GDP, but increasingly, Wall Street is picking winners and losers.
More Wall Street
- Goldman Sachs issues urgent take on stock market for 2026
- Analyst who nailed 2023 bull run sets S&P 500 target for 2026
- Longtime fund manager sends blunt message on P/E ratios
Right now, they’ve decided that one of the big losers from all the data crunching happening on those servers is software. If AI enables enterprises to create their own custom solutions, there’s little need to pony up big money for other solutions.
Investors are also starting to take a hard look at just how much money these data centers are spending and beginning to wonder if the build it and they will come will continue, and for how much longer. These companies have fortress balance sheets, but this capex surge is forcing them to the debt markets, and that’s not free money.
An unfriendly Fed offers (potentially) mixed messages
Kevin Warsh‘s nomination to chair the Fed, replacing embattled Jerome Powell, appears to have toned down rhetoric that the Fed will lose its independence. Warsh is a longtime Wall Street insider (and a partner in hedge fund legend Stanley Druckenmiller’s family office) and a former Fed Governor.
Related: Billionaire Dalio sends 2-words on Fed pick Warsh
Unfortunately for bulls, though, Warsh has a reputation for hawkishness, saying in the past that the Fed has overstepped in propping up Wall Street over Main Street. He’s particularly not a fan of the Fed’s balance sheet, its use of it for quantitative easing, or buying bonds in the open market to lower interest rates.
The Warsh nomination, therefore, throws mixed signals. Warsh has advocated for rate cuts, saying AI is deflationary, so that keeps rate cuts on the table. But his disdain for the Fed’s balance sheet could mean he tries to shrink it by selling bonds and pressuring rates higher.
What’s next for stocks depends on sentiment
We’ve seen the buy-the-dip crowd handsomely rewarded since the stock market lows were put in last April. The temptation to step into the fray will likely mean that selling leads some to step into the fray, particularly in oversold areas, including software.
The big question I always ask at times like this: how durable will that bounce back be?
If we do see bargain hunters buy, will would-be sellers take their fingers off the trigger or pull once stocks head back up to resistance levels? The real tell will be whether stocks make lower highs and lower lows, gaining ground only to roll over and break down.
Earnings are the lifeblood of stock market returns, and forward earnings expectations are broadly bullish, with companies’ guidance coming in quite healthy. Still, at points like this, it’s sentiment that matters most, making it critical to pay attention to whether markets rally off intraday lows (bullish) or sell off from intraday rallies into the close (bearish).
Associated: Goldman Sachs bucks Warsh Fed price minimize fear
