The Orderly Unraveling
The Orderly Unraveling
We spent six months chasing golden unicorns and they turned into lead. Not metaphorically. Literally.
Gold crashed from $5,550 a troy ounce—a new all-time high on Wednesday—to $4,423 by the following Monday. That's an 11% obliteration in one day. Silver fell 31%. Bitcoin tumbled from $83,000 to $74,570, dragging itself further down this week to below $65,000, now down 35% year-to-date. The KOSPI in South Korea, which had ripped 76% last year on the AI boom, sank 5.26% in a single session—its worst day since April. If you were long precious metals, long Korean semiconductors, long software, you've watched your conviction thesis get shredded in 48 hours. This is what happens when a crowded trade realizes it's crowded.
The mechanics are straightforward. Precious metals had become unmoored from fundamentals, gripped by what one analyst called "mania, especially in the age of financialization and gamification." The same retail energy that inflated meme stocks found new objects of worship in gold, silver, and the hottest AI chip plays. Money piled in. Leverage stacked. Risk models got fatter. Then Thursday came and profit-takers started tapping the eject button. By Friday, it was a stampede.
What makes this week genuinely unsettling isn't the volatility itself—2026 opened with promises of it anyway—but the uniformity of the destruction. Software stocks extended a brutal selloff with ServiceNow falling 7.6%, Salesforce down nearly 5%, and the broader S&P 500 posting a 1.23% loss as AI-fueled tech selling met weakness in labor data. Alphabet dropped initially on soaring AI capital expenditure concerns, while Bitcoin tumbled below $65,000. Meanwhile, initial jobless claims rose more than expected, job openings in December fell to their lowest since September 2020, and Challenger reported 108,435 announced layoffs in January—the highest January total since the global financial crisis.
The narrative shift happened between the lines. A week ago, the Fed under Powell had kept rates steady at 3.50-3.75%, signaling policy was "closer" to neutral. Inflation remained sticky. The economy looked resilient. Manufacturing hummed along. Then the labor data started talking. The BLS reported job openings fell to 6.54 million from 6.93 million, a 386,000-job cliff. The ratio of job openings to unemployed workers slid to 0.87-to-1. Two years ago, during the pandemic's labor shortage hysteria, that ratio was running 2-to-1. We've completed a full reversal.
Here's the uncomfortable part: this reversal might not reverse itself anytime soon. The Fed isn't cutting. Trump's transition people nominated Kevin Warsh as the next Fed chair—a man whose intellectual architecture leans toward supporting Trump's pro-growth agenda but who also understands monetary discipline. That's a mixed signal markets are still pricing. But more immediately, if layoffs are really accelerating and job openings are collapsing, then the soft landing narrative gets harder to sustain. You can't engineer an AI revolution without disruption. You can't disrupt without people losing jobs first. The question is whether those displaced workers find new gigs before demand craters, and whether wages stick around to support consumption.
Look at the wreckage more carefully and something else emerges: rotation without direction. DaVita surged 12% on earnings beats, Palantir jumped 11% after blowing out expectations with $1.41 billion in revenue versus $1.33 billion expected, yet most mega-cap tech names like Microsoft and Meta were down over 2%, while Nvidia slumped nearly 3%. Strength in healthcare and defense contractors. Weakness everywhere else. Novo Nordisk tanked 12% after forecasting sales declines of 5-13% in 2026 as it loses market share in weight-loss drugs.
This is what happens when investors stop believing in the narrative and start digging into the p&l. The mega-cap AI story began to look like every other bubble: the companies best positioned to profit from the disruption kept getting more expensive while the companies getting disrupted kept getting cheaper. That arbitrage couldn't last. Markets hate permanent upward slopes. Even in the age of AI and digitization and trillion-dollar tech firms, something has to correct. Turns out it was everything at once.
The real tell might be that we're not seeing panic selling yet. The VIX popped above 18 but that's volatility, not dread. The Russell 2000 has jumped 3% this year even as the S&P 500 slipped 0.7%, and analysts at Wells Fargo specifically identified which retailers would win from Super Bowl spending. People are still making bets. They're just making different ones.
Kevin Warsh's confirmation hearings will matter more than any earnings print in the next few weeks. If he signals the Fed can cut before spring, risk assets will bid. If he sounds like a Powell clone—patient, data-dependent, stubborn—then the pullback in equities might extend. The labor market is the hinge. Everything turns on whether those 108,000 announced layoffs become 200,000, or whether they stabilize and hiring picks up again.
Until we know, expect this: bounces in beaten-down precious metals will be sold into. Software rallies will be viewed with suspicion. The KOSPI and other momentum trades will get bought selectively on dips but won't build conviction until the economic backdrop clarifies. Markets spent six months in vertical rally mode. They're now in question-mark mode. That's a more honest place to trade from, even if it feels messier.
The unicorns didn't disappear. They just got repriced back to what horses actually cost.
Upvoted! Thank you for supporting witness @jswit.