Samsung, the South Korean maker of everything from computer chips to 320,000-ton crude-oil carriers, on Monday reported an almost incomprehensible 1,800% jump in operating profits over last year. So naturally, investors today shouted “SELL!” and drove shares down a not-insignificant 7%.
“The stock had priced in a historic quarter for months, and once the numbers confirmed it was significant but not far beyond what the market had already expected, there wasn’t much to reward anyone stepping in,” eToro analyst Zavier Wong explained. “It acts more like confirmation, and confirmation is what people sell into.”
So the selloff wasn’t all that weird, because investors had already priced in Samsung’s blowout results — but maybe they’re signaling where we are in the AI bubble — assuming for the sake of argument that we’re in one.
Samsung’s massive profits are driven mostly by the memory chip shortage, as AI and other datacenter buildouts consume every chip on the market, before coming back like a psychotic, knife-wielding Oliver Twist and asking for “MORE!”
Or as Wong put it, share prices are “dragged down by concerns that AI infrastructure spending can’t keep growing at the pace that has been driving memory prices.”
Which sounds to me an awful lot like something I’ve written about here a time or two since the AI craze went full Norman Bates. Globally, datacenter capital expenditures (capex) reached $455 billion in 2024. That’s a 51% increase over 2023. That’s expected to hit [dr_evil_voice] ONE TRILLION DOLLARS [/dr_evil_voice] in 2026, and even more every year for the foreseeable future — driving mostly by AI.
That’s just the infrastructure.
At some point, investors will look at all that spending on AI and ask, “When will we see the profits?” If we’re in a bubble — and I’ve read some compelling arguments that we might not be — then that’s how it begins to deflate.
Samsung’s 7% decline indicates that investors might think the big run-up could be coming to an end. But that’s just one indicator, hardly a trend.
How about another? One stock doesn’t make a trend, maybe a little contrary-indicator humor is in order.
So I report the following related item to you half in jest, or maybe just a third.
You must be familiar with TV stock maven Jim Cramer, and how wrong he is. He’s so wrong that there’s an unofficial Inverse Cramer index that tracks returns based on buying what he recommends selling and vice-versa.
According to Yahoo! Finance, in 2025, people investing using the Inverse Cramer strategy enjoyed 60% returns (!!!), while investors following the Do What Pelosi Does strategy earned “just” 25%. Over the last two years, Inverse Cramer performed even better. In other words, betting against Cramer’s buffoonery is somehow more profitable than going with Pelosi’s insider knowledge.
There’s wrong and then there’s Cramer-level wrong.
There’s even an Inverse Cramer parody account on X. I follow it.
I tell you this not just for the amusement factor, but because last week, Cramer told everybody that Meta (parent of Facebook, Instagram, and WhatsApp) is currently a steal.
“I find it difficult to believe that Meta was up only 49 points when it is getting into the most lucrative game, business-to-business at 18x EPS????” Benzinga quoted Cramer on Saturday. The publication explained that Cramer’s outburst came “after reports emerged that Meta is exploring a cloud infrastructure business that would allow developers to access its AI models and compute capacity.”
When Cramer says buy, the immensely profitable Inverse Cramer sits on its cash.
I don’t give financial advice, and this looks like the perfect time to keep not doing that.
Recommended: New Yorkers Are Already Mad at Mamdani, and He’s Just Getting Started
I’m not here to pop your bubble. Just to offer you exclusives at a great price.
PJ Media VIP members get tons of exclusive goodies, including podcasts and video live chats with your favorite writers. You can support alternative conservative news and save 60% with promo code FIGHT.
