Have you ever watched a stock climb so fast that your stomach does that little flip — the one that whispers, “Get in now or regret it forever”? That’s been the AI trade for the past two years. And honestly? I’ve felt it too.
Last week, while the rest of the market was catching its breath, one of the world’s most cautious institutions dropped a quiet bombshell. They didn’t scream “bubble.” They didn’t have to. The message was clear: a lot of what’s pushing these record highs might not be cold, hard fundamentals. It might just be FOMO wearing a very expensive suit.
When Central Bankers Start Talking Psychology
Central banks usually stick to dry numbers — inflation targets, balance sheet risks, the occasional lecture about leverage. So when they start analyzing investor emotions? You sit up straight.
In their latest financial stability report, one major central bank pointed out something we’ve all felt but few dared say out loud: global stocks are insanely concentrated, valuations are sky-high, and at least part of this relentless bid looks like classic fear of missing out.
Their exact words were careful, of course. Markets may be pricing in “optimism that tail risks will not materialize,” but also “fears of missing out on a continued rally.” Translation? People aren’t just buying AI stocks because the spreadsheets look pretty. They’re buying because everyone else is buying — and nobody wants to be the one left holding cash when the music stops.
“Current market pricing does not appear to reflect persistently elevated vulnerabilities and uncertainties.”
– Major central bank financial stability review, November 2025
That line hit me like cold water. Because it’s one thing for Twitter pundits to yell “bubble.” It’s another when the grown-ups in the room — the ones who actually have to clean up when things explode — start highlighting herd behavior.
The Concentration Conundrum Nobody Wants to Talk About
Here’s a stat that keeps me up at night: a handful of U.S. tech giants now make up around 40% of major American indices. Forty percent. Let that sink in.
We’re not talking about a nicely diversified basket of innovative companies spread across sectors. We’re talking about seven names — give or take — that have become the market. When they sneeze, your portfolio catches pneumonia.
And guess what ties most of them together? That’s right — artificial intelligence. Cloud computing, chips, advertising algorithms, autonomous driving, generative models… it all funnels back to the same megatrend.
- One company designs the picks and shovels (the GPUs)
- Three others run the cloud infrastructure
- The rest either build the models or monetize the data
It’s an ecosystem so tightly connected that a single disappointing earnings print can wipe hundreds of billions off the market in hours. We saw it happen just last month — a brief scare, a sharp reversal, then back to new highs. Resilience… or denial?
Are We Repeating the Dot-Com Playbook?
Every bull market gets compared to 1999 eventually. It’s practically a rite of passage. But this time, the parallels are getting harder to dismiss.
Back then, any company with “.com” in its name could raise a billion dollars on a napkin. Today, slap “AI-enabled” on a pitch deck and watch the term sheets rain. Same energy, better PowerPoints.
The central bank report actually nodded to those ghosts. They saw the same pattern: valuations detached from traditional metrics, concentration at extremes, and a narrative so powerful that bad news gets shrugged off like rain on a windshield.
But — and this is important — they also highlighted a key difference. Unlike many dot-com darlings that burned cash for years, today’s leaders are actually delivering explosive earnings growth. Hundreds of billions in real profits. Not promises. Not “eyeballs.” Cold, hard cash flow.
“The current high valuations appear to be underpinned by exceptionally robust earnings performance.”
That’s the bull case in one sentence. And it’s why most strategists I follow aren’t running for the exits.
Where FOMO Is Real — And Where It’s Dangerous
Not all AI stocks are created equal. That’s become my mantra lately.
The hyperscalers and chip giants at the core? They’re printing money. Data center spending is through the roof. Enterprise adoption is accelerating. These aren’t hopes — they’re purchase orders.
But zoom out one ring, and things get spicier. Companies several steps removed from actual revenue — quantum computing plays, speculative software names, infrastructure adjacent firms — some of these are trading at 80, 90, even 100+ times forward earnings.
That’s where I start hearing dot-com echoes loudest. When a company triples because it mentioned “AI agent” on an investor call but hasn’t shipped a meaningful product? That’s speculation, not investing.
- Core AI winners → backed by real earnings growth
- Second-tier hopefuls → backed by PowerPoint and prayers
- Third-tier story stocks → backed by Reddit and vibes
The smart money, from what I’m seeing, is starting to draw those lines very clearly.
What Happens If the Music Slows?
Here’s the scenario that keeps risk officers awake: what if earnings start missing? Not collapsing — just disappointing.
One quarter of “only” 25% growth instead of 40%. One guidance cut. One major cloud provider saying capex might moderate in 2026. That’s all it could take for sentiment to flip.
And because concentration is so extreme, the fallout wouldn’t be pretty. Open-ended funds with liquidity mismatches, hedge funds carrying leverage, private markets loaded with overvalued rounds — the dominoes are lined up.
We’re not there yet. Far from it. But the ingredients for a sharp correction? They’re all on the table.
The “It’s Different This Time” Crowd Has a Point
I’ll be honest — I roll my eyes every time someone says “it’s different this time.” It literally never is. Except… well… hear me out.
AI isn’t just a sector. It’s potentially the most important general-purpose technology since electricity. The companies leading it have network effects, data moats, and pricing power that most dot-com dreamers could only fantasize about.
One well-known tech analyst recently put it bluntly: we’re in year three of an eight-to-ten-year transformation. His analogy? “It’s 10:30 p.m. at the party, and it goes until 4 a.m.”
Crude? Maybe. But directionally, I think he’s right. The buildout phase — data centers, power plants, chip fabs, model training — has barely begun. Trillions will be spent. Someone is going to make that money.
So… Should You Sell? Buy? Hide?
Here’s my personal take after digging through the data, talking to strategists, and yes, staring at my own portfolio with mild panic.
Don’t let FOMO drive your decisions — but don’t let fear of FOMO drive them either.
- Trim winners that have gone completely parabolic without earnings to match
- Keep exposure to the true infrastructure leaders — they’re still in early innings
- Diversify geographically and thematically where it makes sense
- Have cash ready — corrections in concentrated markets can create generational buying opportunities
- Remember that time in the market still beats timing the market over long horizons
The AI revolution is real. The valuations are stretched. The concentration is scary. And yes, some people are absolutely chasing performance because they’re afraid of missing out.
But revolutions are messy. They always overrun on time and budget. The winners get obscenely rich. The spectators kick themselves. And the cautious? Sometimes they miss the whole thing.
The trick — the part I’m still working on myself — is staying invested without getting reckless. Acknowledging the risks without letting them paralyze you.
Because if this really is the biggest technological shift of our lifetimes, being underweight out of fear might be the most expensive mistake of all.
So breathe. Do your homework. And whatever you do — don’t let FOMO be the only reason you own something. But don’t let fear of FOMO be the reason you own nothing.
That’s the balance I’m trying to strike right now. How about you?
