Broadcom just dropped a monster quarter. We’re talking record revenue of $22.18 billion—a 48% jump year-over-year—and earnings that beat expectations. The AI chip maker is basically printing money right now, with AI semiconductor revenue alone surging 143% to $10.8 billion. Their customers? The usual suspects: Google, Meta, OpenAI, and basically every company betting the farm on AI infrastructure.
So naturally, the stock tanked 15% the next morning.
Welcome to the weird world of mega-cap tech stocks, where crushing it can somehow feel like failing.
Here’s what happened: Broadcom’s management came out swinging with guidance that would make most companies blush. They’re projecting Q3 semiconductor revenue around $16 billion—that’s more than 200% growth year-over-year. Total Q3 revenue is expected to hit $29.4 billion, an 84% increase. These aren’t typos. This is the AI boom in real time.
But the market’s reaction reveals something important: expectations for Broadcom had gotten absurdly high. The stock was already up 400% over three years before this earnings report. When you’re trading at 50 times forward earnings (compared to a historical median of 18x), there’s basically no room for anything less than perfection. And even perfection isn’t always enough.
The Bull Case (Why You Might Want to Buy the Dip)
Let’s be real—Broadcom’s fundamentals are legitimately impressive. AI revenue growth in the triple digits isn’t a fluke; it’s a sign that demand is nowhere near saturated. The company has locked in relationships with the hyperscalers that are actually building out AI infrastructure, which means they’ve got visibility into future revenue. Plus, unlike pure-play chip companies, Broadcom has a software division generating recurring revenue, which provides some cushion when hardware cycles get weird.
The “Maybe Pump the Brakes” Case
Here’s the thing though: much of Broadcom’s future depends on a relatively small group of mega-rich customers continuing to throw billions at AI. If Google, Meta, Amazon, and Microsoft suddenly decide they’ve spent enough on AI infrastructure—or if they start building more chips in-house—Broadcom’s growth story gets a lot less exciting. That’s concentration risk, and it’s real.
There’s also the valuation question. Even after the 15% drop, Broadcom isn’t exactly cheap. And margins could compress as AI hardware becomes a bigger part of the mix. The company needs to keep executing flawlessly to justify where the stock is trading.
The Bottom Line
Broadcom’s Q2 results were objectively stellar. Revenue up 48%, AI revenue up 143%, earnings beat, and guidance that suggests the party’s just getting started. But the market’s reaction is a reminder that even the best companies can get ahead of themselves when expectations get too frothy.
Is this a buying opportunity? Maybe. But it’s also a good reminder that “great company” and “great stock” aren’t always the same thing. Broadcom remains one of the best plays on AI infrastructure, but investors might want to stay disciplined about valuation and not get too caught up in the hype.
