A $1.2 Billion AI Guidance Gap Triggered a Valuation Reset
This looked like a valuation reset, not proof that Broadcom's AI story broke.
The reported quarter was still strong: Broadcom posted $22.19 billion in revenue and $2.44 in EPS, both ahead of consensus. But the market was focused on the forward view. The trigger was third-quarter AI chip sales guidance of $16 billion, below the $17.2 billion expected by analysts. In other words, Broadcom did not clear a bar that had already moved higher.
Why the reaction was so sharp
After the stock hit all-time highs on AI enthusiasm, expectations were stretched. That made the response less about a weak quarter and more about a premium multiple getting repriced quickly. The 12% after-hours drop reflected high expectations for companies benefiting from the AI boom.
The core AI infrastructure thesis did not collapse. Broadcom is still positioned to benefit from hyperscaler AI spending this year, and management again said AI semiconductor revenue guidance is in excess of $100 billion. That does not look like a broken growth curve; it looks like a market that is no longer satisfied with "strong" and is now demanding "better than priced."
What still matters
The core AI infrastructure thesis did not collapse. Broadcom is still positioned to benefit from hyperscaler AI spending this year, and management again said AI semiconductor revenue guidance is in excess of $100 billion. That does not look like a broken growth curve; it looks like a market that is no longer satisfied with "strong" and is now demanding "better than priced."
Broadcom's Growth Still Looks Strong, but Expectations Were Even Stronger
The guidance miss was the spark, but the deeper debate is about trajectory. Wall Street is no longer asking whether Broadcom is growing; it is asking whether the company can keep outgrowing expectations as those expectations rise.
Operating momentum remained strong
If you separate execution from valuation, Broadcom's latest operating read was not weak. AI semiconductor revenue in the second quarter grew 143% year over year, and the company guided current-quarter revenue to $29.4 billion versus $28.61 billion consensus. That points to continued strength in the AI infrastructure build-out through Broadcom's custom silicon.
Management also said that momentum should continue into fiscal 2027, with AI semiconductor revenue expected to remain above $100 billion. On its own, that is not evidence of a slowdown.
Why the market sold the stock
The problem was not demand; it was the bar. After the stock added as much as $300 billion in valuation over five sessions, investors were looking for "better than model," not just solid results. In that setup, even a fast-growing company can be sold if it does not beat a very high implicit expectation. That helps explain why the reaction looked harsher than the underlying operating data.
What decides the next move
- Bull watchpoint: AI chip growth stays very strong, revenue beats persist, and fiscal 2027 guidance above $100 billion starts to look conservative.
- Bear watchpoint: AI semiconductor growth decelerates more meaningfully, or Broadcom simply stops raising its outlook fast enough for a crowded trade.
My view is that the market may be early to declare the next leg exhausted. But recovery will likely require more than solid execution; it will require Broadcom to keep compounding faster than Wall Street keeps repricing the story.
What Investors Should Watch Next: Momentum Break or Better Entry?
The setup is straightforward: a pullback can become an entry, but only if the AI infrastructure cycle stays stronger than the market's impatience.
Two paths from here
If the broader build-out holds, this 12% after-hours sell-off may look more like a pressure release than a broken narrative, especially after shares added as much as $300 billion in valuation in five sessions and then dropped sharply.
The bull case still has support from the broader spending trend. Hyperscaler AI spending this year is estimated at $650 billion, and Broadcom remains tied to that spend through major cloud and AI customers including Google, Meta, Anthropic, and OpenAI. If that spending remains robust, Broadcom may still be able to re-rate without another perfect quarter.
What would support a better entry
- Management again sounds more aggressive on AI silicon before the next full quarter.
- Customer capex commentary continues to suggest that the $650 billion spending view is conservative, not peaked.
- The stock stabilizes relatively quickly after the earnings shock, which would suggest the market treated the move as a reset rather than a regime change.
What would argue for waiting
If Broadcom does not raise its AI outlook again, or if the post-earnings move turns into a longer digestion phase, the message is that the market now wants consistency, not just momentum. That would not invalidate the long-term story. Analysts still see it as a great long-term hold. But it would suggest investors should wait for clearer proof that the next leg is real rather than pay up into another euphoria cycle.