Broadcom Shares Tumble Despite Surging AI Revenue. Should Investors Buy the Dip?

Broadcom reported Q2 revenue of $22.19 billion, up 48% year over year, with adjusted EPS of $2.44 versus $2.40 consensus. AI semiconductor revenue surged 143% to $10.8 billion, and management guided fiscal Q3 revenue to $29.4 billion, ahead of the $28.5 billion estimate, though shares fell because investors wanted a higher fiscal 2027 AI outlook. The stock now trades at 22.5x fiscal 2027 forward earnings after the dip.

Analysis

The market is reacting to a change in narrative, not a change in fundamentals. When a stock has been re-rated on an AI scarcity premium, simply reaffirming a long-dated revenue path can trigger a de-risking event because holders were positioned for a faster upward revision; that creates near-term downside even if the core business is intact. The key second-order issue is that Broadcom is becoming a barometer for the durability of hyperscaler/custom silicon capex, so any pause in guidance momentum can bleed into the entire AI infrastructure complex.

The more interesting signal is not the revenue miss, but the mix: networking strength offsetting lower-margin custom silicon implies the company is still monetizing AI buildout, while software growth lag hints at a slower cross-sell/VMware integration cadence. That matters because the market is implicitly paying up for a diversified AI compounder; if software fails to reaccelerate, the multiple can compress even with strong semiconductor growth. In other words, the floor is higher than the headline reaction suggests, but the ceiling depends on whether software becomes an earnings lever rather than just ballast.

Consensus appears to be underweight the supply-chain advantage. In a constrained environment, being component-secure is effectively a demand moat: it can pull forward share from competitors that are forced to wait for packaging, networking, or substrate capacity. That is bullish not only for AVGO, but also for GOOGL and META as their AI deployment timelines become less execution-risky, while it is modestly negative for NVDA/INTC in the sense that every additional custom-silicon dollar broadens the market away from merchant accelerators and toward vertical integration.

The contrarian risk is that the stock may already be pricing in a very high probability of flawless execution through fiscal 2027-28. If hyperscaler budgets normalize, or if one of the six anchor customers rationalizes spend, the implied growth duration collapses and the multiple can de-rate quickly. This is a months-to-years story, but the next catalyst is the next two earnings cycles: if software growth doesn’t improve and AI guidance merely repeats, the post-earnings selloff can become a valuation trap rather than a buying opportunity.

Key Decisions for Investors

• Buy AVGO on weakness with a 3-6 month horizon, but only in tranches; upside comes from multiple re-rating if software reaccelerates, while downside is bounded by AI backlog and supply lock-ins.
• Express a relative-value view: long AVGO / short NVDA over 1-2 quarters if the thesis is custom silicon share gain and hyperscaler diversification; expect AVGO to outperform on durability of earnings, though with lower beta.
• For event-driven positioning, sell near-dated AVGO put spreads into any further post-earnings volatility; the current move looks like a positioning flush rather than a broken thesis, but the stock can stay range-bound until the next guide-up.
• Add GOOGL and META on pullbacks as indirect beneficiaries of more certain AI infrastructure delivery; a 6-12 month horizon captures improved capex execution without paying AVGO’s premium multiple.
• Avoid chasing INTC as a ‘catch-up AI’ trade here; Broadcom’s custom silicon momentum reinforces the structural challenge for merchant/legacy silicon names over the next 12-24 months.