Broadcom: One bad quarter, one market meltdown. Is the AI trade finally running out of road?

Broadcom delivered a strong fiscal second quarter, with revenue up 48% year over year to $22.19 billion and net income up 88%. The company also guided third-quarter revenue to $29.4 billion, above Wall Street expectations, signaling continued momentum. The print was not a blowout, but it was clearly better than expected and should support the stock.

Analysis

The market should read this as confirmation that AI infrastructure spend is still outrunning digestion, not as a clean all-clear for semiconductor cyclicality. Broadcom’s upside matters most for suppliers and competitors exposed to networking, custom silicon, and high-speed interconnects: it reinforces that the next leg of AI capex is shifting from GPUs alone toward the plumbing that makes clusters usable at scale. That tends to support the broader AI stack, but it can also pressure names that depend on a single hyperscaler design win or on ASIC displacement risk.

Second-order, the key issue is concentration. When one of the largest AI infrastructure beneficiaries guides above consensus, it implies hyperscalers are still prioritizing deployment over optimization, which usually extends the spending runway for at least another 2-3 quarters. But that same concentration raises execution risk: if customer mix becomes too dependent on a handful of cloud buyers, a pause in one budgeting cycle can hit growth rates harder than the headline revenue trend suggests.

The contrarian read is that the upside may be less about an acceleration in end demand and more about share gains, pricing, and a favorable product mix. If so, the stock can keep working even if the broader semi complex remains choppy—but the move may be less durable than bulls expect once consensus catches up. The main reversal trigger is not a weak single quarter; it is any signal that AI-related ordering is being pulled forward and that 2025 capex budgets are flattening, which would likely show up first in networking and custom silicon names before it hits the larger AI beneficiaries.

Key Decisions for Investors

• Stay long AVGO into the next 2-4 weeks, but use downside hedges rather than taking profits immediately; the setup supports continued multiple support if hyperscaler AI spend remains intact.
• Pair trade: long AVGO / short a basket of AI-enablement peers with weaker product differentiation or customer concentration risk over the next 1-2 months; this isolates share-gain and mix-upside versus beta.
• Buy short-dated downside protection in the semi complex if AVGO rallies sharply on open; the risk/reward favors hedging against a 'good numbers, crowded expectations' fade over 30-60 days.
• Add selectively to names exposed to AI networking and interconnect demand on a 3-6 month horizon; the signal here is that cluster buildouts are still moving down the stack, not just into accelerators.
• Avoid chasing lower-quality ASIC stories until the next hyperscaler capex readthrough; the second-order risk is that a single customer pause can compress estimates quickly.