Broadcom's Post-Earnings Dip Is a Compelling Buying Opportunity

Broadcom reported 48% year-over-year revenue growth in fiscal Q2 to $22.2 billion, with AI chip revenue jumping 143% to $10.8 billion and net profit margin reaching 42%. Management guided fiscal Q3 revenue to $29.4 billion, implying 32.5% sequential growth, reinforcing the view that AI infrastructure spending is still accelerating. The stock's >20% pullback is framed as a buying opportunity given continued market share gains in ASICs and strong demand from major tech customers.

Analysis

The market is treating this as a single-name miss, but the more important signal is that AI capex is shifting from experimental to industrial. That transition favors suppliers with design-in lock and workload specificity, which is why ASIC-centric exposure should continue to outgrow the broader AI hardware basket even if GPU demand stays strong. The second-order winner is the foundry/advanced-packaging ecosystem: every incremental custom-chip program increases dependency on a small set of manufacturing nodes, which tightens supply and preserves pricing power deeper in the stack.

The near-term setup is less about revenue deceleration than about expectation compression. A stock that rerated on accelerating AI optics can derate 10-15% quickly if investors think the growth curve is still linear rather than lumpy; that means the next catalyst is not just another beat, but evidence of broader customer adoption beyond the largest hyperscalers. If custom silicon starts penetrating more inference workloads, the earnings power expands materially because those deployments tend to be sticky, high-volume, and lower churn than generic accelerator spend.

Consensus is probably underestimating how much of the AI buildout is becoming a procurement arms race rather than a model-quality race. That benefits infrastructure vendors with differentiated economics, but it also creates a timing risk: if hyperscalers temporarily pull forward spend and then pause to digest capacity, the market will misread that as secular slowdown. Over a multi-quarter horizon, the bigger risk to the bullish thesis is not demand destruction but customer concentration—any delayed ramp at one or two platform accounts can create visible quarter-to-quarter volatility even while the year-long trend remains intact.

Key Decisions for Investors

• Add to AVGO on weakness over the next 1-2 weeks, but size for volatility: buy with a 3-6 month horizon and use a 15-20% trailing stop, since the setup is favorable as long as AI capex remains elevated.
• Pair trade: long AVGO / short a basket of more saturated GPU names such as NVDA and AMD for 1-3 months. Thesis: custom silicon and workload-specific inference should capture a larger share of incremental AI capex while GPU multiples remain more sensitive to any digestion pause.
• Initiate a call spread on AVGO expiring in 3-6 months to express rebound potential with defined risk; this is cleaner than outright stock if the market continues to punish guide conservatism.
• Monitor GOOGL and AAPL as leading indicators for the next AVGO leg: if their AI capex commentary keeps improving, that is the highest-conviction confirmation that custom-chip demand is broadening beyond the current hyperscaler cohort.
• Avoid chasing after a 1-2 day relief rally; the better entry is after the post-earnings overhang is fully digested or if the stock retests the next support zone, because the catalyst is multi-quarter rather than immediate.