Back to News
Market Impact: 0.42

Why Broadcom Stock Is Still Going Down

Artificial IntelligenceCorporate EarningsCorporate Guidance & OutlookCompany FundamentalsAnalyst EstimatesAnalyst InsightsInvestor Sentiment & PositioningTechnology & Innovation
Why Broadcom Stock Is Still Going Down

Broadcom reported Q2 2026 EPS of $2.44 on a GAAP profit of $1.91 per share, up 85% year over year, and raised Q3 revenue guidance to $29.4B, implying 32% sequential and 89% YoY growth. However, investors focused on AI-chip revenue guidance of $16B versus the $17.2B Wall Street expected, driving a 12.6% selloff despite the earnings and guidance beat. The article frames the move as an AI guidance miss rather than a fundamental deterioration, with margins still expected to remain stable at a high level.

Analysis

The market is pricing AVGO as if AI revenue is the only metric that matters, but the larger signal is that hyperscaler and enterprise demand is still strong enough to support a very large base-business re-acceleration. The second-order effect is competitive: if Broadcom can monetize custom silicon and networking at this scale while maintaining margins, it raises the bar for peers that rely on generic GPU attach rates and makes NVDA’s ecosystem look less unassailable at the margin. The more important read-through is not near-term absolute growth, but the durability of capex budgets across the AI stack into the next two reporting cycles.

The selloff looks like positioning-driven de-risking rather than a true fundamentals break. In crowded AI winners, any mismatch between sell-side estimates and company commentary tends to trigger mechanical multiple compression, especially when the stock trades on forward earnings that leave little room for narrative slippage. That creates a window where AVGO can underperform for days to weeks even if the business keeps compounding, which argues for distinguishing between a sentiment air pocket and an actual demand inflection.

The contrarian view is that the market may be overreacting to a single-quarter mix issue while underappreciating how sticky custom AI infrastructure spending becomes once deployed. If AI budgets are being reallocated from off-the-shelf accelerators toward tailored networking, interconnect, and silicon, then the ecosystem winners broaden beyond NVDA and may include suppliers with more software-like margin profiles. The risk is that if AI guidance is indeed a leading indicator, then this quarter could mark the first sign that hyperscaler digestion is starting, which would hit the entire semiconductor complex over the next 1-2 quarters.