Broadcom: Buy The Tantrum

Broadcom was reaffirmed as a Strong Buy with a $604 base-case price target, supported by durable AI revenue commitments. The outlook calls for approximately $115B of AI revenue in FY2027 and $230B in FY2028, underpinned by custom-silicon capabilities, a near-monopoly position in AI networking, and VMware-driven software growth. The analysis frames these advantages as reinforcing Broadcom's long-term compounding potential.

Analysis

The key underwriting issue is not AI demand but conversion economics: custom ASIC programs create unusually sticky revenue once designed into hyperscaler platforms, yet they also concentrate bargaining power in a handful of buyers. AVGO can sustain a premium multiple if incremental AI revenue retains semiconductor-like gross margins and networking attach rates rise alongside compute; a mix shift toward lower-margin custom silicon or aggressive customer pricing would challenge that outcome. VMware provides diversification, but it also raises the risk that software execution and customer-retention friction obscure the clean AI narrative.

Near term, positive sell-side framing is unlikely to be a standalone catalyst after a substantial AI rerating. The 1-3 month catalyst path is evidence that booked commitments translate into shipment schedules, customer prepayments, and raised backlog visibility; the critical datapoints are AI segment growth, consolidated gross margin, and management commentary on concentration rather than another long-dated revenue aspiration. A single hyperscaler changing accelerator architecture, delaying capex, or dual-sourcing networking would have outsized implications because AVGO's custom-silicon model has long qualification cycles but lumpy program ramps.

Consensus may underappreciate the second-order constraint: AVGO's success increases dependence on advanced packaging, leading-edge foundry allocation, optical components, and system-level networking deployment. That creates potential read-through beneficiaries in TSM, ANET and COHR/LITE, but also means supply-chain bottlenecks can defer recognized revenue without indicating demand deterioration. Conversely, a broad hyperscaler capex reset would likely compress AVGO's premium valuation faster than earnings estimates fall, given the duration embedded in the AI thesis.

Key Decisions for Investors

• Maintain or initiate AVGO only on post-earnings volatility or a 10-15% pullback, rather than chase bullish target-price revisions; require confirmation of AI growth plus stable-to-up consolidated gross margin. Thesis fails if management cuts AI growth expectations, identifies a material deployment delay, or gross margin declines for two consecutive reported quarters.
• Express the networking attach thesis through a 3-6 month pair: long AVGO / short a diversified semiconductor proxy such as SOXX, sized beta-neutral. This isolates AVGO-specific AI content gains; exit if AI revenue growth decelerates materially while SOXX earnings revisions remain stable.
• Watch TSM and ANET as confirmation trades, not automatic sympathy longs: rising advanced-node/packaging utilization and resilient Ethernet-switching demand would validate AVGO shipment conversion over the next 1-2 quarters. Weak supplier commentary despite AVGO optimism would be an early warning of timing slippage.
• Do not underwrite FY27-FY28 revenue figures into valuation until commitments, customer concentration, cancellation provisions, and margin profile are independently disclosed. If these data remain absent at the next earnings release, treat the long-dated narrative as multiple risk rather than an actionable earnings catalyst.