Broadcom: 4 Reasons To Load Up On The Buys

Broadcom sees over $100B in AI chip revenue by 2027 from custom XPUs, implying roughly 123% growth from a $24.5B base. Management also expects full AI rack deployments to preserve or expand semiconductor gross margins, with Q2 FY26 guidance at 77%. Offseting the bullish outlook, TSMC and component supply bottlenecks could delay revenue realization, and FY26 revenue estimates have already edged lower.

Analysis

The important read-through is that the AI supply chain is transitioning from a compute bottleneck to a systems-integration bottleneck. That shifts value capture upward from generic accelerator vendors toward whoever controls the custom silicon roadmap and the rack-level integration layer, which should keep pricing power unusually sticky even if unit volumes normalize. The market should also stop thinking about gross margin expansion and revenue growth as mutually exclusive here; if deployment density rises faster than component inflation, the mix can stay margin-accretive for longer than consensus expects.

The bigger second-order winner is the rest of the AI infrastructure stack that can absorb complexity without taking custom-design risk: advanced packaging, high-end substrate, power delivery, and network gear. The losers are less about a single named competitor and more about any broad AI hardware exposure dependent on undifferentiated accelerator pricing or spot component availability, because custom systems tend to concentrate share and compress the addressable market for commoditized silicon. On the foundry side, the near-term pressure point is not demand, but throughput; if capacity remains tight into the next two quarters, revenue recognition can lag bookings and create a misleading setup where fundamentals stay intact but estimates still drift lower.

Catalyst timing matters: in the next 1-3 months, the stock is vulnerable to any evidence that shipments are being pushed out rather than canceled, because the sell-side will likely cut near-term numbers before it raises outer-year estimates. Over 6-12 months, the key risk is execution concentration: a few hyperscale customers can both accelerate upside and create volatility if design ramps slip or negotiate harder on pricing once volume scales. The market is probably underestimating how durable the margin profile can be if custom rack deployments internalize more of the value chain; the consensus may be too focused on component scarcity and not enough on structural mix shift.

The contrarian angle is that the supply chain bottlenecks may be misread as a demand problem, which would be a mistake if order visibility is intact. If the company can maintain gross margin guidance while peers struggle with dilution, that implies it has enough pricing power to pass through tightness and preserve returns, making any post-estimate reset a potentially attractive entry point rather than a thesis break. The key tell is whether estimate cuts are confined to the next two quarters or start infecting the outer-year ramp; the latter would signal a real change in AI capex appetite.

Key Decisions for Investors

• Stay long AVGO on any 5-8% pullback tied to near-term revenue delays; use a 6-12 month horizon and favor adding only if gross margin guidance remains intact, because that would confirm pricing power and make estimate cuts a timing issue rather than a thesis break.
• Buy AVGO call spreads 6-9 months out to express upside to the 2027 AI revenue path while capping premium outlay; risk/reward is attractive if the market continues to underwrite only the next two quarters.
• Pair long AVGO / short TSM for the next 1-2 quarters if you want to isolate bottleneck risk: AVGO benefits from pricing and design-win control, while TSM is the cleaner expression of capacity constraint without the same margin leverage.
• Avoid chasing broad AI-semiconductor beta until supply-chain data improve; prefer names with system-level integration leverage over commodity accelerator exposure, since custom rack economics should outperform on a 6-12 month view.
• If AVGO sells off on estimate cuts but forward AI commentary stays firm, scale in aggressively—this is the setup where the stock can re-rate before reported revenue catches up, especially over the next 2 earnings cycles.