Broadcom Q2: Why The FY27 $100B Target May Already Be Obsolete

Bernstein and BofA’s math implies Broadcom’s $100B AI revenue target is a floor, with $20B per GW and 9-10 GW of customer commitments for 2027 pointing to additional upside. Google is committed through 2031, Anthropic added 3.5 GW, and Meta extended through 2029, all signed in April after the last disclosed $73B backlog. The June 3 Q2 earnings report is expected to refresh backlog figures and could lead management to raise FY2027 guidance.

Analysis

The key takeaway is not just that AVGO’s AI narrative is intact, but that the market may still be underestimating the convexity in backlog conversion. If April-only contracts already imply a 2027 runway that can support or exceed the current revenue target, the bigger surprise into the June 3 print is likely a step-up in forward visibility rather than a one-quarter beat. That matters because the stock is still being priced like a hardware beneficiary, when this pattern increasingly resembles a long-duration infrastructure platform with multi-year contracted cash flows.

Second-order winners are the ecosystem names that consume or integrate this capacity: GOOGL and META are effectively signaling that AI compute remains a strategic priority even as capex scrutiny rises. The duration of those commitments reduces the probability of an abrupt hyperscaler capex reset over the next 2-3 quarters, which should support the broader AI complex and custom silicon supply chain. The less obvious loser is any merchant accelerator or networking vendor relying on spot demand; the more demand is locked into AVGO-style bespoke supply, the less room there is for price competition outside committed capacity.

The risk is not demand cancellation, but digestion: after the June print, investors may extrapolate too aggressively and bid the stock ahead of the actual guide revision, leaving room for a sell-the-news reaction if management only confirms rather than raises. A second risk is that backlog quality matters as much as size—if later-stage deployment timing slips into 2028, the headline revenue floor remains intact but near-term multiple expansion could stall. BAC is only indirectly relevant here as a read-through on consensus framing and sell-side model revisions, not as a fundamental driver.

Contrarian view: the market may be too focused on AVGO’s top-line upside and not enough on margin durability. If AI revenue becomes more contracted and less cyclical, the debate shifts from order growth to mix, pricing power, and gross margin normalization; that can cap upside even as revenues surprise. In that scenario, the better asymmetry may be relative value long AVGO versus short a basket of AI beneficiaries with less visible contracted demand and more execution risk.

Key Decisions for Investors

• Add AVGO ahead of the June 3 earnings print; use a 2-4 week horizon and target upside from a backlog re-acceleration re-rate, with a stop if management merely reiterates rather than lifts FY2027 guidance.
• Express a relative-value view: long AVGO / short a high-beta AI hardware basket over the next 1-3 months to capture backlog certainty versus more cyclical order exposure.
• Use GOOGL and META as secondary longs on any post-print weakness in AVGO, on the thesis that committed AI capex reduces near-term downside to hyperscaler spend; time frame 1-2 quarters.
• Consider short-dated AVGO calls into June 3 only if implied volatility remains below the likely earnings surprise range; otherwise prefer common stock to avoid IV crush risk.
• If AVGO gaps sharply higher on the print, trim into strength rather than chase — the highest-risk outcome is a guide raise with no immediate backlog monetization, which could lead to consolidation over the following 4-8 weeks.