Broadcom's Next Earnings Report on September 2 Could Send the Stock Soaring. Here's Why.

Broadcom is positioned for a likely AI-driven “beat-and-raise” catalyst ahead of its fiscal 2026 Q3 report on Sept. 2. In Q2 (ended May 3), revenue rose 48% YoY to $22.2B and adjusted EPS jumped 54% to $2.44, with AI ASIC sales up 143% to $10.8B (nearly half of revenue). Management guided Q3 revenue of $29.4B (+84% YoY) and adjusted EBITDA around $20B (+87%), and the article notes the stock trades at <19x next year’s earnings despite being >23% below its peak.

Analysis

Broadcom is a cleaner way to own AI infrastructure than the market is treating it: the incremental dollar of AI capex is increasingly shifting from generic compute toward custom silicon and the networking fabric that makes clusters usable. That mix matters because it supports both gross margin and visibility, which usually deserves a premium multiple, but the stock still trades like an “AI beta” name rather than a compounder with software ballast.

The immediate setup is event-driven: a beat-and-raise can force systematic buyers back in, while a miss would likely be punished harder than fundamentals justify because positioning is crowded and expectations are high after the recent run. Over 1-3 months, the key question is whether hyperscaler demand is broadening or simply being pulled forward by a few customers; if the latter, growth can decelerate abruptly even if the quarter itself looks strong. Over 6-18 months, continued ASIC share gains are a structural threat to merchant GPU monetization, but that also argues for persistent network spend, which is a tailwind for data-center infrastructure peers.

Contrarian view: the consensus may be underestimating duration, not magnitude. The stock does not need a new AI story; it needs evidence that the current story still has several budget cycles left. If management’s commentary implies digestion in custom silicon orders or softer incremental AI bookings, the multiple can compress quickly despite “good” reported numbers. Conversely, a clean guide-up would likely re-rate the name back toward the high-20s earnings band as investors stop treating AI demand as a one-quarter phenomenon.

Key Decisions for Investors

• Long AVGO into the earnings print via a defined-risk call spread; best risk/reward if implied move is pricing in only a modest beat. Thesis fails if management signals AI digestion or guide-only growth.
• If AVGO gaps up on a beat-and-raise, add to ANET on the first pullback over the next 1-3 sessions; network spend should lag compute by a few quarters but remains the second-order winner of every incremental AI cluster build.
• If AVGO reports any guide-down in AI semiconductor growth or commentary on customer concentration/digestion, short AVGO on the open for a 1-4 week trade; the stock is likely to de-rate faster than fundamentals because positioning is crowded.
• Set a post-earnings alert on AVGO valuation near 19x forward earnings: below that level on strong guidance, the risk/reward improves materially for a 6-18 month core long; above that level after a clean print, trim into strength.
• Use NVDA only as a relative-value hedge, not the primary expression: a broad AI-spend confirmation supports both names, but any hint of custom-silicon share gain should favor AVGO vs NVDA on a 1-3 month horizon.