Morgan Stanley runs to Broadcom's defense as investors worry about competition

Morgan Stanley defended Broadcom as a “core AI winner,” arguing MediaTek’s potential involvement with Alphabet’s TPU chips is “real, but not disruptive,” and reiterated a buy-equivalent rating with a $502 price target. The note highlights Broadcom’s expanding Google chip agreements (through 2013) and expects Broadcom to remain the majority TPU supplier over time with ~80% share, despite prior market concerns. While Broadcom shares are up ~3% Tuesday, they remain down ~22% from the June 2 record close, and the brokerage views recent underperformance as overdone given Broadcom’s AI growth trajectory and ongoing AI compute capacity deals (including with Anthropic via Google chips).

Analysis

The market is still pricing AVGO like a one-customer, one-product story, but the more important mechanism is that custom silicon design wins tend to deepen switching costs once packaging, HBM orchestration, and software/cluster integration are embedded. Even if Alphabet adds a second design partner, that usually expands TPU deployment rather than meaningfully shrinking Broadcom’s content per generation; the incremental risk is margin mix, not a cliff in units. In that setup, the first-order loser is the “displacement” narrative itself, while the second-order winners are the AI infrastructure vendors with the hardest-to-replicate system-level know-how.

The near-term catalyst path is mostly sentiment-driven over the next 1-4 weeks: AVGO can re-rate if investors stop extrapolating MediaTek into share loss and instead focus on multi-year TPU roadmap visibility. The key falsifier is not headlines, but any evidence in the next two earnings cycles that AI revenue inflects below management’s implied path or that Alphabet materially broadens sourcing in a way that compresses Broadcom’s attach rate. If that happens, the multiple compression could be sharper than the revenue impact because AVGO has been trading as an AI quality compounder.

Contrarianly, the consensus may be underestimating how much of the upside is already in the stock’s AI premium, while overestimating how quickly optionality turns into substitution. The bigger risk/reward asymmetry may be in using weakness to add, not chasing strength: if AVGO merely proves stable share rather than gaining share, the stock can still work through EPS growth and balance-sheet return, but not necessarily through multiple expansion. Watch GOOGL’s capex and custom silicon commentary as the real leading indicator; if TPU investment keeps rising, AVGO likely remains the toll collector even in a multi-supplier world.

Key Decisions for Investors

• Buy AVGO on any post-rally fade over the next 1-2 weeks; target a 10-15% rebound if the market pivots from share-loss fears to multi-year AI content growth.
• Structure an AVGO Jan-2026 bull call spread to express upside without paying full volatility; the thesis breaks if the next two quarters show AI revenue growth materially below consensus.
• Pair trade: long AVGO / short a semiconductor basket proxy such as SOXX for 1-3 months if you expect custom-silicon quality to outperform broad AI beta; stop if SOXX outperforms AVGO by 5%+ on renewed GPU enthusiasm.
• Use GOOGL capex/custom silicon updates as a watch item rather than an AVGO short trigger; if Alphabet expands TPU spend again, it is more likely positive for AVGO content than negative for unit economics.
• Avoid chasing NVDA weakness as a direct read-through from this note; the cleaner relative-value expression is AVGO versus the broader semi complex, not AVGO versus NVDA.