Mizuho reiterates Broadcom stock rating on extended Google deal By Investing.com

Broadcom extended its TPU design and supply partnership with Google and Anthropic through 2031, including a five-year TPU agreement with Google and access to ~3.5 gigawatts of TPU compute for Anthropic starting in 2027. Mizuho reiterated an Outperform with a $480 price target and estimates the deal could drive >$80bn cumulative revenue for Broadcom, with Google/Anthropic representing 40–64% of Broadcom’s revenue; Anthropic’s revenue run-rate is cited at $30bn and Broadcom reports a 77% gross margin and 25% y/y revenue growth. Multiple analysts reiterated bullish ratings (BofA $450, Rosenblatt $500, Jefferies $500; D.A. Davidson Neutral $375); market cap is $1.54tn and the stock trades at $326.40, making this a material positive for Broadcom and its role in AI infrastructure.

Analysis

The biggest non-linear effect is not incremental dollar flow to a single supplier but the structural shift in bargaining leverage that comes with multi-year design incumbency. When a vendor becomes the path-of-least-resistance for future AI rack designs, it converts one-off sales into annuity-like revenue with outsized margin retention — that dynamic forces tier-two networking and ASIC vendors to compete on cost or seek niches, compressing their forward multiples and increasing M&A pressure over 12–36 months.

On the supply side, this scale-through-design approach raises a capacity choreography risk: wafer and advanced packaging windows become the gating factor for end-customers' rollout cadence, so semiconductor capacity reallocation (driven by larger GPU/accelerator customers) can turn a 1–2 year growth runway into a 6–12 month timing miss. Equally important are policy and customer-concentration tail-risks — export controls, a pivot to in-house tooling, or a single large buyer re-negotiating economics can erase a multi-year premium fast; treat contractual runway as a probability-weighted cash-flow stream rather than guaranteed revenue.

For portfolio construction the clearest signal is optionality: the market is foreshortening execution risk while paying up for design permanence. That sets up asymmetric payoffs for positions that own the incumbent supplier with limited downside through defined option structures, and for short exposures to smaller networking/ASIC players who face margin compression and potential customer attrition over the next 6–24 months.

Key Decisions for Investors

• Core overweight AVGO (12–18 months): initiate a 3–5% position size into strength; hedge with a 12-month 20% OTM protective put (cost as insurance). Rationale: capture multi-year design premium while capping downside; target asymmetric upside of ~25–40% vs max loss = put premium + stock drop to strike.
• Defined-risk options: buy a 12–18 month bull-call spread on AVGO using ATM / +35% strikes (debit spread). This limits max loss to premium (2–4% of notional) while targeting 3–5x upside if design-led adoption continues — suitable as a directional way to lever the AI-infrastructure rollout without open-ended gamma exposure.
• Pair trade (6–18 months): long AVGO / short MRVL (or small-cap networking peer) equal notional. Expect convergence as incumbent captures higher-margin share; set stop-loss if the spread widens >15% intraday. Reward comes from margin re-rating and customer consolidation; risk is MRVL finding alternative high-growth cloud customers.
• Event hedge (near-term, days–weeks): buy short-dated AVGO puts into next major earnings or product-delivery milestone to protect against delivery/timing disappointment. Keep hedge size to 10–20% of core position to avoid paying continuous carry.