Goldman Sachs reiterates Broadcom stock rating on Meta partnership By Investing.com

Goldman Sachs reiterated a Conviction Buy on Broadcom with a $480 price target after the company expanded its multi-year AI partnership with Meta through 2029. The initial commitment exceeds 1GW and Broadcom’s FY27/FY28 EPS estimates are about 14% above consensus, signaling stronger long-term earnings momentum. The news is constructive for Broadcom and reinforces its position in AI infrastructure, though the article also notes the stock may be slightly overvalued at current levels.

Analysis

This is less about a single supply contract and more about Broadcom hardening its position as the toll collector on hyperscaler custom silicon. The second-order effect is that AI capex is becoming increasingly concentrated in networking, interconnect, and co-design services rather than just compute accelerators; that shifts value away from merchant GPU incumbents and toward the enablers that sit inside the rack and the cluster. If this multi-generation roadmap holds, it validates a longer-duration revenue stream with unusually low churn, which should support multiple expansion more than the near-term EPS beats themselves.

The market is likely still underestimating the operating leverage embedded in these ecosystem wins. Once a hyperscaler commits to a custom silicon stack, switching costs rise nonlinearly because the pain is not chip replacement but system redesign, software retraining, and supply-chain requalification; that creates a durable moat that can compound across future generations. The real competitive pressure falls on smaller ASIC vendors and networking peers that lack the scale to offer end-to-end architecture support, while foundry and advanced packaging capacity likely remains the bottleneck that determines who can actually monetize AI demand.

Near term, the main risk is that investors extrapolate headline-strength faster than the physical rollout can convert into revenue, so the stock can run ahead of shipments for several quarters. A reversal would likely come from any sign of capex digestion, a delay in MTIA deployment, or a broader AI spending pause at the hyperscaler level. Over a multi-year horizon, though, the more important question is whether Broadcom’s share of custom AI infrastructure grows faster than consensus assumes; if yes, earnings power into FY27-FY28 still looks underappreciated relative to the durability of the order book.

Consensus may be missing that this is also a governance signal: moving leadership into an advisory role suggests the relationship is becoming strategic enough to survive personnel changes and board turnover. That lowers execution risk and increases the odds that the partnership becomes a template for additional customers, not just a Meta-specific win. The stock is not cheap, but for a business with rising visibility and high switching costs, valuation may be less about current multiples and more about the market re-rating the quality of future cash flows.

Key Decisions for Investors

• Stay long AVGO on pullbacks over the next 2-6 weeks; use weakness tied to broad semis volatility to add, with a 6-12 month target driven by multiple support from recurring AI infrastructure wins.
• Pair trade: long AVGO / short a basket of GPU-adjacent names with less direct system-level control over custom deployments; this isolates the custom-silicon monetization theme and reduces beta to AI headline risk.
• Buy META on any post-news consolidation over the next 1-3 months; the market may still be underpricing the cost-efficiency and capex leverage of custom silicon versus off-the-shelf compute, which can support margin durability.
• Consider a call spread in AVGO with 6-9 month tenor to express upside while limiting downside from valuation compression; best risk/reward if the market starts to price the FY27-FY28 earnings visibility.
• Avoid chasing suppliers lacking differentiated networking or co-design exposure; the second-order winner is the architecture provider, not the generic component vendor.