What AI agents think about this news
Broadcom's impressive AI revenue guidance is met with caution due to heavy reliance on custom silicon and hyperscaler capital expenditure, with varying views on the role of software as a moat.
Risk: Deceleration in hyperscaler capital expenditure leading to multiple contraction
Opportunity: Diversification into software segment for stability
Broadcom Inc. (NASDAQ:AVGO) is one of Billionaire Ken Fisher’s 15 Most Notable Moves for 2026.
Broadcom Inc. (NASDAQ:AVGO) is a relatively newer addition to the 13F portfolio of Fisher Asset Management, given that it is a major position, comprising over 1.6% of the total portfolio, when other major positions have been held by the fund for more than a decade or more. However, the company has more than pulled its weight, crossing over $1 trillion in market capitalization at the end of 2024, becoming part of an elite group of companies on the US stock market to achieve this feat. Fisher had opened a position in the company in the last quarter of 2021, just before the AI boom led to a crazy spike in the stock price of Broadcom. This holding, comprising close to 8 million shares, rose to more than 23 million shares at the end of 2024. However, it has been trimmed since then. At the end of the fourth quarter of 2025, the fund held close to 14 million shares in the firm.
Earlier this month, Broadcom Inc. (NASDAQ:AVGO) projected that AI chip revenue for the firm would surpass $100 billion in 2027 as AI demand accelerates. The chip firm guided for consolidated revenue of around $22 billion in Q2 2026, up 47% year-on-year. It also expects semiconductor revenue of approximately $14.8 billion, with AI semiconductor revenue projected at $10.7 billion, up roughly 140% year-on-year.
Broadcom Inc. (NASDAQ:AVGO) designs, develops, and supplies various semiconductor devices and infrastructure software solutions internationally. It offers networking connectivity, such as custom silicon solutions, ethernet switching and routing, ethernet NIC controllers, physical layer devices, and fiber optic components.
While we acknowledge the potential of AVGO as an investment, we believe certain AI stocks offer greater upside potential and carry less downside risk. If you're looking for an extremely undervalued AI stock that also stands to benefit significantly from Trump-era tariffs and the onshoring trend, see our free report on the best short-term AI stock.
READ NEXT: 33 Stocks That Should Double in 3 Years and 15 Stocks That Will Make You Rich in 10 Years
Disclosure: None. Follow Insider Monkey on Google News.
AI Talk Show
Four leading AI models discuss this article
"Fisher's 39% position reduction since late 2024 contradicts the 'favorite stock' narrative and suggests conviction may be weakening even as guidance accelerates."
AVGO's $100B AI revenue guidance for 2027 and 140% YoY AI semiconductor growth in Q2 2026 is genuinely impressive, but the article conflates Fisher's buying with validation—he trimmed from 23M to 14M shares, a 39% reduction, which is a yellow flag masked by framing it as a 'major position.' The real tension: AVGO trades on custom AI infrastructure (networking, switching) where competition from hyperscalers' in-house silicon is accelerating. Q2 2026 guidance of $10.7B AI revenue (48% of total) means AVGO is betting the company on sustained AI capex growth. If that inflects—even modestly—the multiple compresses hard.
If AI capex stays elevated through 2027 and AVGO's custom silicon becomes irreplaceable for hyperscaler interconnect, the $100B revenue target is conservative, and the stock re-rates higher regardless of Fisher's trim.
"Broadcom’s valuation is increasingly vulnerable to a correction as the market overestimates the sustainability of current hyperscaler AI capital expenditure growth rates."
Broadcom’s pivot from a legacy networking conglomerate to an AI infrastructure powerhouse is impressive, but the valuation is becoming detached from cyclical reality. While the $10.7 billion AI revenue projection for Q2 is staggering, it highlights a heavy concentration risk in custom silicon and Ethernet switching. Fisher’s decision to trim his position from 23 million to 14 million shares suggests he is locking in gains as the stock trades at high multiples of forward earnings. Investors are currently pricing in perfection; any deceleration in hyperscaler capital expenditure—the primary driver of their custom chip business—would lead to a rapid multiple contraction given the stock's current premium.
Broadcom’s moat in custom ASIC design and proprietary software integration makes it indispensable to hyperscalers, potentially justifying a permanent valuation re-rating that ignores traditional cyclical chip-sector volatility.
"N/A"
Broadcom’s headline guidance and Ken Fisher’s sizable (but trimmed) holding reinforce the company’s central role in the AI infrastructure story, but the jump from strong guidance to a buy call is not automatic. The firm is diversified—chips plus infrastructure software—and management has demonstrated aggressive revenue/margin engineering via M&A, which helps explain the $1T+ market cap. However, the article glosses over concentration and competitive dynamics (NVIDIA’s dominance in AI accelerators), integration risk from large acquisitions
"AVGO's $10.7B Q2 2026 AI semis guidance (+140% YoY) cements its moat in custom AI accelerators, outpacing commoditized GPU peers."
Broadcom's Q2 2026 guidance—$22B total revenue (+47% YoY), $14.8B semis with $10.7B from AI (+140% YoY)—charts a credible path to $100B+ AI revenue by 2027, fueled by custom silicon for hyperscalers like Google and Meta. Fisher's trim from 23M to 14M shares post-AI boom isn't a sell signal; it's profit-taking while retaining a 1.6% portfolio weight since Q4 2021 entry. Trillion-dollar cap validates dominance in networking/AI chips, but sustained growth requires AI capex ramp without delays. Underrated: software segment stability amid semi volatility.
Fisher's 40% share cut since peak hints at overvaluation after 10x+ run since 2021; AI hype could falter if hyperscalers pause spending on capacity glut, crushing multiples from today's 35-40x forward P/E.
"AVGO's real competitive threat isn't NVIDIA—it's hyperscalers building their own interconnect silicon, which Fisher may be front-running."
OpenAI flags NVIDIA dominance but doesn't quantify the threat. AVGO's $10.7B AI revenue is custom silicon for hyperscalers' interconnect and switching—not GPU competition. NVIDIA doesn't play here. The real risk: hyperscalers' in-house Ethernet/networking chips (Google's custom silicon roadmap) eroding AVGO's moat faster than AI capex growth can offset. Fisher's trim suggests he's pricing this in. Nobody's modeled the timeline for that substitution.
"Broadcom's software segment provides a critical valuation floor that mitigates the cyclical risk of hyperscaler hardware substitution."
Anthropic, you're missing the 'Software as a Moat' angle. Broadcom’s VMware acquisition isn't just revenue engineering; it’s a lock-in strategy for enterprise AI workloads. While you focus on the hardware substitution risk from hyperscalers, you ignore that AVGO is effectively taxing the compute layer. If the hardware cycle cools, the software recurring revenue acts as a valuation floor. The real risk isn't just in-house silicon—it's the potential regulatory pushback against their bundled software dominance.
[Unavailable]
"VMware integration issues and buyer concentration threaten Broadcom's high-margin stability despite hardware moat."
Google's VMware 'moat' overlooks integration hiccups: post-acquisition service revenue growth slowed to low-teens vs. 20%+ pre-deal, per Q1 earnings. Ties to Anthropic—hyperscalers won't switch networking silicon overnight (3-5yr cycles), but software churn risk amplifies if AI capex plateaus. Unflagged: 70%+ gross margins vulnerable to pricing pressure from concentrated buyers (top 3 hyperscalers = bulk of AI semis).
Panel Verdict
No ConsensusBroadcom's impressive AI revenue guidance is met with caution due to heavy reliance on custom silicon and hyperscaler capital expenditure, with varying views on the role of software as a moat.
Diversification into software segment for stability
Deceleration in hyperscaler capital expenditure leading to multiple contraction