By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
Despite impressive AI revenue growth, Broadcom's stock faces headwinds due to cyclical semiconductor capex, potential AI demand slowdown, and competitive pressure. The market's repricing of Broadcom's valuation suggests caution is warranted.
Risk: Slowdown in AI capex or increased competitive pressure eroding Broadcom's moat in AI semiconductors.
Opportunity: Successful integration of VMware, driving software margins and providing a valuation floor.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Broadcom (NASDAQ: AVGO) was one of the better-performing artificial intelligence (AI) stocks of 2026, up by almost 40% year to date -- until June 3, when it reported its fiscal Q2 earnings. Following that report, Broadcom's stock heavily sold off and is down nearly 20% from its all-time high. While it's still in positive territory for the year, it has suffered a major reversal.
Was what Broadcom said really that bad? Or is there another force at play here?
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
Broadcom is really a tale of two companies. Broadcom's legacy core business consisted of various software products, along with some computing hardware. However, the part of the operation most investors focus on today is its AI business.
Broadcom's AI exposure comes mostly from two product lines: connectivity switches and custom AI chips. These products are in major demand, and that demand is only continuing to grow. In its fiscal Q2, which ended May 3, AI semiconductor revenue rose 143% year over year to $10.8 billion. In Q3, management expects AI semiconductor revenue to rise 200% year over year to $16 billion. That's significant momentum, and management said expects to see this growth continue into 2027 when it expects this business unit alone to deliver $100 billion or more in annual revenue.
However, that was the same guidance it gave last quarter, and the market wanted more because the stock was priced for more.
Ahead of the Q2 report, Broadcom was trading at about 40 times forward expected earnings. Considering that its overall growth rate was 48% for the quarter, that's a pretty pricey valuation. After the sell-off, Broadcom trades for about 20 times 2027 earnings estimates, which isn't terribly expensive, but it's also far from cheap.
There are still a lot of expectations baked into Broadcom's stock, and if it doesn't continuously set the bar higher, it's going to sell off further following subsequent earnings reports. But I still think that the current price point offers a buying opportunity.
The reality is that there is plenty of growth ahead for Broadcom beyond 2027, as the AI data center build-out is widely projected to last through 2030. There's still plenty of time to make a ton of money selling custom AI chips between now and then. As a result, I think Broadcom is an OK buy right now, but investors will need to continue monitoring it to see if expectations pick up over the next few quarters.
Four leading AI models discuss this article
"Broadcom’s optimistic 2027 AI revenue path of ~$100B hinges on an uninterrupted, multiyear AI capex cycle, which is uncertain and vulnerable to cyclical slowdowns, making near-term risk-reward skew bearish."
Broadcom’s AI push appears big but is highly contingent on a multi-year, uninterrupted data-center capex cycle. Q2 AI revenue was $10.8B (+143% YoY) with Q3 guided to $16B; management even pointed to a $100B AI-revenue target by 2027. That path requires sustained pricing power, continued AI demand, and execution across chips and switches. The stock has swung from ~40x forward earnings to ~20x 2027 estimates, leaving little margin for error if growth slows or margins compress. In a cyclical semis backdrop, any deceleration or competitive pressure could trigger meaningful multiple compression, not just a growth slowdown.
Proponents would argue the AI data-center build-out is secular and Broadcom has pricing power and software-driven profitability. If that view proves true, the stock may not re-rate meaningfully to the downside.
"Broadcom is currently transitioning from a growth-at-any-price darling to a cyclical industrial play, making its valuation highly sensitive to hyperscaler capital expenditure volatility."
Broadcom’s 20% pullback isn't a failure of execution, but a violent repricing of a momentum trade that had outpaced its own fundamental reality. While management’s guidance for $100 billion in AI revenue by 2027 is staggering, the transition from a 40x forward P/E to a 20x multiple suggests the market is finally discounting the cyclicality of semiconductor capex. The core risk isn't the AI demand itself, but the 'lumpiness' of custom ASIC (application-specific integrated circuit) orders. If hyperscalers like Google or Meta shift their internal development timelines or consolidate their infrastructure spend, Broadcom’s revenue concentration becomes a liability rather than a moat.
Broadcom’s reliance on a handful of massive hyperscaler clients creates a 'customer concentration' risk where a single delay in a data center rollout could trigger a multi-quarter earnings miss that the current 20x multiple has yet to price in.
"Broadcom's stock repriced not because AI demand collapsed, but because management's unchanged $100B+ 2027 guidance signals growth deceleration, and at 20x forward earnings, the margin of safety is thin if hyperscalers or competitors erode market share."
The article frames a 20% pullback as a buying opportunity, but misses the core problem: Broadcom guided the same $100B+ AI revenue for 2027 two quarters running. That's not caution—it's stalling. At 40x forward P/E pre-earnings, the market priced in acceleration. At 20x post-correction, you're not getting a bargain; you're getting repriced to reality. The real risk isn't near-term AI demand (which is real) but whether Broadcom can actually grow into these numbers or if it becomes a mature 15–18x compounder. The article also glosses over competitive pressure: NVIDIA's custom chips, AMD's MI300, and hyperscalers building in-house—all eroding Broadcom's moat in AI semiconductors.
If Broadcom's AI business truly hits $100B+ by 2027 and sustains 30%+ growth thereafter, even 20x 2027 earnings is cheap and the stock re-rates higher. The article's point—that the data center buildout extends to 2030—is legitimate and undervalued by a market spooked by one flat guide.
"The post-earnings valuation reset may stall or reverse if AI capex growth decelerates before the 2027 target is reached."
Broadcom's AI semiconductor revenue hit $10.8B in fiscal Q2 with 143% YoY growth and Q3 guidance of $16B (+200% YoY), yet the stock fell 20% from highs because management repeated its prior $100B+ annual run-rate target for 2027 without raising it. At 20x 2027 earnings the multiple looks reasonable against 48% overall growth, but the sell-off highlights how much of the prior 40x valuation priced in continuous upward revisions. The article underplays execution risk in custom ASICs and potential hyperscaler capex pauses after 2025-26 builds.
Even if 2027 revenue lands at $100B, any delay in the broader AI data-center cycle through 2030 could compress multiples back toward 15x and erase the apparent value in the current dip.
"A $100B AI run-rate by 2027 does not guarantee a high multiple; a slower growth path, margin compression, or price competition could push the multiple down to 12-15x, capping upside."
Gemini correctly flags lumpiness, but the bigger flaw is assuming $100B AI revenue by 2027 sustains a high multiple. If AI capex slows or hyperscalers push price competition, Broadcom may not sustain AI-margin strength; markets could re-rate toward 12–15x even with that revenue target. The article underplays a possible multi-year pause in data-center spending, which would cap upside and pressure multiples.
"VMware's software-driven margin expansion provides a valuation floor that mitigates the cyclical risks of the custom ASIC business."
Claude, you’re missing the VMware integration tailwind. Broadcom isn't just a hardware play; they are aggressively stripping costs from VMware to drive software margins that buffer the 'lumpiness' Gemini and Grok fear. While you focus on the $100B AI target stalling, you ignore that VMware’s recurring revenue provides a valuation floor that standard semiconductor peers lack. This isn't just about AI chips; it's about a software-defined margin expansion that justifies a higher multiple than a pure-play cyclical.
"VMware's recurring revenue provides downside protection, not multiple expansion—a critical distinction the panel hasn't clarified."
Gemini's VMware margin-buffer argument assumes software integration proceeds flawlessly while custom ASIC lumpiness persists. But VMware's $8.6B revenue base grows ~5% annually—immaterial to a $100B AI target. Software margins don't insulate against hyperscaler capex cycles; they just smooth earnings volatility. If AI revenue decelerates 2026–27, VMware's recurring revenue becomes a floor at ~12x, not a re-rating catalyst. The valuation floor argument conflates stability with upside.
"VMware margins cannot insulate against hyperscaler ASIC delays that would still trigger 12-15x compression."
Gemini, VMware's ~5% growth and cost cuts add a modest earnings buffer but won't offset ASIC revenue lumpiness if hyperscalers delay 2026-27 builds. That same concentration risk Claude flags could force a re-rating toward 12-15x even with recurring software revenue, since the $100B AI target remains hardware-driven. The integration tailwind is real yet too small to change the multiple compression scenario.
Despite impressive AI revenue growth, Broadcom's stock faces headwinds due to cyclical semiconductor capex, potential AI demand slowdown, and competitive pressure. The market's repricing of Broadcom's valuation suggests caution is warranted.
Successful integration of VMware, driving software margins and providing a valuation floor.
Slowdown in AI capex or increased competitive pressure eroding Broadcom's moat in AI semiconductors.