Broadcom’s 143% AI Growth Still Triggered a 12% Selloff

Broadcom just reported 143% AI semiconductor revenue growth, guided for more than 200% growth in the next quarter, and the market’s response was to sell the stock down 12% in after-hours trading. On the evening of 3 June 2026, Broadcom released Q2 FY26 results that exceeded analyst expectations on revenue and delivered forward guidance that surpassed even the most optimistic pre-release projections. The after-hours share price reaction was a sharp decline. The same night, CrowdStrike fell 9% on a modest earnings beat, and Palo Alto Networks dropped 5.6% despite above-forecast guidance. The pattern was not noise. What follows is an analysis of what Broadcom’s results reveal about how AI-driven earnings are being priced, why exceptional numbers can still disappoint a positioned market, and what investors following high-growth AI hardware names need to understand about the gap between reported results and market expectations.

Broadcom’s numbers were exceptional: here is exactly what they showed

Broadcom’s Q2 FY26 revenue reached US$22.19 billion, a 48% year-over-year increase that edged past analyst estimates of approximately US$22.0 billion. The beat was modest on the headline. Where the quarter distinguished itself was in the AI semiconductor line: US$10.8 billion, up 143% year-over-year, confirming the growth trajectory that has made Broadcom one of the most closely watched names in the AI hardware build-out.

The forward guidance moved the goalposts further. Broadcom projected Q3 FY26 total revenue of US$29.4 billion, implying 84% year-over-year growth, and AI semiconductor revenue of US$16.0 billion, implying growth exceeding 200% compared to the prior year period.

Q3 FY26 AI semiconductor revenue guidance: US$16.0 billion. This figure surpassed even the most optimistic pre-release projections from market participants, according to Investing Live and SEC filings by Broadcom (AVGO).

The strength of these results matters for a specific reason: the selloff that followed was not a response to disappointment in the numbers themselves.

Broadcom’s position within the AI chip supply chain is structurally distinct from Nvidia’s: where Nvidia dominates discrete GPU accelerators, Broadcom occupies the custom ASIC and networking layer, supplying hyperscalers including Google and Meta with bespoke silicon that does not compete directly with merchant GPU architectures.

What “sell the news” actually means and why it keeps happening in AI stocks

The term “sell the news” gets used loosely. It deserves a more precise definition, because what happened on the evening of 3 June 2026 was not a single-stock anomaly. Three companies reported positive results or issued above-consensus guidance, and all three saw sharp after-hours declines:

• Broadcom (AVGO): Reported 143% AI revenue growth and Q3 guidance above estimates; shares fell approximately 12% after hours
• CrowdStrike (CRWD): Delivered a modest earnings beat, raised FY27 net new ARR guidance by 5.2% at the midpoint, and announced a 4-for-1 stock split; shares fell approximately 9% (Source: The Street)
• Palo Alto Networks (PANW): Issued above-forecast guidance citing AI-driven cybersecurity demand; shares fell 5.6% (Source: CNBC)

“Sell the news” is not a verdict on business quality. It is a positioning unwind. Traders and funds accumulate shares ahead of expected catalysts, and once those catalysts are confirmed, the buying pressure that drove the run-up reverses. The result confirms the thesis; the marginal buyer has already acted.

Why AI hardware names are especially prone to this pattern

AI hardware and adjacent names have attracted outsized pre-earnings positioning because growth expectations are high, momentum is strong, and the sector narrative remains dominant. The share price entering an earnings report already reflects a significant probability of a strong beat.

The broader growth stock repricing cycle that preceded Broadcom’s June 2026 report included technology equities declining sharply in Q1 2026, with Broadcom named among the concentrated losers alongside Nvidia, Meta Platforms, and Microsoft, as growth stocks reached a 21% discount to fair value, a level Morningstar data shows has occurred less than 5% of the time since 2011.

This creates what market participants sometimes call the “whisper number,” the implicit expectation that sits above the published analyst consensus. A result that beats the formal estimate can still fall short of the threshold required to sustain the pre-earnings price level. Once the event passes, the marginal buyer who would sustain the stock has already bought. Profit-taking follows without a new incremental buyer to absorb it.

Valuation math and the expectations gap that explains the paradox

The disconnect between Broadcom’s results and its share price response becomes less paradoxical when viewed through valuation mechanics. The logic follows three steps:

1. A high-growth stock enters earnings priced at elevated forward multiples. The share price already embeds expectations for a strong result, meaning the multiples can only compress if guidance materially exceeds what the market has modelled.
2. Guidance beats the published consensus but lands within the range implicitly priced by active market participants. The formal analyst estimate is one benchmark; the price itself is another, and it often implies a more aggressive assumption.
3. No new information exists to push the multiple higher. The pre-earnings price level, which reflected optimism about AI revenue acceleration, becomes difficult to sustain once the guidance merely confirms rather than exceeds what was already embedded.

Applying the framework to Broadcom’s specific guidance

Broadcom’s Q3 AI semiconductor revenue guidance of US$16.0 billion implies growth exceeding 200% year-over-year, a figure that is extraordinary by any measure for a company of this scale. The sequential step-up from Q2’s US$10.8 billion is significant.

Yet in a market that had run hard on AI infrastructure conviction, this figure may have landed within the range that sophisticated participants had already incorporated into their positioning. No post-release analyst price-target revisions or named institutional commentary is available from current sourcing to confirm or deny the specific multiple dynamics at play. The analysis here is structural: when guidance confirms an embedded expectation rather than exceeding it, the stock’s pre-earnings valuation carries no new catalyst to sustain it.

The capex-to-revenue lag in AI semiconductors, estimated at 18-24 months by Morningstar analyst Dennis Li, creates a structural tension: hyperscalers are committing capital at record rates, but the revenue those investments generate for chipmakers arrives on a delay that makes it difficult to validate current forward multiples from reported results alone.

The broader picture: what one night of AI earnings reveals about sector positioning

Zooming out from Broadcom, the full picture of 3 June 2026 strengthens the structural argument. Three companies across two distinct subsectors of the AI trade reported positive results, and all three declined in after-hours trading.

The convergence matters. A single stock selling off after a beat can be dismissed as company-specific. Three stocks selling off on the same evening, across AI semiconductors and AI-adjacent cybersecurity, points to a sector-wide positioning unwind.

The Broadcom Q2 FY26 results confirmed total revenue of US$22.19 billion and AI semiconductor revenue of US$10.8 billion, with Q3 guidance set at US$16.0 billion, figures released directly by Broadcom on 3 June 2026 and forming the basis for the positioning analysis that followed.

Alphabet’s contrasting signal: On the same evening, Alphabet upsized its equity offering to US$84.75 billion (from US$80 billion), including a US$10 billion private placement from Berkshire Hathaway, with all proceeds designated for AI compute investment (Source: Reuters). Institutional conviction in AI infrastructure spending remains high, even as individual earnings positioning plays unwind.

Tesla’s 40% year-over-year rise in China-manufactured EV sales in May 2026 (Source: CNBC) provides a further distinction: not all tech-adjacent names moved in the same direction on the same reporting cycle. The selloffs were specific to AI-positioned names, not a broader technology rotation.

What investors should actually do when a strong AI earnings report triggers a sharp selloff

The diagnosis is clear. The next question is practical: how should an investor respond? The first step is distinguishing between two fundamentally different types of post-earnings selloff.

A positioning unwind occurs when the business thesis is intact and the selloff is mechanical, driven by profit-taking from participants who bought ahead of the catalyst. A genuine guidance miss or deterioration in business quality, by contrast, warrants a re-evaluation of the investment case. These require different responses.

Three diagnostic questions help separate one from the other:

1. Did the company beat on both revenue and guidance? Broadcom reported US$22.19 billion in revenue against estimates of approximately US$22.0 billion and guided Q3 AI revenue to US$16.0 billion, surpassing the most optimistic pre-release projections. The answer here is yes.
2. Did the guidance miss only the “whisper number” rather than the published consensus? The published consensus was beaten. The implicit market expectation, embedded in a stock price that had run on AI momentum, may have required even more. The answer here is consistent with a positioning unwind.
3. Is the selloff concentrated in one stock or spread across peers in the same sector? CrowdStrike and Palo Alto Networks declined on the same evening on positive catalysts. The selloff was sector-wide.

Investor response options follow from the diagnosis:

• Hold and monitor if the diagnostic questions confirm a positioning unwind and no fundamental deterioration is visible
• Staged accumulation on continued weakness if the long-term thesis remains intact and the price decline creates a more attractive entry; note that leveraged positioning unwinds can persist for several days
• Reduce exposure if a genuine fundamental concern is identified, such as decelerating sequential growth, margin compression, or a guidance miss on the published consensus

No post-release analyst consensus data is currently available for Broadcom’s Q2 FY26 report. Investors should seek updated price-target commentary from their own research sources before acting.

For investors wanting to build a more structured process before the next AI earnings cycle, our dedicated guide to post-earnings decision frameworks walks through pre-release decision rules, the optimal timing for acting on analyst revision drift, and the short-term capital gains tax thresholds that make the 12-month holding period a concrete variable in any trim calculation.

The 12% drop does not change the AI infrastructure story, but it should change how investors read earnings night

Broadcom’s underlying business delivered exceptional results. AI semiconductor revenue of US$10.8 billion, up 143% year-over-year, and Q3 guidance implying 200%+ growth to US$16.0 billion are objectively strong by any historical benchmark for a company of this scale.

The after-hours selloff reveals something important about market structure in the current AI cycle. The speed and scale of capital rotation into AI names has created a gap between reported results and the price expectations already embedded in share prices. Navigating that gap, distinguishing between a positioning unwind and a fundamental re-rating, is now a core competency for investors in this sector.

The structural signal from 3 June 2026: Broadcom delivered 143% AI revenue growth and still fell 12%. CrowdStrike and Palo Alto Networks confirmed the pattern. Yet on the same evening, Alphabet raised US$84.75 billion for AI compute, with US$10 billion from Berkshire Hathaway. Institutional conviction in the AI infrastructure build-out has not wavered. What has changed is the price at which that conviction gets expressed on earnings night.

The quality of the business and the mechanics of how that quality is priced are two separate questions. Broadcom’s results answered the first. The selloff answered the second.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results, and forward-looking statements, including revenue guidance figures, are subject to change based on market conditions and various risk factors.