After the market closed on June 3, Broadcom (AVGO) delivered what can only be described as a "blockbuster" earnings report—revenue surged 48% year-over-year to $22.2 billion, and AI semiconductor revenue soared 143% to $10.8 billion, both beating market expectations. Yet the next day, the stock price plunged over 13%, with a two-day cumulative drop nearing 20%. At the same time, Bitcoin fell to around the $60,000 mark, while the Nasdaq and S&P 500 continued to hit new highs.
This isn’t an anomaly. Instead, it’s a repricing of "expectations" and a clear signal of institutional capital flows—funds are moving out of crypto assets and into the AI infrastructure sector. Understanding AVGO’s "post-earnings drop" may be the key to grasping the current underlying logic of the market.
Earnings Beat Across the Board: Why Did "No Upward Revision" Become the Last Straw for the Stock?
Let’s start with the basics of Broadcom’s Q2 performance. According to the company’s financial report and multiple media outlets, Broadcom’s fiscal Q2 2026 revenue hit a record $22.2 billion, up 48% year-over-year; non-GAAP diluted EPS was $2.44, beating the consensus estimate of $2.40. AI semiconductor revenue reached $10.8 billion, a staggering 143% increase year-over-year; total bookings topped $30 billion, and order visibility extended from "2027" three months ago to "2028." Free cash flow also hit a record high of $10.3 billion.
By any single metric, this is a set of results most companies would be proud of. But for Broadcom, the market’s logic was entirely different.
The real negative signal came from two "non-revisions":
First, Q3 AI semiconductor revenue guidance fell short of the market’s highest expectations. Broadcom projected Q3 AI semiconductor revenue of $16 billion, up more than 200% year-over-year, but this figure was below the analyst consensus of around $17.2 billion. Some analysts pointed out that Citi’s Atif Malik had expected as much as $17.5 billion, and some buy-side models were even more aggressive. In a market that had already priced in extreme growth, $16 billion—while very strong—failed to clear the psychological bar set by the highest expectations.
Second, the long-term AI revenue target was not raised. Broadcom reiterated its fiscal 2027 outlook for "over $100 billion" in AI chip revenue, unchanged from last quarter, rather than raising it as the market had hoped. Meanwhile, the full-year fiscal 2026 AI chip revenue forecast of about $56 billion also fell short of some institutional estimates of $57.6 billion. Against the backdrop of competitors like Marvell having recently raised guidance, Broadcom’s "conservative stance" created a clear psychological gap for the market.
At its core, this is a game of "expectation management." When the stock price had already risen over 15% in the five trading days before the earnings release—adding about $300 billion in market cap—an earnings report that merely "beats expectations" simply wasn’t enough.
The AI Semiconductor "Domino Effect": From AVGO’s Plunge to $1.3 Trillion in Chip Sector Losses in Two Days
Broadcom’s sell-off wasn’t an isolated event. As one of the most important bellwethers in the AI compute chip space, AVGO’s plunge quickly rippled across the entire industry chain.
On June 4, Broadcom shares tumbled about 13%, the biggest single-day drop in over 16 months, wiping out about $286 billion in market value in one day. The next day (June 5), Broadcom fell another 7.9%, bringing the two-day cumulative decline to nearly 20%.
The contagion spread rapidly:
- NVIDIA (NVDA): Down about 6%, with more than $300 billion in market value lost in a single day;
- Micron Technology (MU): Plunged 13%, losing about $150 billion in market value in one day;
- Marvell Technology (MRVL): Fell about 17%;
- AMD and Intel (INTC): Both dropped around 11%;
- Optical communications sector: Ciena fell nearly 16%, Lumentum dropped nearly 6%, and Coherent also saw a major pullback.
At the sector level, the Philadelphia Semiconductor Index (SOX) plunged 10.26% on June 5, the biggest one-day drop since the outbreak of COVID-19 in March 2020. According to data from Zhitong Finance, the chip sector lost about $1.3 trillion in market value over two trading days.
A noteworthy distinction: Not all tech stocks were battered in this storm. Google rose 3.82% against the trend, TSMC ADR climbed 1.88%, and NVIDIA also rebounded to close up about 1.8%. This divergence reveals two signals:
The long-term demand for AI infrastructure remains intact—TSMC’s status as the sole foundry "node" gives it defensive qualities; the market’s "punishment" is targeted, not indiscriminate or sector-wide: core AI infrastructure beneficiaries whose valuations haven’t been excessively stretched haven’t actually been abandoned by capital.
Meanwhile, capital didn’t exit the US stock market as a whole, but rotated from semiconductors into financials, consumer stocks, and other Dow components—on June 5, the Dow Jones Industrial Average actually surged 874 points, hitting a new all-time high.
The Real Logic Behind Institutions’ "Selective Reduction" of Crypto Holdings
If the chip stock plunge can be attributed to a mismatch between valuation and expectations, then Bitcoin’s decline during the same period points to a deeper structural shift.
From June 2 to June 5, 2026, Bitcoin’s price slid from around $72,000 to $59,895, briefly falling below the $60,000 threshold and hitting its lowest level since October 2024.
The "decoupling" from traditional risk assets is intensifying. During the same period, the Nasdaq 100 just hit a record high, with the index up about 41.5% over the past 12 months, while Bitcoin is down about 37% year-over-year and still about 48% below its all-time high last year. Such a stark divergence in performance is rare in history. K33 Research analysts stated in a report: "Bitcoin’s weakness reflects persistently cooling institutional demand, with many market participants seeing the opportunity cost of holding BTC as too high, while anything AI-related continues to soar."
The scale of capital outflows from crypto assets in this round can be confirmed by ETF data:
- Since mid-May, US spot Bitcoin ETFs have seen a cumulative net outflow of about $4 billion, with 17 of the past 19 trading days recording net outflows;
- As of early June, Bitcoin ETFs saw a cumulative outflow of 62,794 BTC over the previous three weeks—the second-largest outflow on record;
- Since mid-to-late May alone, BlackRock’s IBIT saw a single-day outflow peak of $528 million;
- CME Bitcoin futures open interest has dropped to its lowest level since October 2023—a classic signal of waning institutional participation.
At the same time, large amounts of capital are flowing into AI infrastructure. MicroStrategy founder Michael Saylor stated on X: "About $400 billion has been invested in AI projects over the past six months. This isn’t a devaluation of Bitcoin, but a capital rotation." Wintermute’s latest market analysis in early June echoed this, noting that the correlation between crypto and US equities is experiencing its sharpest break this year, with significant capital outflows from crypto and AI semiconductor stocks becoming the new destination for institutional funds.
Curve founder Michael Egorov’s recent analysis on Gate Plaza also supports this view, arguing that the core factor in the current market isn’t a deterioration in crypto fundamentals, but a temporary shift in capital preferences—AI stocks have become the main market theme in 2026, a view echoed by multiple independent market studies.
An easily overlooked supporting fact: In the same week that Bitcoin continued to fall, US nonfarm payrolls for May added 172,000 jobs, far exceeding expectations and reinforcing the prospect that the Fed might maintain high rates or even hike again. Interest rate futures show the market’s expectation for a December Fed rate hike rose from 48% to 63%. For crypto assets, a tighter rate environment means higher holding and opportunity costs, providing further macro rationale for institutions to reduce crypto exposure.
"Homegrown, Globally Admired": Wall Street’s Contrarian Vote
In sharp contrast to the market’s emotional swings, major Wall Street institutions almost unanimously raised their price targets and reiterated buy ratings for Broadcom after the plunge.
- Goldman Sachs: Raised its price target from $500 to $525, maintained a "Buy" rating, kept Broadcom on its "Conviction List," and projected nearly 30% upside from the pre-market price of about $408.
- BNP Paribas Exane: Raised its target from $600 to $640, maintained an "Outperform" rating, and called the post-earnings drop "shortsighted."
- Morgan Stanley: Raised its target from $470 to $485.
- TipRanks average target: AVGO currently has an average target price of $512.88 based on 24 "Buy" and 3 "Hold" ratings, implying about 28.45% upside.
Wall Street’s consensus: The root cause of this sell-off was that the market had priced too much "perfection" into AVGO’s valuation—by the earnings release date, Broadcom’s P/E ratio had topped 90x, leaving little room for error. When the company issued its usual conservative guidance, the market chose to take profits, not reject Broadcom’s long-term AI thesis. Jefferies analyst Blayne Curtis attributed the drop to the market "pricing in huge AI growth signals and expecting management to further raise guidance"—when Broadcom’s response was to "reiterate" rather than "raise above expectations," the valuation was repriced.
Forecasting the Market Through Capital Flows: Two Frameworks and Key Indicators
The core variable in today’s market is no longer whether the "AI boom" will continue, but how capital is being reallocated between crypto, AI infrastructure, and macro sectors. The following three logical frameworks can help market participants build their own judgment systems:
First, how long will the "rotation pressure" on crypto assets last?
The current outflows are a structural signal, not just a short-term sentiment blip. AI chip capital expenditures are in a historic expansion cycle. According to Goldman Sachs, Broadcom now has six custom chip clients (including Google, Meta, Anthropic, and OpenAI), with several multi-gigawatt data center deployments scheduled over the next two years. Against this scale of industrial expansion, it’s unlikely that crypto assets will regain the "high-growth narrative" lead in the short term. However, it’s worth noting that large institutions are also building positions via OTC trades, meaning long-term holders haven’t completely exited.
Second, how will AVGO’s "valuation reset" play out?
The current roughly 14% decline in AVGO reflects a return from nearly 100x P/E to a more reasonable range, not a repudiation of the AI business logic. Broadcom management has extended AI order visibility to 2028—a hard metric that’s hard to ignore in any market cycle. Thus, the key variable going forward isn’t whether Broadcom is "still growing," but how much of a growth premium the market is willing to pay. If Q3 guidance is meaningfully raised when results are delivered, the current gap could mark the start of a new repricing.
Third, how will the correlation between Bitcoin and the AI sector evolve?
The two are currently in a phase of "decoupling," a sharp contrast to Bitcoin’s historical role as a "tech risk sentiment gauge." But if institutional returns on AI infrastructure investments fall short of expectations (for example, if the profit realization cycle after large-scale capex is extended), capital may reassess crypto as an alternative allocation tool. At that point, the current divergence in price trends could become a "leading indicator" for the next narrative shift.
Conclusion
Broadcom’s Q2 earnings case highlights a pattern that’s being repeatedly validated in 2026’s tech asset pricing: When valuations inflate to a "perfection" stage, the market’s desire for "upward revisions" outweighs the absolute quality of the earnings themselves. Beating expectations is no longer enough to drive stock prices higher—the market demands "blowout" results.
AVGO’s decline isn’t a turning point for the AI investment thesis. Broadcom’s AI orders are booked through 2028, and the six major clients’ data center expansion plans are proceeding as scheduled. What’s really being repriced is the valuation level that was pushed above 90x P/E by extreme optimism—this is a reasonable correction, not a wholesale rejection of the AI infrastructure investment logic.
Meanwhile, the $4 billion outflow from crypto ETFs and Bitcoin’s drop below $60,000 shouldn’t be viewed in isolation. These phenomena point to a common narrative: capital is flowing out of the crypto market and into the AI infrastructure value chain, and this flow is systemic at the institutional level.
For investors, distinguishing between a "secular downturn" and a "phase of capital rotation" is crucial at this stage. When the tide of sentiment recedes, identifying what’s simply a valuation reset and what signals a deeper shift in industry logic may be the "key" to navigating the current market divide.
