Summary
- Broadcom earned a record $22.2 billion in Q2 FY2026 revenue - up 48% year-over-year - and the stock still sold off. The culprit isn't Iran; it's AI guidance that wasn't infinite enough.
- CrowdStrike beat revenue estimates with 26% growth and dropped 9% in the same session. Earnings beats don't cure valuation anxiety.
- The S&P 500 fell roughly 1% on June 3, with chipmakers leading the decline. Geopolitical headlines provide cover, but the mechanics are entirely earnings-driven.
- MicroStrategy continues its eight-month collapse, now down roughly 70% from its 52-week high. That is a different problem: a bitcoin proxy that lost its premium.
- When the AI growth bar becomes mathematically impossible to clear every quarter, the false narrative is that anything is wrong with the companies. The real issue is expectations priced for perfection.
I've been very surprised that the market's response to Broadcom's Q2 results has been framed as geopolitical risk at all. The Iran ceasefire was agreed in early April. The hostilities that briefly spiked oil prices and rattled supply-chain nerves are behind us. What actually happened on June 3 was far more banal, and far more revealing: Broadcom reported record quarterly revenue and the stock punished it anyway because the Q3 guidance wasn't quite infinite.
Let's be precise. Broadcom's fiscal Q2 2026 - which ended March 31 - delivered $22.2 billion in revenue, up 48% year-over-year. AI semiconductor revenue alone exceeded $10.7 billion. For context, that AI run rate in a single quarter now surpasses the total annual revenue of most S&P 100 companies. The adjusted EPS came in at $2.44. By any measure that matters - cash generation, growth, competitive moat - this is a dominant quarter.
However, Broadcom guided Q3 AI semiconductor revenue at approximately $16 billion. The street's implicit expectation, extrapolated from the 48% pace, was closer to $28.7 billion in total revenue. The gap between what Broadcom delivered and what the market's arithmetic demanded was the gap between reality and the AI expectations trap. The stock fell because investors had already priced in a trajectory that no capital expenditure cycle in history has sustained.
This is not a Broadcom problem. This is a market-wide false narrative: the idea that AI-driven semiconductor revenue growth can compound at 50% or more, quarter after quarter, in perpetuity. It can't. That doesn't mean Broadcom is a sell. It means the bar has moved to a level where even excellent results are disappointing.
That being the case, let me look at the fundamentals that matter beneath the noise. Broadcom's free cash flow for fiscal 2025 was $26.9 billion, up nearly 39% year-over-year. In Q1 of the current fiscal year alone, free cash flow hit $8 billion. The company generates the kind of cash that funds both dividend growth and aggressive share repurchases - in Q1, Broadcom authorized $10 billion in buybacks. The price-to-free-cash-flow multiple sits around 77x, which means the stock's valuation has absorbed years of flawless execution. The FCF yield of roughly 1.4% tells you everything about what the market is demanding: no risk, no stumble, no deceleration. When that demand meets a guidance number that represents real-world constraints on fab capacity, customer adoption curves, and the natural limits of sequential AI buildout, the reaction is mechanical. Not rational - mechanical.
Now consider CrowdStrike. The company reported Q1 FY2027 revenue of $1.39 billion, up 26% year-over-year, beating analyst estimates. Subscription revenue grew 26% to $1.32 billion. Net new annual recurring revenue hit a record $255.8 million. The company turned positive GAAP net income for the first time in the quarter - $27.8 million. And the stock dropped 9% after hours, closing at $767.48.
Compare that to Broadcom's reaction, and the pattern is the same. CrowdStrike's stock is up over 60% year-to-date. At $767, the market has priced in flawless execution through every competitive threat - agentic AI automating cybersecurity, Microsoft bundling its own defenses, the endless fear that cloud-native security is a race to the bottom. The earnings beat wasn't bad. It just wasn't "beat the consensus that already assumes perfection" bad enough to clear the bar.
This is where the false narrative tightens its grip. Headlines will say the Nasdaq fell on chipmaker losses and geopolitical worries. But the Iran ceasefire is months old. The actual mechanism is simple: when a stock has run 60% in a year, every earnings report is a referendum on whether the next 60% is already priced in. CrowdStrike's 26% growth, while excellent for a company of its size, signals the same reality as Broadcom's guidance - the hyper-growth phase is settling into very-good growth. The market hates the transition even when the destination is still strong.
That being the case, what does this mean for an investor who actually holds these stocks, rather than watching them from headlines?
For Broadcom, I believe the sell-off represents an irrational overreaction. The company's dual revenue engine - custom AI chips for hyperscalers and VMware-based enterprise software - gives it structural advantages that most of the market underestimates. The AI custom-silicon business is not cyclical; it's a multi-year infrastructure buildout where Broadcom's relationship with the hyperscalers gives it sticky, high-margin recurring revenue. VMware adds software-like margins on top. The Q3 guidance disappointment is real, but it's a speed bump on a trajectory that still compounds free cash flow at an exceptional rate. At current levels, the valuation is still stretched, but the cash generation profile supports a Hold, and I'd consider adding on further weakness.
For CrowdStrike, the calculus is different. The company is profitable now, growing revenue at 26%, and the sell-off has created breathing room after a 60% year-to-date advance. The agentic AI threat that spooked investors earlier this year hasn't materialized; if anything, AI-driven attack surface expansion makes cybersecurity more mission-critical, not less. However, CrowdStrike doesn't pay a dividend, doesn't generate the same free cash flow density as Broadcom, and trades at a premium that demands continued flawless execution. I rate it a Hold at current levels - not because I think the business is weakening, but because the valuation still requires more beats from a stock that has already run hard.
Meanwhile, MicroStrategy continues to bleed. The stock has fallen for eight consecutive months and is down roughly 70% from its 52-week high. Its multiple to net asset value has compressed to roughly 1.09x, meaning the market is no longer paying any premium for the bitcoin proxy strategy. When bitcoin itself has been in a multi-month decline, MSTR's leverage to that outcome works in reverse: it's a leveraged short on confidence. The structural problem here isn't cyclical timing; it's a business model that has proven vulnerable when its one asset class underperforms. The market has already spoken on this one.
So here's the frame. The headlines will keep using geopolitical tension as a backdrop because it provides drama where the actual story - AI growth expectations colliding with reality - is less cinematic but far more important. The companies that are selling off today are not broken. They are being punished for being finite in a market that priced them as infinite. That's not a structural thesis. It's a sentiment dislocation.
In my opinion, Broadcom's free cash flow engine and dual revenue moats make it the stronger position among the names being punished today. The Q3 guidance miss is a speed bump, not a wall. CrowdStrike is solid but more exposed to continued multiple compression after its run. For investors who can tolerate the volatility that comes with high-valuation growth stocks, these sell-offs create entry points for companies that are structurally strong and being misread as cyclical.
I rate Broadcom a Buy on the dip, with the understanding that the valuation still demands patience. I rate CrowdStrike a Hold until the multiple stabilizes below its recent expansion. For the broader portfolio, when the AI trade corrects on guidance expectations rather than fundamentals, defensive positioning in dividend-generating energy and value stocks provides ballast while tech resets its growth math.
