SpaceX and AMD Both Plunge in After-Hours Trading Post Earnings, Potentially Signaling the Next Phase of U.S. Stock Market Trends
However, post-market sentiment was quickly dragged back into tension by two heavyweight earnings reports — SpaceX and AMD both released their latest quarterly results almost simultaneously after the close. Each set of numbers had its own highlights, yet the market reaction was strikingly uniform: both stocks fell.
SpaceX dropped over 7% after hours, while AMD fell more than 9%. Revenue beat expectations in both cases, and guidance wasn’t bad either — so why did investors still choose to vote with their feet?
The answer comes down to the same reason — the market’s “passing grade” for the 人工智能 narrative has been raised to a near-unforgiving standard.
1. SpaceX: Rapid Growth, But Even Faster Cash Burn
SpaceX delivered its first quarterly earnings report since going public.
Revenue came in at $7.8 billion, beating market expectations. As a “debut,” the number itself is respectable. But the problem lies on the capital expenditure side.
The report showed that SpaceX’s Q2 capital expenditures reached $18.369 billion, with approximately $16 billion allocated to AI computing infrastructure. That figure is significantly higher than the $13.2 billion analysts had previously expected.
$16 billion is nearly equivalent to SpaceX’s combined revenue over two quarters. In other words, for every dollar the company earns, it burns more than two dollars on AI infrastructure.
The market’s concern over this “burn rate” was reflected directly in the stock price. Despite Musk maintaining his characteristically aggressive long-term vision during the earnings call — moving the company’s $1 trillion annual revenue target up from 2031 to 2030 — these long-dated narratives clearly failed to soothe investor worries about near-term cash returns.
Musk also emphasized that the market is severely underestimating Starlink’s potential. He stated that within the next 10 years, Starlink could provide primary internet services across most countries where it’s permitted to operate. SpaceX’s long-term growth trajectory is anchored on three pillars: Starlink’s expansion, Starship’s commercialization, and the return on AI infrastructure investment.
But here’s the cold reality: if capital deployment doesn’t quickly translate into quantifiable revenue growth, the trillion-dollar revenue target is nothing more than a number on paper.
What’s more unsettling for the market is that shortly after the earnings release, the company will face the unlocking of 910 million shares — a figure that even exceeds the current free float. Such a massive release of shares will almost certainly impose heavy supply pressure on the stock price.
In short, SpaceX’s position is awkward: full-throttle long-term narrative, excessive near-term cash burn, and poor share structure. It possesses nearly all the elements of a “great company,” but at this moment in time, it still has some distance to go before becoming an “ideal investment target.”
2. AMD: Numbers Beat Across the Board, But the 市场 Wanted “Big Beats”
AMD’s earnings report, on paper, was nearly flawless.
Data center revenue hit $6.72 billion, up 107% year-over-year and approximately 15.6% quarter-over-quarter, serving as the core engine driving overall performance. Total revenue reached $11.54 billion, up 50% year-over-year to an all-time high, surpassing the market’s expectation of $11.28 billion. Adjusted EPS came in at $1.66, beating the analyst consensus of $1.62.
Guidance was equally positive: Q3 revenue is expected to be approximately $13 billion, above the market’s $12.52 billion estimate, with adjusted gross margin of around 56%.
On the AI front, AMD is accelerating its challenge to Nvidia’s dominance. CEO Lisa Su announced that the second-generation Helios AI servers, featuring MI455X accelerators and TSMC’s “Venice” processors, are now in full production and will begin shipping over the coming months. Reports indicate that AMD will start delivering related AI servers to Meta and OpenAI by the end of this quarter.
Nearly every metric pointed to a rosy picture, yet the stock plunged over 9% after hours.
Why? Because AMD’s stock price has already doubled this year, pushing the market’s valuation premium to an extreme level — the DCF fair value is only around $130, while the actual stock price sits at $400–$500. Under this valuation structure, “beating expectations” has become the bare minimum. What investors want is “beating by a wider margin” — higher AI growth rates, stronger guidance, and more aggressive long-term targets.
When the market’s expectations for a company are pushed to the ceiling, even “good” earnings can be punished for not being “good enough.”
3. The New Standard for AI Earnings: Spend Smart, Don’t Let Core Revenue Slip, and Guide Extremely High
The post-market moves of SpaceX and AMD together reveal a paradigm shift taking place in the market.
AI remains the hottest trading narrative in U.S. equities, but the bar for “participating in AI” has been raised significantly. In the past, merely mentioning AI in an earnings call or showing capex growth was enough to send the stock higher. Now, the market demands a more complete framework:
- No mindless spending — capital expenditures must have a clear path to returns, not a burn rate that’s twice the revenue like SpaceX
- Core revenue must not be impacted — AI investment can’t come at the expense of core business profitability
- Forward expectations must be sky-high — guidance, growth rates, and market share targets can’t afford to be low on any front
Only by meeting all three criteria simultaneously can a company earn the market’s reward. Miss any one of them, and it faces the risk of “good news exhausted, turning into bad news.”
4. Final Thoughts: U.S. Stocks Are Becoming “Meme-ified” — 工具s Matter More Than Predictions
When the bar for earnings evaluation is pushed to such an extreme, the market’s price action characteristics are also undergoing subtle changes.
Any positive news is prone to sharp spikes — because capital is extremely crowded in the AI narrative, with high-density directional positioning. The moment there’s a genuine upside surprise, buying floods in rapidly.
But any slight miss on expectations can also trigger violent pullbacks — equally because of the crowding, any disappointment gets amplified into panic selling. The post-market moves of SpaceX and AMD are textbook cases of this “high expectations + low error tolerance” environment.
This price action pattern increasingly resembles the volatility logic of meme coins in 加密 — not driven by fundamentals, but by sentiment and expectations. In such an environment, risk management has become more important than ever.
Currently, BIT Securities has launched an options buying feature that allows investors to insure their existing positions, strictly capping downside risk, and helping investors navigate frequent large-scale volatility with confidence.
[Risk Disclosure] The analysis and views contained in this article are solely for market information consolidation and personal observation, and do not constitute any investment advice, offer, or solicitation, nor do they constitute a recommendation for the purchase or sale of any related securities. The financial data and forecasts cited herein are as of the time of publication; for any discrepancies, please refer to the company’s official financial reports and publicly disclosed documents. Options trading involves significant risk and is not suitable for all investors; purchased options may expire worthless, resulting in the loss of the entire premium. Please refer to the options trading terms for specific risks. Investors should make their own decisions based on their financial condition and risk tolerance, and consult independent professional advisors. This article is contributed by an external guest author; the views expressed herein are solely those of the author and do not represent the official position of BIT. BIT makes no guarantee as to the accuracy or completeness of the content.
Related: The cheaper AI gets, the more expensive chips become
On June 30, Anthropic released Claude Sonnet 5. This is a mid-tier model, the “most capable worker” in the Sonnet series. On the agentic capability benchmark SWE-bench Pro, it scored 63.2% — just 6 points shy of the flagship Opus 4.8’s 69.2%. On another dimension, the graduate-level reasoning test GPQA-AAA v2, Sonnet 5 actually outperformed Opus 4.8. Pricing is even more critical. During the promotional period, it charges $2 per million input tokens and $10 for output. Opus 4.8’s corresponding prices are $5 and $25 — Sonnet 5, at 40% to 60% of the price, delivers over 90% of the flagship’s capability. This news can be interpreted in two ways. The first: AI is getting cheaper again. Cost reductions benefit everyone, the Chatbot war continues, and model vendors are locked…
简体中文 
English 
繁體中文