US macro numbers yesterday:
- the trade deficit is narrowing - exports are holding firm - labor demand remains healthy with limited layoffs - factory and capital spending are picking up. Together, these signals suggest an economy that continues to support growth, investment, and employment. Source: Daniel Lacalle, Bloomberg
$AMD crashed 12% after hours despite beating on every headline number.
- Revenue: $11.54B vs $11.28B expected - Adjusted EPS: $1.66 vs $1.62 expected - Data center revenue: $6.72B, up 107% year over year - Q3 guidance: $12.7B to $13.3B vs $12.5B expected First, capex came in at $808 million against a $298 million estimate, nearly three times what analysts modelled. The market is no longer rewarding AI spending. It wants to see the revenue. Second, valuation. The market is asking whether today's price already reflects years of exceptional growth. Here's an interesting case made by Oguz Erkan @oguzerkan on X ➡️ Assume AMD can sustain 40% annual revenue growth through 2030. That would imply roughly $200 billion in revenue. With a 40% net margin, net income would reach around $80 billion. Applying a 20x earnings multiple results in a company worth approximately $1.6 trillion. Discounted back to today at a 10% annual rate, that equates to a present value of roughly $990 billion. With AMD already valued at around $845 billion, investors see relatively little room for execution mistakes or weaker-than-expected growth. Third, another concern is that AMD's rapid expansion is closely tied to the current AI infrastructure spending boom. As customers increasingly diversify beyond NVIDIA, AMD has emerged as one of the biggest beneficiaries. However, many investors expect AI capex growth to moderate after 2027, raising questions about how long this tailwind can last. In short, AMD's outlook remains compelling, but its valuation leaves little margin for disappointment. Source: Bull Theory, Oguz Erkan @oguzerkan on X
Big oil is profiting big
Chevron, $CVX, had its best quarter ever. Exxon, $XOM, its best since 2022. Shell, $SHEL, second-highest on record. Source: Hedgeye @Hedgeye Bloomberg
The Stock Market is approaching a major top and a possible 1987-like crash, warns Michael Burry, the man who has predicted 60 of the last 2 market crashes
Source: Barchart
The Nasdaq 100 has surged +9.3% in just 4 sessions since last Thursday.
The rally already ranks among Tech’s sharpest 4d rebounds around major market shocks of the past decades; from the GFC and Covid to the 2022 hiking cycle and Liberation Day. Source: HolgerZ, Goldman, Bloomberg
SpaceX's first print as a public company (Tue 4 Aug, after the close) was a clean operational beat that the market rejected on cash flow. That gap is the whole story.
The numbers Revenue $7.81bn, +92% YoY from $4.1bn, against ~$6.9bn consensus; EPS: loss of $0.09/share vs -$0.26 expected; net loss narrowed to $541m from $1.0bn; Adjusted EBITDA $3.5bn vs $2.0bn consensus. All three segments beat 🚀 And then: capex $18.37bn, more than six times year-ago levels, of which $15.83bn went to AI — against a $13.22bn FactSet estimates 🔥. Segment split: -> Connectivity $4.29bn (+66% YoY), 1.7m net adds with ARPU held flat at $66, enterprise/government revenue +108%, Starlink now at 12 million subscribers. -> AI $2.6bn (+247% YoY), first quarter of positive segment adjusted EBITDA at $1.1bn, helped by $1.6bn of incremental Colossus cloud services revenue. -> Space only $962m with a $205m EBITDA loss — Starship remains an R&D line, not a business. 🤔 Why the stock reversed Shares closed +9.43% at $125.33, then fell to ~$114.6 after hours — the entire day's rally erased. ➡️ Reuters framed it precisely: the concern is cash flow, or the absence of it, with analysts fearing the burn rate forces a return to markets for equity and/or debt; the bonds have already been weak, and 911.5 million insider and employee shares come free on Thursday. ⚠️ Management guided Q3 and Q4 capex to broadly match Q2 — so roughly $55bn of capex in a single year against a $100bn cash pile, most of which is IPO proceeds. Set against that, Johnsen's claim of a sub-one-year payback on compute deployments, $6.7bn of additional cloud contracts signed in the first weeks of Q3, a $100bn ARR (Annual Recurring Revenue) target for December, and the $1trn revenue projection pulled forward from 2031 to 2030. 😎 Classic Musk: the promise lengthens as the cash burn steepens. With 34% short interest on the float, the after-hours move is also positioning, not just fundamentals. ‼️ Market drivers to carry forward - Nvidia exclusivity is the cleanest read-across. Musk committed SpaceX to Vera Rubin processors exclusively, and NVDA rose ~2% after hours. - A target of 15–20GW of power and cooling online by end-2027, deliberately built ahead of GPU supply — that is a bid for turbines, transformers and grid capacity as much as for silicon. - Memory is the bottleneck, and the tape agrees. The Kospi added 4.0% overnight led by SK Hynix and Samsung, shrugging off SpaceX and AMD. 👉 My read: SpaceX is now the purest listed proxy for the AI capex question — a business with genuine cash-generative assets (Starlink at 40%+ segment margins) funding an option on compute at a burn rate no cash flow currently supports. Thursday's lockup is the near-term technical risk; the $100bn December ARR figure is the near-term fundamental test. Anyone who owns this at 1.6trn is underwriting Starship reusability and Starlink V3 economics, not this quarter's numbers. App Economy Insights
An AI U-turn ?
JP Morgan's Mark Schilsky argues that investors are becoming increasingly convinced that hyperscalers' massive AI investments are generating returns well above their cost of capital. After reading the latest quarterly earnings transcripts from Amazon, Microsoft, Alphabet, and Meta, that view is becoming difficult to dismiss. Management teams sounded more confident than ever that today's unprecedented AI spending will translate into sustainable, long-term earnings growth. If Schilsky is correct, the valuation multiples of the Magnificent Seven may have already reached their lows. As confidence builds that AI capex is creating durable value rather than simply driving higher costs, investors could once again be willing to pay premium valuations, potentially well before free cash flow bottoms and begins to accelerate. Source: TME
Why So Many Companies Are Failing by Avoiding Risk Instead of Pursuing Opportunity
Source: Marketoonist
