Bianca Botes, Managing Director at Citadel Global
The United States (US) technology reporting season has so far followed a consistent pattern; companies have largely met or beaten their numbers while the market has sold off a good portion of them anyway. The reason is that investors are no longer paying for the quarter just ended but are instead pricing what each company intends to spend and earn, over the next two years.
The NASDAQ 100 gained 2.2% this week on the strength of space exploration company, SpaceX and information technology company, Super Micro Computer, yet as at the time of writing its futures were 1.7% lower, with chip stocks leading the retreat after electric vehicle company, Tesla, tech giant, Meta and semiconductor capital equipment company, Lam Research, reported late on Wednesday, all with fundamentally strong results.
SpaceX still increasing its value, but Tesla battles
The most consequential disclosure of the week came on a results call. SpaceX, in only its second quarter as a listed company, reported revenue of $14.58 billion, up 161% year-on-year and net income of $4.4 billion against a loss a year earlier, while lifting fourth quarter capital expenditure (CAPEX) guidance (the forecast of its future financial performance) to between $35 billion and $40 billion to accelerate its compute build-out – the process of purchasing and scaling the hardware, facilities and power infrastructure needed to run large-scale data processing or artificial intelligence (AI) workloads. Asked about Tesla, CEO, Elon Musk, confirmed that the boards of both companies have formed special committees to examine a merger between SpaceX and Tesla, which at current valuations would be the largest corporate union on record and on the news, SpaceX closed 7.8% higher.
What exists today is a formal process rather than a transaction, since Musk himself cautioned against treating the outcome as certain, approval would require a majority of all outstanding Tesla shares and the Tesla proxy filed on Monday contains no merger proposal. United Kingdom (UK) banking company, Barclays, frames the rationale as being around vertically integrating the AI ambitions of both groups, while Tesla’s independent shareholders will weigh a governance question in which Musk holds roughly 45% of SpaceX’s equity but around 82% of its voting power, raising questions over how much influence Musk will have over the merged company.
Tesla’s own results beat on earnings, with adjusted earnings per share of $0.48/share against $0.45/share expected on revenue of $28.05 billion that was flat year-on-year and automotive gross margin excluding regulatory credits recovered to 18.1%. Free cash flow of $1.53 billion against expectations of an outflow of about $4.5 billion looked like a clear positive surprise, but part of it came from lower spending and management also cut its 2026 CAPEX plan to about $20 billion, which sits awkwardly with the long-term investment in robotics and autonomy that underpins the share price. The shares, having risen 3.4% during the session on the merger headlines, fell roughly 3% after hours, as operating income declined 56% year-on-year to $718 million and Tesla pushed out its timelines for unsupervised self-driving, Robotaxi revenue and Optimus volumes.
Meta struggles to convince investors
Meta’s revenue rose 26% to $64.6 billion, ahead of the roughly $63.4 billion expected, as average price per digital advert climbed 14% and the company narrowed full-year CAPEX to between $135 billion and $145 billion, raising the floor from $130 billion, while free cash flow turned negative at $2.12 billion on third quarter CAPEX of $34.9 billion. Operating income fell 37% as a $9.89 billion legal charge, stemming from a legal agreement to resolve United States (US) nationwide litigation concerning teen mental health and social media addiction and higher infrastructure costs absorbed the revenue gain. The shares fell after an after-hours gain of more than 6% as attention turned to how the AI build-out will be funded and the debt market sent a similar signal on Thursday, with Meta’s five-year credit default swaps rising as much as 8.9 basis points to around 119 basis points, their highest level since the contracts began trading, but pointing to caution rather than stress. In this instance the direction the default swaps are moving in matters more than the number, because a strong set of results was not enough to push the cost of protection against a default lower.
Tech peers also lose
The semiconductor results showed where the market’s sensitivity lies. Lam Research reported record revenue of $8.22 billion, up 54% year-on-year, yet the shares fell about 7% after hours because its fourth quarter guidance was capped by its own supply constraints and falling revenue in China, whereas semiconductor company, Texas Instruments, guided above consensus and traded higher.
A Bloomberg report this week, on how Wall Street banks are encountering soaring AI costs, helps explain why each new CAPEX figure is met with suspicion. These banks’ chief financial officers who funded early adoption with few constraints are now restricting staff access to the most powerful models, testing smaller ones and considering alternative providers including Chinese models, because agentic workloads consume unpredictable amounts of computing power and benchmarking firm, Evident, found that only 12 of the 50 largest global banks have publicly attributed returns to AI. Hyperscaler spending on the scale Meta and SpaceX are guiding to, earns its return only if corporate customers keep paying rising compute bills, so a shift towards cost discipline among those customers would pressure pricing at the moment new capacity arrives.
Microsoft and Alphabet both report on 28 October, the same day as the US Federal Reserve’s (Fed’s) rate decision, with Amazon following on 29 October and Apple on 2 November. Each will be judged on the two measures applied to Tesla and Meta, namely the CAPEX guide and the evidence that customers are paying for what is being built, and the experience of this week suggests that a higher return on the quarter will buy very little unless revenue is visibly keeping pace with the spending.
ENDS
