Tesla's earnings report today revised its full-year capital expenditure guidance down from 'over $25 billion' to about $20 billion. Q3 actual capex came in at $4.7 billion, far below analysts' expectations of around $7.5 billion, and free cash flow unexpectedly turned positive to $1.53 billion. This sparked a lot of discussion, with some saying it could be a precursor to big tech cutting capital expenditures. But since SpaceX and Meta are still increasing their capex, that argument doesn’t fully hold water. The official explanation is 'more efficient capital deployment,' but from a personal perspective, it feels more like preparation for a merger between the two companies. After all, many of Tesla's capital expenditures overlap with SpaceX. For instance, Tesla is building its own data centers: these clusters are collectively called Cortex and are specifically used for training autonomous software for FSD (Full Self-Driving), Robotaxi, and the Optimus humanoid robot. Cortex1, with about 120MW, went online in 2024. Cortex2, currently under construction, has already reached 250MW as of the latest update, with a year-end target of 400MW and a maximum plan of 500MW. The additional 250MW alone would cost at least $5 billion—it might make more sense to just use SpaceX’s data centers, especially since data centers are already SpaceX’s largest business segment. Then there’s the Tesla-SpaceX joint project, Terafab. At this point, Tesla doesn’t really need to keep spending on it. Lower expenditures also strengthen Tesla’s cash position, while SpaceX itself is capital-intensive (especially with AI). Elon himself has repeatedly mentioned the 'increasing overlap' between the two companies (particularly in areas like Terafab, AI chips, robotics, and orbital computing). Just my personal speculation, for reference only. #Tesla #SpaceX #AI #FSD #Robotaxi #Optimus #Tech #Finance