Elon Musk, ever attuned to the political zeitgeist, has updated Tesla Inc.’s mission to “amazing abundance.” It is the kind of hyperbole beloved of investors in the company he runs. Before the amazing variety arrives, however, another type of abundance was announced on Wednesday evening’s earnings call: Tesla’s investment budget will more than double.
For the bulls, Tesla is finally unleashing its financial firepower to own the future of autonomous vehicles, robots and artificial intelligence. Yet the news came alongside weak fourth-quarter results and the bombshell announcement that Tesla is investing about $2 billion in xAI, Musk’s own artificial intelligence venture. Amid all the plans and targets thrown out on the call, the one certainty is that Tesla will burn a lot of cash this year.
The financial results themselves were messy and underwhelming. The closely watched metric of auto gross profit margin, adjusted for regulatory credits, came in at 17.9% — surprisingly high given a collapse in vehicle deliveries, even factoring in foreign exchange gains. Tesla’s energy business performed well, although gross profit was essentially flat with the third quarter. In any case, none of this seeming strength trickled down. Tesla’s overall operating margin fell to just 5.7%. “Other” costs, possibly reflecting swings in crypto values, soared. Fourth-quarter GAAP earnings slumped by 60%, year over year.
None of which matters, of course. Tesla’s stock is determined less by reported numbers, more by a complex, if nebulous, function that multiplies planned initiatives with the level of faith in Musk. Both factors are high. Musk announced that Tesla will retire two of its premium priced, and oldest, models, the S and X, next quarter — an acknowledgement of sliding sales, perhaps, but cast as symbolizing the company’s shift toward fully autonomous vehicles such as Cybercabs. Production of those is planned to begin by the end of June. Tesla also plans to unveil its third-generation version of the Optimus humanoid robot soon, with mass production “planned” to begin by year-end. Big things are also planned in solar power, batteries, chargers and chips.
This narrative setting is all par for the course with Tesla, and Musk’s targets should be treated with skepticism. When asked for specifics on how many Optimus robots are working today at Tesla factories, and what they are doing, Musk demurred. He said the technology was still in the research and development stage, which sits rather oddly with the idea that mass production will begin in less than 12 months (and that the S and X are being scrapped to switch over their production lines to making robots). Musk’s expectation of having fully autonomous vehicles operating in a quarter-to-half of the US by year-end should be put in the context of his expectation of reaching half the US by year-end — last year.
What can’t be doubted is the money Tesla is spending to do all this. The mooted capex figure of $20 billion-plus for 2026 is not only more than the prior two years combined, it is substantially higher than Tesla’s best ever annual cash flow from operations, $14.9 billion in 2024. That declined slightly in 2025 and the consensus is for it to fall again this year. Unsurprising, since Tesla is essentially de-emphasizing its main source of profits, EV manufacturing, and plowing billions into nascent businesses that won’t earn a profit for a while, even assuming success.
Remarkably, that $20 billion is equivalent to half of the entire value of the property, plant and equipment assets carried on Tesla’s balance sheet. It signals a radical expansion in a very short space of time.
Based on consensus forecasts, the capex budget implies Tesla burning around $6 billion of cash this year. With $44 billion on its balance sheet, Tesla can afford it. But this would make 2026 Tesla’s first year of negative free cash flow since 2018, before a ramp-up of sales of then-new models and the pricing tailwind delivered by the pandemic’s disruption flipped free cash flow positive.
This echo of the 2010s, for much of which Tesla was a kind of publicly listed start-up, dovetails with the xAI investment. Recall that Tesla shareholders were given a non-binding vote on this at November’s meeting. While there were more ‘yeas’ than ‘nays’, a high number of abstentions meant that it technically counted as a no. Tesla, however, appears to have focused more on the yeas and Musk said offhandedly on Wednesday’s earnings call that “we’re just doing what shareholders asked us to do, pretty much.”
