SpaceX IPO in 2026? Tesla Stock's 'Next Moment' and the AI Investment Bubble Debate
The ARK monthly market update recorded on December 11, 2025, connected the dots from current macro factors (employment, inflation, interest rates) to AI capital expenditure (CapEx) and monetization, the US-China AI race, power (nuclear energy), and the 'next break' for SpaceX and Tesla. The conclusion is simple.'While concerns remain, 2026 could be the phase where productivity pushes down inflation and winners justify their CapEx with earnings.'—that is the outlook.
1. From 'Rolling Recession' to 'Stability'? But the pain point is employment
Cathie Wood stated that the economy has 'moved from a rolling recession to more stability,' while citing the unemployment rate for new graduates (college graduates) exceeding 10% as the statistic she is most concerned about. It rose from 9.3% to 10.2% in the September data, and the fact that it takes an average of 25 weeks to find a job—a prolonged period—is also heavy.
On the other hand, while acknowledging that AI is putting pressure on entry-level jobs, she encourages people by saying, 'Learn AI. Learn vibe code. The barrier to starting a business is lower than it was in 2014.' True to ARK's style, she did not end the discussion on worsening employment as a 'reason for pessimism,' but connected it to the new opportunities created by technology.
2. The logic that 2026 could be a 'Goldilocks' scenario
Inflation is the Fed's biggest challenge, but rate cuts will continue—the basis for this is that the unemployment rate is a lagging indicator, suggesting that 'this time, it may not take a year to recover.' Furthermore, what she emphasizes is the view thatProductivity = the strongest anti-inflation factor.
Specifically, she argues that falling oil prices and lower rents (due to population migration) will affect the CPI with a time lag, potentially leading to a 'Goldilocks' scenario in 2026 (strong growth, low inflation). In addition, the market has weathered storms like 'tariff chaos,' 'government shutdowns,' and 'hawkish Fed remarks,' so expectations have begun to be priced in—that was the summary.
3. The biggest risk for 2026: 'Division' in monetary policy and 'wandering' in fiscal policy
In response to the question, 'What is the biggest fiscal and monetary risk for 2026?', she expressed concern that on the monetary side, factions within the Fed that believe 'growth = inflation' are increasing, leading to divided opinions. On the other hand, she also mentioned the nuance that if a certain person is in charge, the problem might be minor.
The fiscal side is more politically charged. While she evaluates 'popular tax policies' such as tax cuts and deductions, the scenario is that if the economy is weak, it will lead to short-term handouts (temporary spending) due to anxiety ahead of the midterm elections, and the bond market will view the deficit as a problem. In short, it is dangerous when the premise of 'diluting the deficit through growth' collapses.
4. Is AI CapEx a 'Dot-com bubble redux'?—The difference lies in funding sources and valuations
Regarding market anxiety that AI investment is becoming similar to the late 90s as a percentage of GDP, ARK breaks down the 'similarities' and the 'decisively different points'.
Hyperscaler CapEx could total over $500 billion next year (analyst forecast)
However, looking at the funding sources for CapEx, in the late 90s, CapEx as a percentage of S&P 500 operating cash flow peaked at 75%, whereas now it is 46%
Multiples are also different; the NASDAQ 100 had a P/E ratio of 89x in 2000, compared to 28x today (not the same level of speculative fever as back then)
And regarding investor sentiment, ARK suggests that 'the market is too afraid of a bubble, making it actually difficult to reach extreme valuations.' Cathie cites a comparison by Altimeter's Brad Gerstner, noting that back then, money was poured into 'dark fiber optics,' but now GPUs are in short supply, and every GPU produced is 'being used.'
5. The impact of 'AI where earnings catch up': OpenAI, Anthropic, and the coding sector
Most striking are the examples of revenue growth for AI-native companies (all with a nuance of estimation).
OpenAI: Annualized run rate of approximately $20 billion, a 10-fold increase in two years from approximately $2 billion in 2023
Anthropic: Approximately $9 billion run rate within the year, a rapid expansion from approximately $1 billion at the beginning of the year
Coding assistance: Cursor is estimated to have grown 10x to over $1 billion in annual revenue, while Replit is estimated at over $150 million.
The important point here is that even with massive CapEx, they are "proving with numbers that demand is real and there is a willingness to pay". To borrow Kathy's words, this is not an era like Amazon during the bubble period where "stocks rise by boasting about expanding deficits." Now, investments require "proof of monetization."
6. US-China AI Competition: Models are closing in, but computing resources are the bottleneck
China is increasing its presence on the low-cost, open-weight side, with the lag in frontier performance explained as being about "six months." On the other hand, they lack the computing resources to release continuously, and in terms of manufacturing capacity, it was estimated that TSMC's advanced fabs for China have "about 38 times the total transistor count."
The conclusion is realistic: the key is (1) whether they can access NVIDIA GPUs through deregulation, and (2) whether they can rapidly expand domestic supply capacity. If they cannot, the view is that Western models will regain market share even in the open domain.
7. Electricity: Transmission and distribution (utilities) are in a regulatory cage, while power generation is returning to a "learning curve"
In the context of AI power, transmission and distribution have yield caps due to regulations, making innovation difficult, while the generation side (especially nuclear) is seen as having favorable tailwinds.
The story is that regulatory support, 24/7 power demand, and SMRs (Small Modular Reactors) could "re-ignite" the cost-reduction trend for nuclear power that once broke down in the 70s. Furthermore, if there is a "five-year wait for grid connection," hyperscalers may move toward "behind the meter" (self-contained power sources), potentially bypassing traditional utilities.
8. Tesla's next moment: Austin "operator removal" is the turning point
Tesla positions the statement that in-car operators will no longer be needed for robotaxis (possibly within a few weeks in Austin) as a catalyst. Waymo is advancing urban deployment but with only thousands of vehicles, whereas Tesla has overwhelming production capacity; the structure is that Waymo paves the way on the regulatory side, and Tesla wins on scale.
ARK's argument is consistent, even mentioning a target where robotaxis will account for the "major part" of Tesla's corporate value in 2029.
9. Healthcare (ARKG): Will deregulation and the resumption of M&A break the "long stalemate"?
While attributing ARKG's long-term struggles to the fact that the market became "cash-flow focused" due to post-COVID uncertainty, they cite large-scale deregulation and the revival of M&A as the turning point. In particular, the point that "without strategic price discovery, it is easy to be hit by short sellers" is practical.
Furthermore, the view was presented that large pharmaceutical companies will face a revenue gap of $200 billion to $300 billion over the next five years and will be forced to fill it through acquisitions and partnerships.
10. Abbott x Exact Sciences: $21 billion indicates a "re-evaluation of diagnostics"
Abbott's acquisition of Exact Sciences for approximately $21 billion is treated as a case that confirms the "size of the TAM for cancer diagnostics." The focus is on pipeline investment that extends not only to Cologuard (stool tests) but also to blood tests (MRD, multi-cancer). The fact that peers rose after the acquisition symbolizes the revival of "strategic price discovery."
11. ARK Venture Fund: Liquidity design and avoiding the "premium trap"
ARKVX is explained as an interval-type crossover fund (target: 80% private equity / 20% public) that allows redemption of up to 5% of NAV quarterly, with listed stocks serving as the source of liquidity.
And what Kathy emphasized is the "premium several times the NAV" problem that occurs in other closed-end funds. She warned against the structure of "losing at the entrance," where another fund holding a lot of SpaceX trades at about 6 times its NAV, meaning the price could fall even if the NAV grows.
12. SpaceX IPO rumors and "space data centers": The idea of "escaping to the sky" to solve power constraints
The final climax is SpaceX. In addition to rumors of a 2026 IPO, we treat the space-based data center concept discussed by Elon Musk as a 'new option value.' In contrast to the terrestrial constraints that force years of waiting for grid connections, completing the reusable Starship to 'launch power and infrastructure into space whenever needed'—this was the conclusion that the West, currently clogged by regulations and NIMBYism, could potentially 'leapfrog' instead.
