AI Hedge Fund by Former OpenAI Researcher Loses 92% of Value
Leopold Aschenbrenner's AI hedge fund lost 92% of its $20.24 billion June value, now holding $1.58 billion, after a heavy leverage bet on AI stocks collapsed.
Microsoft and Alphabet plan hefty capex increases, supporting Nvidia's AI chip demand. But supply constraints and rising costs may dampen near-term results.
Microsoft plans to invest over $50 billion in capital expenditure during the December quarter of 2026, compared with $45.2 billion in the prior quarter. Alphabet also intends to raise spending significantly in 2027. These projections point to ongoing demand for the infrastructure that powers artificial intelligence services, benefiting Nvidia. However, the chipmaker's revenue in the upcoming quarter will also hinge on how much equipment it can actually ship.
Nvidia is due to report its fiscal third-quarter earnings on 17 November, covering the period through 25 October 2026. The September results from cloud companies offer insight into that same timeframe. Their future spending intentions give traders a longer view of whether investment will persist beyond the quarter Nvidia is about to report.
The plans reveal several insights.
Microsoft's December-quarter projection offers a near-term view of its investment plans. Additional spending is anticipated in the first half of 2027, which includes some computing capacity delayed from earlier deliveries. These outlays encompass central processing units (CPUs) and graphics processing units (GPUs), as well as increased memory and storage costs.
Alphabet kept its capital expenditure guidance at $195–205 billion for 2026 and indicated a significant rise in 2027. Amazon also confirmed its cash capital expenditure estimate of approximately $220 billion for 2026. These company-wide budgets employ different accounting methods and extend beyond Nvidia gear.
The timing of these figures limits their relevance to Nvidia's upcoming earnings announcement. Microsoft's December-quarter plan is more indicative of the near-term outlook than Alphabet's capital spending trajectory for all of 2027. The longer-range plans instead demonstrate how far cloud providers anticipate continued expansion.
A look at the breakdowns helps clarify how much spending actually flows to Nvidia.
The September quarter's detailed figures shed light on where the money is directed. Microsoft devoted roughly three-quarters of its capital expenditure to short-lived assets, mainly CPUs and GPUs. The $45.2 billion total included $7.4 billion in finance leases, meaning the figure does not represent cash spent on new processors during that period.
Separately, companies can lease computing capacity from providers such as CoreWeave, which runs Nvidia-based infrastructure. In this model, the business pays for usage while the provider purchases the hardware. Thus, cloud service spending can indirectly support Nvidia sales.
Alphabet posted $51.4 billion in spending, largely on technical infrastructure. Around 65% of that infrastructure investment went toward servers and networking gear, with the remaining 35% allocated to data centres. The equipment segment is more directly tied to Nvidia, as it supplies computing systems, networking technology, and processors.
Cloud companies also have other options. Alphabet uses its own tensor processing units (TPUs) alongside Nvidia chips, while Amazon has noted strong demand for its Trainium processors. Consequently, their budgets do not reveal how spending is divided among suppliers. Nvidia's broader customer base includes specialized cloud providers, corporate clients, and government-backed artificial intelligence initiatives, so these three tech giants represent only a portion of overall demand.
Near-term shipments are constrained by supply.
During its August earnings call, Nvidia indicated that demand outstrips available supply. In October, it disclosed additional capacity expansion plans with manufacturing partners, encompassing advanced packaging and other production steps.
Analyst Parker Tian highlighted two bottlenecks: high-bandwidth memory (HBM), which rapidly delivers data to AI processors, and advanced packaging, which links processors and memory within a chip package.
“I read Big Tech’s capex as a signal of how long this cycle can last, not a forecast of Nvidia’s next quarter, which is still set by HBM memory and advanced packaging capacity,” says Parker Tian, Head of EBC AI Lab at EBC Financial Group.
Nvidia has also noted strains across its broader supply chain. Since expanding capacity is time-consuming, larger cloud budgets do not automatically translate into an immediate rise in equipment available for sale.
Rising costs also influence the outlook.
Supply constraints impact both profitability and delivery volumes. Nvidia reported that memory prices had increased more than anticipated. Therefore, selling more systems can boost revenue even as elevated component expenses lower the margin retained as gross profit.
Nvidia had already factored in some pressure. It guided third-quarter gross margin at 74%, plus or minus 0.5 percentage point, versus 75% in the previous quarter. Management also forecast a subsequent decline to 71–72% in the fourth quarter. Thus, any margin decrease should be assessed against those projections before being seen as an additional negative.
Key items to watch on 17 November.
On 26 August, Nvidia issued revenue guidance of $108 billion, plus or minus 2%, with no assumed data-centre compute revenue from China. As of 13 November, Investing.com reported an analyst consensus of $108.8 billion. Revenue at management's midpoint would satisfy its own guidance while falling marginally below that published estimate.
The guidance for the subsequent quarter will indicate the revenue Nvidia expects to generate, while its margin projection can reveal how component costs are affecting profitability. Should revenue guidance fall short, management's explanation will be crucial to determine whether production delays or softening orders are the cause.
“The warning sign could be management describing orders, rather than supply, as the constraint, which would mean the AI build-out is shifting from a supply problem to a demand question,” says Tian.
Share to
Disclaimer: this article comes from third-party media and is provided for reference only. It does not constitute investment advice. Crypto and other financial products carry significant price volatility risk, so please make your own decisions carefully.
Leopold Aschenbrenner's AI hedge fund lost 92% of its $20.24 billion June value, now holding $1.58 billion, after a heavy leverage bet on AI stocks collapsed.
Positive Volume Delta means more trades were initiated by buyers, but prices can still fall if sellers absorb demand.
S&P 500 futures face resistance around 7,425 ahead of Nvidia earnings. The bearish outlook persists but buyers have a recovery path.
Ripple-backed Evernorth Holdings reported about $505 million in paper losses on its XRP holdings due to accounting rules that prevent fair-value treatment.