据报道,英伟达将披露一笔创纪录的 5000 亿美元表外特殊目的实体(SPV)交易,旨在资助全球最大规模的“互吹互捧”活动。
Nvidia Set To Reveal Record $500 Billion Off-Balance Sheet SPV Deal To Fund The World's Biggest Circle Jerk: Report

原始链接: https://www.zerohedge.com/markets/nvidia-set-reveal-record-500-billion-balance-sheet-spv-deal-fund-worlds-biggest-circle-jerk

据报道,英伟达正与包括阿波罗(Apollo)、黑石(Blackstone)、贝莱德(BlackRock)和 KKR 在内的一批大型金融机构组建财团,筹集 5000 亿美元资金,用于人工智能基础设施建设。这一大规模行动旨在为维持人工智能热潮所需的各类数据中心、芯片和电力生产提供资金。 批评人士认为,此举是一种孤注一掷且带有“循环”性质的融资方案,令人想起互联网泡沫时期的供应商融资模式。通过协助合作伙伴举债来购买自家技术,英伟达实际上是在为自身的营收增长提供动力。这种对表外特殊目的载体(SPV)的依赖引发了人们对金融风险集中的担忧,导致英伟达股价在消息传出后下跌。 尽管私募股权巨头将预计 8 万亿美元的人工智能建设规模视为历史性的投资机遇,但怀疑论者警告称,该行业正变得杠杆过高。这一模式的可持续性仍存巨大疑问,特别是如果更廉价的开源人工智能模型使得当前昂贵的前沿模型变得过时。最终,这项交易凸显了流入人工智能领域的前所未有的资本规模;如果预期中“曲棍球棒式”的自由现金流增长未能实现,许多人担心这将对信贷市场产生潜在冲击。

相关文章

原文

In what is shaping up as a record-breaking, off-balance sheet SPV deal, one which will send CDS and credit yields surging far beyond where the recent explosion in AI debt has already pushed them, and as shown in the chart below, it's pretty, pretty far...

... the FT reports that Nvidia - unable to generate fund the old-fashioned way by earnings - is working with the world’s largest financial groups are to assemble a $500bn funding package for AI infrastructure development, in what will be the biggest AI lending efforts to date of any kind.

A consortium of groups including private credit (and equity) giants Apollo, Blackstone, BlackRock, Brookfield, Goldman and KKR is entering a partnership with Nvidia to invest in the AI build-out, the FT cited six people briefed on the talks.

The "partnership" as the FT calls it as it sounds a bit better than "hail-may, off-balance sheet arrangement" underscores Nvidia’s increasingly desperate efforts to raise capital for itself and its clients (because everyone now admits that circular financing is absolutely critical to keeping the AI bubble going) to continue assembling the chips, power production and data centers at the heart of the AI boom.

The $5.25tn company, whose GPUs underpin most of the leading US AI models available today, has positioned itself at the center of the AI boom, providing chips, infrastructure and software to a wide array of partners developing the technology.

More ominously, as it brings back fresh memories of the vendor financing that was in part responsible for the collapse of the dot com bubble, the chipmaker traditionally provides financial backing to help its AI partners raise debt in capital markets, which then helps boost Nvidia’s own revenue, in what has become AI's trademark circle jerk deal.

As we have discussed here for the past year, the circular nature of such transactions has raised concerns about concentrated risks in the sector, and it's why NVDA stock slumped immediately after the FT news broke, sliding as much as 2% on the FT report and wiping out more than $70bn in market capitalisation.

Separately, Nvidia was in talks to provide a massive guarantee for a 10-gigawatt data centre project in Ohio leased to OpenAI, according to a person familiar with the matter.

The deal shows how Nvidia is building relationships with the giants of the private capital industry, which are collectively preparing to invest trillions of dollars of their insurance, retail and institutional investor assets into AI infrastructure. 

In recent years, private capital groups such as Apollo and Blackstone have structured off-balance sheet SPV, pardon, AI infrastructure deals to assist companies like Anthropic finance their heavy spending on chips and data centers. We discussed this extensively two months ago  in "The $1.8 Trillion Off-Balance Sheet Time Bomb At The Heart Of The AI Supercycle", a topic which the FT also caught up with today. 

Nvidia, whose market cap has grown 15x since the launch of the first chatGPT in November 2022, is not the only beneficiary of the gargantuan circle-jerk deals that define the AI bubble: the biggest cloud-computing companies, including Meta, Oracle, Microsoft, Alphabet and Amazon, have also dramatically increased their spending on AI infrastructure as they look to win the race to dominate the emerging technology. Morgan Stanley projects so-called hyperscalers will spend $3.5tn between 2026 and 2028.

That need for capital has forced technology groups to tap every source of cash they can find, including public equity, investment-grade and high-yield bonds, securitised debt, private credit and project finance markets.

“[The] sheer size of the AI infrastructure build-out is unprecedented,” Jim Zelter, president of Apollo, said on an earnings call earlier this month. “More than $8tn of capital is expected to be invested, a staggering sum. We see an enormous opportunity for private capital to finance a portion of this along with public capital.”

While all that is great, it appears nobody has done the math on what happens if traditional, expensive frontier models lose to vastly cheaper, open-weight/source models. One thing is certain: the free cash flow hockeystick projected in the chart below will never, ever happen should open-weight models win the arms race. 

And now we sit back and wait to see just how far into the stratosphere this record SPV deal will send hyperscaler CDS and bond spreads. 

联系我们 contact @ memedata.com