New Ask Claude or ChatGPT about any stock, using Bullpen. Bullpen in Claude or ChatGPT. 7 days free

Saturday, November 21, 2026

Back to the front page

From X

SouthernValue on X

The $NVDA guide for 75% growth in FY28 depends on $SPCX buying 10GW of GPUs next year, their largest order by far as other large customers shift budget to ASICs. Folks long $NVDA should watch $SPCX CFO at TomorrowX. They don’t have capital to buy 10GW of GPUs but think they can do it with customer prepayment. Except, $NVDA doesn’t have customers when building a DC, they operate in the spot market. They’re basically hoping $ANT / OpenAI / $GOOG / $META end up with way more demand than they planned for and have to buy in the spot market at elevated prices. How does customer prepayment work in the spot market? Makes no sense, and if SPCX wants to use this form of financing, they will need to accept significantly lower prices and longer contract terms, like 1/2 the price they’ve indicated to investors. Btw, $GOOG calls their existing SPCX compute “bridge capacity” until they bring more TPUs online next year. $ANT is telling investors they intend to churn $SPCX in 2H27. I don’t expect $META to have the financial resources to add to their FY27 plan and they may have over budgeted for Muse. There are rumors OAI may try to get out of some of their existing compute commits as their ARR trends below plan to fund it. All this before potentially some compute currently used by Chinese labs comes available, perhaps ~20% of installed GPU compute. Who is SPCX going to sell 10GW to next year? As well as I can track, the capital plans of $AMZN $MSFT $META $GOOG $ANT are reasonable well funded by earnings and secured funding sources, and demand is reasonably balanced between internal and external uses and a variety of customers. Frankly I do not see a glut here. A glut, should one arise, would most likely come from $SPCX who intends to add ~30% incremental capacity growth with no committed customers, while existing customers are telling us pretty clearly they aren’t planning to stick around. Because most XPU/ASIC capacity is well funded, balanced by internal and external use, secured by the most well capitalized participants, sold in take-or-pay long term contracts, and because customers earn better economics on XPUs vs GPUs, I expect this part of the supply chain to have significantly less risk of downward estimate revisions than the GPU ecosystem should a glut emerge. GPUs are the high cost of compute, represent most of the spot market, and have lower customer quality on avg (Neos) and therefore I expect GPUs to absorb the majority of downward revisions should a glut emerge. GPU players who play for a continued tight spot market do have more financial upside as Xai has shown in 2026, but there is a reason that exists and a real downside risk tradeoff. It is odd to me that nobody on the sellside is writing about this while many on the buyside recognize it.

Sledgehammer Time

Nvidia's earnings and uncertainty over post-Jensen AI investment are reverberating across markets as the stock comes under pressure.

Grand Expectations

Grand Theft Auto VI launched to a 99 Metacritic score after estimated $1.5 billion to $2 billion development and marketing spending,…

INTU no longer

Intuit’s $3.68 billion Mailchimp sale removes a noncore unit as TurboTax shifts to assisted filing and Credit Karma faces volume pressure.

Google

Google has narrowed OpenAI’s chatbot lead through Gemini distribution across Search, Android and Chrome, backing its AI stack with custom…

Key Man Risk

Meta and Nvidia show how founder-led strategy makes key-man risk central to long-term AI and Reality Labs spending.

PSN · long

Parsons CORP

Parsons generates roughly half of its $6.4 billion revenue from Federal Solutions, serving defense and intelligence customers.

Meta Earnings, Muse and the Frontier, Musebook

Meta’s earnings hinge on whether investors back Zuckerberg’s Muse vision and the compute spending required to build it.

Anthropic Deep Dive

Anthropic’s $80.7 billion first-half loss was mostly fair-value marks, while $16.3 billion revenue and compute economics showed a $974…

My Last GFL Update - Hopefully

GFL’s take-private structure hinges on BC Partners rolling 60% of its stake to preserve debt portability under the 40% permitted-holder…

Publications we follow

Every publication on this site, hand-picked.

130 publications

Newsletters 110

Seeking Alpha 98 of 303

X accounts we follow

Every account on this site, hand-picked.

94 accounts

Investor letters we follow

383 firms

Sectors

Ordered by how often our editor reaches for them.