FEATURED

The Tab Nobody Can Pay Right Now

Here is the thing about the OpenAI IPO delay. It is not really a story about one company’s timing. It is a story about the single largest unverified assumption holding up a multi-trillion dollar market.

OpenAI confidentially filed its S-1 with the SEC on June 8. Then, last Thursday, the New York Times reported the company is leaning toward pushing its debut to 2027. The reason is simple and uncomfortable: Sam Altman has told advisers that any valuation below $1 trillion is a non-starter, and public markets, right now, are not offering that price. The company’s last private round valued it at $852 billion. The gap between that and a trillion-dollar public float is the whole problem.

What matters is that OpenAI sits at the center of the AI demand chain. The company has signed an extraordinary string of supply and infrastructure deals that underpin a massive share of the industry’s expected spending. A delay in going public does not cancel those deals. But it does raise the question of how a company burning cash at this scale funds those commitments without a major capital raise in the near term.

The numbers from the leaked 2025 financials, verified by the Financial Times, are worth sitting with. OpenAI posted $13.07 billion in revenue last year against $34 billion in total costs, a $20.92 billion operating loss, and a $38.5 billion net loss attributable to the company after accounting for a $41.55 billion fair-value adjustment tied to its conversion from a nonprofit to a for-profit entity. Internal projections put 2026 losses at roughly $14 billion on a non-GAAP basis, though some analysts tracking the company’s burn rate put the GAAP figure considerably higher. Revenue is growing fast. The losses are growing faster in absolute terms.

That is a company with a lot to prove before it asks public markets to underwrite a trillion-dollar valuation.

The Market’s Reaction Said Everything

Chip stocks got hit hard. Arm Holdings and Marvell Technology fell around 4%. Advanced Micro Devices was off 3.5%, Intel dropped 3%, and Micron gave back more than 5% of the 16% gain it had logged just one day earlier on blowout earnings. The Philadelphia Semiconductor Index sank 4.8% on Friday alone. For the week, it was down nearly 8%, its worst stretch in months.

The broader indexes told the same story in slower motion. The S&P 500 slid roughly 2% on the week. The Nasdaq fell 4.6%, its worst week in over a year. The Dow held up, rising 0.6% as money rotated toward more defensive corners of the market. That divergence is not noise. It is the same rotation that has been building all year, suddenly accelerating.

JPMorgan flagged it cleanly: the IPO delay raised concerns about the sustainability of AI infrastructure spending given the delay in funding from capital markets. That is the actual risk. Not sentiment. Structure.

SoftBank Has the Most to Lose

If OpenAI is the center of the trade, SoftBank is the most concentrated single expression of it. And the math there deserves more attention than it is getting.

SoftBank’s cumulative investment in OpenAI is on track to total approximately $64.6 billion by October 2026, representing an ownership stake of roughly 13%. That makes it the largest external shareholder. When the New York Times story hit Thursday evening, SoftBank shares fell 13% in Tokyo the following morning, dragging the Nikkei 225 down more than 4% in a single session.

Slight tangent, but it matters: SoftBank CFO Sarah Friar has reportedly told associates the company is aiming for a 2027 listing. That warning, coming from inside one of the most valuable private companies in history, is the kind of thing that gets buried in AI optimism but should probably be one of the first things serious analysts are asking about.

The position is not a write-down. A 2027 IPO at a trillion-dollar valuation could still deliver a strong return. But SoftBank borrowed heavily to fund its OpenAI tranches, and the carry cost of delaying capital recycling for a year in a higher-for-longer rate environment is not trivial. SoftBank’s final $10 billion tranche is due in October 2026, with no near-term path to price discovery in public markets. That is a company with a ticking clock.

Kashkari Flipped. That Changes the Math.

Layer in what happened on the macro side and the picture gets more complicated.

Minneapolis Federal Reserve President Neel Kashkari, speaking at the Aspen Ideas Festival on Friday, said he has revised his full-year policy outlook from “one rate cut by year-end” in March to “one rate hike by year-end.” Kashkari is a voting FOMC member this year. His shift makes him the first core official in the current cycle to explicitly move to a hike stance, and it reflects what nine of the 18 FOMC officials now project. The median Fed forecast for policy rates has moved to 3.8%, up from 3.4% in March.

What drove the change? Inflation. The Fed’s preferred gauge, the all-items PCE, hit 4.1% in May, its highest since spring 2023. Core PCE came in at 3.4%, its highest since October 2023. Inflation has now missed the Fed’s 2% target for five consecutive years.

Here is where it gets interesting. Kashkari explicitly stated that investment in AI data centers is itself inflationary in the near term. The same capital spending that the entire semiconductor sector has been treating as a permanent tailwind is now being cited by a Fed voter as a reason rates might go higher. That is not a small idea. The discount rate applied to every high-multiple AI name moves in one direction if that argument gains traction inside the central bank.

Read that again slowly.

What the Market Is Missing

Most of the coverage this week framed the OpenAI delay as a confidence shock. A sentiment event. But the more important question is structural.

The Q1 earnings season gave the bulls everything they needed to stay comfortable. With 97% of the S&P 500 reported, FactSet puts blended year-over-year earnings growth at 28.6%, up sharply from the 13.1% analysts expected at quarter-end. That would be the highest growth rate since Q4 2021. Information Technology led all sectors with 54.3% earnings growth. But strip out Nvidia and Micron, and that figure drops to 30.1%. Strip Alphabet and Meta from Communication Services’ 48.9% growth and it flips to a 4.1% decline.

The AI earnings story is real. It is also extraordinarily narrow. A delay in the primary funding mechanism for the world’s largest AI buyer puts pressure on the width of that story in a way the index has not fully absorbed yet.

The consensus still reads this week as a healthy pullback inside a structurally intact AI spending cycle. That may be right. But there is a version of this where OpenAI’s delay raises questions about how a company burning billions annually funds hundreds of billions in commitments without tapping public markets. And if the answer involves renegotiating some of those commitments, even the strongest AI infrastructure stocks face a demand revision that has not been modeled yet.

The Honeywell Angle: A Different Kind of Spinoff Story

While the AI complex was unwinding, one of the most interesting structural events of the quarter completed quietly in the background. Today, June 29, Honeywell Aerospace begins regular-way trading on Nasdaq under the ticker HONA, completing its formal separation from Honeywell International, which continues as Honeywell Technologies under the HON ticker.

This one deserves more attention than it is getting. For 2026, Honeywell Aerospace expects sales growth of 7% to 9%, adjusted EBIT of $4.6 billion to $4.7 billion, and second-half free cash flow of $1 billion to $1.5 billion. The company is targeting $6.5 billion in adjusted earnings by 2030, supported by 6% to 8% annual sales growth. Its order backlog sits at $19 billion, up 20% year over year, with a book-to-bill ratio of 1.1.

The GE Aerospace comparison is the one to watch. GE gave investors a working example of what happens when a major industrial conglomerate becomes a focused aerospace company. That rerating from conglomerate discount to pure-play premium took GE’s aerospace multiple significantly higher in the years after separation. Honeywell Aerospace launches with content on roughly 90% of in-service aircraft, 36,000 employees, and a backlog growing at 20% year over year.

The first number to watch when HONA starts trading is the implied multiple against 2026 EBIT of $4.6 billion to $4.7 billion. If it trades at a premium aerospace peer multiple, the market is telling you Honeywell’s aerospace business was undervalued inside the conglomerate. The primary risk: a broader defense sector rotation reversal or a post-spinoff sell-off in either entity as index rebalancing shakes out. That second scenario, if it happens, is worth watching closely from a value standpoint.

Options Framework for This Week

Two structures worth considering against this week’s developments:

On the AI infrastructure fragility theme: Implied volatility across semiconductor names has spiked after five consecutive down sessions for the Nasdaq. For traders who believe the OpenAI delay is a near-term overhang rather than a fundamental break in AI capital spending, a defined-risk bull put spread on NVDA or AVGO in the August expiration captures elevated premium while limiting downside to a specific level. If you believe the spending commitments remain intact and this is a 12-month timing issue rather than a cancellation, the risk-reward on selling near-term fear is asymmetric. The caveat: if Kashkari’s hawkish pivot accelerates and the broader market begins pricing in a rate hike before year-end, growth multiples across the sector face additional compression regardless of IPO timing.

On Honeywell Aerospace (HONA): The first two to four weeks of trading create a structural buyer through index inclusion mechanics. Pure-play aerospace peers trade at materially higher multiples than Honeywell did as a conglomerate. For options traders, defined-risk call structures in the first 30 days of trading capture both the index-buying tailwind and the potential rerating to peer multiples, with the primary risk being a broader defense sector reversal.

The Risk Wall Street Is Underweighting

The Nasdaq was down 4.6% this week. The Dow was up 0.6%. The S&P 500 slid roughly 2%. That divergence is not random. It is the same rotation that has been building all year, and this week it accelerated.

The earnings are real. The AI spending cycle is real. But the market may have been pricing the peak version of those earnings against the lowest possible discount rate, and right now both of those assumptions are moving in the wrong direction at the same time. Higher rates. A delayed funding event for the world’s biggest AI buyer. And a Fed voter explicitly arguing that AI data center spending is itself part of the inflation problem.

The question now is whether this week was a rotation or something more durable. Nobody has a clean answer yet. But the people who are most exposed to that question are not talking about it loudly enough.