MARKET RECAP — FRIDAY, JUNE 26, 2026 By Nicholas Thomas | TrendyVest.com

The Nasdaq closed lower for the fourth consecutive session Friday, chip stocks fell for the second straight day, and the most powerful AI company in the world quietly admitted it cannot go public at its own asking price. None of these things are unrelated. Together, they tell the story of a market arriving at a question it has been avoiding for months: who actually pays for all of this?

The S&P 500 fell 0.75% on the day, the Nasdaq dropped 1.25%, and the Dow Jones shed 0.44%. The Russell 2000, a proxy for smaller domestic companies with no AI exposure, gained 0.71% — a signal, not a coincidence. For the week, the Nasdaq fell nearly 4% while the Dow rose 0.6%. That divergence is the most important data point of the week, and most coverage buried it.

The OpenAI story is bigger than an IPO delay

Late Thursday, the New York Times reported that OpenAI is leaning toward delaying its public offering until 2027. The company's advisers had presented CEO Sam Altman with two options: wait for a $1 trillion valuation, or accept a lower number and list sooner. Altman, according to people involved in the deliberations, called any compromise on the trillion-dollar figure a nonstarter. What followed was treated as IPO gossip. It was not.

JPMorgan's trading desk was direct about what the delay actually signals, writing that it raises serious concerns about the sustainability of AI infrastructure spending given the delay in funding from capital markets. Buried even deeper was the disclosure that OpenAI's CFO Sarah Friar is pushing internally for the delay because the company is sitting on roughly $600 billion in compute infrastructure commitments through 2030 — commitments it needs runway to manage before facing the scrutiny of quarterly public reporting. This is a company burning cash at a record pace, carrying an enormous forward obligation, and unable to achieve the market capitalization it needs to justify it. That is not a footnote. That is a systemic concern about the economics of the entire AI buildout.

Micron's blowout quarter made chip stocks worse

The paradox at the center of Friday's tape was Micron Technology. The company reported what can only be described as the greatest quarter in semiconductor history — revenue of $41.46 billion against an estimate of $35.84 billion, adjusted earnings of $25.11 per share against expectations around $20. The stock surged 17% Thursday. By Friday's close, Micron was down 4%, and Sandisk, Lam Research, and Western Digital had fallen as much as 10%.

The reason reveals a structural fracture forming inside tech. Micron winning means memory prices are rising. Rising memory prices mean higher costs for every company that builds products using chips — which is every major technology company. Apple raised MacBook and iPad prices this week. Microsoft raised Xbox prices. Both cited higher memory costs. The chip winners are now a cost burden for the companies that buy from them, and investors are beginning to price that dynamic with precision. Schwab's market team captured it plainly: good news for Micron is not necessarily good news for tech. The PHLX Semiconductor Index is now verging on correction territory, down nearly 10% from recent highs, and the selling has been methodical rather than panicked.

Seoul needed a circuit breaker. Americans weren't watching.

Friday morning in South Korea, the Kospi index plunged more than 8% in intraday trading, triggering a marketwide circuit breaker that halted all trading for twenty minutes. The index closed down 5.8%. The Korea Exchange noted that Samsung Electronics and SK Hynix — the two largest memory chip producers in the world — together make up more than half the Kospi's total market value. When a single-theme pullback hits an index that concentrated, it doesn't stay contained to one country. Friday's chip sell-off in New York was not independent of what happened in Seoul. It followed from it.

The rotation story is healthier than the headlines suggest

Despite the Nasdaq's losses, something notable was happening underneath the surface. Even on the week's worst days, advancing stocks outnumbered declining ones across the broader market. Money wasn't leaving equities — it was being reallocated with conviction. Visa and Walmart each gained more than 2% on Friday. Eli Lilly surged 7% after the European Union approved its leukemia therapy. Healthcare, industrials, and financials carried the Dow to a weekly gain while tech bled. This is not what a broken market looks like. This is what a repricing looks like.

SpaceX joins the Russell 1000 tonight. $22 billion in forced buying starts Monday.

One story that went largely unnoticed in the noise: after Friday's closing bell, FTSE Russell formally added SpaceX to its U.S. indexes as part of a newly inaugurated semi-annual reconstitution. The changes go live for trading Monday, June 29. Every ETF and mutual fund benchmarked to the Russell 1000 must now hold SpaceX shares — not by choice, but because the index mandates it. Estimates for the passive buying required from Russell 1000-tracking funds alone range from $22 to $27 billion. That demand is non-discretionary. Fund managers don't vote on it. They execute it at Friday's close to minimize tracking error. SpaceX's stock has retreated sharply since its June 12 IPO, falling from a high near $225 to around $153. The forced index buying represents a structural tailwind at exactly the moment the stock needed one. The Nasdaq 100 inclusion comes next.

A Fed hawk called for a rate hike — on a Friday afternoon

Minneapolis Federal Reserve President Neel Kashkari said Friday that he has altered his economic outlook and now expects one rate hike this year. The timing was notable: a relatively quiet afternoon, no major data release, and a statement that shifts the rate path. Meanwhile, the University of Michigan's final consumer sentiment reading came in at 49.5, slightly above the 49.0 estimate, with five-year inflation expectations falling sharply to 3.3%. Consumers are more optimistic about the long run than they were a month ago. The Fed, or at least one of its members, is not. That contradiction will define the bond market's behavior heading into Q3.

Oil fell 4% on a drone attack. That is the strangest data point of the day.

President Trump posted Friday that Iran launched at least four one-way attack drones at a cargo ship in the Strait of Hormuz, with one striking the vessel's upper deck. He called it a foolish violation of the ceasefire agreement. WTI crude fell 4.1%. Brent fell 4.5%. Tankers kept moving. The market's interpretation is that the ceasefire is holding and this incident will not escalate. That may be correct. It may also be the most dangerous complacency trade of 2026. There is no moderate read on a drone strike hitting a cargo ship.

What Friday actually means

The AI trade has not collapsed. But it has fractured into two narratives moving in opposite directions. The infrastructure layer — memory, chips, power, storage — is winning on genuine demand. Micron's numbers proved it beyond argument. The application layer — the companies assuming AI will transform their products, justify their valuations, and generate returns on the spending — is being interrogated in a way it has not been before. OpenAI's inability to go public at its target price is the clearest expression of that interrogation. The capital markets are asking whether the economics of this buildout actually work.

That question did not get answered Friday. It rarely does in a single session. But the tape is accumulating evidence in one direction — and next week, with earnings season restarting in July and a Fed decision looming, the market will be asked to answer it more directly.

— Nicholas Thomas | TrendyVest.com | June 26, 2026