Everything About AI
NextFin News - US stock futures fell on Thursday as the benchmark 10-year Treasury yield climbed to 5.135%, its highest level since July 2007, and oil prices extended a summer rally that has pushed West Texas Intermediate back above $100 a barrel - a one-two punch that is forcing the market to reprice the cost of capital for an entire investment cycle.
Contracts on the S&P 500 were down 0.6% at 7:44 a.m. in New York, while Nasdaq 100 futures dropped 1%, setting the tech-heavy gauge up for another sharp session after it notched its first record close in months earlier in the week. The premarket decline extends a two-day selloff: the S&P 500 fell 0.75% to 7,706.03 on Wednesday, the Nasdaq Composite shed 1.13% to 26,936.04, and the Dow Jones Industrial Average lost 352.10 points, or 0.68%, to 51,511.59.
The trigger was a fresh surge in Treasury yields after the latest purchasing managers' index readings came in hotter than expected, reviving bets that the Federal Reserve - which raised its benchmark rate by a quarter point last week to a target range of 3.75% to 4% - will keep policy restrictive rather than pivot. The 10-year yield's jump was its largest single-day move since April 7, 2025, while the 2-year yield touched 4.947%, its highest since May 2024. Oil compounded the pressure: crude plunged to $70 a barrel in early July during a temporary ceasefire in the Middle East, then ratcheted back up as hostilities expanded, including attacks on Saudi Arabia's energy infrastructure.
The combination is doing something that neither development alone would: it is compressing equity valuations through the discount rate at the same time it is raising the inflation floor that keeps the Fed's hands tied. The result is a selloff that is broad rather than confined to rate-sensitive corners of the market.
Why Yields and Oil Hit Equities Through the Same Channel
The transmission mechanism runs through the risk-free rate, which is the denominator in every discounted-cash-flow model on Wall Street. When the 10-year Treasury yield moves from 4.2% to above 5% - the level it averaged over the 12 months ending February 2026 - the present value of a dollar of earnings ten years out falls by roughly a fifth, even if nothing changes about the company actually earning that dollar. That is why utilities and consumer discretionary stocks led Wednesday's declines, each falling more than 1%: these are long-duration businesses whose valuations are mathematically sensitive to the discount rate.
Oil works through the same channel from the other direction. Higher crude prices raise input costs for transport, manufacturing, and chemicals; they show up in gasoline prices that consumers see at the pump; and they feed inflation expectations, which is what keeps bond investors demanding a higher yield in the first place. The correlation between Treasury yields and WTI has been tight since late February 2026, when the bombing campaign in Iran commenced - Haver Analytics notes the two series have moved in lockstep through the spring and summer.
This is the mechanism that makes the selloff self-reinforcing. Hot economic data pushes yields up; yields push equity multiples down; oil pushes inflation expectations up; and inflation expectations push yields up again. Breaking the loop requires one of the links to fail - and so far none has.
The AI trade, which powered the Nasdaq's record earlier in the week, offered only partial insulation. Meta Platforms' rollout of its new AI assistant, Muse, drew a strong reception that analysts say could benefit technology infrastructure stocks. But enthusiasm for one company's product cannot offset a 5% risk-free rate that is now embedded across the entire market's discount curve. A stock can have the best product in the world and still trade at a lower multiple when the alternative of a risk-free government bond pays investors five percent to do nothing.
Cyclical Trigger, Structural Floor: Two Regimes Are Overlapping
The most important question investors need to answer is whether this is a cyclical spike that will mean-revert, or a structural shift in the regime for interest rates. The honest answer is that both are happening at once, and confusing the two is the fastest way to misread the next quarter.
The cyclical leg is real and identifiable. Oil spiked on geopolitics, not on a permanent supply shortfall. WTI traded at $65 a barrel at the end of February 2026, jumped to the high-$90s by the end of March, touched $108 in May, then collapsed to $70 in early July when a ceasefire held. That volatility is the signature of a war premium, not a structural supply deficit. If de-escalation takes hold, the oil component of the yield move can unwind quickly - and with it, some of the pressure on equities. On September 21 alone, West Texas Intermediate fell 3.8% to $96.46 a barrel on hopes for diplomatic progress, and the 10-year yield fell with it.
But the floor beneath yields has moved up for structural reasons that a ceasefire alone will not fix. Three forces underpin the shift from a 4.2% average to a 5%+ trading range. First, the AI buildout is capital-intensive and increasingly debt-financed, raising the demand for long-duration capital at the same time the Treasury is supplying more bonds. Second, persistent fiscal deficits mean the government must issue more debt into a market that is already absorbing war-driven energy inflation. Third, the term premium - the extra compensation investors demand for holding long-dated risk - has reappeared after years of suppression by central-bank buying programs that no longer exist.
This is why the conventional wisdom that "yields will fall back once oil settles" is only half right. Oil settling takes the cyclical leg off the table; it does not restore the old regime of cheap long-duration money. The structural floor remains because the supply of bonds has changed, the demand for capital has changed, and the central bank that used to absorb duration risk has stepped back.
"This will probably be not a one-and-done, but a series of rate increases. It will be dependent on oil; that is really the source of inflation and it's starting to seep into other parts of the market," said Paul Nolte, a senior wealth adviser and market strategist at Murphy & Sylvest in Elmhurst, Illinois.
Nolte's point cuts to the policy bind: the Fed raised rates last week precisely because inflation remains elevated, and oil is the swing factor that determines whether that single hike is enough. The central bank's own statement after the September 16 decision said "inflation remains elevated" and framed the action as support for "a timelier return to the Committee's 2 percent goal" - language that leaves the door open to further tightening if energy costs keep feeding through.
The Second-Order Effect: AI Capex Runs Into a Higher Cost of Capital
The first-order effect of rising yields is mechanical and already well understood: discount rates up, valuations down. The second-order effect is more consequential and less discussed. The same companies leading the AI buildout - the ones buying graphics processors, building data centers, and signing multi-year power contracts - are now facing a materially higher cost of capital just as their spending plans reach their steepest phase.
A 100-basis-point move in the risk-free rate does not simply reduce a stock's present value on a spreadsheet. It forces capital-allocation committees to re-underwrite projects that were approved when money cost 3%. Marginal data-center builds get delayed. Leveraged acquisitions of AI-adjacent businesses get repriced or abandoned. Equity issuance needed to fund capex gets done at lower multiples, diluting existing shareholders. The companies with the deepest balance sheets keep building; the ones that relied on cheap debt discover that their business plans were subsidized by a monetary regime that has ended.
The irony is sharp: the AI boom is part of what is pushing yields higher, and higher yields are now the single biggest threat to the AI boom's valuation math. Haver Analytics describes the link plainly - real rates have risen alongside "the booming AI buildout that is involving more debt financing." The sector that drove the Nasdaq to its record is now the sector most exposed to the repricing that record required.
There is also a cross-asset transmission that equity investors are only beginning to price. Higher Treasury yields make government bonds genuinely competitive with equities for the first time in years - a 5% risk-free return forces active managers to justify why they should own earnings yields of 4% or 5% with substantially more risk. That rotation does not require a recession or an earnings collapse. It only requires the alternative to look attractive enough, and at 5.135% on the 10-year, it does.
The Counter-Thesis: Why the Bears Could Still Be Wrong
The strongest case against the view that yields have structurally repriced is straightforward and widely held: the Fed just raised rates, and tightening cycles historically end in cuts, not in a permanent 5%+ 10-year yield. If the economy slows - if the hot PMI print proves to be a final flare of demand before a downturn - the bond market will front-run the pivot and yields will fall back toward 4.5% or lower, restoring the multiple support that equities lost this week.
This is the base case for many institutional investors, and it is not a weak argument. The 2-year yield at 4.947% is already pricing in a restrictive policy path; if inflation rolls over, the curve could bull-flatten quickly, as it did on September 21 when the 10-year fell 2.46 basis points to 4.9715% on oil-driven disinflation hopes. A ceasefire in the Middle East would remove the oil shock entirely, and a softening labor market would hand the Fed a reason to pause rather than continue hiking.
The rebuttal is that the evidence for a structural floor is accumulating faster than the evidence for a cyclical top. The term premium does not vanish because one PMI print runs hot; it reflects a multi-year reassessment of fiscal sustainability, supply-chain fragmentation, and the capital intensity of the energy and technology transitions. For the cyclical view to win, two conditions must hold simultaneously: inflation must fall back toward target without a recession, and oil must stay down. For the structural view to win, only one needs to fail.
The falsifying signal is concrete and observable: if the 10-year Treasury yield closes below 4.50% for five consecutive trading sessions while oil holds below $85 a barrel, the structural-repricing thesis is wrong, and this week's selloff was a cyclical overshoot that has already played out. Until that combination prints, the burden of proof rests on the bears who are betting on a swift return to cheap money.
Earnings in the Crosshairs: What the Premarket Movers Signal
Thursday's session also carried company-specific catalysts that will test whether corporate America can grow into valuations at a higher discount rate. Darden Restaurants was due to report first-quarter fiscal 2027 results before the bell at 8:30 a.m. Eastern, after a fiscal 2026 in which it posted $13.2 billion in revenue, diluted earnings per share of $10.38, and a 9.1% net profit margin. The question for Darden is whether diners keep spending at Olive Garden and LongHorn Steakhouse as energy and borrowing costs rise - a live test of consumer resilience.
Dropbox traded in a $33.30 to $34.31 range with a market capitalization of $7.44 billion and a price-to-earnings ratio of 19, after a second quarter in which adjusted earnings of $0.75 a share beat the $0.69 consensus. Everpure - which has risen roughly 62% since April after reporting second-quarter earnings of $0.70 a share against expectations of $0.19, a revenue jump of 37.7% to $1.19 billion - carried a consensus price target of $129.10, implying about 15% upside. MGM Resorts, whose second-quarter adjusted earnings of $0.59 a share topped the $0.56 consensus on revenue of $4.45 billion, was among the names in focus after its shares pulled back toward the lower end of a 52-week range of $29.19 to $51.59.
These names matter because they span the economy's exposure to the two forces driving the selloff. Darden and MGM are consumer-facing businesses that feel higher energy costs and tighter credit immediately. Dropbox and Everpure are technology companies whose valuations are most sensitive to the discount rate. If they can all show pricing power and durable demand, the equity market can grow into its multiples even at a 5% risk-free rate. If they cannot, the multiple compression has further to run.
What Comes Next: Three Signals and Three Time Horizons
The near-term path for equities now depends on three observable signals. First, oil: a sustained move below $90 a barrel would relieve the inflation channel and give the Fed room to sound less hawkish. Second, the 10-year yield: a failure to hold above 5% would signal that the bond market is not committed to the new regime. Third, earnings: if companies can show they are passing energy and financing costs to customers without destroying demand, the market can absorb a higher discount rate.
Across time horizons, the picture splits. In the short term, sentiment and liquidity dominate - the next PMI print and any escalation or de-escalation in the Middle East will drive headlines, and both can reverse within days. In the medium term, fundamentals take over: third-quarter earnings will show whether margins can absorb higher energy and financing costs without cutting investment. In the long term, the structural question decides the regime: whether the 2020s are a decade in which cheap capital briefly returned, or a decade in which 5% on the 10-year Treasury is the new normal.
Three scenarios frame the path ahead. The base case is a grinding range-bound market in which yields hold between 4.75% and 5.25%, oil stays elevated but does not spike, and earnings growth is positive but slow - a market that trades on stock selection rather than multiple expansion. The upside case requires oil to fall below $85 and the 10-year to break below 4.5%, which would restore multiple support and reignite the AI-led rally. The downside case is a stagflationary squeeze: oil above $110, the 10-year pushing toward 5.5%, and earnings guidance being cut as financing costs bite - the environment in which the 2007 yield levels last seen become a warning rather than a milestone.
Thursday's selloff is the market's answer to a question it has been avoiding all year: what happens to the AI boom when money is no longer free? The yield curve, not the earnings report, is now the story - and until the bond market blinks, the equity market will keep adjusting to a world where the price of capital has changed for good.
Data as-of: premarket trading, Thursday, September 24, 2026; prior-session closes as of Wednesday, September 23, 2026; Federal Reserve decision as of September 16, 2026.
Explore more exclusive insights at nextfin.ai.
