Everything About AI
NextFin News - The world's most-watched stock indexes are hitting record highs on artificial-intelligence euphoria, even as the bond market behind them flashes a warning last seen more than two decades ago. The 10-year US Treasury yield touched 5.34% this week, its highest level since 2002, while Brent crude hovered above $100 a barrel and borrowing costs across the developed world raced to multi-decade peaks. The S&P 500 can climb anyway — but only because a handful of AI darlings are carrying it.
The Rally That Hides a Retreat
The divergence is the story, and it has rarely been this wide. On one side, the S&P 500 closed at 7,818.93 on October 6, an intraday peak of 7,844.52 and its first record close since August; the Nasdaq Composite posted a second consecutive record, and the index is up 14% this year while the Nasdaq is up 18.78%, on track for a fourth straight year of double-digit returns. On the other side, the price of money is repricing faster than at any point this century: the third quarter of 2026 was the biggest quarterly rise in US Treasury yields since 2000, and the 30-year Treasury bond reached a 24-year peak before settling back to 5.61%.
The pressure is not American. In the United Kingdom, the 30-year gilt yield crossed 6% for the first time since 1998, and the 10-year gilt climbed 56 basis points in the quarter to 5.43%, its highest since 2007. France's 10-year yield reached its highest point since 2002 and drew close to 5%, while the spread between French and German borrowing costs stretched toward levels last seen during the eurozone debt crisis of the 2010s. Germany's own 10-year Bund approached its 2009 high at 3.61%, and Japan's 10-year yield hit its highest level since the mid-1990s.
Two forces are driving the same move. Elevated oil prices — Brent futures near $100.58 a barrel on October 6, with the US Energy Information Administration forecasting an average of $105 a barrel in the fourth quarter — are stoking a fresh inflation impulse. And the Federal Reserve has answered it: on September 16, the Federal Open Market Committee voted 12-0 to raise the federal funds target range by 25 basis points to 3.75%-4.00%, the first increase since 2023.
"We removed a dose of accommodation," Chair Kevin Warsh said at the post-meeting press conference. "Inflation is the problem. Stable prices have been the problem for now more than five and a half years."
Yet the equity market is behaving as if none of this matters. That is not because bonds are wrong. It is because the stock market is no longer a broad market.
Why the Stock Market Can Ignore the Bond Market
The mechanism is concentration, and the numbers make it almost mechanical. Nvidia, Apple and Microsoft together account for 21.1% of the S&P 500's weight — a share large enough that their gains can pull the index to records while the majority of stocks lag. Roughly 70% of the stocks in the major indexes are trading below their 50-day moving averages, a breadth reading that looks like a correction hiding inside a breakout.
This is not a normal bull market.
"Every bull market has a dominant theme, and technology and AI remain this market's dominant theme," said Keith Lerner, chief investment officer and chief market strategist at Truist Advisory Services. "Technology and communication services were the only two S&P 500 sectors to rise last month, while the other nine declined."
The capital flows explain the paradox. Hyperscalers are pouring hundreds of billions of dollars into data centers, and investors are betting that spending will earn a high return.
"They are betting that the money pouring into data centres will earn a good return, and so far, that belief has outweighed rising interest rates, more expensive oil and a 10-year bond yield above 5 percent," said Lochlan Halloway, a senior equity strategist at Morningstar Australia.
Through this channel, higher yields do not simply compress valuations the way textbook finance says they should. They also signal an economy strong enough to justify the spending — and a policy response that is preventive rather than reactive. The Fed's own Summary of Economic Projections shows the median policymaker expecting real GDP growth of 2.3% in 2026 and 2.4% in 2027, with unemployment at 4.1% and PCE inflation easing from 3.7% this year to 2.3% in 2027. When a central bank tightens into strength, the equity market can treat it as a vote of confidence in earnings power, even as the bond market prices the higher discount rate.
There is a second layer to the mechanism that most commentary misses. The AI buildout is not just an equity narrative; it is a force in the bond market too. Data centers are power-hungry, capital-intensive, and long-duration. They draw capital away from bonds, they lift forecasts for long-run growth and therefore for the long-run level of short-term rates, and they add to electricity and infrastructure spending that keeps demand hot. AI is both the rally and part of the repricing. The bond market is not wrong to demand more; it is pricing the fiscal and inflation consequences of a world that is building data centers as fast as power grids can supply them.
That is the first-order story. The second-order story is uglier, and it is where the rate shock actually lands.
The Second-Order Shock: Who Actually Pays for the Money
The bond-market stress is not abstract. Last week, 30-year mortgage rates crossed 7% for the first time since early 2025, tightening financial conditions for households. Higher yields raise financing costs for businesses and force governments to spend more on debt service. The Washington-based Institute of International Finance estimated that advanced economies paid more than $3.3 trillion in interest on government bonds traded internationally in the past year — surpassing global spending on AI ($2.6 trillion), defense ($3.1 trillion), or clean energy ($2.3 trillion). The same quarter that delivered record stock highs also delivered a record global debt load, which rose by $10 trillion in the first half of the year to top $365 trillion.
There is a bitter symmetry in those numbers. The AI boom that is lifting the indexes is being financed by a debt burden whose interest bill now exceeds the AI investment itself. If the return on data-center capital does not materialize, the bill still comes due. That is the gap between the stock market's story and the bond market's story: one is pricing a productivity revolution, the other is pricing the cost of borrowing to build it.
"As yields have crept higher, that is going to tighten financial conditions and could increase the risk of a slowdown," said Danny Zaid, a portfolio manager at TwentyFour Asset Management. "But fundamentals right now — the broader economy — still look very strong."
That sentence contains the entire tension. If fundamentals stay strong, the Fed can keep tightening without breaking the economy, and the AI earnings story can validate the concentration. If financial conditions tighten faster than earnings grow, the narrow leadership has nowhere to hide. The market is effectively betting that the Fed's reaction function is preventive — tightening because the economy can take it — rather than reactive, tightening because inflation is escaping. History suggests the difference matters: preventive hikes can accompany rising stocks; reactive hikes, the ones that follow an inflation surprise the central bank failed to foresee, tend to end with the bond market winning.
The Counter-Thesis: Bonds Are Right, Stocks Are Wrong
The strongest case against the equity rally is that it is a mirage sustained by a handful of names. Jeffrey Gundlach, chief executive of DoubleLine Capital, warned this week that stocks are deteriorating beneath the surface even as major indexes avoid the deep declines that bond-market stress has historically triggered. The historical record is uncomfortable for bulls: previous episodes in which the 10-year yield surged to multi-decade highs while breadth deteriorated ended with equities repricing lower, not with the bond market blinking.
The bear case has a clean transmission chain. Oil stays above $100 a barrel; inflation stays above the Fed's 2% target — the Fed's own projections call for PCE inflation of 3.7% in 2026 before easing to 2.3% in 2027; the Fed keeps hiking, with markets pricing up to 100 basis points of additional tightening by September 2027; the 10-year yield grinds toward 5.5% and beyond; and eventually the discount-rate math catches up with even the most dominant mega-caps. At that point, concentration becomes fragility: a stumble in any one of the three names that make up 21.1% of the index would drag the benchmark down with it.
The bull case, fairly stated, is that this time the supply side of the economy is being rewritten. AI is not just demand; it is a productivity shock that can grow the economy faster than inflation, letting the Fed hold rates higher without a recession. The September projections already lean this way: the median unemployment forecast fell by 0.2 percentage points even as the rate outlook moved up, and 12 of 18 policymakers project at least one additional hike this year — a signal of confidence, not panic. If AI-driven productivity shows up in the output data, the bond market's inflation fear is the thing that will prove wrong.
So which is it — cyclical noise or a regime shift? The answer has to be split, and conflating the two is the most common error investors make here. The rate shock itself is cyclical: it is a policy reaction to an oil-price spike and a hot inflation print, and it will mean-revert if Brent falls back and core inflation cools toward 2%. Markets price a policy path, and policy paths change when the data does. But the concentration of the equity rally has a structural leg: the data-center buildout, the power constraints, and the winner-take-most economics of AI infrastructure are not going away on their own. The cyclical leg can pass without touching every stock; the structural leg does not protect the 70% of the market trading below its 50-day average. Treating the rate shock as permanent means missing the mean reversion; treating the AI concentration as temporary means missing the regime shift. Both are mistakes.
What to Watch: The Signal That Breaks the Thesis
The base case is continued divergence: the S&P 500 and Nasdaq grind higher on mega-cap AI earnings while the 10-year yield trades in a 5%-5.5% range and breadth stays weak. The upside case is that AI monetization accelerates faster than expected, earnings lift the entire technology complex, and breadth improves — pulling the majority of stocks off their 50-day lows and validating the rally as broad-based. The downside case is that oil stays above $100, inflation prints stay hot, and the 10-year yield sustains above 5.5%; at that level, financial conditions tighten enough to crack earnings expectations, and the narrow leadership reverses.
The falsifying signal is specific: if the 10-year Treasury yield holds above 5.5% while the S&P 500 advance-decline line makes new highs — meaning breadth improves even as borrowing costs rise — the "narrow rally masking stress" thesis is wrong, and the AI productivity story is stronger than the bond market believes. Conversely, if the 10-year yield breaks 5.5% and breadth deteriorates further, the bond market's warning was the leading indicator all along.
Three data points carry the signal. First, the monthly core PCE inflation print: a reading at or above 0.3% month over month for two consecutive months keeps the Fed hiking and pushes the 10-year toward the 5.5% threshold. Second, the weekly 10-year yield level itself. Third, the percentage of S&P 500 stocks above their 50-day moving average — a move back above 50% would confirm broadening participation; a drift toward 20% would confirm the divergence is worsening.
The market is not telling one story. It is telling two, and they contradict. The indexes say the AI revolution is here and earnings will justify every multiple. The bond market says the price of money has changed, and eventually every asset has to answer to it. For now, the few are carrying the many — but the cost of carrying them keeps rising, and the bill for the debt that built the rally now exceeds the investment in the rally itself. That is the arithmetic the stock market is betting it can outrun.
Explore more exclusive insights at nextfin.ai.
