#Gate广场万圣节 Inflation’s Return and AI’s Demystification: Global Markets Are Experiencing a “Reset in Pricing Logic”

In late October, global markets shattered a long-held asset allocation principle in an almost brutal fashion: when the threat of war looms, gold is not necessarily a safe haven. Brent crude hit $117.45 per barrel intraday on Friday, breaking above that level for the first time since April, and gaining more than 10% for the month; spot gold fell below the key psychological level of $4,000 per ounce, losing more than 4% for the week and hitting a more than three-month low; the Nasdaq 100 posted a 5.5% weekly decline the previous week, its biggest weekly drop since April 2025, while the Philadelphia Semiconductor Index fell for eight consecutive trading sessions. Surging oil prices, plunging gold, and retreating tech stocks—three seemingly independent moves—converged in the same window, pointing to the same underlying variable: markets are simultaneously reassessing the inflation outlook and the sustainability of AI assets’ earnings. This is not an ordinary shift in risk appetite, but a reset in pricing logic from “safe-haven trades” to “inflation trades,” compounded by the difficult transition of the AI theme from “narrative premium” to “fundamental scrutiny.”

01 How Oil “Killed” Gold’s Safe-Haven Function!
Under the traditional framework, military escalation in the Middle East should benefit gold. But the transmission chain in this market move has been precisely the opposite. At a rally on Friday, Trump said the US might again launch a “small-scale ‘Epic Fury’ operation” against Iran. Iran’s Revolutionary Guard announced that two supertankers had struck mines in the Strait of Hormuz, involving a combined total of more than 4.5 million barrels of crude oil, while a senior Iranian security official also revealed that plans had been drawn up to target undersea communications cables in the Gulf.
The tail risk of supply disruptions is being seriously priced in by markets. Rapidan Energy President Bob McNally warned that if the conflict spirals out of control and causes major damage to key oil infrastructure, oil prices could quickly surge toward $150.
The problem with high oil prices is not the prices themselves, but that they reopen the gates to inflation. The US five-year breakeven inflation rate rose 8 basis points to 2.47% on Friday, while the 10-year US Treasury yield climbed to 4.99%. Rising inflation expectations strengthen the case for the Federal Reserve to keep rates high, pushing up real rates in turn and directly increasing the opportunity cost of holding gold, a zero-yield asset. A more hidden source of pressure comes from liquidity: concentrated selling in US AI stocks triggered margin calls, forcing investors to liquidate the most liquid assets to replenish collateral, turning gold into a de facto “cash machine.”
In other words, gold has not “failed”; it has simply been suppressed by a stronger chain of logic—rising oil prices → higher inflation → central bank tightening → rising real rates. This chain has far greater pricing power than the mechanical reaction of “buying gold during geopolitical conflict.”

02 AI’s Retreat: From “Theme-Based Buying” to “Calculating Returns”!
If the oil shock is squeezing all risk assets from the denominator side (interest-rate expectations), the reversal in AI trading is applying pressure from both the numerator side and the funding side.
Goldman Sachs Prime brokerage data is emblematic: from October 8 to 14, hedge funds were net sellers of global AI stocks for five consecutive trading sessions, wiping out in just one week all their cumulative net buying since the start of 2025. The combined market value of the world’s 20 largest AI and semiconductor stocks fell by more than $1 trillion, South Korea’s stock market entered bear market territory and triggered circuit breakers, and the Philadelphia Semiconductor Index lost about $2 trillion in market value over eight trading sessions.
Goldman Sachs Prime’s Marco Laicini noted that a third “once-in-five-years” extreme market event in 10 months suggests hedge funds’ existing value-at-risk models may systematically underestimate extreme risks. The immediate trigger for the exodus of funds was concern over the sustainability of AI infrastructure financing, but the deeper disagreement is whether computing power is actually a “good business.” At this moment, Silicon Valley and Wall Street are giving diametrically opposed answers.
On an earnings call, Musk said the payback period for computing power could be shortened to “as soon as six months next year, which cannot be ruled out,” and planned to add 10 GW of online computing capacity in 2027; Google CEO Pichai confirmed that securing computing capacity for 2027 is “extremely difficult” for customers, and that the entire industry hopes to have twice the capacity it originally planned. GPUs are highly analogous to airlines, with extremely low marginal operating costs; once supply catches up with demand, spot rental prices will “fall off a cliff.” Meanwhile, the market is shifting from “theme-based buying” to genuine fundamental research: “This test is only just beginning.” —Citadel founder Ken Griffin. The fundamental disagreement between the two is not about the strength of demand today, but a fundamental question about the long-term value of computing power: is it a persistent premium on a scarce resource, or infrastructure that will eventually become commoditized?
As hedge funds withdraw from tech stocks at the fastest pace in at least a decade, the market’s interim answer is that, constrained by rising financing costs and uncertainty over the pace of earnings delivery, “forward-looking stories” on the numerator side must give way to real interest rates on the denominator side.

03 The Point of Convergence: The Two Threads Meet in Liquidity!
The US-Iran situation and the retreat in AI may seem to belong to two unrelated dimensions—geopolitics and the tech industry—but they converged at one key point: liquidity. Higher oil prices are pushing up inflation expectations and narrowing the Federal Reserve’s room to shift toward easing. That means the valuation denominator for risk assets cannot be “rescued” by falling interest rates.
The concentrated unwinding of AI positions, meanwhile, occurred precisely among hedge funds with high leverage— the VIX surged about 30% at one point in mid-October, breaking above 21 and reaching a more than four-month high; market rumors swirled that a large hedge fund was in trouble, and Goldman’s trading desk described conditions at the time as “a genuine capitulation-style selloff and deleveraging.”
When falling AI stocks trigger margin calls, investors are forced to sell all liquid assets, including gold, closing the loop where the two threads meet in a liquidity squeeze.
Nanhua Futures’ summary is accurate: risk assets broadly fell, while crude oil and US Treasuries rose, mainly because the situation in the Middle East pushed up oil prices, increased inflationary pressure, and strengthened expectations of rate hikes. At the same time, tech earnings reports raised concerns about the sustainability of earnings, and a liquidity squeeze intensified selling pressure on metals and precious metals.

04 After the Reset in Pricing Logic
The real legacy of this convergence is not a handful of price levels, but a pricing framework that is being rewritten.
The first shift is the replacement of safe-haven trading by inflation trading. In the face of geopolitical shocks, mechanically “buying gold to hedge against war” is no longer reliable—the money flows toward the most urgent problem, and when inflation once again becomes the main enemy, safe-haven funds will abandon gold without hesitation. This does not mean gold has lost its long-term allocation value: global central banks bought 79 tonnes of gold in August, more than four times the average level before the 2022 Russia-Ukraine conflict, and central bank buying is still accelerating around $4,000. But the source of short-term pricing power is now clear: real rates dominate everything.
The second shift is the narrowing of AI investment from a “broad theme” to “bottleneck assets.” In its fourth-quarter outlook, BlackRock said the AI investment theme had shown no clear warning signs, but opportunities were shifting from “AI broadly” to bottleneck areas such as power, grids, chips, and data centers. Morgan Stanley, meanwhile, believes valuation adjustments for most stocks are nearing an end, and that prices, orders, and earnings trends for core AI infrastructure chips have not deteriorated to a corresponding degree. Disagreements remain, but a consensus is forming: the phase of broadly betting on “AI concepts” is over, and returns in the next phase will belong to areas whose fundamentals can be verified.
For investors, the core challenge in the current environment is not to judge whether oil will reach $150 or whether the Nasdaq has bottomed, but to accept a more uncomfortable reality: the asset-pricing paradigm of the past two years, jointly supported by “expectations of easing + the AI narrative,” is being replaced by “inflation constraints + earnings scrutiny.”
In this new paradigm, gold may not rise and tech stocks may not fall, but stories that generate no cash flow will find it harder than ever to command generous valuations.$NAS100