1Gold is 20% below its January high in the middle of the oil shock that should be lifting it
Gold trades near $4,300 an ounce, roughly 20% below the January record around $5,600, and silver is 30–40% off its own high. That is happening while Israeli strikes on Iran and Lebanon push oil up $3 in a session, Iran's envoy describes the Strait of Hormuz as open but subject to transit fees, and global airlines cut their 2026 profit forecasts on the fuel bill. A haven falling into a war is the most awkward fact on the tape.
Felix, on Felix & Friends, argues the fall is the first leg of the oil-shock template rather than a refutation of it. In 1973, 1979, 1991, 2001 and 2022 the sequence ran the same way: the shock lifts oil, oil lifts inflation expectations, the Fed cannot cut, yields and the dollar rise, and gold dips before making higher highs. On that reading the drawdown is the mechanism working, not failing. His supporting evidence is positional. Both metals are trading well below their 200-day moving averages, which is rare, and on the last ten occasions gold touched that level six were profitable and all ten were higher twelve months later. Central banks are net buyers again, with Asian and Polish buying and Middle-Eastern selling slowing; silver is in supply deficit; JPMorgan and Deutsche Bank carry targets 35–40% higher, near $6,000.
His preferred expression is miners and royalty companies, for the operating leverage built into them — a 10% move in the metal is roughly 30% in the miners. (Agnico Eagle) carries all-in sustaining costs near $1,400 an ounce against a ~$4,300 price, a margin of roughly $3,300 an ounce; (Franco-Nevada) and (Wheaton Precious Metals) are the royalty versions of the same exposure. The caveat travels with the call: he has not bought yet, because his read of the selling pattern says it is late-stage but unfinished. Directional conviction, entry deliberately unmade.
Set against that, Reuters reports gold extending losses on US rate-hike fears, and it fell another 0.86% overnight. That is the same two facts — an oil shock and a Fed that cannot cut — running to the opposite conclusion. One frame says that is precisely why gold dipped and why it recovers; the other says that is why it keeps falling. What would settle it is whether a recovery leg appears while yields are still rising, which is the one thing neither frame has yet observed. The bull side deserves a discount for its packaging, too: the segment sits inside heavy lead generation — an app trial, a paid academy, a webinar funnel and a report giveaway — which does not make the framework wrong, but the certainty is being sold alongside it.
So what: an oil shock that stops the Fed cutting is a headwind for gold's rate story and a tailwind for its inflation story at the same time. Which of the two dominates decides whether $4,300 was the low or a way station.
2The AI rally lost its leaders, and what replaced them at the top doesn't turn a profit
The tape did the work. fell 6.2% and 11.3% in the semiconductor selldown, with dragging the group on guidance, while fell 12% and 26%. Reuters ran the session as "Tumbling tech darlings slam brakes on AI rally"; CNBC placed the hyperscalers at the epicentre of the bear case. Volatility repriced with it — the VIX rose 39.7% to 21.51 and the KOSPI fell 8.15% overnight. Index futures, meanwhile, were near flat, the S&P down 0.06% and the Nasdaq up 0.06%, so the damage sat in names rather than in the market.
Excess Returns supplies the structural version. The Mag 7 is now being beaten by both small-cap tech and unprofitable tech, which Jim Paulsen, speaking there, calls a late-1990s-style change of stripes — leadership passing to the lower-quality end of the same theme, with Goldman Sachs' AI index parabolic on the way. The strain shows first in what led. sits at 65 times sales and has stalled despite 72% revenue growth, so the ceiling is the multiple rather than the business, and trades above 100 times earnings with the standing caveat that fundamentals have never been what set its price.
The same episode reframes what this means for value investing, and the reframe is more useful than the obituary. The factor's underperformance comes from value traps in the industries technology actually disrupted — retail is the standard example — while in insulated sectors value has worked fine. Cheapness was never the problem; being cheap in front of a disruption was.
One fact points the other way on the same day. Naver said it will build gigawatt-scale AI factories on Nvidia hardware — an order, not a sentiment reading, landing while the shares fell 6%. Japan's first-quarter GDP, revised weaker on soft capital spending, cuts back toward the bearish side. So what: when leadership passes from the profitable end of a theme to the unprofitable end, the breadth that looks like a rally getting healthier is the same data a late-cycle read uses, and this session did not distinguish between them.
3Three funds are sold as the quantum trade, and the biggest one is mostly a chip fund
Brian, on Business with Brian, ranks the quantum ETFs by how much quantum they actually hold, and that turns out to be the whole distinction. (WisdomTree) comes first as the genuinely pure-play option, with the four listed leaders — , , and — as top holdings and enough operating history behind it to hold long term. (Corgi) is second: active, about 18 names, roughly half the fund in IonQ and D-Wave, with a post-quantum-encryption angle. It is also only weeks old, which is why the call is "one to watch" rather than a core position — the caveat is the call.
(Defiance) is the biggest and oldest of the three and only 4–6% pure quantum. About a third of it is semiconductors — , — overlapping a Nasdaq-100 fund like that a holder probably owns already. Its distinct use is access to foreign semiconductor names, MediaTek and Tokyo Electron among them, which a US-listed chip fund does not provide.
Two caveats travel with the ranking. The episode carries a disclosed paid-sponsor segment for Telescope Innovations , plus a Patreon and newsletter push — the sponsored name sits outside the day's recommendations, but the ranking arrives from a channel carrying a commercial relationship in the same video. And the legal risk is named rather than buried: a securities-fraud suit against is flagged, and the name is held anyway, at about 2.5% alongside the other three pure-plays, sized as a deliberately asymmetric bet where a zero would not matter.
The one thing ruled out is the leveraged version. (carried in places as ), the 2x daily long quantum product, decays — the worked example turns $1,000 into roughly $180 — and buying it is described as catching a falling guillotine. So what: past a single session, that product's path matters more to the outcome than its direction does, which is a different risk from the one the theme is being bought for.
4The most expensive large-cap listing on record is arriving without an S&P 500 seat
SpaceX comes to market at roughly $1.75trn and about 100 times sales, described on Excess Returns as the most expensive large-cap listing ever. S&P declined to waive its seasoning and profitability requirements, so index inclusion is not close, while Nasdaq and Russell will add it. Dave Nadig, on that episode, says to sit it out — no edge long or short — because a 30-day window of index front-running collides with a very small free float, and that combination sets the price for reasons that have nothing to do with the company. So what: whatever the first month does, it is an index-mechanics artefact, and it will not settle the valuation question underneath.
What the sources recommended
Twelve names on the buy side, all of them from two of the day's three videos, and they collapse into two theses plus a pair of favourite funds. The single sell is a product structure, not a company.
The buys, on quantum. first for purity; second and explicitly provisional at only weeks old; third with the caveat that most of it is not quantum; and the four pure-plays , , and held directly at about 2.5% each, with the securities-fraud suit against disclosed in the same breath. and are named as favourite chip funds rather than fresh initiations.
The buys, on metals. , and for leveraged exposure to the metal, plus gold and silver themselves on a months-long view. Every one of those comes from Felix, who says he has not bought yet, so the whole set reads as directional with the timing left open.
The sell. (also carried as ), the 2x daily long quantum ETF, ruled out on decay rather than on the theme — by the same person recommending the unleveraged quantum funds. Avoiding a structure is not a view on what sits inside it.
The AI read-through arrives midweek, and until then oil sets the tape
reports Wednesday after the close — the week's cleanest read on cloud and AI order flow, which is the exact question behind the leadership problem in the second section. follows Thursday after the close, with the same evening for housing, and on Wednesday and on Tuesday for the consumer. Today's calendar is minor: and . Until Wednesday the swing factor is the war — another strike near Hormuz moves oil, and oil is the input the gold argument in the first section runs on.
