1The war premium is being paid in the bond market, not the stock market
The weekend's escalation around Iran barely touched equities and moved the long end a long way. Wire coverage from Reuters, Bloomberg and CNBC carried drone attacks on the UAE, Saudi intercepts of launches out of Iraqi airspace, Bessent calling on the G7 to follow US sanctions on Iran, and a running bill to global companies already above $25bn — with Capital Economics' extreme case putting crude at $150/bbl through 2027. The equity reaction was a shrug: S&P futures −0.27%, Dow futures −0.58%, Nasdaq futures −0.04%, and the VIX up 3.15% to 19.01, which is elevated rather than alarmed.
The long bond took it more seriously. The 30-year Treasury yield topped 5.1%, its highest in nearly a year, on what the wires framed as flaring, oil-driven inflation rather than a growth signal. April producer prices running at 6% year on year sit underneath that move, and a CNBC trader survey put the odds of stagflation by the end of 2026 near 40%.
That is the awkward pairing: inflation arriving through the energy price while the growth impulse softens. The Fed's May inflation forecast deteriorated in the same week, which narrows the room for cuts, and the argument about cutting has become a personnel question as well as a data one — Governor Miran resigned and threw his support behind Warsh as next chair, in what CNBC described as a family fight over rates.
Goldman Sachs, quoted in Yahoo Finance's markets coverage, takes the other side of the equity shrug and says Q1 earnings strength fundamentally supports the index at these highs. Both can be observed at once, and they resolve on the oil price: earnings can justify the level while the discount rate keeps moving against it. If the crude bid persists, the long end goes on repricing whatever the profits do.
2The AI build-out is flattering profits through the depreciation schedule, and that flattery expires
Jeremy Grantham of GMO, in a clip on Excess Returns' weekly roundup, makes an accounting argument rather than a valuation one. A capex boom lifts reported profits in the near term because the buyers of the equipment do not depreciate it immediately: the revenue lands with the seller now, the cost lands with the buyer across years. Aggregate profits therefore look better while the spending accelerates. The second step is the one that bites — once the spend is table stakes rather than an edge, it becomes the cost of doing business and stops lifting aggregate margins meaningfully above normal. Note the level he is arguing at. This is a claim about margins in aggregate, which can be right while individual winners keep compounding.
The same episode surrounds it with two other structural readings. Andy Constant, formerly of Bridgewater and Brevan Howard, frames bubbles as untimeable but recognisable, requiring "fertile soil" — cheap money, a narrative, and a new technology — and says the same pattern has recurred across his career. A private-credit segment, credited on the show only to Mark, argues that middle-market dependence on private-credit financing is a real knock-on risk if regulatory intervention combines with redemption gating and the fraud cases already surfacing, while explicitly sizing it below 2008: problems in private credit, but probably not enough to bring down the financial system.
Two things temper all of it. The four voices in the hour — Grantham, Constant, Ed Chancellor and Mark — are institutional bubble-watchers assembled by one show, so the agreement between them is an editorial selection rather than four independent readings. And Constant's own framing marks the limit: recognisable, not timeable. A claim about aggregate margins resolves in aggregate data, not in any one quarter, which is why the nearest thing the tape offered was atmospheric — Bloomberg and Yahoo both flagged "borderline mania" and concentration risk in the chip-sector leadership of the S&P rally, with off 4.42% and off 6.18% intraday on Friday.
3An executive order created an FDA priority category, and every way to own it is pre-revenue
Felix, on Felix & Friends, builds a single-name thesis on a policy change. An 18 April executive order fast-tracked psychedelic-assisted therapy into a national-priority FDA category with $50m of research funding attached; three small biotechs jumped 20–50% the following day, and the news cycle then moved on. His framing is that the setup now sits in "the boring middle" ahead of a possible catalyst-driven re-rate.
The precedent he leans on is Johnson & Johnson's Spravato — esketamine, approved in 2019, running at roughly $2bn of annual revenue at $590 a dose even after a cost-effectiveness watchdog declined to back coverage. He reads that as proof both that the patient population pays and that insurers cover, against a population he puts at 280m with serious depression globally and 85m treatment-resistant. He holds no position in it: he sold at +40% and cites it purely as a case study.
The structural problem is his own. He screens candidates on three filters — gross margin near 70%, patent life of fifteen years or more, and a drug already selling — and all three names fail the third. That is what makes this a category trading on approval risk rather than on execution: with no revenue anywhere in the group, an FDA decision window in late 2026 or early 2027 is the whole of the thesis. He names the ways it breaks — rejection, dilutive raises, demand failure, political turnover around RFK Jr., and lobbying to protect Spravato — and he hedges the upside in the same breath as he states it, with the 10x figure qualified as "what could happen if everything goes sort of half right" and "not a forecast." Read the video title, "Trump Just Opened Up A $400 BILLION Market (Get In Now)?", against those hedges: the packaging is promotional and the risk language underneath it is unusually explicit.
What the sources recommended
Three buys came out of two videos, all three from one person, and they are one trade rather than three — capped by the person making it at 1–3% of a portfolio and labelled "calculated speculation, not investment."
(Compass Pathways) is the primary pick: phase 3 complete on COMP360, a single dose with 26-week durability, and an FDA approval target of late 2026 or early 2027 that would be the first approval of a classic psychedelic. (ATAI Beckley) is the platform version — five candidates after December's Beckley merger, Peter Thiel among the backers — but it carries partial exposure through a licence, so it is not an independent second bet. (GH Research) is the technology bet: one compound, 5-MeO-DMT, with a roughly 30-minute session against six to eight hours for the alternatives, which is an argument about clinic throughput rather than about efficacy.
Wednesday's chip print is the first hard test of the margin argument
reports Wednesday after the close, with consensus near $1.79 of EPS on $80.2bn of revenue, and it lands into a tape that has just flagged chip-sector concentration as the rally's main structural risk. Before that, China internet reports this morning — against a $32.0bn revenue estimate, plus and . Retail and housing follow on Tuesday with ($42.8bn estimated) and ($24.9bn), and reports Thursday morning against $176bn, which is the cleanest read available this week on whether energy-led inflation is reaching the consumer — the mechanism the first section rests on.
The geopolitical calendar runs alongside it: Putin and Xi meet in China on 19–20 May, Axios reported a Trump security-advisor meeting for Tuesday, and the Fed's May inflation refresh is the macro item behind the rate argument. If crude stays bid through those meetings, the yield printed alongside Wednesday's result does more to the multiple than the result does.
