The 56 basis points that used to be 175

RE: The 56 basis points that used to be 175

To whoever's still long duration on hope,

Nine nights. That's how long the strikes on Iran have run as of this morning, and Brent just punched through $90.97, up 3.3% before New York's even had coffee. The Strait of Hormuz — you know, the one everyone insisted was a "tail risk" back when WTI was sitting comfortably in the high $60s in early July — is now the single largest input into every rates model on the Street. Congratulations to whoever built a 2026 forecast that didn't have a war-premium line item. You're going to need a new tab.

Here's the number that should ruin your week: swaps traders have taken implied Fed easing for the rest of 2026 down to 56 basis points, from 60 on Friday. Read that twice. We spent the first half of this year debating whether Kevin Warsh would cut twice or three times once inflation "normalized." Now the conversation isn't about how many cuts — it's about whether the hold at 3.50%-3.75% survives the July 28-29 meeting at all, because nine of eighteen dots already wanted a hike before this latest escalation. Add a Brent print near $91 and a core PCE base effect from energy that's about to un-normalize itself, and you have a Fed that walked into the building citing "well-anchored expectations" and is now watching the anchor drag.

The dollar knows what's happening even if the equity desk hasn't fully priced it. The Bloomberg dollar index jumped as much as 0.8% overnight to its highest level since early February. That's not a flight-to-quality bid on US growth exceptionalism — that's a rates-differential trade, pure and simple, because a hawkish-forced Fed sitting on top of a global oil shock is about the only asset class left that pays you to wait. Everyone who spent Q2 shorting the dollar on "Fed cuts are coming, deficits are ugly, de-dollarization is structural" just got run over by the exact geopolitical risk that de-dollarization was supposed to protect against. Oil shocks don't weaken the dollar. They never have. They just make everyone else's import bill worse first.

Now overlay this onto the equity tape, because the timing here is almost insulting. We are nine trading days removed from the worst week for the Nasdaq Composite since the Q2 close, semiconductors leading the way down — SOXX shed more than 10% week-to-date, SMH close behind at nearly 9%, and IBM posted its worst single session in company history after telling clients that software budgets are getting reprioritized. That's the AI capex trade, the one thing propping up S&P 500 earnings growth estimates north of 20% for the quarter, cracking at the exact moment the macro backdrop stops cooperating. You cannot run a valuation model that assumes both unlimited hyperscaler capex AND a Fed that's on hold-to-hike with oil at $91. Pick one. The market, as of this morning, is starting to pick neither, which is its own kind of answer.

Let's be honest about what actually happened here, because the sell-side narrative will try to smooth this into "geopolitical volatility" language by Thursday. What happened is that an entire consensus trade — long duration, short dollar, overweight semis, betting on a Warsh pivot — got built on the assumption that the Hormuz situation was a controllable variable. It was never controllable. It was a live war with a ceasefire that had already broken once before this latest round of strikes, sitting on top of the one waterway that moves roughly a fifth of the planet's oil. Pricing that as background noise instead of the dominant risk factor was not a modeling error. It was wishful thinking wearing a modeling error's clothes.

So where does that leave the desk. Short duration remains the only rates trade that isn't fighting the tape, and I'd rather be early on that than early on cuts again — we've already been burned once flipping from 175bps of expected easing down to whatever fraction the swaps market wants to call "56" this week, and there's no floor yet visible under that number if the ninth night becomes a tenth and an eleventh. Gold hasn't caught the bid you'd expect from a proper inflation scare, which tells you real yields are still doing the heavy lifting against it — worth watching for a break, not worth chasing yet. And anyone still treating semiconductor weakness as a rotation story rather than a capex-versus-cost-of-capital story is going to keep getting surprised by IBM-style prints.

The uncomfortable truth nobody on this desk wants to write in a note that goes to clients: the Fed's credibility problem and the Strait of Hormuz's shipping-lane problem are now the same problem. Warsh cannot cut into a supply-side oil shock without looking like he's ignoring the inflation print staring him in the face, and he cannot hike without accelerating exactly the equity drawdown that a fragile capex cycle can't absorb. There is no clean exit here. There's just positioning for which mistake he makes first.

File this one under: told you the tail risk had a mouth.