The summer is not over; it is half over. There is a habit of treating the first week of September as the moment the year resumes, but that is a London habit, and it mistakes one clock for the whole clock face. The world's market has two centres of roughly equal weight, and they do not return together. The City came back this week, after the August bank holiday. Wall Street does not come back until next, after its own Labour Day — the serious American risk-setters, the ones who move size rather than mind it, are away until the eighth. So this weekend is not the end of the interval. It is the seam in the middle of it: one desk manned, the larger one still dark. Everything this publication wrote a week ago — the deferrals set down in July, the crowded diary waiting past the holiday, the assets that reward patience — remains true this weekend, because the event that would test it has not yet arrived. It arrives when the second centre opens.

That distinction is not pedantry; it is the correct reading of the week's tape. What looked, to a London eye, like the year resuming with a bang was in truth a thin-market week — one centre trading, the other absent, and liquidity correspondingly shallow. Moves made on that liquidity should be read for what they are: real prices, but struck in half a market, and therefore provisional until the other half returns to confirm or fade them. This is the discipline the seam demands. The evidenced layer — the prices — is genuine. The framing layer says: do not mistake a move made while Wall Street sleeps for the verdict Wall Street will deliver when it wakes.

The Strait ran to its extremes, and extremes are the medium these two parties bargain in. The week's headlines were violent. The administration framed a fresh economic push against Iran's trading partners as an "economic D-Day"; Central Command struck Iranian air-defence and mine-laying sites and, on the President's approval, adopted a "tanker for tanker" doctrine, striking Iranian tankers in retaliation rather than in blockade enforcement; by mid-week three merchant vessels had been hit in the corridor after the President urged tankers to "show some guts" and transit it. Tehran, for its part, said two tankers had struck its naval mines — a claim Washington disputes — and answered strike with strike. Read literally, this is escalation. Read in the context of everything that has preceded it, it is the negotiating style of two counterparties who both bargain by going to the brink and staying there until the other blinks. The President's method across the spring was pressure to the edge, then a claimed deal timed for effect; Iran's was defiance to the edge, then a concession extracted for a price. The theatre is loud precisely because the stakes are real and the deal is not yet struck. Brinkmanship of this kind is not the opposite of a settlement. It is the overture to one.

So the two-readings posture this publication has held on the Strait is unchanged, and this week strengthens rather than unsettles it. Relief is not resolution — the reopening, when it is claimed, will move prices before it moves a tanker, and whether it holds is the separate question. But the reopening thesis is not dead; it is being negotiated, in the only register these two use. The corridor's clock still runs to the Xi visit of the twenty-fourth: the interval in which a reset can be staged for maximum effect is measured in weeks, and a fortnight of loud strikes is exactly the leverage a claimed handshake would be built on. What the "economic D-Day" framing does change is what the summit carries. The threat of secondary sanctions falls hardest on Iran's largest trading partner, which is China itself — so the visit becomes a negotiation with the Strait on the table rather than a photo-opportunity with cheap oil as the backdrop. That is the one firm shift. The rest is theatre, and theatre is the tell that the parties are still at the table.

The City came back to a long end that had lost its footing. On Tuesday the thirty-year Gilt yield reached roughly 5.89 per cent intraday, its highest since 1998, the ten-year a nineteen-year high — and not in isolation: Japan's ten-year touched three per cent for the first time in a generation, the German Bund its highest since 2011, the French its highest since 2008, the long end of the developed world repricing together on the familiar collision of fiscal premium, energy-driven inflation and thin returning liquidity. And here the seam matters. The spike came into a half-open market, one financial centre still on holiday; by Friday it had substantially unwound, the ten-year Gilt closing back at 5.135 per cent and the thirty-year at 5.775, alongside a US Treasury thirty-year that itself settled at 5.245, below its own early-week high. A yield that spikes to a multi-decade high and gives most of it back within the week is behaving like a dislocation in a thin market, not a verdict. The reckoning this publication has long expected at the long end may well come; it did not come this week, in half a market. The true test is the Budget on the twenty-eighth of October, and the fuller market that will price it.

And the premium has a politics behind it, worth naming plainly because the market is pricing it. This is the first autumn Budget of a government at open war with the Treasury's own orthodoxy. Andy Burnham took office in July pledging to break four decades of what he calls trickle-down economics, built a "No 10 North" in Manchester as a counterweight to Whitehall, and set out an agenda — reindustrialisation, devolution, utilities and mass house-building — that no Gilt investor of the past generation has had to price. Against it stands his Chancellor: John Healey, who called the Treasury's culture "a dead hand on dynamic government," yet made fiscal discipline his first priority and pledged to meet the fiscal rules "in lockstep." The market has shown which it fears and which it trusts — sterling fell when Burnham floated fiscal-rule flexibility, and recovered when Healey, the "safe pair of hands," was named to hold them. Both readings are live and this publication holds both: Healey the anchor, the rules surviving the ambition; or a Prime Minister with a mandate to spend and a base built to circumvent the Treasury, a standing invitation to the bond market's oldest reflex — the one it exercised, without ceremony, in the autumn of 2022. The Gilt funds neither ambition nor restraint on faith; it prices the distance between them, and this week that distance widened. The arithmetic sharpens it: on the week's yields, Deutsche Bank's Sanjay Raja put the Chancellor's headroom against the fiscal rule at around £13.8bn, down from the roughly £26bn scored in the spring — almost all of it lost to debt interest, not new spending. A single house's estimate, but the direction is not in dispute, and it does not reverse when yields ease for a week. Nor does the Bank of England lean against it: markets now fully price a rate rise by year-end. A long end this high, easing only because half the market is absent, remains the price of borrowing the economy must service before it does anything else.

And the shape of the curve, read across its full length, says something the headline yield conceals. At Friday's close the Gilt curve rose the whole way out to a peak at the thirty-year, 5.775 per cent — and then turned down: the forty-year at 5.738, the fifty-year all the way back at 5.340, below even the twelve-year. The very back of the curve inverts. This is not the fiscal premium fading at the fifty-year mark; the Treasury's borrowing does not become sounder because it is longer. It is the signature of structural demand — the pension funds, liability-matchers and insurers who must own duration and have nowhere else to source it, bidding the ultra-long end down even as the twenty- to thirty-year sector carries the full weight of the fiscal worry. The curve is drawing, in its own hand, the distinction this publication has pressed all year: the fiscal premium is loudest in the belly-to-long, where the supply is heaviest and the buyer discretionary; the very long end is spoken for. For the investor whose instrument is the low-coupon, long-dated Gilt, that inversion is not a warning. It is the demand base, made visible.

That structural demand has a name, and it is the instrument this publication has returned to all year: the half-per-cent Treasury Gilt of 2061, TG61. The longest conventional Gilt in issue, carrying one of the lowest coupons in the register — and those two facts together make it the most convex bond the Debt Management Office has ever sold — and, for the same reason, the most volatile: convexity felt from the holder's side is simply the size of the swing, and TG61's price moves furthest, and more than proportionately, for any given move in yield. A half-per-cent coupon in a five-per-cent world prices it far below par — a Gilt is quoted as a percentage of its £100 nominal, and TG61 closed the week offered at 22.92 — under a quarter of the face value it will repay at maturity — and it is that discount which loads the convexity: if the demand-destruction thesis is right and long yields eventually fall, no conventional Gilt delivers more capital gain, and its gain is exempt for the higher-rate taxpayer who holds it direct. It is the sharpest expression of the thesis — and, this publication will not pretend otherwise, the instrument that has punished impatience hardest. Many bought it early, on the right thesis but before the fiscal premium had finished repricing, and have sat through a drawdown to 1998 yields. Being early is the cost of being positioned: the property that hurt on the way up is the identical one that pays on the way down. It is a reason to hold, not to fold. The instrument was never the error; the calendar was, and the calendar corrects. But it is not an instrument to hold carelessly: the same volatility that will reward the thesis will test the nerve holding it, and it wants sizing, patience and a clear head about the swings to come — which is to say it suits some investors and not others, a matter for each and their own adviser.

The overseas third of the register — around a third of all gilts on the latest official reckoning, and on the Debt Management Office's own distribution data now the largest single sector, having drawn clear of a Bank of England holding that quantitative tightening has run down to some eighteen per cent, from a peak near thirty-four in 2022 — extended the summer its patience through August. Tuesday was the sound of it testing the exit; Friday was the sound of it deciding, for one more week, to wait. Which way it settles is a question for the fuller market, and for the Budget it is pricing toward.