The G10 tightens and redraws the hierarchy against the dollar

The DXY gained 0.09% on the week, while USD/CAD lost 1.17%, AUD/USD rose 0.91% and GBP/USD 0.25%. The greenback rose only against the euro (EUR/USD −0.53%), whose weight in the index is enough to offset the rest. Meanwhile, the US 2-year stayed between 4.65% and 4.84% throughout October. The week confirms the thesis of the FX Daily of 21 October, a dollar that is "losing its monopoly on tightening", and invalidates our bearish bias of 19 October on the loonie.

The BoC saw the largest repricing. The Canadian September CPI (3.1% against 3.2% expected, 2.2% excluding gasoline) did not justify such a move; August retail sales (+1.2% on the month) and the preliminary September estimate (+0.7%) triggered it. Swaps now give a 78 to 80% chance of a move from 2.25% to 2.50%, and 19 of the 34 economists polled by Reuters expect it. USD/CAD has fallen in every session since 12 October, from 1.4262 to 1.3995, posting its largest weekly fall since April. BBH regards this pricing as excessive. We share the reservation about what follows: Wednesday's hike is priced in, and the Monetary Policy Report will have to justify a second move.

The RBA is on the same path. After Hauser's speech on 21 October, which judged the inflation outlook worse than expected, the probability of a hike to 4.85% rose from around 20% to 50-60%, depending on the source. The AUD rose despite the risk-off of 22 October, helped at the margin by Chinese third-quarter GDP at 4.5% year on year, above the 4.4% expected. Our bearish bias of 16 October on the AUD was wrong: the rate differential weighed more than equity beta. The quarterly CPI on 28 October will show whether this pricing still has room to run.

The Fed will not move before December. In the Reuters poll conducted from 16 to 20 October, 94 of 100 economists expect no change on Wednesday and 100 of 100 a hike in December, against fewer than half the previous month. Until then, the BoC, the RBNZ, the RBA and the BoE can all hike, the last on 5 November: each decision narrows the greenback's rate advantage against these four currencies.

On Wednesday the US curve priced a growth scare that the data did not confirm. After the sell-off in AI-related stocks that followed Meta's results, the 2-year lost 16 bp, to 4.65%, and the 10-year 17 bp, to 4.96%. The next day, S&P Global's flash composite PMI came in at 57.6, its second-highest level since March 2022, and manufacturing at 56.0, its highest since May 2022. Williamson, at S&P Global, sees it as consistent with annualised growth close to 5% in the fourth quarter, with employment rising at its second-fastest pace in more than four years. The 2-year regained 4 bp, to 4.69%.

For the Fed, the most useful figure in the survey concerns selling prices, which are rising at the fastest pace since August 2022. The final October Michigan survey points the same way: sentiment at 48.0, a five-month low, and one-year inflation expectations of 4.7% according to early reports, a figure we have not been able to verify against the university's document. Consumers worried about prices and companies raising them suggest second-round effects in the making. The Fed will have to gauge them in December, which Warsh, together with October, counts among the moments of "crystallisation".

The thesis of the FX Daily of 23 October, a dollar "decoupling from its rates", comes out half validated. The first half holds: the greenback did not follow the fall in the 2-year. The second, a curve pricing a lower path beyond December, lacks support, since the PMIs point the other way. The 2-10-year slope flattened to +31 bp, against +35 bp on 16 October; we see it as duration buying rather than expected rate cuts.

The September PCE, published on Thursday, the day after the FOMC, will settle it. The consensus expects 3.6% year on year for the headline, after 3.42% in August, and 3.1% for the core, after 3.01%. Headline PCE would then stay above 3% for the seventh consecutive month.

September HICP was confirmed at 3.8%, the highest since September 2023, but the ECB remains priced for a pause on 29 October, at 2.50%, then for a final hike on 17 December. The October flash composite PMI fell to 51.9 against 52.7 expected; France slipped back into contraction, at 48.4, while Germany held at 52.8. The gap with the US composite reaches 5.7 points. With an ECB on hold until December and a BoC, an RBA and a BoE that can hike this week or next, traders are selling the euro to fund positions in those currencies: EUR/GBP fell in every session, from 0.8438 to 0.8373.

The French premium has added to this rate disadvantage. An aggregator puts the 10-year OAT-Bund spread at around 170 bp on 23 October, against 135 bp on the 15th and 95 bp in mid-September; no primary source corroborates the week's jump, and we treat it as an order of magnitude. On Friday evening, after our data cut-off, Moody's kept France's rating at Aa3 with a negative outlook, whereas most analysts expected a downgrade to A1. The agency makes this affirmation conditional on a 2027 budget that reduces the deficit, projected at 5.4% of GDP in 2026 and targeted at 5.0% in 2027.

The reprieve removes a tail risk, but the budget has yet to be voted. The formal vote on the revenue section of the 2027 finance bill, postponed to Tuesday 27 October with more than 2,000 amendments outstanding, will be the first test: a rejection would reopen the spread despite Moody's, facing a government without an absolute majority. Against the dollar, positioning limits the downside, since the net speculative short on the euro reached 138,000 contracts as of 13 October, a ten-year high. The euro's weakness should therefore show first against sterling, the loonie and the AUD.

The Brent futures contract settled on Friday at $116.72, around 6% above the previous week and at its highest since late April. The rise reflects an accumulation of threats: Houthi strikes on Saudi territory, the departure of non-essential US personnel from Saudi Arabia, the evacuation of Aramco's headquarters in Dhahran, and a statement by Iran's first vice-president Aref on the need to impose Iranian control over Hormuz. The Revolutionary Guards claim seven incidents involving tankers; UKMTO has confirmed only two. Yet Bloomberg reported on 19 and 20 October an increase in physical flows leaving the Gulf. Oil buyers are therefore paying a premium for a resumption of US strikes on Iran, which the press sees as increasingly likely, more than for any observed loss of supply.

The IEA and OPEC describe two different markets. The IEA expects global demand to fall by 2.6 mb/d on average in 2026, with supply still below demand in September and European diesel stocks not rebuilt before the end of 2027. OPEC, on the contrary, sees demand growing by 0.2 mb/d and a balanced market. The IMF, in its October WEO, projects global growth of 3.1% in 2026 and assumes Hormuz reopens in March 2027. With a gap of nearly 3 mb/d between the IEA and OPEC on demand, neither provides an anchor, and prices will follow the headlines on Iran.

On the currency side, crude did not set the week's ranking. The CAD gained the most, but we attribute its rise first to the BoC repricing. Among energy importers, the euro, already weakened by France, paid for the rise in oil, while the yen rose in spite of it.