The Daily Brief

Thirteen weeks. That is how far back you have to look to find sterling lower than the 1.32 it touched in yesterday's session, and it got there on a day when the dollar's own case was weakening. US hike expectations fell sharply and oil eased, both of which would normally hand sterling a lift. The fact that it slid instead tells you the pressure was coming from the UK side of the pair, and specifically from the one support sterling has leaned on all month: the expectation that the Bank of England will soon follow the Fed.

That expectation took its first real challenge of the week. Alan Taylor, one of the six who voted to hold this month, said the case for further increases is not compelling unless energy prices stay high for an extended period and start to feed broader inflation persistence. He described evidence of second-round effects as scant, pointed to pay growth consistent with the target, and questioned whether a single insurance hike is even practical without the market reading it as the start of a cycle. Set against pricing that still puts a November hike above 80% and roughly four increases by mid-2027, that is a reminder that the committee is more divided than the curve implies.

With oil now retreating, his argument got stronger on the same day he made it. The domestic data added texture rather than direction. Consumer credit jumped by nearly £2.5bn in August against expectations near £1.9bn, while mortgage approvals slipped to around 55,000. Rising unsecured borrowing alongside a cooling housing market reads more like households absorbing higher costs than a consumer boom, which is exactly the distinction Taylor is asking the committee to draw. The gilt market was more forgiving, with 10-year yields easing to around 5.37% from Monday's close near 5.43% as oil fell, although a 2036 gilt sale still cleared around 23 basis points above its previous auction.

This morning's final Q2 GDP print, expected at 0.4% quarter on quarter, is a test of how much growth cushion the hawks actually have. Sterling opens near 1.323, only about 0.2% above yesterday's 13-week low around 1.320 and roughly 0.7% above June's 1.3142, with the 1.325 quarter-end model estimate sitting just overhead. The balance of risk has shifted in a subtle but important way: the dollar's support is now anchored in long-end US yields rather than Fed expectations, so softer US data no longer does much for the pound, while any further erosion of BoE hike pricing removes the one prop that has kept this decline orderly. On that setup, a test of 1.3142 looks more likely than a return to 1.33, and it would probably take a firm GDP revision this morning and a soft PCE this afternoon, together, to change that.

This afternoon's August core PCE is now the most consequential number for the dollar this week, because the market it lands in has changed shape overnight. The probability of an October Fed hike has fallen to around 49%, from roughly 75% a day earlier, after New York Fed President John Williams said there is no need for urgency and that one more increase this year may be enough. Job openings slipped to 7.08 million against expectations of 7.23 million, and Conference Board consumer confidence dropped to 81.9, its weakest reading since 2014, as fuel costs and borrowing rates bite.

What did not change is the dollar. The index sits at 101.3 this morning, still near a two-month high after touching 101.5 yesterday. The reason is visible in the Treasury curve: two-year yields slipped just below 4.90% as hike odds fell, but the 10-year held near 5.23% and the 30-year reached its highest level since 2002. That is a steepening driven by investors demanding more to hold long-dated US debt, a mix of inflation risk, heavy supply and fiscal concern, and it supports the dollar through a channel that has little to do with the Fed. Add a haven bid that has not gone away, and the currency now has a floor that softer data alone does not remove.

That makes today's reaction function unusually asymmetric. Consensus looks for core PCE of 0.3% on the month and 3.3% on the year, with headline at 3.7%, alongside the final Q2 GDP estimate and ADP private payrolls. With October now a coin flip, a 0.4% core print would likely push hike odds straight back above 70% and add front-end support on top of the long-end support already in place. A 0.2% print would build on Williams, but it would need Friday's payrolls, where consensus sits near 90,000, to confirm it before the market fully abandons October.

At 101.3, the Dollar Index is about 2.6% above its four-month low near 98.7 and within 0.2% of yesterday's 101.5 high. Yesterday's session showed that a 25-point collapse in October hike odds could not dent it, which means the path lower now runs through the long end of the Treasury curve, not the Fed. The downside on a soft PCE therefore looks modest, while a hot print would stack both sources of support together, and that leaves the balance of risk tilted towards a break above 101.5 rather than a meaningful retreat.

The rand was handed an unusually helpful set of conditions in yesterday's session and chose not to trade on it. Oil fell sharply, US hike expectations halved and gold recovered from near $4,100 towards $4,180, a combination that in most months would have pulled USD/ZAR well below 16.30. Instead the rand closed near 16.40, barely changed on the day and still close to its weakest level since early August. When a currency ignores its usual tailwinds, the explanation is usually positioning, and the evidence points to investors who had already been cutting South African exposure and are not yet willing to rebuild it.

The bond market carried the clearest signal. Tuesday's auctions of the 2038, 2039 and 2042 bonds cleared at roughly 9.19%, 9.30% and 9.45%, each around 25 basis points above their previous sales, which means the Treasury had to pay materially more to place long-dated debt. The benchmark 10-year has eased to just below 9.0% this morning as oil fell, but a primary market that demands a quarter point more in a single auction cycle is a market being cautious about duration, and that caution feeds straight into how much support foreign flows are willing to give the currency.

Oil's slide also arrived too late to rescue October at the pump. The fuel adjustment is due to be announced on Monday and take effect on 7 October, and on the latest projections petrol is still tracking an increase of close to R3 a litre, with inland wholesale diesel heading above R33. Because the formula averages international prices and the exchange rate across the whole review window, a sharp fall in the last few days does little for this month. If Brent holds near current levels, the relief belongs to November, and that gap in timing is part of why the rand is not being rewarded for cheaper oil today.

This afternoon's August trade figures add the domestic test. Consensus expects the surplus to narrow from R20.1bn to around R3.5bn as the oil import bill swells, with the August budget balance due alongside. A narrow surplus is priced; a deficit would remove one of the few structural supports the rand still has at a moment when its cyclical ones are not being trusted.

At 16.39, the rand sits at the top of its 16.05 to 16.44 September range and just below last week's high near 16.50, with the 16.98 peak from early August marking the upper edge of the band it has held since. What has changed is the nature of the asymmetry. Yesterday proved that external good news is no longer enough on its own to pull the rand lower towards the 16.29 quarter-end estimate; that now needs domestic confirmation, starting with today's trade print. Without it, a drift through 16.50 remains more likely than a return to 16.29, even with oil below $100.

Middle Eastern crude exports have recovered to around 17.5 million barrels a day on a 10-day average, roughly 98% of their pre-war level, and that single figure did more to move oil yesterday than anything from the negotiating table. Saudi Arabia has restored around half of the capacity on its East-West pipeline, a steady flow of cargoes continues to slip through Hormuz, and Washington offered up to 40 million barrels from the Strategic Petroleum Reserve as the final tranche of its international commitment. Together they shifted the oil story from how long the disruption lasts to how much of it has already been worked around.

The price response needs reading carefully. WTI fell almost 4% to around $89, and Brent's December contract, now the benchmark, is trading near $96 after touching $100 early on Tuesday. The $105 to $107 quotes of earlier this week belonged to the expiring November contract, and the gap of roughly $10 between the two tells its own story: the market is paying a steep premium for barrels now and expects that tightness to ease. It is also why the retest of $108 that yesterday's brief judged more likely did not come. The prompt squeeze is real, but it is being priced as temporary.

The relief has not reached the bond market. The 30-year Treasury yield touched its highest level since 2002 yesterday and the 10-year holds near 5.23%, while UK 10-year gilts sit around 5.37% and Japan's 10-year near 3.1%. Wall Street closed slightly lower on those yields, although US futures are up around 0.2% this morning and the Nikkei has rallied more than 1%. Gold has recovered to around $4,180 after testing $4,100, helped by the softer US data. Today is also quarter-end, and after a quarter in which bonds sold off hard and energy outperformed, rebalancing flows into the 16:00 London fix could exaggerate moves in either direction.

Brent near $96 now sits about 8% below its $104.5 quarter-end model estimate and only just above the $95.5 area where it closed in early September. The supply data argues for a test below $95, but the depth of the backwardation and a reserve heading for its lowest level since 1982, with little appetite in Washington for a further drawdown, mean the policy buffer is close to spent. That leaves a modest downward bias in the near term against a larger upside tail: a fresh disruption in the Gulf would find fewer cushions left to absorb it than at any point since the war began.