2026-08-26 → 2026-08-28

The monthly headline is expected to swing from -0.1% to +0.9%, a full point of acceleration in a single print. Most of that arithmetic is base and energy: June's soft reading unwinds and fuel turns from a drag into a contributor, so the eye-catching monthly number overstates any genuine change in inflation's trend.

The board it feeds is not a comfortable one. The Reserve Bank held the cash rate at 4.35% this month, disclosed that it discussed a hike rather than a cut, and continues to describe inflation as too high; the next decision lands on 29 September. A monthly indicator this noisy is exactly the kind of release a committee already split can read either way.

Take the m/m as texture, not signal. A +0.9% that is all fuel and base effects changes nothing for a board that looks through energy; the version that matters is a hot month with the trimmed mean firm alongside it, which would say momentum is rebuilding into a meeting the RBA has already admitted was a hike debate.

For the Australian dollar the asymmetry runs one way this week. With the board leaning hawkish, a firm composite lifts AUD and the front of the curve; a soft print simply validates the pause and leaves the currency trading the US dollar and Friday's Jackson Hole tape.

The annual rate is the cleaner read, and consensus has it decelerating to 3.3% from 3.8%. That is genuine disinflation, the latest step down this year, but it still sits above the top of the 2-3% target band, which is the whole of the RBA's problem in one number.

This is the last inflation read before 29 September, and a board that discussed hiking needs the trend to keep bending. At 3.3% the disinflation story holds and the hold is safe; at 3.5% or above the deceleration stalls above target and the hike talk stops being rhetorical.

The uncomfortable case is a headline that falls while the core stays sticky, because the RBA steers by the trimmed mean, not the top line. A 3.3% headline that masks firm underlying prices is the print that keeps 4.35% from being the peak the market assumes it is.

The trimmed mean is the measure the Reserve Bank actually targets, and it is expected unchanged at 0.3% on the month. Compounded, 0.3% a month runs near a 3.7% annualised pace, still above the band and the reason the board keeps a hike on the table even as the headline falls.

Unchanged is not the same as benign here. A stable 0.3% core says underlying inflation has stopped decelerating, which is precisely what a hawkish board does not want to see three weeks before it meets. The comfortable outcome is 0.2%; the one that moves the September conversation is 0.4%.

Read the three releases as one picture. A falling annual headline with an unchanged core is disinflation on the surface and stubbornness underneath, enough to justify the hold and not enough to close the hike debate the RBA has already said it is having.

This is the inflation number the Fed writes its mandate around, and it lands two days before its chair speaks at Jackson Hole. Consensus has core PCE firming to 0.2% from 0.1%, a modest re-acceleration that would still be consistent with the disinflation July's CPI began to show.

The context is a committee that held at 3.50-3.75% in July over three dissents that wanted a hike, and a market that has since marked a September move down to a minority probability. Core PCE is the gauge that either confirms that dovish drift or interrupts it, and it is the last reading of it before the 16 September decision.

The scenarios split cleanly. An in-line 0.2%, or a 0.1% surprise, lets the tame-inflation narrative carry into Warsh's keynote and keeps the front end where it is. A 0.3% print, especially with the firmer service components that a hot core PPI flagged feeding through, revives the September-hike case and hands Warsh a hawkish backdrop he did not necessarily choose.

Because it prints the same morning as the GDP revision and two days before Jackson Hole, the PCE is the week's cleanest standalone catalyst. Everything after it is interpretation; this is data.

The second estimate of Q2 growth is expected to confirm the advance reading of 1.5% annualised, down from 2.1% in the first quarter. A revision that holds at 1.5% says the economy is slowing but not stalling, the soft-landing pace that neither forces cuts nor justifies hikes.

The headline is the least interesting part; the composition is where the read lives. A downgrade driven by weaker consumption would strengthen the doves' case that restrictive policy is finally biting; an upgrade led by final demand argues the economy is absorbing 3.50-3.75% with room to spare, which is the hawks' entire premise.

On its own this rarely moves the front end, and it shares the morning with a core PCE that will command the desk's attention. Treat it as a quiet cross-check on the growth side of the mandate rather than a market event in its own right.

June GDP is the last growth read before the Bank of Canada decides on 2 September, and consensus expects a +0.2% month after +0.3%. That would extend the picture of an economy that resumed growth through the second quarter after a soft patch earlier in the year.

The Bank has held the policy rate at 2.25% since July and looks, on core inflation near target and a steadying economy, like the closest thing the G10 has to a finished cycle. Another hold on 2 September is close to fully priced; this print would have to miss badly to disturb it.

For the loonie the release is low-voltage. An in-line number leaves CAD trading US rates and the Jackson Hole tape, as it has most of the summer; only an outright contraction, or a surprise well above 0.2%, gives the currency a domestic story ahead of the Bank's decision.

This is the week, and everything else is prologue. Kevin Warsh delivers his first Jackson Hole keynote as Fed chair on Friday, around 10am in Wyoming, three weeks before the 16 September meeting and at the last unscripted moment before the committee goes quiet. The symposium's theme is financial innovation and payments, not the rate path, which is itself a tell about how much the chair intends to pre-commit.

Warsh arrives with a hawkish reputation and a communication style built on saying as little as the moment allows. The market wants a September signal; the theme gives him cover to withhold one. That tension is the trade: the keynote can move rates as much by what it declines to rule out as by what it says.

The scenarios are asymmetric around a crowded dovish lean. A keynote that leans on unfinished inflation, or that pointedly keeps a September hike alive, sells the front end and bids the dollar into a market positioned the other way. A keynote that stays on theme and offers no rate signal disappoints the hype and lets this week's soft data set the tone by default.

Whatever the July FOMC minutes revealed about the dissents, Warsh now owns the message. This is the last high-profile Fed communication before the pre-meeting blackout, and the first chance to hear the September framing in the chair's own voice rather than the committee's record.

At the same 10am hour as Warsh, the BLS publishes the preliminary estimate of its annual benchmark revision, the recount that anchors the establishment survey to the QCEW tax records covering nearly every job through March 2026. It measures one thing: how many payroll jobs the monthly headlines over-counted.

The number matters because the labour narrative is already softening. The July report showed payrolls falling 23K, and a large downward benchmark would say the deterioration started earlier and ran deeper than the monthly prints admitted. Last year's preliminary revision came in at -911K; another figure of that magnitude reframes the whole cycle as weaker than reported.

The scenarios feed straight into September. A deep negative revision strengthens the dovish case that policy is already too tight, just as Warsh takes the stage; a modest one takes air out of that argument and leaves the labour market looking sturdier than the bears claim. Backward-looking as it is, it colours the ground Warsh is speaking on.

The Kansas City Fed's symposium runs 27-29 August at Jackson Lake Lodge under the banner of financial innovation, payments and policy. It is the calendar's annual set-piece for central-bank signalling, and this year it doubles as the debut of a new Fed chair, which is why a conference nominally about payments has become the market's whole week.

The venue's history is the reason it is priced. Jackson Hole is where policy turns have traditionally been trailed, and a room full of global central bankers guarantees that any shift in tone travels. The formal agenda points away from the rate path; the Q&A, the corridor remarks and Warsh's own emphasis are where a signal would actually leak.

For everything that reports before it, core PCE and GDP on Wednesday and the benchmark revision on Friday morning, the symposium is the lens. Each release will be traded less for itself than for how it stacks the deck for what Warsh, and the other speakers, choose to say about the autumn.