2026-09-21 → 2026-09-25

Governor Michele Bullock appears at an 'In conversation' event at the Fullerton Hotel in Sydney at 1.10pm AEST - 05:10 CEST - with a question-and-answer session, open to the media, per the RBA's own media programme. No forecast attaches to a speaking engagement, but this one has a date behind it: the RBA board decides on 29 September, seven days later, and this is the Governor's last scheduled public appearance before it.

The cash rate is 4.35%, held in August. CPI rose 3.5% over the twelve months to July, down from 3.8% in June, while trimmed mean inflation was unchanged at 3.6% - both above the 2-3% target band. On 18 September, before a parliamentary committee, Bullock said that although growth in the Australian economy is slowing, some upside risks to inflation appear to be materialising, naming the Middle East conflict, the AI boom and extreme weather as pressures on energy, food and technology prices. Those remarks were enough to move Westpac's call from November to September. NAB already looked for a 25bp hike to 4.60% this month; ANZ and CBA remain at November.

The hawkish path from here is that she repeats or sharpens the 18 September framing - that the board must decide whether 4.35% is sufficient to cool prices - and Thursday's labour force report becomes a formality rather than a hurdle. In that case a September move goes towards fully priced, the front of the Australian curve cheapens, and the Australian dollar rallies against a US dollar that has already had its hawkish news. The dovish path is that she leans on the slowing growth half of her own sentence, points to July's employment contraction, and tells the audience the board would rather see the Q3 CPI in late October before moving. September pricing unwinds, the Australian dollar gives back ground, and November becomes the live date again.

What makes this worth watching is that Australia is the clean test of the question every central bank is now facing: whether an energy-driven price shock justifies tightening into slowing domestic demand. The Fed answered yes last Wednesday, the ECB answered yes on 10 September, the Bank of England answered not yet by 6-3, and the Bank of Japan answered yes by 7-2. Bullock speaks with all four answers already on the table and with her own labour market data arriving thirty-six hours later. In isolation a conversational Q&A is not a policy signal. Seven days from a live decision, with the market split between two meeting dates and one major forecaster having switched sides on her last set of remarks, it is the closest thing this week has to one.

The ABS Labour Force release for the August 2026 reference month, scheduled by the Bureau for 24 September. Consensus looks for +20.9K against July's -15,800, a contraction that took total employment to 14,807,200 in seasonally adjusted terms. The internals of that July fall matter more than its sign: full-time employment rose 16,300 while part-time fell 32,200, so the headline decline was entirely a part-time story with the better-paid, higher-hours category still adding.

The RBA's own characterisation is that various indicators continue to suggest labour market conditions remain close to, but a little tighter than, full employment. Against that description a single negative month is noise. Two consecutive negative months would be the first evidence that the slowdown Bullock referenced on 18 September has reached employment, and it would change the complexion of the 29 September decision considerably more than any inflation print could at this distance.

A strong number - at or above the +20.9K consensus, and particularly one led by full-time work - confirms July as a blip and hands the hawks their argument: a labour market still adding jobs while headline and trimmed mean inflation sit above the band is a labour market that can absorb 4.60%. Expect the front of the Australian curve to cheapen and the Australian dollar to firm. A soft number, near zero or negative again, does the opposite and does it forcefully, because it would be the second in a row; the board would be tightening into a visibly deteriorating labour market on the strength of an oil price it does not control.

Two caveats belong on this print. The forecast implies a 36.7K swing from the previous month, which is large relative to the month-to-month noise in this survey, and the ABS is mid-way through a programme of modernising the Labour Force Survey - reason enough to treat any single seasonally adjusted headline with care. Read the full-time/part-time split and the participation rate before reading the headline. This is the last significant Australian data before the board meets.

Released at the same moment as the employment change, from the same survey. Consensus is 4.5%, unchanged from July, when the rate rose to 4.5% from 4.4% in June as the number of unemployed people increased by 4,000 against a 16,000 fall in employment.

One tenth of a percentage point is not a deterioration, but it is a direction, and it is the second piece of evidence in the same direction after the employment contraction that accompanied it. A stable 4.5% remains consistent with the RBA's description of conditions as close to but a little tighter than full employment. A print at 4.6% would not be, and it would arrive five days before a decision at which a meaningful minority of forecasters expect a hike.

The useful information is in how this number agrees or disagrees with the one published alongside it. Employment strong and the rate rising means participation is climbing - a supply-side story, benign for wage pressure, and an argument for patience. Employment weak and the rate flat or falling means people are leaving the workforce rather than joining the queue, which tightens the labour market on paper while the economy softens underneath. That second combination is the least comfortable outcome for a board that has to justify its September choice in public either way.

Taken alone, a forecast of no change is the least interesting line on this week's calendar, and it deserves to be said plainly rather than dressed up. Its value is as the denominator for Thursday's headline. At 4.5% the rate is a tenth above June and still inside what the RBA calls a little tight; unchanged is a green light for whichever way the inflation argument points, and 4.6% is the one outcome that would take a September hike off the table on its own.

The SNB sets policy four times a year, so this is one of only four opportunities in 2026 for the Bank to change its stance or its published view. The assessment carries the conditional inflation forecast and is released at 09:30 CET alongside the rate decision, with the press conference following at 10:00. The forecast, not the rate, is the reason to pay attention.

The June assessment put average annual inflation at 0.6% for 2026, 0.6% for 2027 and 0.7% for 2028, with growth of around 1% expected for 2026. August CPI has since overtaken the near end of that path: consumer prices rose 0.8% year on year, against expectations of 0.5% and after 0.4% in July, the fastest rise since August 2024. Core inflation accelerated to 0.4% from 0.3%, its first increase this year. Higher rents were the largest contributor, with gas, diesel, heating oil and inpatient hospital services also rising, and a weaker franc lifting the price of imported goods.

So the question is whether the conditional forecast is revised up, and across how much of the horizon. At 0.8% inflation is still comfortably inside the 0-2% target band - less than half the ceiling - so an upward revision would not in itself be a tightening signal. It would be a statement about whether the Bank reads the energy and import pass-through as a one-off level shift or as something with persistence. Nomura expects a hold, noting stronger GDP growth, the rise in inflation and the franc's depreciation, and sees no hikes before 2028 at the earliest.

A forecast revised up across the whole horizon, with the 2028 figure drifting towards 1%, is a hawkish document even with the policy rate at zero, because it shortens the distance to a first move and the franc should firm on it. A revision confined to 2026 and framed as energy passing through leaves the 2028 anchor intact and is neutral. Of Thursday's three Swiss entries this is the only one carrying genuinely new information; the rate is priced for no change and the press conference is where the document gets explained.

The headline number, expected at 0.00%. The rate has been at zero since at least the March 2026 assessment and was confirmed there again in June, so a hold would be the third consecutive confirmation rather than a decision.

Switzerland is the outlier in this brief and the contrast is worth stating. Every other central bank here has either just tightened - the Fed to 3.75-4.00%, the Bank of Japan to 1.25%, the ECB to a 2.50% deposit rate on 10 September - or is arguing about it in public, as the Bank of England's 6-3 vote showed on Thursday. The SNB sits at zero with inflation at a two-year high that is nonetheless 0.8%. The same global energy shock that is forcing hikes elsewhere shows up in Switzerland as a number that has merely returned to normal.

Because a hold is the overwhelming expectation, delivering it moves nothing. The tail risk sits in the currency rather than the rate. The Bank retains an increased willingness to intervene in the foreign exchange market to counter rapid and excessive appreciation of the franc, and Nomura flags a possible shift back to standard intervention language. The franc has now weakened enough to be contributing to imported inflation, which inverts the problem the SNB has spent years managing: for the first time in a long while its currency discomfort could plausibly run in the other direction.

In isolation this release matters almost not at all, and that is the honest assessment rather than a failure to find an angle. A zero rate expected to remain at zero is not a market event. What it does is set the terms for the assessment published with it and the press conference thirty minutes later, where the intervention language and any forecast revision are actually explained. The instruction for the week is to trade the document and the conference, not the number.

Thirty minutes after the assessment, the Bank presents its decision and takes questions. With the policy rate a foregone conclusion at 0.00%, this is where whatever Swiss risk the week contains actually sits, and it is the only scheduled opportunity before December to hear the reasoning rather than read it.

The first thing to listen for is the foreign exchange language. The June formulation retained an increased willingness to intervene against rapid and excessive appreciation of the franc. A currency that has depreciated enough to push up import prices makes that sentence awkward to repeat verbatim. Whether it is retained, softened, or replaced with the older standard formulation - which is what Nomura suggests may happen - is a live question, and the franc will trade on the answer more readily than on any number in the document.

The second is the zero floor. At 0.00% the Bank has one conventional direction remaining and no stated intention of using it soon. Questions will press on what conditions would produce a first hike and on whether negative rates are genuinely off the table should the franc turn sharply. The answers define the market's read of the SNB reaction function for the next two quarters, which is worth more than a single decision given the Bank meets only four times a year.

The hawkish outcome is an upward forecast revision defended as broad-based rather than energy-driven, intervention language softened, and a first hike discussed as a 2027 rather than a 2028 question: the franc firms and the front of the Swiss curve cheapens from a very low base. The dovish outcome is the 0.8% print attributed to rents and petroleum, the intervention stance reaffirmed, and no hike horizon offered at all. Given how little is priced in either direction, the asymmetry favours the hawkish tail for the simple reason that none of it is in the price.

A caveat before anything else. The ForexFactory calendar carries this slot at 11:15 CEST on Friday, but the Bank of England's own upcoming-events page does not list an engagement for the Governor on 25 September at the time of writing. That page shows Sasha Mills on Tuesday, Ruth Smith on Wednesday, and Swati Dhingra, Sarah Breeden and Clare Lombardelli on Thursday, and nothing at all on Friday. Treat the entry as provisional. If it goes ahead, it is the Governor's first scheduled public appearance since the decision.

That decision was a hold at 3.75% by a 6-3 majority on 17 September, with Megan Greene, Catherine Mann and Huw Pill voting to raise Bank Rate to 4.00%. The Monetary Policy Summary said CPI inflation increased to 3.1% in August and is likely to rise further over coming quarters, and that on current energy prices inflation was expected to reach around 3.75% in Q4 2026 and slightly above 4% in Q1 2027. August's rise was driven by transport, where annual inflation went to 4.6% from 3.6%: petrol rose 9.1 pence a litre to 161.3 pence and diesel 14.2 pence to 181.8 pence, taking motor fuel inflation to 23.0% from 15.5%.

A 6-3 hold, with the Governor in the majority and the Committee's own forecast showing inflation above 4% next spring, is an unstable position. It needs one further defection to become a hike, and the Committee has publicly told the market to expect inflation to keep rising for two more quarters before it turns. The stated reason for leaning against is that the risk of second-round effects grows the longer higher energy prices persist or are more volatile - which makes the decision a function of the oil price, and nobody on the Committee sets that. Brent was around $95.52 and WTI around $91.02 in mid-September.

If Bailey defends the hold as durable - energy as a level shift, second-round effects absent so far, the 2% target reachable without moving - the gilt curve richens further and November pricing softens. If he emphasises instead how narrow the vote was and how far the Committee is prepared to lean against persistence, the market starts pricing the defection and the front end cheapens. The gilt is already the highest-yielding 10-year in this brief by a wide margin, and how much of a hike is embedded in that number is precisely what Friday could clarify - if it happens.