US inflation surprises higher, UK GDP posts another solid monthly gain, and Australia’s labor market sends a yellow light as attention shifts to inflation and the RBA’s next policy move remains open.
Four-month high
Three-month high
Driven by gasoline prices
Stronger than expected
Growth driven by production expansion
Strong
Balanced full-time rising to the same tune, but still worrying.
More rises possible
Jumpy
The inflation data this week produced a trifecta of upside surprises, starting with the consumer price index on Wednesday, producer price index on Thursday, and (to a much lesser extent) import prices on Friday. This briefly pushed rate hike expectations for the October 28 meeting to almost 50% before turbulence in financial markets helped bring that down to 15% Friday morning.
Consumer prices rose 0.6% MoM in September, driven heavily by a 9.1% MoM jump in gasoline prices that lifted the energy complex by 5.2% MoM. By contrast, food prices were unchanged, which is good news but unlikely to last. Overall goods prices rose a brisk 1.2% (double the August pace) and are now up 4.6% YoY. Services prices rose 0.4% MoM and are up 3.2% YoY. Within services, airfares posted another sizable gain and are 22.4% higher than a year earlier; this is where the energy price shock is unmistakable.
The report’s most troubling development was a 0.4% monthly increase in shelter costs. Apart from an anomalous reading in April, this was the largest increase since August 2024 and the clearest blemish in the report. There were some mitigating details. Hotel rates inflated the category, while rent of primary residence rose by a more modest 0.2%. Even so, a meaningful slowdown next month will be needed to ease concerns that the shelter disinflation process has stalled or ended.
All these lifted the headline inflation rate three tenths to 3.7% YoY and the core rate two tenths to 2.6% YoY. At higher precision, the two numbers were 3.738% and 2.564%, respectively.
Producer prices surged 1.3% MoM, lifting the YoY comparison to 6.3%, the highest reading since December 2022. Goods PPI inflation accelerated 2.2 percentage points (ppt) to 9.8% YoY, and services PPI inflation accelerated 0.2 (ppt) to 4.9% YoY. Both core metrics (ex-food and energy; and ex-food, energy, and trade services) advanced two tenths to 5.0% YoY.
Import prices also rose a little faster than expected in September, but this was to a large extent attenuated by a downward revision to the prior month, so it should really be seen as a wash in the bigger picture. Import prices are up 8.1% YoY. Notably, export prices are up 11.6% YoY, highlighting the energy-related terms of trade improvement for the US.
GDP expanded by 0.1% MoM in August, taking three-month growth to 0.7% after June was revised slightly higher. Production accounted for the monthly gain, while services were flat. Across the three-month period, services led growth, notably professional services, ICT, and administration. However, this momentum is unlikely to persist amid residual seasonality, elevated energy costs, rising inflation, and pressure on household spending.
Growth will probably weaken further toward year-end as persistently high oil and gas prices push inflation above 4%, potentially prompt an interest-rate hike, and squeeze households. Restrictive fiscal policy will compound the drag. We expect a recovery to take hold from late 2027 as energy prices ease and inflation retreats.
The labor market delivered another downside surprise in September, historically a seasonally positive month. Employment fell by 39.3k (sa), while the unemployment rate held at an upwardly revised 4.7%. More concerningly, participation fell 0.2 percentage points to 66.8%, meaning that, unlike last month, the weakness cannot simply be attributed to an increase in labor supply.
Still, we see this as a yellow light for the Reserve Bank of Australia (RBA), not a red one. Full-time employment rebounded by 34.4k, while the weakness was concentrated in part-time employment. Hours worked also edged higher by 0.1%. This looks, at least partly, like some payback after a long period of strength in more marginal forms of employment, rather than clear evidence that the labor market has rolled over.
September also marked an important change in how we can read the Australian labor market. With the Australian Bureau of Statistics (ABS) completing its Labor Force Survey modernization, industry employment information is now collected monthly rather than only quarterly. The first monthly breakdown adds some breadth to September’s yellow light: manufacturing (+28.7k) and public administration (+23.5k) posted strong gains, while employment fell in transport, professional services, mining and retail. Regular readers will know that we pay particular attention to retail as a leading signal when the data are appropriately standardized. Importantly, these new industry estimates are original rather than seasonally adjusted and currently cover only August and September. They therefore point to softer labor demand, but not yet a broad-based rollover.
We also particularly welcome the changes to the rotation-group framework, an issue we spent considerable time on last year. The roughly 50,000-person monthly ABS survey is divided into eight rotation groups, with one normally leaving and another entering each month, meaning the characteristics of those groups can sometimes influence the headline reading. That is why we have consistently looked beneath the unemployment rate itself. In August and September 2025, our rotation-group analysis was already pointing to upside unemployment risks even while the headline rate remained relatively benign. Those risks subsequently became visible when unemployment rose to 4.5%, a surprise that materially shifted the market narrative, with the probability of a November RBA rate cut jumping to 71% from 40% before the release. In our September 2025 quarterly, for example, we specifically highlighted an unusually low 3.3% unemployment rate in the incoming group, versus its five-month average of 4.5%, and warned that the headline unemployment rate was likely to rise.
The new framework makes this analysis cleaner. Supplementary questions are now confined to the outgoing eighth group and are asked only after the core labor survey has been completed, removing the supplementary-survey effect that previously influenced core responses and required statistical adjustment. September’s rotation-group numbers also provide some context for the latest weakness: August’s outgoing group had an unusually low 3.5% unemployment rate, compared with 4.6% for September’s incoming group and 4.5% for the overall original estimate. That suggests some of the increase in unemployment reflected the changing composition of the sample rather than a sudden deterioration in underlying conditions.
Put together, September is a yellow light rather than a red light for the RBA. The employment decline, lower participation and broader industry weakness deserve attention, but stronger full-time employment, stable hours and the rotation-group dynamics argue against treating one month as confirmation of a labor-market rollover. The focus now shifts squarely to inflation. We expect the upcoming CPI report to show a sizeable quarterly increase of around 0.9%**–**1.1%, which, if realized, would keep the tightening debate alive. For now, there is little urgency for the RBA to hike in November, but we continue to expect at least one further increase over the next year, taking the cash rate closer to, or slightly above, 5%.
There's more to the Weekly Economic Perspectives in PDF. Take a look at our Week in Review table – a short and sweet summary of the major data releases and the key developments to look out for next week.