Australia's Trimmed Mean Came In at 1.0 Per Cent. By the Bank's Own Test, That Is a Rate Rise

The Australian Bureau of Statistics published two inflation readings on Wednesday and both went the wrong way for anyone hoping the Reserve Bank could hold on 3 November. The monthly CPI rose 0.8 per cent in September, 0.6 per cent seasonally adjusted, taking annual inflation to 4.3 per cent from 4.0. The monthly trimmed mean rose 0.3 per cent and 3.6 per cent over the year, from a revised 3.5. The quarterly figures, which the Bank still treats as the cleaner read, showed the CPI up 1.3 per cent in the September quarter and 3.9 per cent over the year, and the trimmed mean up 1.0 per cent in the quarter and 3.7 per cent over the year, against a market consensus of 3.6 and the Bank's own August projection of 0.8 per cent for the quarter.

The composition is the problem. Automotive fuel rose 9.4 per cent in September alone and 23.6 per cent over the year, and it is excluded from the trimmed mean. What remains is housing at 5.2 per cent, with electricity up 12.8 per cent as Commonwealth rebates ended and rents up 3.6 per cent; recreation and culture at 4.5 per cent, with domestic holiday travel up 8 per cent in a month of school holidays and major sporting fixtures; and food at 3.2 per cent. Services inflation ran at 1.3 per cent for the month and 4.1 per cent for the year, goods at 4.6 per cent. Stripping out fuel, fruit and vegetables and holiday travel leaves annual inflation at 3.6 per cent, which is where the Bank's target band ends.

The Bank publishes new forecasts with its decision at 2.30pm Sydney time on 3 November.

From the analysis

A week ago deputy governor Andrew Hauser said the outlook was "materially worse" than the August forecast and that the trimmed mean for the quarter would likely come in above it. It did, by two tenths. Market pricing for a quarter-point rise on 3 November moved from about 60 per cent before the release to about 85 per cent after it, having been 27 per cent before Hauser spoke. The big four banks had all forecast 1.0 per cent for the quarter, so the number itself was no surprise; the surprise is that the annual rate kept climbing and that services, the component rates can actually reach, did most of the climbing.

Sox set out a test for this print last week: if the trimmed mean came in at or above the path implied by the August forecast, the board should raise and say why; if below, hold and fix the communication. It came in above. The case for holding, which Sox made on 15 October when unemployment reached 4.7 per cent, rested on the labour market doing the Bank's work and on the inflation being imported fuel. The fuel part is still true, and the trimmed mean removes it. The rest of the basket is now running at a 4 per cent annual pace with a fifth consecutive quarter of underlying inflation between 0.8 and 1.0 per cent. That is persistence, by any definition the Bank has used.

We think the board should raise the cash rate to 4.85 per cent on Tuesday and attach to it the one thing four previous rises have lacked, a plain statement of what would make it stop. The expectations argument Hauser made is the right one: a growing share of businesses now treat 3 to 3.5 per cent as the new normal, and this print tells them they are right. The labour market argument has not gone away. Unemployment at 4.7 per cent is already above the Bank's year-end forecast and a fifth rise will push it higher. The honest framing is that the Bank is choosing more unemployment now over more inflation later, and the board owes the people who will lose jobs an explanation of the trade and a threshold for reversing it.

The Treasurer's statement on Wednesday recorded the 3.7 per cent trimmed mean without comment; in recent weeks he has pointed to the Middle East war for fuel and the end of rebates for electricity, both true, and neither relevant to a trimmed mean that excludes the first and has already absorbed the second. The Bank publishes new forecasts with its decision at 2.30pm Sydney time on 3 November. The October monthly CPI follows in late November. The question for Tuesday is whether the board treats 3.7 per cent as a number it can still talk down, or as the fifth quarter of evidence that talking has not worked.

The Semaform

The news

The ABS reported on 28 October that Australian trimmed mean inflation rose 1.0 per cent in the September quarter and 3.7 per cent over the year, with monthly headline inflation at 4.3 per cent, lifting market odds of a fifth RBA rate rise on 3 November to about 85 per cent.

Sox’s view

The print passed the test Sox set last week, so the board should raise to 4.85 per cent on Tuesday, and should pair the rise with an explicit statement of the conditions under which it would stop, given unemployment is already above the Bank's forecast.

Room for disagreement

CBA's economists still expect a hold, arguing four rises are working through, the economy is slowing and much of the overshoot is fuel and rebate effects; the Treasurer attributes the rise to the Middle East war and the end of electricity rebates.

Sources

A human editor is accountable for this analysis. Read our editorial standards, how we use AI, what the content labels mean, and how to ask for a correction.

Australia's Trimmed Mean Came In at 1.0 Per Cent. By the Bank's Own Test, That Is a Rate Rise

The Semaform

Sources

More from the AU Edition

Snowy 2.0 Will Cost $25 Billion to Save $7 Billion. Canberra Says Stopping Would Cost More

The RBA's Deputy Has Named the Only Good Reason for a Fifth Rate Rise. It Has Nothing to Do With Oil

A Dreaming Story Arrived After the Report. That Is Why the McPhillamys Ban Fell