Interest rate hikes remain on cards as Australia’s underlying inflation climbs, economists warn

The split the headline hides

The Australian Bureau of Statistics reported two opposing prints in the same release:

• Headline CPI: 4.0% YoY (consensus had expected 4.4%)

• Trimmed mean CPI: 3.6% YoY, up from 3.4%

• Home building costs: +0.9% MoM, strongest monthly print since late 2022; annual pace 5.6%

• Food and drink inflation: accelerated to 3.3% from 2.8%

• Restaurant and takeout meals: +4.0% YoY

Fuel did the heavy lifting on the headline. Strip it out, and core pressures worsened. Treasurer Jim Chalmers called the print "much better than the market expected." The trimmed mean says otherwise.

Market pricing vs economist call

Two senior economists, two opposing verdicts on the same data:

• NAB's Sally Auld: next RBA move is a cut. Headline will peak below the 5% rate embedded in the federal budget. Mortgage relief, in her view, is still roughly a year away.

• AMP's Shane Oliver: fourth hike in August. Underlying print came in hotter than expected, and the RBA is focused on preventing an "inflation psychology" from anchoring.

• Market-implied odds: 32% probability of a hike at the 11 August RBA meeting; 56% probability of at least one hike by year-end.

The split is a function of which series each economist weights. Auld anchors on headline and slowing activity. Oliver anchors on trimmed mean and services stickiness. Both are reading the same ABS release.

Verdict: what we do with this

For readers running money through Australian equities, AUD ETFs, or AUD fixed income:

• Hawkish-hold bias in the curve. A 56% year-end hike probability and rising trimmed mean argue against fading RBA hawkishness through Q3. We hold duration short on the short end of the AUD curve.

• Banks and rate-sensitive sectors stay bid. Mortgage books reprice on a 3-6 month lag. If Auld is right and the next move is down — but delayed — the major Australian banks retain the carry we logged last quarter. No change.

• Headline-driven risk. A 12% monthly drop in fuel is a one-off. It will roll off the base effect by Q4, at which point the headline will catch up to the trimmed mean. Position for that mechanical repricing, not for the current print.

• Verdict on consensus: fail. The 4.4% consensus forecast and the "much better than expected" framing both miss the point. The RBA's preferred gauge deteriorated. A 32% hike probability on 11 August is not consistent with an inflation print where core accelerated and services inflation remains sticky at 3.3%+.

We treat the August meeting as a live hike risk, not a hold. Watch the next two monthly trimmed mean prints. If core holds above 3.5%, the 32% number is mispriced low.