ABS CPI May 2026: Headline Eases to 4.0% but the Trimmed Mean Climbs to 3.6%

At a Glance: April vs May 2026

Source: ABS Monthly CPI Indicator, May 2026; advertised-rent comparison per SQM Research.

The Australian Bureau of Statistics released its monthly CPI indicator for May 2026 on 25 June 2026, with the cash rate at 4.35% — held in June after three rises earlier in the year — and the major banks openly disagreeing about what happens next. On the surface it reads as good news: headline annual inflation eased from 4.2% to 4.0%. But the more important number went the other way. The trimmed mean — the RBA's preferred measure of underlying inflation — rose from 3.4% to 3.6%, a second straight monthly increase and a level above the top of the 2–3% target band.

Why Headline Inflation Fell While Underlying Inflation Rose

Headline inflation captures the whole basket, including volatile and policy-affected items; the trimmed mean strips out the largest moves in both directions to expose the persistent trend, which is why the RBA targets it. The monthly CPI fell 0.7% in original terms (−0.1% seasonally adjusted), and much of the annual easing is energy washing through the comparison: the ABS reported electricity up 21.1% over the year, but just 3.9% excluding the unwinding of government rebates. The eye-catching electricity line is a base effect, not a fresh shock.

The trimmed mean rising to 3.6% is the release's most consequential number: a second straight increase, above target, moving the wrong way just as the RBA judges whether it has done enough. Crucially, because the trimmed mean excludes the extremes by construction, a higher reading points to pressure that is broadening rather than concentrated in a few items — and breadth is what central banks fear most, because it signals demand- and wage-driven inflation rather than a passing supply shock. That is why a 0.2-point rise in the trimmed mean carries more weight than a much larger move in any single headline line.

What Surprised Economists

The headline easing was broadly expected — the energy base effects were well flagged. The surprise was the trimmed mean rising again rather than stabilising. Consensus had leaned toward underlying inflation flattening as the 2026 hikes bit; instead it has climbed for two months, and the firmer-than-expected new-dwelling and services components suggest the domestic pulse is more stubborn than the easing headline implied.

That gap — soft headline, firmer core — is what moved markets. Rather than reading 4.0% as the start of a return to target, money markets kept an August hike priced as a live possibility, and the most hawkish forecaster (Westpac) treated the trimmed mean as vindication. When economists are "surprised" by a 2026 CPI print, it is almost always the core, not the headline, doing the surprising — and it is the core the RBA acts on.

The Housing Breakdown Investors Should Read Closest

Housing is both the largest CPI contributor and the group most relevant to investors. The main contributors to the May annual rise were electricity, new dwellings and rents (ABS).

New dwellings +5.6% — the line that accelerated

New dwelling purchase costs rose 5.6% over the year, up from 4.7% — an acceleration, not a cooling. It feeds the trimmed mean (helping explain the firming core) and it is the supply story in real time: when build-cost inflation re-accelerates against sale prices that are nationally flat-to-falling, marginal projects get shelved and completions slip — thinning new supply while population growth still adds demand. For investors, established stock keeps a scarcity premium over hard-to-deliver new builds, and the rental shortage that supports yields is unlikely to be solved by new supply soon. It also reframes the 2026 Budget's new-build incentives as swimming against a strong cost tide rather than with it.

Rents +3.6% in the CPI — still lagging the market

CPI rents rose 3.6%, well below advertised-rent growth of around 7% per SQM Research (see our SQM April 2026 vacancy analysis, national vacancy 1.2%). The gap is structural: the CPI series reflects the whole stock of existing leases, so it lags the market as in-place leases reset. With vacancies tight, the rent contribution to CPI likely has further to run — a slow but persistent upward force even as energy effects fade.

Services Inflation: The RBA's Stickier Problem

Beyond housing, the more persistent challenge sits in services — insurance, healthcare, hospitality, education — where labour-intensive pricing adjusts slowly and stays elevated. Two forces keep it firm: wage growth, which though past its peak remains embedded in services prices (the next Wage Price Index is a key checkpoint), and insurance, among the faster-rising lines, feeding landlord, building and strata premiums.

The goods-versus-services split is the key to the whole release. Goods prices — fuel, electricity, imported items — are volatile and policy-affected, and can swing the headline without telling you much about the trend. Services are domestically generated, labour-heavy and slow to turn, so when the trimmed mean rises while the headline falls, services are usually doing the work underneath — the configuration the RBA fears most.

What It Means for the August RBA Decision

The RBA held at 4.35% in June. The May CPI does not force its hand, but a rising trimmed mean removes the easy case for cutting and keeps a hike on the table — producing the widest split among the major-bank forecasters in some time.

The split reflects genuinely different reads. Westpac has been the most hawkish of the majors for months and weights the persistence of services and the firming trimmed mean most heavily, reading May as confirmation. NAB revised the other way in June, judging the tightening already done is enough and the cycle has likely peaked. ANZ sits on the steadiest line — no further moves and a flat 4.35% well into 2027 — while CBA pencils in cuts once the core is clearly back in the band. Three of four expect a hold in August; the credible range now runs from "hold" to "two more hikes," with no major calling a near-term cut. That range is the planning envelope.

Why monthly CPI matters less than the June quarterly read

The monthly CPI indicator is timely but directional — historically it captured only part of the basket each month, so it can be noisier than the full quarterly read the RBA has long treated as its preferred inflation dataset. The next release — the ABS complete monthly CPI for June, on 29 July 2026 — doubles as the quarterly benchmark and is the more comprehensive, decision-grade read. May sharpens the August debate; the 29 July print is the one most likely to settle it.

What could change the call

The outlook is two-sided. Pointing toward a hold or cut: a softer June quarterly CPI, weaker wage growth, a weaker labour market, soft retail sales, or a stronger Australian dollar easing imported-price pressure. Pointing toward a hike: a firm June CPI confirming the trimmed-mean trend, a re-acceleration in wages, or higher global commodity and energy prices. Several of these are external and outside the RBA's control, which is part of why the Bank has kept a tightening bias — but the burden of proof now sits with the doves, not the hawks.

Three Scenarios for August — and What Each Means for Borrowers

Figures are illustrative, on a representative $600,000 interest-only investment loan, assuming the lender passes any move through in full.

Approximately $125 per 0.25% on a $600,000 interest-only loan, assuming full pass-through. Actual repayments depend on your rate, balance and loan type.

The Read-Through for Property Investors

Borrowing capacity stays constrained

With the cash rate at a restrictive 4.35% and underlying inflation firming, the serviceability environment that has compressed borrowing capacity through 2026 won't loosen soon. Lenders assess at the actual rate plus a buffer, and higher measured living costs feed the expense assumptions in that test — so a firm CPI bites capacity twice. A borrower who could service around $700,000 in early 2025 may find their maximum meaningfully lower today on the same income. We walk through the mechanics in our borrowing capacity guide.

Holding costs are still climbing

Construction-linked costs, insurance and utilities all feed the cost of owning property. A property advertised on a 5.0% gross yield ($30,000 rent on $600,000) can quickly become a 3.5–4.0% net yield once management, insurance, rates and maintenance are deducted — before any loan interest. Model holding costs on current, not historical, assumptions.

Tight rentals remain the offset

With vacancies near 1% and advertised rents still rising around 7% (SQM) — well ahead of the lagging 3.6% CPI line — income returns keep firming even as values plateau nationally. Because the CPI series captures the whole stock of leases, the gap to advertised rents is a pipeline of increases still to flow through as tenancies reset, supporting in-place rent growth over the next year or two even if advertised growth moderates.

The two-speed market

These national figures sit over a divergent market — Perth and the mid-sized capitals still growing while Sydney and Melbourne soften (see our May 2026 monthly review). A higher-for-longer rate environment bites hardest where affordability is most stretched and yields thinnest (the larger southern capitals) and least where prices are lower and yields higher. The macro signal is national; the opportunity set is regional.

The Bottom Line for Investors

The May CPI's message is simple: the disinflation investors banked on in late 2025 has stalled, and the measure that matters most to the RBA is firming. A 4.0% headline reads well but is propped up by energy base effects; the 3.6% trimmed mean — rising for a second month, on a broad basis — is the honest read, and it is what shapes rates, capacity and holding costs through 2026. Plan for a plateau with risk skewed up, buffer to the 4.85% scenario, model costs on current numbers, and lean on the one confirmed tailwind — a tight rental market with a long lease-reset pipeline. The investors who do well here are not those who time the next move, but those whose positions work whether August is a hold or a hike.