Full Beds, Thin Margins: Occupancy Alone Is No Longer the Lever in Residential Aged Care
Full Beds, Thin Margins: Occupancy Alone Is No Longer the Lever in Residential Aged Care
Occupancy rose to 95.2% and the share of homes losing money rose to 62%. When your best operational number improves and your result gets worse, the question stops being "how do we fill beds" and becomes "which line moved, and by how much".
Two numbers in the same survey moved in opposite directions, and that is the whole story.
StewartBrown's aged care financial performance survey for the nine months to 31 March 2026, released in late July, covered 1,173 residential aged care homes — roughly 96,000 beds, a substantial share of the sector. Over that period, occupancy rose to 95.2%, up from 94.2% a year earlier. Over the same period, the proportion of homes recording an operating loss rose to 62%, up from 49%.
If occupancy alone still drove the result, that would be hard to explain. A sector fills more beds and loses more money at the same time only when something else — funding, cost or mix — has moved faster than occupancy did. Occupancy still matters, because an empty bed earns nothing against fixed costs. It has simply stopped being enough.
|
62%
of surveyed homes recorded an operating loss in the nine months to 31 March 2026, up from 49% a year earlier.
|
95.2%
average occupancy, up from 94.2%. The operational metric improved while the financial result deteriorated.
|
|
−$9.16
average operating result per bed per day, against a small surplus of $0.91 the prior year.
|
$5.91
direct care result per bed per day, compressed from $18.46 a year earlier — the largest movement among the survey's segment results.
|
The number that moved isn't occupancy
The direct care result fell from $18.46 per bed per day to $5.91. Roughly two-thirds of that contribution disappeared in twelve months. Nothing else in the survey moved on that scale, and nothing else needs to in order to explain a swing from a $0.91 surplus to a $9.16 deficit.
That matters because direct care is the funded part of the business. It is the segment with a dedicated funding mechanism attached to it, and it is the segment most providers assumed would at least cover itself. On these numbers, it barely does.
At sector level the survey is clear about why: direct care revenue rose 6.6% over the year while direct care costs rose almost 11.7%, as providers staffed up towards mandated care minute targets and absorbed award wage increases. A sector average cannot tell an individual provider which of two things happened to them, though. Either the cost of delivering care rose faster than the funding attached to it, or the mix of what is being delivered changed in a way that the funding model does not follow. Those two causes look identical on a consolidated profit and loss statement and call for completely different responses, and only a provider's own ledger can say which applies.
Which is exactly the problem.
Why most providers can't answer "which line moved"
Ask a finance team in a mid-sized provider for the direct care result per bed day by home, this year against last, and the honest answer is usually that it would take a fortnight. Not because the data is missing, but because of how it sits.
Three things get in the way, and they are remarkably consistent across mid-sized providers:
The profit and loss statement is at entity level. The board sees one consolidated result. Direct care, everyday living (hotel services) and accommodation are the three segments where the funding and the economics actually differ, and most management reporting packs do not separate them.
Cost centres were built for payroll, not for margin. Where segment reporting does exist, it was usually mapped once, years ago, to suit rostering or award interpretation. Agency staff, allied health, and clinical management sit in different places in different homes, so the segments are not comparable between sites.
Nothing is normalised. A larger home will always show a bigger number. Until every line is expressed per bed day, you cannot tell whether a home is performing worse or is simply bigger, and you cannot compare this year to last year through a period where occupancy itself moved.
That last point is the one worth fixing first. Per bed day is the unit the whole sector reports in, and converting to it strips out both size and occupancy. It turns a set of unreadable profit and loss statements into one comparable table.
The one-page view worth building this month
The deliverable is a margin bridge, not a variance report. It starts with last year's result per bed day, ends with this year's, and every line in between is one driver with a dollar amount against it. For a residential provider it has four blocks:
Revenue per bed day, split between the funded care component, resident contributions and accommodation. Direct care cost per bed day, split between permanent staff, agency, and allied health. Everyday living, accommodation and administration cost per bed day. And occupancy as its own line, so that the effect of filling beds is visible and separate from everything else.
Built per home and sorted by movement, the bridge usually produces an uncomfortable finding within a page: two or three homes carrying most of the deterioration, and a driver that nobody had named. Agency reliance in one building. A clinical management structure that was fine at 90% occupancy and is not at 95%. A hotel services contract repriced mid-year.
None of that is visible on a consolidated result, which is why "we have an occupancy problem" survives as an explanation long after it has stopped being the whole story.
Where AI genuinely helps, and where it must not be let near
This is a good task to put AI on, for an unglamorous reason: most of the work is reconciliation and normalisation, not judgement.
Mapping inconsistent cost centres across several homes to one common segment structure is a genuine strength — the model reads the account names and descriptions, proposes a mapping, and flags the ones it is unsure about. Converting every line to a per bed day basis using each home's actual occupied bed days is arithmetic. Building the bridge and sorting it by movement is arithmetic. Drafting the first description of each variance is a reasonable first pass.
What cannot be delegated is the part that matters: deciding whether a movement is structural or timing. An agency spike in one quarter because two staff were on extended leave is not the same finding as an agency spike because the roster can no longer be filled from the permanent pool, and no model can tell you which one you are looking at. That requires someone who knows the building.
The practical rule is the one that applies to most finance automation: the model proposes the mapping and does the arithmetic, a human approves the mapping before anything is built on it, and a human writes every conclusion. The mapping approval step is not optional. A silently wrong cost centre mapping produces a bridge that is clean, confident and wrong, and it will be believed because it looks like analysis.
Before any of this goes near an AI tool: cost centre exports, rosters and payroll extracts are staff data, and care minute records sit close to resident data. Confirm whether your vendor retains customer inputs or uses them to train its models, and check the setting rather than assuming the default. The safest posture is a tool where your data is not retained for training. De-identify where you can, and keep resident-level records out of it entirely — a margin bridge needs occupied bed days and cost totals, not names.
This post is general commentary based on publicly available information and does not constitute legal or tax advice. Always seek independent professional advice before acting.
The number that should worry the board most
One figure in the survey is less discussed than the deficit and matters more over a five-year horizon. Average operating EBITDA came in at about $4,835 per bed per annum. The range the survey indicates is needed to attract investment in new development is $20,000 to $22,000 per bed.
That gap is not a trading problem, it is a capital problem. At less than a quarter of the required return, a provider cannot fund a refurbishment from operations, cannot raise external capital on the strength of its earnings, and will struggle to justify a new build to a board that has read the numbers. The practical consequence shows up well before any of that: deferred maintenance, ageing buildings, and a slow drift in the resident mix a provider can attract — which feeds back into the same margin.
For a finance lead, the useful move is to put that comparison in front of the board now, while the capital question is still hypothetical. A board that knows how its own homes compare with a sector earning less than a quarter of what new development requires will ask different questions about the next five years than a board looking at a loss per bed day.
What to do in the next fortnight
Three things, in order. Separate direct care, everyday living and accommodation in your management reporting, even roughly, even if the mapping is imperfect at first. Convert every line to a per bed day basis so the numbers are comparable between homes and between years. Then build the bridge from last year's result to this year's and name every driver with a dollar amount against it.
Two cautions on the figures above. They describe the sector to 31 March 2026 and were released in late July, so they are a reference point rather than news. StewartBrown has since said it expects the share of loss-making homes to rise in the full-year result, which had not been published at the time of writing — so check for newer figures before quoting these to a board. And a sector average says nothing about any individual home. The point of the exercise is to find out whether you are the 62% or the 38%, and if you are in the 62%, which building and which line is responsible.
Occupancy at 95.2% is close to a practical ceiling — StewartBrown expects it to sit above 96% for the next five to ten years, which leaves perhaps a point to gain. There is not much more to win there, which means the next improvement has to come from somewhere you can currently see only in aggregate.
Can your reporting tell you which home and which line moved?
If the answer takes a fortnight to assemble, the reporting is the problem before the margin is. PFL provides senior-level outsourced finance, management reporting, and AI automation for Australian NFP, NDIS, and SME organisations — including segment reporting built so the margin question can be answered in an afternoon.
Talk to PFL →Inside Ageing — More aged care homes fall into the red despite rising occupancy, StewartBrown warns
The Weekly Source — StewartBrown sounds alarm on residential aged care margins in latest survey
Australian Ageing Agenda — Resi care profit falls as occupancy hits 95 per cent
The Weekly Source — StewartBrown warns more aged care homes will fall into loss (September 2026)
Comments
Post a Comment