Care home funding decisions turn on occupancy modelling more than any other single variable. Lenders, investors and acquirers all stress-test the same three things: the achievable steady-state occupancy, the time taken to reach it, and the resilience of the model to a regulatory event. Operators who present a single best-case forecast invite immediate scepticism. Operators who present a base, downside and stress case usually get faster decisions on better terms.
In this article
- Why does occupancy modelling carry so much weight?
- Which metrics actually move a funder's view?
- How is the model stress-tested?
- What are the common operator mistakes?
- How are CQC and local authority risks built in?
- What does a defensible model actually look like?
Why does occupancy modelling carry so much weight?
Care home economics are dominated by fixed cost. Staff rotas, regulatory overheads and property finance are largely the same whether occupancy is 75% or 92%. That means each additional bed sold contributes most of its fee to cash flow. Funders know this, so the occupancy assumption is the lever the model is most sensitive to. Get it wrong by three points and the entire investment case can flip from comfortable to marginal.
This is why funders test the assumption rather than accept it. They want to see the working, what occupancy is now, how it has trended over the last 24 months, what the local market is achieving, what the operator has done that justifies projecting higher.
Which metrics actually move a funder's view?
Three metrics carry the most weight. Trailing 13-week occupancy gives a current run-rate that is more honest than a quarterly average. Average length of stay matters because it drives the number of admissions required to maintain occupancy. Self-pay versus local authority mix matters because the fee differential is material, typically £350 to £500 per resident per week, and a model assuming a sudden shift in mix is a model worth challenging.
Beyond those three, funders look at deposit history (a leading indicator of incoming residents), enquiry conversion rates (a check on the operator's commercial discipline), and the trend in agency staff cost as a proportion of total payroll (a leading indicator of margin pressure).
How is the model stress-tested?
The standard stress tests are predictable but not always applied. A 5% reduction in steady-state occupancy. A 12-month delay in reaching steady state. A 10% increase in staff cost over 18 months. A regulatory event that suspends new admissions for three months. Each test is run individually and in plausible combinations. The model that survives a single-variable stress but breaks under a two-variable combination is the one that worries lenders most.
According to the 2024 Knight Frank UK Healthcare Capital Markets report, lenders to the care sector reported that 38% of operator-supplied models presented in 2024 failed the basic two-variable stress test on first pass. The fix is rarely structural, it is usually a more honest base case combined with explicit downside scenarios, rather than a single optimistic forecast.
What are the common operator mistakes?
Five mistakes recur. Modelling steady-state occupancy at 95%+ when the realistic ceiling for the home's catchment is 88%. Assuming self-pay mix can be lifted from 40% to 65% within 18 months without a credible repositioning plan. Treating CQC ratings as static when they are reviewed regularly. Underestimating the time and cost of replacing the manager. Ignoring the impact of a single competing new home opening within five miles.
The common thread is that each error feels small in isolation but compounds. A 95% occupancy assumption with a 65% self-pay mix and a static CQC rating produces a model that looks attractive on paper and is almost guaranteed to underperform in practice. The model loses credibility, the operator loses negotiating leverage, and the funding either falls through or comes on worse terms.
How are CQC and local authority risks built in?
CQC risk is built in through scenario rather than through a single number. The model should show the cash impact of a Requires Improvement rating triggering a six-month admissions hold, the cost of bringing in turnaround support, and the recovery period to return to a Good rating. Most operators have not modelled this because they have not experienced it; the funder is unconvinced precisely for that reason.
Local authority fee risk is more about timing than amount. Most authorities now publish indicative fee uplifts annually. The risk is the gap between what the operator assumes and what the authority actually grants. A defensible model uses the published indicative figure, not a wished-for one.
What does a defensible model actually look like?
A defensible occupancy model has a clear base case derived from trailing actuals and local benchmarks, a downside case showing what happens if steady state lands three points lower than base, and a stress case showing the cash position through a six-month admissions suspension. Each scenario is documented with its underlying assumptions and the source of each material number.
The model is also internally consistent: occupancy, staff cost, fee mix and capex all move together in plausible ways. Funders distrust models where one input changes and others mysteriously do not. The discipline of internal consistency is more important than the absolute numbers.
Written by Bharat Varsani FCCA. Director of Key Ledgers Global, currently operating at CFO level within a 15-entity care and property group including a 117-bed nursing home and 140 supported living beds. 20+ years of forensic, audit and CFO experience across regulated care.
Sources: Knight Frank UK Healthcare Capital Markets 2024 report for the cited stress-test failure rate; Care Quality Commission published inspection methodology and rating decision frameworks.
Frequently asked questions
What occupancy figure should a care home model use as base case?
It should be derived, not assumed. Take the trailing 13-week occupancy, adjust for known one-off factors, then benchmark against local competitors and historical performance. The base figure is whatever survives that triangulation, not a number the operator wishes to see.
How does CQC risk affect care home valuations?
Materially. A Good rating is the baseline for a normal valuation. Requires Improvement typically reduces multiple by 0.5x to 1.5x EBITDA. Inadequate effectively halts marketability until rating is recovered. Outstanding adds modestly but not dramatically because most buyers expect Good as standard.
Should self-pay mix be modelled to grow?
Only with a credible plan. Lifting self-pay mix requires repositioning, refurbishment, and usually a 12 to 24 month lead time. A model assuming it happens automatically without that investment is the model funders distrust most.