Golden House: How Much of Occupancy Is Truly Permanent, and How Much Is Still a Bridge
The 95% occupancy headline at Beit Gil Hazahav looks strong, but underneath it still sit 59 bridge units, 20 trial apartments and a cash model that also depends on deposit churn. This follow-up breaks down the gap between near-full physical fill and economics that are not yet fully normalized.
What This Follow-Up Is Isolating
The main article already made the compressed point: Golden House's 2025 recovery is real, but part of the jump still rests on occupancy that is not permanent assisted-living residency. This continuation isolates that gap, because in assisted living it is not a cosmetic distinction. It is the difference between a house that looks almost full and a house that is already producing stable economics from permanent residents rather than from a transition pipeline.
The first headline number is 281 apartments occupied on average in 2025, equal to about 95% total occupancy. But the same passage says that around 59 apartments on average, about 20% of the stock, were occupied by tenants who were not permanent assisted-living residents, and that revenue from this type of occupancy reached about NIS 2.7 million. That is not a footnote. It is the core difference between physical occupancy and permanent occupancy.
The appraisal reinforces the same reading rather than contradicting it. At the cut-off date it shows 296 apartments, 59 units rented as holiday stays, 20 trial apartments that still pay maintenance fees only, and 17 apartments actually vacant. So even the value document itself is not describing an asset that has already reached its clean end-state. It is describing an asset still in transition.
| Layer | What the documents show | Why it matters |
|---|---|---|
| Total 2025 occupancy | 281 apartments occupied on average, about 95% occupancy | This is the physical headline for the year |
| Non-permanent fill | 59 apartments on average, 52 at year end, and about NIS 2.7 million of revenue | This is the bridge layer underneath the headline |
| Year-end appraisal picture | 59 holiday units, 20 trial apartments and 17 actual vacancies | The appraisal itself shows that the property is not yet sitting on a clean permanent base |
| Cross-document economic read | Cross-checking 281 occupied apartments against 59 bridge units leaves about 222 permanent-equivalent units, roughly in line with 75% occupancy on a 296-unit stock | This is why 95% and 75% are not conflicting numbers |
95% Does Not Describe the Same Economics as 75%
This is not a contradiction. These are simply two numbers describing two different layers of the same house. The 75% average occupancy describes the economics of the assisted-living activity as reported in the annual data table. The 95% figure describes a wider occupancy picture that also includes apartments filled for short periods instead of remaining empty.
What matters is that the two documents almost reconcile the gap on their own. If 59 non-permanent units are deducted from the 281 apartments occupied on average, what remains is roughly 222 units. Against the appraisal's 296-apartment inventory, that converges to about 75%. In other words, the annual report is already saying that the house looks fuller than its permanent base really is, and the appraisal shows which transition layers are being used to hold that bridge together.
The year-end picture is even sharper. The appraisal does not stop at 59 holiday lets. It adds another 20 trial apartments, which during the initial three-month period pay maintenance fees only and have not yet chosen a full contract route. So even inside the year-end occupancy figure there is a layer whose economics are not yet normalized. Anyone reading 94.3% physical occupancy and instantly translating it into full stabilized economics is skipping over that detail.
This is also the difference between a real recovery and an end-state that has already been reached. The recovery is real because the property is no longer standing with the same depth of empty inventory seen in the weaker years. But the permanent occupancy target has still not been reached, because part of that distance has been closed through a bridge solution. That bridge is legitimate. It is just not the same thing as a permanent resident entering a full contract route and staying there.
Contract Routes Change Cash Quality, Not Just Revenue
To understand normalized cash generation in assisted living, counting apartments is not enough. The contract route matters. The company uses four main routes: entrance fees, declining deposits, special deposits and monthly rent. In the first three, an upfront deposit-like amount comes in, erodes over time and part of it is later refunded. In the monthly route, there is no such upfront payment, only ongoing monthly income.
The annual report explicitly says that 2025 revenue growth also came from a higher share of residents on the monthly-rent route instead of deposit or entrance-fee routes. That sharpens an important point: shifting toward the monthly route can improve the current revenue line, but it does not create the same upfront cash float that a deposit route creates on entry.
The company also says outright that the main funding sources of ongoing operations are operating profitability plus the difference between deposits received and deposits repaid. That is a material disclosure. It means normalized cash generation here does not rest only on NOI or management fees. It also rests on resident turnover and on the spread between incoming and outgoing deposits.
That is why the right cash frame here is normalized / maintenance cash generation, not all-in cash flexibility at the full-company level. The documents do not provide one clean final normalized-cash number for shareholders, but they do show what drives it: operating profitability, route mix and deposit churn. Inside that frame, the good news is that the company estimates only about NIS 4 million of deposit repayments in 2026 rather than the full deposit balance shown as a current liability. The less comfortable part is that this estimate is based on a 14% to 18% turnover assumption and an average route mix, and is explicitly labeled as unaudited forward-looking information.
So occupancy quality matters twice. First through the income statement, because the monthly route lifts current revenue. Second through cash, because the house still depends in part on a net-deposit engine. The key question is therefore not only whether the 59 temporary units disappear, but what replaces them and what payment route the next resident cohort chooses.
| Route | What comes in now | What comes back later | Why it matters economically |
|---|---|---|---|
| Monthly rent | Higher ongoing current income | No refundable entry deposit | Better for the current revenue line, but it does not create the same entry cash float |
| Deposit or entrance-fee routes | Upfront payment with erosion over time | Part of the balance is later refunded | Supports cash on entry, but also creates a future refund burden and churn dependence |
| Trial apartments | Maintenance fees only during the trial period | No full route yet | Useful for marketing and pipeline fill, but not yet normalized economics |
The Appraisal Also Does Not Treat 2025 as Fully Permanent
This is the center of the issue. The appraisal does not simply look at 2025, see 95% occupancy and conclude that the asset already operates like a fully stabilized assisted-living property. It explicitly builds a transition model. It says the valuation should include cash at resident replacement, income from monthly-rent apartments, income from trial apartments, income from vacant apartments, normative management surplus, rent paid to Koptash and economic depreciation.
The numeric split makes the transition dependence very clear. Under a base assumption of 15% annual churn, a 4.25-year period for first-turnover income, a 7.6-year period for future churn, an 8.5% discount rate for deposit cash flows, a 7.0% discount rate for apartment-rent income and a 10% rate for the operating stream, the appraisal assigns about NIS 32.1 million to first-turnover income, about NIS 16.0 million to trial apartments, about NIS 43.0 million to vacant inventory and about NIS 30.2 million to future resident churn. Together, that is about NIS 121.2 million of value layers tied to conversion, marketing and turnover.
That does not mean the valuation is necessarily aggressive or conservative. It means something simpler: even the document meant to anchor the property's value is effectively acknowledging that the house is still midway through the transition. Alongside those transition layers, the appraisal capitalizes about NIS 99.9 million for monthly-rent apartments and about NIS 63.9 million for commercial leases, then subtracts about NIS 25.6 million for management activity and other burdens before renovations, building rights and the deposit balance.
The implication is that the appraisal does not read 2025 as a year that already stabilized the property's permanent economics. It prices the path toward that result. So the right interpretation is not "95% occupancy, problem solved," but "the bridge is working, yet the bridge is still a bridge."
What Has To Happen Before This Can Be Called Permanent Occupancy
The first checkpoint is straightforward: the number of bridge units has to keep falling without reopening the vacancy gap. If the holiday lets and trial apartments give way to permanent residents, the 95% headline will begin to look like a real 95%. If they are simply replaced by another wave of temporary fill, the headline will stay attractive while the economics remain unfinished.
The second checkpoint is the route mix of the next resident cohort. More monthly-rent residents may continue to support reported revenue, but that should not be read as identical to deposit routes from a cash-engine perspective. The larger that share becomes, the more important it will be to see the operating line itself carry the story rather than apartment-fill optics alone.
The third checkpoint is deposit behavior in 2026. The company estimates about NIS 4 million of repayments, while the appraisal's future-churn model rests on 15% annual turnover and an 80% refund rate. If those figures hold, the deposit engine can still be viewed as controlled. If actual repayments come in materially higher or if churn becomes less healthy, the normalized-cash read will look less clean very quickly.
Bottom Line
The issue at Golden House is not that the 95% occupancy figure for 2025 is fake. The issue is that it is still not permanent occupancy. Part of the distance was closed through holiday lets, another part through trial apartments, and part of the economics still runs through deposit-turnover mechanics rather than through fully stable recurring income alone.
That matters because it explains how readers can simultaneously see a real recovery, a higher appraisal and still have a valid reason to resist the conclusion that the asset is already normalized. The path to permanent occupancy clearly exists, and the appraisal even builds it in. But as of year-end 2025, it is still a path, not the finish line.
Disclosure: Deep TASE analyses are general informational, research, and commentary content only. They do not constitute investment advice, investment marketing, a recommendation, or an offer to buy, sell, or hold any security, and are not tailored to any reader's personal circumstances.
The author, site owner, or related parties may hold, buy, sell, or otherwise trade securities or financial instruments related to the companies discussed, before or after publication, without prior notice and without any obligation to update the analysis. Publication of an analysis should not be read as a statement that any position does or does not exist.
The analysis may contain errors, omissions, or information that changes after publication. Readers should review official filings and primary sources before making decisions.