Isras Holdings: Where the Office Risk Really Sits Inside Isras Investments and Why It Still Sets the Discount
The office headline inside Isras Investments is real, but the weakness is not evenly spread. It sits mainly in Ramat Hahayal and the Brosh building at Har Hotzvim, and for now those two pockets are enough to keep the holding-company discount in place even as other assets look materially stronger.
Where The Risk Really Sits
The main article identified offices as the weak layer inside Isras. This follow-up isolates a narrower question: is that risk really spread across the whole portfolio, or is it concentrated in a few pockets that end up shaping the entire reading of the subsidiary and of the holding company above it. The answer, based on Isras' own disclosure, is fairly sharp. The "42% offices" headline is true at the portfolio level, but the weakness is not uniform. It sits mainly in Ramat Hahayal, and second in the Brosh building at Har Hotzvim.
That is still not a comforting accounting nuance. Offices are 42% of the fair value of Isras' income-producing real estate and 37% of NOI, but in 2025 they contributed only NIS 4.6 million of revaluation gains, roughly 3% of total revaluation gains in the income-producing portfolio. At the same time, average office occupancy fell to 82%, year-end occupancy stood at 79%, and actual yield slipped to 5.1%. This is too large a layer in value, too soft in occupancy, and too weak in value creation for the market to treat it as just another use within a diversified portfolio.
In other words, the market is not only looking at the size of the exposure but at its quality. In retail, industry, and technology parks, Isras is showing year-end occupancy of 96% to 98% and actual yields of 6% to 6.4%. In offices, the same company looks different. That is why, even with a comfortable cash position at the parent, the layer that still sets the discount sits lower down, inside the assets where occupancy and NOI do not yet look clean enough.
That chart explains why offices continue to set the tone. It is not just a large layer, it is the layer where occupancy, yield, and revaluation contribution look weakest together. That is exactly the type of data point the market tends to extrapolate to the whole group.
The "42% Offices" Headline Hides Two Very Different Stories
Once the analysis moves down to the asset level, it becomes clear that the word "offices" is grouping together two very different stories. One is a classic office asset with weak occupancy, eroding NOI, and a thin current yield. The other is a strong campus whose headline looks weak because one large building is still in lease-up. For now, the market is not really separating those two stories.
| Asset | Fair value 2025 (NIS m) | Year-end occupancy 2025 | NOI 2025 (NIS m) | Actual yield | What it means |
|---|---|---|---|---|---|
| Jerusalem Technology Park, Malha | 1,089.0 | 99% | 78.4 | 7.0% | Strong and stable park, not the source of the discount |
| Har Hotzvim | 1,253.9 | 66% | 61.4 | 5.0% | The weak headline is driven mainly by Brosh |
| Har Hotzvim excluding Brosh | 728.9 | 91% | Not separately disclosed | Not separately disclosed | The standing core is much stronger than the headline |
| Kiryat Hatikshoret, Neve Ilan | 362.4 | 85% | 19.8 | 6.3% | A decent asset, not perfect, but not the point that scares the market |
| Ramat Hahayal | 610.1 | 65% | 24.9 | 4.08% | The classic weak pocket inside the office layer |
Ramat Hahayal: This Is Where The Classic Office Risk Sits
If one asset best explains why the market remains cautious on the whole office layer, it is Ramat Hahayal. Its fair value fell in 2025 to NIS 610.1 million, average occupancy fell to 66%, year-end occupancy fell to 65%, NOI slipped to NIS 24.9 million, and actual yield dropped to just 4.08%. This is not an asset sitting on the verge of a breakout. It is an asset whose operating profile has weakened for three straight years.
The more interesting point is the kind of weakness involved. Ramat Hahayal had only 21.5 thousand square meters leased at the end of 2025, down from 25.9 thousand in 2023. Even more telling, the line for average rent in contracts signed during the period shows a dash for 2025. That does not prove there was no leasing effort, but it does mean the disclosure gives investors no fresh operating proof that the read is changing.
The more important clue sits inside the valuation. In the year-one assumptions of the valuation model, Ramat Hahayal is still assessed on 65% occupancy and average monthly rent of NIS 85 per square meter. In other words, even the annual valuation is not built around a fast return to full occupancy. That is the heart of the story. The market sees an office asset worth roughly NIS 610 million, about 19% of the office layer by value, with weak current yield and no immediate valuation assumption of a sharp operating turn. It is hard to ask the market to ignore that.
Brosh At Har Hotzvim: A Weak Headline That Is Concentrated In One Building
Har Hotzvim is a very different story. Its headline looks weak, but the headline does not tell the whole truth. The fair value of the campus rose in 2025 to NIS 1.254 billion, NOI rose to NIS 61.4 million, and revaluation gains jumped to NIS 82.3 million. Yet reported average occupancy stayed at only 66%, and year-end occupancy was also 66%. At first glance, that looks like another clean proof of office weakness.
This is where the data has to be split properly. In the same table, Isras also discloses the metrics excluding Brosh. Excluding Brosh, Har Hotzvim's fair value stands at NIS 728.9 million, average occupancy in 2025 stands at 94%, and year-end occupancy stands at 91%. In other words, the gap between the weak headline and the much stronger operating core is explained almost entirely by Brosh. The valuation disclosure shows the same thing: in 2025 Brosh itself is assessed on 53% occupancy, while the standing buildings excluding Brosh are assessed on 91%.
That leads to an important quantitative conclusion. The gap between total Har Hotzvim fair value and the value excluding Brosh is roughly NIS 525 million. That is not a marginal number. One large building is enough to drag down the headline of an asset worth more than NIS 1 billion. After the balance-sheet date, an additional lease for roughly 6,000 square meters was signed, lifting total occupancy for Har Hotzvim to roughly 74%. That is still not a comfortable level, but it already points to a lease-up risk in a new layer rather than a structurally weak campus.
That chart matters because it separates two different types of risk. Brosh is still a real drag, so Har Hotzvim is not a clean asset. But it is equally wrong to describe Har Hotzvim as a weak office campus without explaining that the weakness is concentrated in one large building rather than across the entire site.
What Is Not At The Center Of The Problem
The other side of the story matters just as much. Jerusalem Technology Park in Malha ended 2025 with fair value of NIS 1.089 billion, 99% average occupancy, 99% year-end occupancy, NOI of NIS 78.4 million, and actual yield of 7%. The asset also has a major tenant from the government housing field that accounts for roughly 40% of the tenant-linked area and more than 20% of asset revenue. That does not eliminate concentration risk, but it does mean the asset is much more anchored than the broad "office" headline suggests.
Kiryat Hatikshoret in Neve Ilan also does not look like a distress point. Yes, year-end occupancy slipped to 85% in 2025 and NOI fell to NIS 19.8 million. But fair value still rose to NIS 362.4 million, revaluation gains were NIS 10.3 million, and actual yield remained 6.3%. That looks more like an asset that still needs active management and leasing work, and less like an asset that should by itself force an outsized discount on the whole value stack.
The message of that chart is straightforward: the market is not really afraid of "workplaces" or "office-like assets" as one uniform bucket. It is afraid of the places where occupancy is weak, current cash yield is thin, and proof of improvement is still incomplete. That is obvious in Ramat Hahayal. It is still open in Brosh. It is much less obvious in Malha and Kiryat Hatikshoret.
Is The Discount Too Harsh, Or Still Rational?
The answer here is more nuanced than the headline. The March 2026 presentation showed NAV of NIS 464.9 per share on a market-value, pre-tax basis, against a market price of NIS 286.3 on March 27, 2026. That gap does not stay open only because Isras Holdings is a holding company. It also stays open because the market still sees an office layer inside Isras that has not fully cleaned up.
On one hand, the discount is still understandable. Ramat Hahayal is sitting on just 4.08% current yield and 65% occupancy, and Brosh is still a drag on Har Hotzvim even after the lease signed near the report date. Those two pockets alone create enough discomfort for investors to say that the parent's NAV may exist, but it still sits above an operating layer that needs proof.
On the other hand, the broad "42% offices" headline also simplifies too much. It causes the market to punish all of Isras through the same lens even though the numbers show a much more split picture: Malha is almost full and yields 7%, Har Hotzvim excluding Brosh looks much stronger, and Kiryat Hatikshoret is far from looking distressed. So the discount is still rational directionally, but it is probably broader than the places where the real risk sits today.
What Will Decide The Next 2-4 Quarters
First test: Ramat Hahayal needs to stop deteriorating, not only in narrative but in occupancy, in actual signed lease terms, and in NOI. As long as those numbers stay close to 2025 levels, the market will keep treating the asset as proof that office risk is still alive.
Second test: Brosh needs to move from a leasing promise to visible NOI. The post-balance-sheet lease lifts Har Hotzvim occupancy to roughly 74%, but what matters is whether that turns into stable cash generation and a smaller gap between reported Har Hotzvim and Har Hotzvim excluding Brosh.
Third test: The office layer needs to contribute to value creation again rather than mainly weighing on it. In 2025 it generated only about 3% of the total revaluation gains in the income-producing portfolio despite representing 42% of value. As long as that mismatch remains, it is hard to argue that the discount has already done all the work.
Bottom Line
The office risk inside Isras is real, but it is not spread evenly across the portfolio. It sits mainly in Ramat Hahayal and in the Brosh building at Har Hotzvim. Ramat Hahayal is the classic weak office asset, with low occupancy, thin current yield, and no fresh proof of improvement in 2025. Brosh is a different story: not weakness across an entire campus, but partial lease-up of one large building that is still too big for the market to ignore.
So the right conclusion is neither that the market is simply wrong nor that it is fully right. The market is correctly identifying the problem layer, but it is still pricing that layer across the whole system too bluntly. Until Ramat Hahayal stabilizes and Brosh turns into visible NOI, offices will continue to set the holding-company discount at Isras Holdings even if other parts of the portfolio already look materially stronger.
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.