Mega Or 2025: the core is strong, but 2026 will be decided by funding and data center delivery

Introduction to the Company

At first glance, Mega Or at the end of 2025 still looks like a near-classic yield real estate story: 99% occupancy, owner-share NOI of NIS 392.3 million, management FFO of NIS 291.7 million, equity of NIS 3.9 billion, and an LTV of 46%. That is already too simple. Mega Or's economics now sit on three different layers: a very strong legacy investment-property core, a Data Centers platform that still produces very little cash but is already producing revaluations and capital needs, and a holdings-and-securities layer that adds value and also quite a lot of noise to reported profit.

What is working now is real. The legacy logistics and commercial portfolio continues to run at near-full occupancy, same-property NOI rose 6.4% to NIS 309.7 million, and the group also holds roughly NIS 1.45 billion of liquidity, including cash and marketable securities, plus NIS 175 million of unused credit lines. The newer operating layer has also crossed the concept stage: the first Data Center in Modiin was fully delivered for use, and during the period and up to the report date the company signed agreements totaling 174 MW IT.

But the first read still misses the real bottleneck. In 2025 the Data Centers segment generated only NIS 13.3 million of revenue and only NIS 10.1 million of NOI, while group net profit jumped to NIS 844.3 million. That gap did not come from a parallel jump in cash. It came from a mix of NIS 430.1 million of real-estate revaluations, NIS 379.4 million of gains from fair-value changes or sales of financial assets, and NIS 61.1 million of impairment-reversal profit in DSKSH. That is not necessarily a negative sign. It is simply a reminder that the year can no longer be read only through the current rent roll.

The active friction in 2026 sits elsewhere: funding and delivery. Mega Or has already proved that it can generate NOI, refinance debt, and keep a strong balance sheet. What it has not yet proved is that it can execute a very large Data Centers build-out in parallel without becoming too dependent on the bond market. That matters now because precisely where the legacy business looks cleanest, the debt schedule, the shareholder loans into Mega D.C, and the fact that part of the value has already been booked ahead of cash, are what will determine the quality of the next year.

Mega Or's economic map looks like this:

These first charts sharpen the central point. Mega Or is still funded mainly by logistics and commercial real estate, not by server farms. In other words, the company has already earned much of the growth narrative around Data Centers, while its actual cash engine still sits mostly in the legacy layer.

Events and Triggers

Data Centers moved from a slide deck to a delivery schedule

The first trigger: 2025 was the year in which the Data Centers activity moved from strategy to contracts and delivery. Stage A of the Modiin site, with roughly 9.5 MW IT, was fully delivered during the year for customer use. At the same time, agreements totaling 174 MW IT were signed during the period and up to the report date. In December 2025 the company signed a package of agreements for 19 MW IT in Modiin Phase B and Masmiya, with another 50 MW IT of expansion optionality. In January 2026 two more agreements were signed for a total of 80 MW IT in Masmiya and Beit Shemesh.

The good news is that demand is no longer only a theoretical thesis. The less comfortable part is that the next step is still not yet sitting in the 2025 numbers. Delivery for Modiin Phase B and Masmiya is scheduled for the third quarter of 2026, and delivery for Beit Shemesh is scheduled in phases from the third quarter of 2026 through the first quarter of 2027. That makes 2026 an execution year, not a full harvest year.

This chart matters because it prevents a common mistake. Not all 314 MW IT under construction carry the same risk profile. Part of the capacity is already backed by contracts, while another part is still waiting for customers or for a more advanced delivery stage. Anyone reading the entire pipeline as one uniform block will miss where the real risk sits.

Funding became part of the story, not a footnote

The second trigger: during the year the company expanded three bond series, increased its commercial-paper layer, and after the balance-sheet date also received funding for three projects totaling about NIS 19.9 million. That was not accidental. The company explicitly says that given the significant expected increase in investment in the Data Centers projects, it plans to issue new bond series, expand existing ones, and raise bank financing.

The third trigger: the collateral-release notice for Series H, published on the same day as the annual report, is not a technical side event. The company said that after the March 31, 2026 principal repayment it intends to release two pledged Modiin assets, Nisko Management Center and Plot 61 Management Center. The meaning is that Mega Or is no longer managing debt only at the macro level of ratings and amortization. It is already actively working the secured layer through collateral substitution and asset release.

The fourth trigger: there is also real organizational expansion behind the new story. Mega D.C moved from 3 employees at the end of 2024 to 17 at the end of 2025, and by the report date it already had 29 direct employees. That is not cosmetic. It shows the company building an operating base, control-room teams, and electro-mechanical staffing before the new business becomes materially cash-generative.

Efficiency, Profitability, and Competition

The main takeaway from 2025 is that Mega Or's core business really improved, but reported profit improved by far more. The right way to read the year is therefore to separate a strong rent engine from a profit line that already includes a meaningful amount of value booked ahead of cash.

What actually carried the year

The legacy business continues to work. Same-property NOI rose to NIS 309.7 million from NIS 291.0 million, and owner-share NOI rose to NIS 392.3 million from NIS 331.9 million. The increase came from a combination of occupancy gains, acquisitions, and indexation: Big Mega Or Migdal HaEmek, Jumbo Beer Sheva, the Iskor complex, Counterop Management Center, and also the partial contribution from the first Modiin Data Center.

The portfolio mix also remains supportive. In the capital-markets presentation, logistics and industrial contributed NIS 239 million of actual NOI, commercial assets contributed NIS 117 million, yielding land NIS 12 million, photovoltaic systems NIS 10 million, and Data Centers only NIS 10 million. That is a critical point. Even after all the headlines of 2025, the current economics of the company are still carried by the older property layers, not by the new growth engine.

That picture becomes even clearer when looking at occupancy. Logistics and industrial are at 100%, commercial at 98%, and yielding land at 100%. This is a very strong operating base, and it is the reason the market still gives the company credit even as it opens a large new investment front.

Where profit looks stronger than recurring economics

This is the heart of the story. Net profit attributable to shareholders reached NIS 844.3 million, but to get down to management FFO of NIS 291.7 million the company had to strip out NIS 430.1 million of investment-property fair-value changes, NIS 379.4 million of gains from fair-value changes or sales of financial assets, and NIS 61.1 million of impairment-reversal profit in DSKSH. In other words, 2025 was a good year, but it was also a very noisy one.

This chart is not meant to say the profit is unreal. It is meant to show that 2025 profit is not equivalent to recurring cash. That matters especially in a company that is exactly at the point where it needs to convince the market not only that its assets are worth more, but that it can finance the next leap on a stable base.

The Data Centers segment makes that gap even sharper. In 2025 it contributed NIS 13.3 million of revenue and NIS 10.1 million of NOI, but also NIS 233.5 million of revaluation gains. That makes economic sense in a development-driven company advancing land, power, and contracts. Still, it means the market has to hold two different pictures at once: value created, and cash already flowing.

DSKSH belongs to that same middle layer. In 2025 DSKSH contributed NIS 90.4 million to Mega Or, but NIS 61.1 million of that came from impairment reversal and only NIS 29.3 million from the ongoing equity-method profit share. That is legitimate accounting value. It is still not core NOI.

The new competition is not against another mall, but against execution

Mega Or's commercial and logistics properties still operate under competition the company understands very well: stable end tenants, relatively low rent burden versus tenant sales, and large-format properties in locations that allow attractive rents to tenants. In Data Centers, the rules are different. The company itself lists the critical success factors: relationships with Hyperscale customers, access to large and reliable power, connectivity, professional teams, and delivery schedules.

That is exactly where the risk also sits. Supply chains for electrical equipment, backup systems, cooling systems, and communications equipment have stretched out, the number of alternative suppliers is limited, and the local market already includes both international and domestic operators with existing capacity. Mega Or's real-estate advantage is real, but it is not enough on its own. The edge will be proven only if the company delivers on time, under reasonable funding terms, and at a service level that satisfies significant long-term customers.

Cash Flow, Debt, and Capital Structure

This is the most important section in Mega Or's annual report because this is where the difference between accounting value, operating value, and cash value stops being theoretical.

The right cash bridge here is all-in cash flexibility

Here it is not enough to stop at FFO or NOI. The real question is how much cash remains after all the year's actual uses. On the all-in picture, the group generated NIS 253.9 million of operating cash flow in 2025. That is a solid improvement from NIS 167.7 million in 2024. But it sat alongside NIS 914.0 million of investment in investment property and development, NIS 131.5 million of interest paid, NIS 130 million of dividends, NIS 730.7 million of bond repayments, NIS 189.5 million of commercial-paper redemptions, NIS 81.4 million of bank-loan repayments, and NIS 7.3 million of lease-principal payments.

The conclusion is straightforward: the operating business generates cash, but it does not fund the build-out on its own. Year-end cash rose to NIS 1.098 billion not only because of the underlying property engine, but also because of NIS 809.0 million of bond issuance, NIS 589.4 million of commercial-paper issuance, NIS 198.7 million of long-term loans, NIS 822.6 million of securities sales, and NIS 249.4 million of restricted-cash release.

Put differently, 2025 was not a distress-funding year. But it was also not a self-funding year. It was an investment year financed by a combination of a strong legacy operating base and an open debt market.

Even at the parent layer, the cash is moving downward before it moves upward

The picture becomes sharper still when looking at the company-only statements. Cash flow from operations attributable to the company itself was only NIS 150.0 million, while the company extended NIS 428.6 million of loans to subsidiaries and invested NIS 181.8 million in equity-accounted companies. That connects directly to the company's explicit statement that Mega D.C is currently funded through shareholder loans provided by Mega Or.

That point should not be missed. A meaningful part of the growth leap is not yet sitting in a self-contained project-finance structure below the listed entity. At the end of 2025 it is still resting on the parent layer and on its public debt stack.

The macro debt picture is good. The instrument-level picture is less even

At the wide-angle level, Mega Or still looks strong. Net financial debt stood at about NIS 4.1 billion, LTV fell to 46%, debt-to-cap fell to 51%, the company had about NIS 5.1 billion of unencumbered real estate and investments, and the LTV on pledged assets stood at 62%. The broad corporate covenants also still look comfortable: equity of about NIS 3.9 billion is far above the series-level minimums, and the 39% equity-to-balance-sheet ratio remains well above floors of 21% to 24%.

But this is precisely where the reading needs to get more disciplined. Not every debt layer looks the same. In the bondholders' appendix, Series XII ended the year at 79.4% versus an 80% collateral threshold. Series H stood at 74%. So at the group level the picture is comfortable, but at the level of some secured instruments the headroom is much thinner.

This chart surfaces another non-obvious point. The company says it tries to structure recurring annual principal amortization, excluding secured components, around annual FFO. In practice, the disclosed 2026 schedule shows NIS 1.227 billion of principal excluding secured debt versus NIS 291.7 million of 2025 FFO. The practical implication is clear: 2026 depends materially on refinancing, not only on recurring internal cash generation.

The Series H collateral-release notice fits directly into this logic. It shows that the relatively low pledged-asset LTV does in fact allow the company to actively manage the secured layer, release assets, and restore flexibility. But it also reminds investors that this flexibility is being used in practice. It is not a side benefit.

Guidance and Outlook

2026 looks like a clear bridge year. The legacy operating base already knows how to run, but the story the market will measure next year is not another 1% move in occupancy. It will measure whether Mega Or can turn signed capacity, land, power, and presentations into actual delivery without eating too much into its funding flexibility.

Finding one: a meaningful part of the growth is already signed, but not all of it. At the end of 2025, MDCIL-2 had binding lease contracts on 75% of the asset, Beit Shemesh Phase A on 100%, while Beit Shemesh Phase B and Bnei Shimon Phase A had no binding contracts yet. That means the company has already moved beyond the first demand question, but it has not yet closed the full risk layer of the pipeline.

Finding two: management is asking the market to look far out. In the capital-markets presentation it frames a scenario of NIS 1.6 billion of projected NOI and NIS 1.2 billion of projected FFO by the fourth quarter of 2028, versus a representative base of NIS 500 million of NOI and NIS 314 million of FFO in the fourth quarter of 2025. That is a very large gap. It is also exactly why 2026 cannot be judged through a 2028 lens.

Finding three: the practical test in the coming year is much narrower. The company needs to deliver Modiin Phase B and Masmiya during the third quarter of 2026, begin Beit Shemesh delivery from the third quarter of 2026 through the first quarter of 2027, and keep the debt market open enough to migrate part of the shareholder-loan burden into longer funding.

Finding four: even if everything moves according to plan, 2026 still will not be a clean harvest year. Data Centers cash flow needs to grow from a base of only NIS 10.1 million of NOI in 2025, while the company is dealing with expected project costs of NIS 1.0 billion for MDCIL-2, NIS 1.5 billion for each of the first two Beit Shemesh phases, and NIS 1.78 billion for MDCIL-5 Phase A. In its presentation the company also says that across projects under construction it has already accumulated NIS 1.586 billion of construction cost and still expects another NIS 7.18 billion.

This chart is not there to present a conservative target. Quite the opposite. It is there to show the gap between the promise and the near-term proof burden. If the market starts pricing the 2028 numbers in 2026 before seeing actual delivery, actual NOI, and a cleaner funding structure, it will be moving too far ahead of the evidence.

What must happen over the next 2 to 4 quarters for the thesis to strengthen? First, timely delivery of the sites already promised to customers. Second, meaningful NOI growth from the Data Centers business above the roughly NIS 10 million level of 2025. Third, preservation of LTV and covenant room without leaning too heavily on collateral release and commercial-paper rollovers. Fourth, continued strength in the legacy core so that logistics and commercial assets keep supplying FFO even as funding costs and project count rise.

Risks

The first risk is a funding risk, not a demand risk. 2025 showed that the company knows how to find customers and keep legacy assets highly occupied. 2026 will test whether that success can continue to sit on a debt structure that remains comfortable. Any disruption in refinancing, any further material rise in funding cost, or any temporary closure of the debt market, would hit flexibility first and profit later.

The second risk is a quality-of-earnings risk. 2025 net profit includes a large layer of revaluations, financial-asset gains, and an impairment reversal in DSKSH. If capital markets or valuation assumptions move in a less favorable direction, that same layer can reverse much faster than NOI.

The third risk is execution and supply chain in Data Centers. The company itself flags availability of electrical equipment, cooling systems, communications equipment, and contractors as potential friction points. In a business where a large customer expects delivery on time and at a high service level, any delay quickly moves from an execution issue to a reputational issue.

The fourth risk is customer concentration in the new business. The company is explicitly targeting a Wholesale and Hyperscale model, which means a customer base that is smaller, larger-ticket, and more demanding. That can become a quality layer if delivery works. It can also become concentration if one site slips or one major customer changes pace.

The fifth risk is value accessibility. In 2025 Mega Or created real operating value, real accounting value, and also value that still sits in land, revaluations, and tradable assets. Those layers are not equally accessible to common shareholders.

Against those risks, it is also worth watching the market's actual positioning. Short interest as a share of float stood at 1.07% at the end of March 2026, versus a sector average of 0.55%. In January 2026 it had already climbed to 1.81% with an SIR of 7.02, before falling back to an SIR of 2.81. That is not an extreme short setup, but it is also not indifference. Skepticism has eased since the peak, not disappeared.

Conclusions

Mega Or finishes 2025 with a stronger operating core, a new growth engine that has already moved beyond the first demand proof point, and a balance sheet that still looks impressive even after another investment-heavy year. What supports the thesis right now is rising NOI, rising FFO, near-full occupancy, real collateral flexibility, and a debt market that is still open to the company. What blocks a cleaner thesis is the sharp transition from a veteran yield real-estate business into something that also behaves like a development and infrastructure platform, with delivery schedules, funding costs, and revaluations that move ahead of cash.

In the short to medium term, the market is likely to spend less time asking whether Mega Or is a good real estate company, and more time asking whether 2026 will prove that Data Centers are moving from a valuation engine into an NOI and cash engine. That is where the next market reading will be written.

Current thesis: Mega Or's real-estate core is already proving its strength, but the center of gravity has shifted to the company's ability to fund and deliver the Data Centers leap without leaning too heavily on revaluations and refinancing.
What changed relative to the earlier read: 2025 is no longer a year in which Mega Or is judged only by NOI, occupancy, and asset multiple. It is a year in which the funding structure and the quality of the value being booked matter almost as much as the operating core.
Counter-thesis: the cautious read may be too strict because the company still has 99% occupancy, roughly NIS 1.45 billion of liquidity, NIS 5.1 billion of unencumbered assets and investments, and a material part of the new capacity already covered by binding agreements.
What could change the market interpretation in the short to medium term: on-time delivery of Modiin Phase B and Masmiya, new funding at a reasonable cost, and a visible rise in Data Centers NOI beyond the 2025 level.
Why this matters: because Mega Or is no longer being judged only as a stable income-producing real estate company, but as a company that now has to prove that the gap between value created, value booked, and value arriving as cash for shareholders does not widen too much.
What must happen over the next 2 to 4 quarters: the thesis strengthens if delivery happens on time, if NOI from the new activity starts to scale, and if the funding layer remains open without acute pressure on collateral. It weakens if 2026 remains another year of revaluations and debt expansion without a parallel step-up in NOI and cash.