Dalia Energy 2025: The Existing Fleet Funds the Next Leap, but Ashkol Value Is Still Not Yet Accessible

Getting To Know The Company

Dalia Energy can look, at first glance, like a large power-generation company that simply opened another growth leg. That is true, but only partly. In practice this is an Israeli power platform with three very different layers sitting inside the same group: the existing Dalia station that still generates most of the cash, a 75% holding in the Ashkol site with its generation units, land and planning option, and two new H Class projects, Dalia 2 and Avshal, that are meant to push the platform toward roughly 3.3 GW by the end of the decade. A superficial read of 2025 through net profit of NIS 154.4 million, cash and deposits of NIS 925.1 million, and the real-estate valuation at Ashkol can make this look like a clean maturity story. That is too shallow a read.

What is working now is clear. Dalia remains a strong engine, operating cash flow stayed high at NIS 713.1 million, Dalia 2 moved after year-end from signed financing into first drawdown and full repayment of the bridge facility, and Ashkol gives the group a scale base it did not have before the site was delivered in June 2024. This is no longer a thesis built on "if financing arrives" or "if a license is granted." Large parts of the platform already exist, already operate, and already have an identifiable funding path.

The active bottleneck sits somewhere else: how much of that value is actually accessible at the listed-company layer, and how quickly. Ashkol can look strong at site level, Dalia 2 can look funded, and the data-center option at Ashkol can look attractive, yet for the company’s creditors and capital providers the real test still runs through distribution restrictions, shareholder loans, 75% ownership, and the ability to execute two heavy new builds in parallel without turning all that value into something that stays locked behind a thick financing stack for years. That is the core of the 2026 read.

There is also an actionability constraint worth stating early. This is a public issuer with a bond-only trading surface, not a listed equity. So the public market is giving a credit read much more than a classic equity rerating read. That does not reduce the importance of the projects, but it does make questions of liquidity, distributions, covenant headroom and execution discipline more important than any abstract discussion of "value unlocking."

The quick economic map looks like this:

All of this sits in Israel only, and that matters. There is no dispersed country risk or a long list of operating currencies. But it also means the entire story is exposed to the Israeli power market, local regulation, local tariffs, and the pace of approvals from the Electricity Authority, Noga, the Israel Land Authority and planning bodies.

That chart sets the thesis early. Dalia Energy is not trying to add one more small asset next to an existing station. It is trying to replace part of Ashkol’s older base, preserve the Dalia cash engine, and build two new H Class units in parallel. That is why 2026 to 2029 are not decorative years. They are years of shifting the platform from today’s base into tomorrow’s base.

Events And Triggers

Dalia 2 has already moved from a financing story into an execution story

The first trigger: Dalia 2 is no longer only a project with a license and financing documents. The conditional license arrived in April 2025, the main equipment agreement with Siemens was signed in June, the EPC contract followed in August, senior financing documents were signed in December, and by March 2026 the company had received the building permit, the Electricity Authority’s confirmation that the financial-close conditions were satisfied, and the first senior drawdown that fully repaid roughly NIS 773 million of bridge debt.

The meaning is two-sided. On the positive side, Dalia 2 is no longer a thesis about whether financing will show up. It is now an execution project. On the other side, once bridge financing is replaced with senior financing, the key question shifts from year-end liquidity optics toward schedule discipline, budget control, and compliance inside a project whose total investment is estimated at about NIS 3.8 billion, excluding financing costs during construction.

One more detail matters. Dalia 2’s financing already looks like long-cycle infrastructure financing, not a short bridge. Roughly 75% of the debt is expected to transition into a fixed-rate euro tranche based on the euro swap curve plus a margin of 2.5% to 3.1%, alongside a shekel portion of roughly 25%. That improves long-term funding visibility, but it also pushes the project deeper into execution, coverage-ratio and hedging discipline.

Avshal is still one stage behind

The second trigger: Avshal has advanced, but it is still not where Dalia 2 is. In January 2025 the company signed a non-binding MOU with Bank Hapoalim for up to NIS 5.3 billion of senior financing, including about NIS 0.9 billion of refinancing of existing obligations. A positive grid survey arrived in April, a bridge loan of up to NIS 600 million was taken in May, and the project received a conditional license for 850 MW in February 2026.

The interesting part is not the license by itself, but the structure around it. The company says that about NIS 700 million of equity has effectively already been injected through the Ashkol acquisition as shareholder funding, and that under the current financing outline no additional owner injection should be required. If that converts into binding financing documents, it is a real strength. If not, it can quickly become an open funding question. That is why Avshal should not yet be read as if it already sits at the same de-risking stage as Dalia 2.

There is also a project-level friction point that can be missed at first glance. Avshal’s conditional license is for 850 MW, but export capacity is capped at 670 MW until early July 2035. So even if the project closes on time, that is not necessarily the economics of a full 850 MW output from day one. The installed capacity on paper and the deliverable output in the early years are not quite the same thing.

Ashkol has a real value option, but not yet a contract

The third trigger: In February 2026 Ashkol signed an exclusive MOU, for up to six months, regarding the construction and operation of a data-center project at the site. The numbers are meaningful: about 50 dunams of land, at least 130 IT MW, a lease term of 24 years and 11 months, and initial annual rent estimated at 2026 prices in the NIS 30 million to NIS 50 million range. Ashkol Energies may also hold up to 30% in the joint entity, and electricity is expected to be sold by one of the group’s licensed suppliers at an agreed discount to tariff.

That sounds like a strong real-estate and infrastructure option, but this is exactly where discipline matters. The document is not commercially binding. It remains subject to negotiations, more information and an eventual binding agreement actually being signed. And the option is not free to monetize: the company estimates roughly NIS 80 million of soil-treatment costs and about NIS 65 million of demolition and evacuation costs at the site. So there is no reason to give the data-center thread the same weight today that one would give to contracted cash flow from an operating power asset.

Efficiency, Profitability, And Competition

The main 2025 story is not a clean profit growth year. It is a transition between value engines. Revenue declined 2.2% to NIS 2.646 billion, operating profit fell 8.6% to NIS 425.7 million, yet net profit still edged up to NIS 154.4 million mainly because finance expense fell to NIS 208.3 million as CPI rose more moderately than in 2024. That matters a lot. It means the bottom line improved not because the group is already harvesting Dalia 2 or Avshal, but because the existing stations kept working and the funding environment was somewhat less heavy.

Dalia remains the core engine

Dalia station still funds the story. On the internal-management view, the bilateral segment posted adjusted EBITDA of NIS 748.7 million in 2025, versus NIS 786.9 million in 2024. That is lower, but still a very high profitability base relative to the listed-company layer. Just as important is how that number was built. Revenue from private customers rose to NIS 2.446 billion from NIS 2.201 billion, while revenue from the system operator fell to NIS 199.3 million from NIS 503.3 million. Electricity sold to private customers held almost flat at 5,723 GWh versus 5,771 GWh in 2024, while sales to the system operator fell sharply to 444 GWh from 1,987 GWh.

That chart says something important about revenue quality. 2024 included a more unusual window of sales to the system operator and market-model exposure, while 2025 looks much more like a return to Dalia’s real core, selling to private customers and suppliers. That is positive for the business read, because it puts weight back on the engine the company actually built around, but it also reminds readers that the group still relies very heavily on one station to carry the transition years of the next projects.

Ashkol adds scale, but not in a straight line to listed-company economics

The easiest mistake in reading Dalia Energy is to take Ashkol’s EBITDA and assume it flows almost automatically into the listed company. That is not what happens here. On the management view, the group’s share of relative EBITDA from the associates stood at NIS 429.9 million in 2025, versus NIS 250.9 million in the comparison period, and Ashkol Yitzur’s EBITDA on a 100% basis stood at NIS 573.3 million. In the fourth quarter alone, Ashkol Yitzur’s EBITDA excluding one-offs reached NIS 119.4 million versus NIS 69.8 million a year earlier.

But the IFRS view is more complicated. The market-model segment is shown in internal reporting on a proportional-consolidation basis, while under IFRS the Ashkol entities are treated under the equity method. That means there are heavy adjustments, share-of-loss lines, and top-holding-company costs sitting between site-level EBITDA and the consolidated line.

That is the heart of the distinction between value created at the asset layer and value accessible at the listed-company layer. Ashkol does add generation scale, EBITDA and land value. But not every shekel created at the site reaches the top in the same form or on the same timetable. There is 75% ownership, project financing, distribution restrictions, a separate real-estate layer, and an accounting move from internal proportional reporting to equity-method treatment in the audited statements.

The fourth quarter was strong, but it proves only part of the case

The fourth quarter of 2025 suddenly looked much cleaner: revenue rose to NIS 578.0 million from NIS 490.8 million, operating profit climbed to NIS 88.3 million from NIS 2.7 million, and net profit swung to NIS 63.0 million from a loss of NIS 41.3 million in the fourth quarter of 2024.

That does matter, because it shows Ashkol already entering the quarterly read more constructively and finance costs becoming less heavy. But it does not solve the bigger question. A strong fourth quarter still does not change the fact that 2026 to 2029 will require the company to fund and build two new units at once while keeping the existing base available and productive.

Competition is shifting too. On one side, opening the Israeli power market to more competition supports the expansion of private customers and private suppliers, which helps Dalia’s bilateral business. On the other side, that same opening increases the number of private generators and suppliers and can pressure margins. In Ashkol’s case, the tariff formula also makes the economics more sensitive to the US dollar, so the economics of a large market-model plant are not only a function of electricity demand but also of tariff structure and FX exposure.

Cash Flow, Debt, And Capital Structure

This is the right place to use an all-in cash flexibility lens, meaning how much cash is actually left after real cash uses, not a normalized picture that strips out construction spending as if it no longer belongs to the story. On that basis, 2025 looks less comfortable than the bottom line. The group generated NIS 713.1 million from operating activity, but spent NIS 845.3 million in investing activity and received NIS 374.1 million from financing activity.

That chart captures what 2025 really was, a carrying year rather than a harvest year. The existing stations still produce a lot of cash, but that cash is almost fully absorbed by investment, refinancing and capital moves. That is exactly why the group cannot yet be read as if the new platform already funds itself.

Year-end liquidity looked weaker than the post-balance-sheet reality, but it is not fully resolved

At year-end 2025 the company held NIS 537.5 million of cash and cash equivalents, NIS 139.1 million of short-term deposits and NIS 248.6 million of restricted deposits, or about NIS 925.1 million of cash and deposits. At the same time, current liabilities jumped to NIS 1.500 billion from NIS 754.3 million a year earlier, mainly because Dalia 2’s bridge loan sat inside current liabilities.

That means year-end optics were tighter than the economic position after March 2026. And indeed, once the financial-close conditions were completed, the Dalia 2 bridge was fully repaid out of the first senior drawdown. That removes the classification problem and reduces the immediate liquidity pressure. But it is important not to stretch that conclusion too far. The issue did not disappear. It simply moved from a bridge-financing test into an execution, budget and simultaneous-build burden.

The debt frame is manageable, but distributions are still behind glass

The company has three public bond series, and the bond rating stands at A3.il with a stable outlook. At year-end 2025, current bank and institutional maturities stood at NIS 913.0 million and current bond maturities at NIS 221.3 million, while non-current bank and institutional debt stood at NIS 1.313 billion and non-current bonds at NIS 2.096 billion. Equity increased to NIS 2.608 billion.

At the framework level, this does not look like a near-edge condition. But the more important layer is the distribution restriction. The company reported distributable profits of NIS 304.5 million at year-end 2025, yet did not distribute a dividend, and the board had already stated in early 2024 that it did not intend to distribute in the coming years in order to preserve debt cushion following the Ashkol acquisition and the development of the new units. That is an important signal. It says explicitly that even when value exists, it is not currently meant to flow upward.

Ashkol value is real, but not yet free

The clearest example of the gap between created value and accessible value sits at Ashkol. The Ashkol Energies real-estate layer was valued at NIS 1.736 billion at the end of 2025 on a 100% basis. There is also existing rent in that layer, about NIS 30 million per year from Ashkol Yitzur’s use of the site, although the company’s investor presentation says that amount is expected to fall to about NIS 15 million after the steam units close. So even inside the real-estate layer itself, part of the current economics still needs to be replaced by new economics.

And above that sits a financing structure that does not allow the valuation to be treated as if it were fully free. The refinanced real-estate loan stands at NIS 780 million, with an additional NIS 50 million facility, and its financing documents prohibit distributions until the loans are repaid. The company also says explicitly that it depends on cash flows from subsidiaries, whether through repayment of shareholder loans or through dividends, and that its ability to receive those flows can be limited by distribution restrictions embedded in senior debt documents. That is exactly the difference between "there is value" and "there is accessible cash."

Outlook

Before going into the details, these are the four non-obvious conclusions that should lead the 2026 read:

• Dalia 2 is no longer a search-for-financing story. After March 2026 it is an execution, schedule and budget story.
• Avshal is not there yet. It has a conditional license and a financing outline, but not a full financial close.
• Ashkol’s non-power value is real, but still optional. The data-center thread remains an MOU, not a binding contract.
• 2026 is a bridge year. The existing fleet is expected to carry the group while two new units consume capital and management focus.

2026 is a bridge year, not a harvest year

The company is targeting Dalia 2 for commercial operation by the end of 2028 and Avshal for commercial operation in the first half of 2029, after a financial close it hopes to reach by mid-2026. That means the market is being asked to look three layers forward at once: what generates cash today, what is now under funded construction, and what still needs to move from MOU and conditional license into a full project.

That is exactly why 2026 reads like a bridge year. The existing stations, above all Dalia, are expected to keep carrying the group’s service cash generation. In parallel, Dalia 2 is supposed to move from financing completion and first drawdowns into deeper execution. And Avshal is supposed to prove that the story around already-injected equity, non-recourse financing and a mid-2026 close is actually real.

What has to happen at Dalia 2

At Dalia 2 the key elements now look relatively clear. The conditional license is in hand, senior financing is signed, the financial-close conditions were completed, and the first drawdown has already been made. From here the market will focus mostly on execution signals: construction pace, budget discipline, hedging of euro and interest-rate exposure, and the ability to hold the end-2028 target without material cost slippage.

The good news is that the biggest financing hurdle is already behind the company. The less comfortable part is that once financing is in place, it becomes much harder to hide behind bridge mechanics or conditions precedent. The next requirement is simply to build.

What is still open at Avshal

Avshal is still the more open thread. On one side it has a financing MOU, a conditional license, a bridge facility, a completed asset split, and a company that says it is working toward financial close by mid-2026. On the other side it still needs permits, tariff approval, full financing documents, and proof that the structure meant to avoid another owner-equity injection really holds.

The implication for credit readers is straightforward. If Avshal closes on time and on terms close to what the company presents, 2025 will look in hindsight like the year in which the funding bridge to the next stage was built. If Avshal slips, the group will remain for longer much more dependent on the existing stations and on debt already embedded in the structure.

What the market is likely to test in the near and medium term

In the near term, three things can shift the market read quite quickly. The first is Dalia 2, any execution update that shows schedule and budget discipline will be read as direct support for the thesis. The second is Avshal, the move from conditional license and financing MOU into full financial close would remove a large part of the open skepticism. The third is Ashkol beyond power, if the data-center MOU turns into a binding agreement or a clearer monetization path for building rights emerges, the debate over site value will sharpen quickly.

There is also a smaller quarterly catalyst, but it is not thesis-changing. In March 2026 Ashkol Yitzur signed a settlement with its insurers under which about NIS 23.7 million will be recognized in the first quarter of 2026 in connection with steam unit 9, which had been down from June 2024 to June 2025. That helps the headline of the quarter, but it does not solve any structural question.

Risks

The first risk is dual execution load

Dalia 2 and Avshal are not just two names on the same slide. They are two separate projects, under two different financing layers, with different permitting paths and different capital needs, progressing almost in parallel. Even if each one looks manageable on its own, the combination raises execution risk, management load and sensitivity to delay.

The second risk is FX, CPI and regulation asymmetry at Ashkol

The company describes the asymmetry clearly. At Dalia there are partial built-in protections against dollar and CPI changes through the production component and the availability payments. At Ashkol, by contrast, a weaker dollar can erode profitability because the margin element in the regulated tariff is dollar-linked, while part of debt service and fixed costs are shekel-denominated and CPI-linked. On top of that, a stronger euro raises maintenance and construction costs, while higher CPI increases the burden of CPI-linked debt.

The third risk is availability of the older units

Ashkol still includes four older steam units, and the company itself says spare-part availability for them is limited and depends on ad hoc orders. Unit 9 already entered unplanned maintenance shortly after site delivery and remained down for almost a year. That ended in an insurance settlement, but it also demonstrated the risk: the existing base has to stay available precisely during the years in which the group is building its future.

The fourth risk is trapped value

The company depends on cash flow from subsidiaries, and project and real-estate financing documents include distribution restrictions. That means even when Dalia, Ashkol or the site’s land layer create value, that value may not be available when and where the listed company’s creditors and capital providers need it. That is especially important in a public bond issuer with no traded-equity layer.

The fifth risk is gas and dependence on critical suppliers

Dalia station relies fully on its long-term operation and maintenance agreement with General Electric, and Ashkol depends in part on General Electric and Siemens for core equipment support. In addition, Dalia and Ashkol are committed to minimum gas-purchase volumes, and in an extreme scenario of unusually low consumption the maximum TOP exposure could run at roughly $160 million to $190 million per year. As of year-end 2025 actual consumption remained materially above TOP commitments, so this is not an immediate threat. But it is a good reminder that a sharp drop in availability or demand is not only an operating risk, it is also a contractual one.

Conclusions

Dalia Energy ends 2025 as a much more tangible and much larger power platform than the one investors were looking at before the Ashkol acquisition. That supports the thesis. The main blocker is that the next few years will be judged less by the mere existence of assets and more by the company’s ability to move value through financing, ownership and execution layers into the listed-company layer. In the near and medium term, the market read will be driven mainly by Dalia 2, by Avshal financial close, and by whether Ashkol remains mostly paper value or starts becoming a real cash path.

Current thesis in one line: the existing fleet still funds the group, but 2026 is a bridge year in which the value being built at Dalia 2, Avshal and Ashkol still has to prove it can move toward accessible cash instead of staying locked behind debt, regulation and ownership structure.

What changed: the read on Dalia 2 changed materially after year-end, from a project still working through conditions into one that has already drawn senior debt and repaid its bridge loan. At the same time, Ashkol has shifted from being mainly a big acquisition event into a more complex question of site profitability, capital structure, and monetization of value outside the power core.

The counter-thesis: the cautious read may prove too conservative. If Avshal closes on time, if Dalia 2 stays on schedule without major cost slippage, and if the Ashkol data-center thread advances faster toward a binding agreement, the market may discover that more of the value already sits in a much less speculative place than it currently assumes.

What could change the market interpretation: a sequence of milestones, not one number. Execution updates from Dalia 2, tariff and financial close at Avshal, and binding progress on the server-farm option, these are the events that can clean up or burden the read far more than another quarter of decent net profit.

Why it matters: in a power and infrastructure platform, value is not determined only by MW, EBITDA or a land valuation. It is determined by how much of that value can actually move up through project financing and holding-company layers and become real, accessible cash.

What must happen over the next 2 to 4 quarters: Dalia 2 has to stay on track on execution and budget, Avshal has to move from conditional license and financing outline into signed financial close, and the existing stations have to keep generating cash without major availability issues precisely while the group burns capital on the next stage. What would weaken the thesis is delay at Avshal, cost slippage at Dalia 2, or a continued situation in which Ashkol looks excellent at asset level but remains largely closed off at the company layer.