Afcon Holdings 2025: Less Volume, More Quality, and the Energy Question

Getting to Know the Company

Afcon is classified in the market as a holding company, but that label misses the actual economics. In practice, this is a technology infrastructure and execution group with five main engines: control and technologies, trade, EPC and building systems, multimedia and communications, and renewable energy. A superficial read of the annual report could focus on the top line and conclude that 2025 was a slowdown year. That is not the right read. Revenue fell 5.6% to NIS 1.683 billion, but operating profit rose 50.1% to NIS 121.3 million, EBITDA rose 33% to NIS 200.1 million, and net profit rose 30.8% to NIS 83.2 million.

What is clearly working now is the quality of the operating mix. Afcon deliberately cut back weaker civil construction work, added adjacent activities such as Magma and IMS, and improved profitability in a year when revenue was lower. What is still not clean is the energy layer and some older tail risks. The renewable portfolio is progressing, but much of the value still sits at the project, partnership, and financing level rather than as simple, accessible value at the listed-company level. At the same time, the assisted-living dispute is still sitting inside the cash story.

That matters now because Afcon is no longer being screened only as a contractor whose value rises and falls with project volume. In 2025 it tried to prove that it can trade some volume for better quality. If that holds even after a large hi-tech project winds down, and if the energy layer moves from project math to shareholder-accessible value, the way the market reads the company can change. If not, 2025 will look more like a strong transition year than a durable reset.

What is still missing for a fully clean thesis is straightforward: core margins need to remain high without the help of a few specific projects, the assisted-living dispute needs to stop consuming cash, and the energy portfolio needs to move from forecasts, MOUs, and 100%-basis project tables into commissioning, monetization, or real cash visibility at the public-company level.

Four things worth holding from the start:

• The revenue decline does not reflect broad-based weakness. It mainly reflects a deliberate retreat from lower-quality civil construction.
• The profitability improvement is real, but part of it comes from mix and exiting losses, not from broad demand acceleration.
• The renewable portfolio may be valuable, but much of that value still sits above the common-shareholder layer.
• The market is still not fully buying the story, and the short data shows it.

Events and Triggers

The first trigger: 2025 was a real strategic reshaping year, not just a better reporting year. Afcon completed the acquisition of control in Magma in February, bought 52% of IMS in March, increased its holding in Sherut Hogen, and kept pushing the group toward higher-value activities in control, services, multimedia, and technology infrastructure rather than broad civil construction. The revenue drop in EPC and building systems, from NIS 916.8 million to NIS 747.0 million, is the price of that shift.

The second trigger: the banks amended Afcon’s financial covenants in March 2025 to better fit both the core operating business and the company’s renewable-energy development activity. The point here is not that Afcon was facing an immediate covenant problem. The point is that lenders are already looking at the group through a different mix of operating activity and development exposure, which is exactly how investors need to read the business now.

The third trigger: the energy layer moved forward on two tracks that do not fully line up. On one hand, the Spanish Arjona project secured a 10-year financial hedge for about 60% of expected power production, financing was closed in Spain, and financing was also closed in October 2025 for a roughly 100 MW solar project in Poland. In November 2025 Afcon also signed a US joint venture with Namco, where it will hold 25% and seek to develop projects across up to 20 properties in the first phase, with a total expected cost of up to about $185 million subject to roughly 65% financing. On the other hand, the December 2025 immediate report only extended the Sunflower MOU for another 60 days and explicitly said there was still no certainty the deal would close.

The fourth trigger: the balance-sheet date was followed by meaningful order intake. The report itself says large jobs of roughly NIS 400 million were received after year-end. It also gives the building blocks: a January 2026 substation project worth about NIS 310 million, February 2026 government work of about NIS 70 million, and a January 2026 DM scope increase of about NIS 58 million. That does not solve the revenue question by itself, but it does show that the lower 2025 revenue base did not come with an empty pipeline.

The fifth trigger: capital allocation moved in two directions at once. Afcon issued סדרה ה', expanded סדרה ד', and also completed a private equity and warrants placement in December 2025 for about NIS 74.1 million. At the same time, it paid a NIS 20 million dividend in April 2025 and another NIS 50 million dividend after the balance-sheet date in March 2026. That signals confidence, but it also uses cash while not all value layers have yet become liquid or fully accessible.

Efficiency, Profitability and Competition

The main 2025 story is quality, not growth. Gross margin rose from 12.4% to 15.7%, operating margin rose from 4.5% to 7.2%, and EBITDA margin rose from 8.4% to 11.9%. That is a sharp change for a group that used to be read mainly through execution volume.

Less Volume, Better Margins

The clearest proof is inside EPC and building systems. Revenue in that segment fell by NIS 169.7 million, yet operating profit jumped from NIS 17.8 million to NIS 46.2 million and EBITDA rose from NIS 39.8 million to NIS 68.9 million. That is the strategic shift in one line. Afcon explicitly links the revenue decline to lower civil engineering and construction activity, while explaining that profitability improved due to a better outcome in a large international hi-tech project, improvements across the rest of the segment, stronger maintenance and service activity, and lower losses in civil construction.

That is real progress, but it is not yet the same as proven recurrence. Part of the improvement came from what was removed and part from where a specific project sat in its execution curve. To keep the new read intact in 2026, Afcon needs to show that the next backlog produces similar economics rather than merely benefiting from the cleanup of the old one.

The Core Segments Did Improve

Control and technologies grew revenue to NIS 464.8 million. Operating profit there slipped slightly to NIS 42.8 million, but the driver mix matters more than the headline. Fire detection was weaker, while security-related activity and the first-time consolidation of Magma supported the business. This is a mix shift toward activity with stronger technological depth and better strategic fit.

Trade grew to NIS 211.8 million of revenue and NIS 27.3 million of operating profit, mainly due to first-time consolidation of two companies in high-voltage electricity and air-handling systems. This is not the largest segment, but it is profitable, useful to the rest of the group, and consistent with the move toward higher-quality adjacent activity.

Multimedia and communications was more mixed. Revenue was essentially flat at NIS 321.1 million, while operating profit slipped to NIS 20.8 million. Multimedia improved, communications weakened. That means this segment still matters, especially around government and multi-year service activity, but it is not the single next leg of the thesis on its own.

Backlog Fell, but the Mix Improved

Group backlog fell 20.1% to NIS 1.867 billion. On the surface that is a warning sign. In context, it mainly reflects the deliberate reduction in civil engineering and construction. EPC backlog still stood at NIS 1.107 billion, control at NIS 385 million, and multimedia and communications at NIS 354 million. That is still a substantial base of work, and the report also says roughly NIS 400 million of large orders were received after year-end.

What matters is not just how much backlog is left, but what kind of backlog is left. Afcon is trying to replace lower-quality or riskier project volume with service, systems, substations, multimedia, electrical infrastructure, and other technology-heavy work. If that transition is real, smaller backlog can still mean better economics.

Cash Flow, Debt and Capital Structure

all-in cash flexibility: that is the right cash frame here. The central question is not how much the business can generate before every use of cash, but how much flexibility remains after real uses of cash are absorbed. Afcon ended 2025 with NIS 282.6 million of cash and cash equivalents, up from NIS 176.6 million a year earlier. That came after NIS 62.8 million of operating cash flow, NIS 53.4 million of investing outflow, NIS 97.6 million of financing inflow, and a NIS 20 million dividend during the year. This is not a maintenance-cash-generation bridge. It is the full cash picture.

The good news is that Afcon looks better inside that framework than a quick headline might suggest. The less good news is that one large unresolved drain still sits inside the cash story: the assisted-living project. The 2025 operating cash flow included a negative NIS 48 million drag from that project. That cuts both ways. It implies the underlying core is stronger than the headline operating cash flow suggests, but it also means the cleanup is not complete until the drain stops or some of that cash comes back.

The legal dispute is therefore not a footnote. In May 2025 Afcon received a partial removal notice from the project, a demand to call guarantees of roughly NIS 17 million plus about NIS 4 million of indexation, and a NIS 57 million claim. The guarantees were called in June 2025. The company believes, based on legal advice, that the odds of recovering the guarantee cash are high and that the claim is more likely to be dismissed than accepted. Maybe so. But cash that already left the system is a fact, not an assessment.

From a balance-sheet perspective, Afcon ends the year in a reasonable place. Equity rose to NIS 691.3 million. Working capital rose to NIS 460.6 million. The company had about NIS 100 million of unused cash credit lines and NIS 861 million of guarantee lines, of which NIS 557 million were used at year-end. Gross financial debt stood at NIS 695.9 million, but net financial debt excluding non-recourse debt was only NIS 279.4 million. Net financial debt to balance sheet, excluding non-recourse, stood at 13.1%.

Just as important, this is no longer a covenant-tension story. At year-end, tangible equity stood at NIS 534.9 million versus a NIS 200 million minimum. The tangible-equity-to-tangible-balance-sheet ratio stood at 28.28% versus an 18% minimum. Debt coverage stood at 1.94 versus a ceiling of 5.4. Net debt plus weighted guarantees to tangible equity stood at 0.85 versus a ceiling of 3.0. The bank question is no longer whether Afcon is close to a cliff. It is whether it will allocate capital and execute growth with enough discipline.

That said, the operating finance load is still real. Average customer credit days were about 130, supplier credit days were about 112, and average open customer credit was about NIS 285 million versus NIS 270 million to suppliers. This remains a business that requires working capital, guarantees, and tight execution management. It just enters 2026 from a more comfortable position than before.

Some tail risks also remain outside the assisted-living story. In the wind-farm financing, the company had already made early repayments, absorbed an interest step-up due to rating and turbine issues, received a reservation-of-rights letter from lenders in January 2026, and classified NIS 13 million as current. That is not a group-defining crisis, but it is a reminder that the “other” bucket has not stopped being noisy.

Forward View

Before looking at 2026, five non-obvious findings matter:

• 2025 proved that Afcon can earn better on a different mix. It did not prove that the group has already returned to a durable revenue growth path.
• A meaningful part of the margin improvement came from mix and from reducing weaker civil-construction exposure, not from broad demand acceleration.
• The energy portfolio may be real, but a large share of the attractive numbers still sits at the project and partner level, not directly at the public-company level.
• The financing story has moved away from covenant stress and toward capital-allocation discipline and monetization discipline.
• The short data says the market is still skeptical enough to demand proof.

2026 Looks Like a Proof Year

The right label for 2026 is a proof year. Not a rescue year, and not yet a breakout year. Afcon enters it with a stronger base, still-large backlog, meaningful post-balance-sheet orders, improved profitability, and a decent cash cushion. But it also enters it after a year in which part of the improvement came from removing weaker work rather than from broad operating acceleration. That distinction matters. Stopping bad activity lifts margins, but it does not automatically build the next growth engine.

In the Core Business, the Test Is Conversion

In the core business, Afcon now needs to show that 2025 margins were not overly dependent on one large project entering a favorable late stage or on weaker civil construction fading out. The new backlog, including substations, service work, multimedia, government contracts, and electrical infrastructure, has to convert into revenue without rebuilding the old problems of weak pricing, heavy cash drag, or execution slippage.

The fact that after year-end Afcon added work worth roughly NIS 310 million, roughly NIS 70 million, and another roughly NIS 58 million shows there is business to execute. The next question is whether there is enough pricing discipline and execution quality to keep the 2025 economics intact.

In Energy, the Question Is Not Whether There Is Value, but How It Reaches Shareholders

This is where the easy reading fails. In the project tables, Afcon shows projected revenue, EBITDA, and FCF for projects in Poland and Spain, and also outlines a broader future portfolio in the US and Israel. But the report itself makes clear that project revenue is shown on a 100% basis, including partner shares, and that project EBITDA and FCF are project-level metrics. That is not the same as cash flowing directly into Afcon’s consolidated revenue line or directly to listed-company shareholders.

This is the key bridge in the thesis. Energy may become a strong value engine, but it still needs to pass through several layers before that value becomes accessible: construction completion, commercial operation, output assumptions, pricing assumptions, debt service, partner economics, and possibly a sale or some other monetization path. That is why the fact that the Sunflower MOU was merely extended rather than closed matters a lot more than it might first look. It tells you that as of late 2025 Afcon was still searching for the right way to crystallize that layer.

There is real substance behind the portfolio. The two Grabno projects in Poland are in advanced construction and are expected to start commercial operation by the end of 2026. The Spanish Arjona project has a 10-year hedge covering roughly 60% of expected power output and contractually expected revenue of about EUR 21 million over the hedge period. Pepino and El Lobo are also meant to move toward operation in 2026, although El Lobo has already seen exclusivity lapse and the parties are now in talks over a reduced price. In the US, the Namco JV can be interesting, but the need to begin construction by mid-2026 to benefit from tax incentives highlights how tight the execution clock already is.

What the Market Is Likely to Test Next

In the near term, the market is unlikely to reward another ambitious narrative on its own. It will test three things. First, whether core margins hold as the mix of active projects changes. Second, whether the assisted-living dispute stops showing up as a real cash drain. Third, whether the energy layer gets closer to commissioning, monetization, or some other concrete path to shareholder value.

Risks

The first risk is recurrence risk. Good margins generated by a better mix and the runoff of weaker construction can still fade if the next backlog carries lower quality or if the large hi-tech project contribution is not replaced by economics of similar quality.

The second risk is value-access risk in energy. Afcon may be building a valuable portfolio, but shareholders do not own a presentation. They own the route through which project value becomes accessible, whether via commissioning, upstream distributions, refinancing, or a sale. As long as the Sunflower process remains non-binding and most projects are still in development or construction, there is still a gap between value created and value accessible.

The third risk is operational cash-flow risk. Customer days of 130, used guarantees of NIS 557 million, and a recurring need for project support mean that even a profitable group can feel tight if several projects stall at the same time. The assisted-living dispute is the clearest proof of that.

The fourth risk is that tail issues remain louder than management would like. The “other” segment is still loss-making, and the wind-farm financing still carries friction with lenders. That is not the central problem today, but it does make it harder for the market to assign Afcon a clean, higher-quality multiple.

Short Read

The short data does not tell a story of extreme bearish positioning, but it does show growing skepticism. Short float rose from 0.20% in mid-November 2025 to 2.22% by late March 2026. That is not extreme in absolute terms, but it is a sharp increase. More importantly, SIR rose from 1.4 to 8.8, well above the sector average of 3.138. In other words, the outright short position is still moderate, but liquidity makes that position meaningfully tighter.

My read is that the short case is not built around an immediate balance-sheet crisis. It is built around the idea that the market may be too quick to extrapolate the 2025 margin improvement before proving recurrence and before the energy layer or legacy issues are truly resolved.

Conclusions

Afcon ends 2025 as a better company than it was entering the year. The core operations are more disciplined, profitability is materially better, the balance sheet is nowhere near a covenant squeeze, and post-balance-sheet order intake says the pipeline is still alive. The main block remains the same one: not all value created is yet accessible to shareholders, especially in energy, and not all legacy friction is gone, especially in assisted living and in the “other” bucket.

Current thesis in one line: Afcon has shown it can earn more with less volume, but 2026 still needs to prove that the upgrade can repeat without the help of a few specific projects and with a clearer route for energy value to reach shareholders.

What changed in the way Afcon should be read is the focus. This is now less a story about raw scale and more a story about business quality. The strongest counter-thesis is that 2025 profitability was helped by a favorable project stage and by the runoff of weak construction, and may therefore fade as the backlog turns over. What can change the market reading over the next few quarters is the ability to convert new orders into profit, move energy closer to monetization, and remove the cash overhang from the assisted-living project. That matters because if Afcon locks in a better mix without losing control of working capital and financing discipline, it stops looking like a plain execution-volume contractor and starts looking like a technology infrastructure platform with real optionality.

Over the next two to four quarters, Afcon needs to pass three clear tests: keep core margins firm even as the project mix evolves, stop the assisted-living dispute from draining cash, and move the energy layer from projected value toward commercially accessible value. What strengthens the thesis is more quarters with high margins, visible commercial and financing progress in energy, and a cleaner cash picture. What weakens it is a return to chasing volume, a renewed erosion in margins, or another year in which energy remains mostly a promise layer above the listed company.