Elad 2025: The Balance Sheet Is Cleaner and AI Is Advancing, but the Real Test Still Sits in Service Productivity

Company Overview

Elad is still, first and foremost, a large Israeli IT services and integration platform, not a pure product company. That sounds obvious, but it is the right starting point for reading 2025. Almost all of the big numbers still come from projects, outsourcing, CRM, digital, data, and expert services. The AI story is real, and Chameleon is starting to move toward cloud delivery and a more product-like layer, but the economics of the group still rest mainly on people, execution, backlog, and collections.

What is working now? Revenue rose 9.6% to ILS 535.1 million, backlog rose 15.6% to ILS 293.0 million, and net profit climbed 23.2% to ILS 27.9 million. After the IPO, cash increased to ILS 109.9 million, short-term bank credit fell to ILS 6.0 million, and the equity-to-assets ratio reached 57.7%. That is a real change in financial flexibility.

What is less clean? The second half of 2025 already told a more complicated story. Revenue in H2 rose to ILS 272.3 million, but operating margin fell to 7.5% from 9.2%, and adjusted EBITDA margin excluding IFRS 16 fell to 7.2% from 9.7%. At the same time headcount rose 13.7% to 1,532 employees, faster than revenue growth. On an end-of-year basis, revenue per employee fell to roughly ILS 349 thousand from roughly ILS 362 thousand at the end of 2024. That does not prove structural weakness, but it does show that AI has not yet translated into visible group-level productivity.

That is why 2026 looks less like a breakout year and more like a proof year. If Chameleon’s cloud migration, LayerX, and the broader AI push begin to change mix, strengthen recurring revenue, and improve productivity, Elad can be read differently. If not, it remains a solid and growing company, but one whose valuation is still constrained by labor intensity and service economics rather than software-like scale.

At the closing price of April 6, 2026, market value was around ILS 510 million. On that same trading day volume was only about ILS 27 thousand. That does not determine the business thesis, but it does remind readers that stock actionability is a separate issue.

Events and Triggers

The IPO fixed the balance sheet, not the profit engine

Trigger one: the February 2025 IPO brought in about ILS 80 million gross, and the impact on the balance sheet was immediate. Share capital and premium rose to ILS 111.5 million, total equity rose to ILS 169.5 million, and year-end cash reached ILS 109.9 million versus ILS 15.3 million a year earlier. This is not cosmetic. The company suddenly has a real cushion, and it no longer looks like a smaller IT house funding growth mainly through short-term bank lines.

But this needs a second sentence. It is not the same company from a balance-sheet perspective, but it is still the same company from an economic-engine perspective. Growth in revenue, headcount, AI infrastructure, and partner management still belongs to a services business. The IPO mainly solved the funding-flexibility question. It did not yet solve the mix-quality question.

The Elad Campus sale is portfolio cleanup, not just a one-off event

Trigger two: in April 2025 the company sold Elad Campus, its training and course business, as part of a sharper focus on core technology solutions for business customers and against a backdrop of weaker demand for open B2C courses. That is the right move. Not all growth is good growth, and exiting a weaker B2C activity improves quality.

This matters because the “Other” bucket shows a useful paradox: revenue fell to ILS 26.4 million from ILS 29.8 million, but segment profit jumped to ILS 8.1 million from ILS 2.7 million. The presentation explicitly says part of that improvement came from removing the loss-making B2C training activity from 2024. In other words, the improvement is real, but it is not full proof that AI or Chameleon already transformed group economics. First, a weak piece was removed.

Chameleon is beginning to move up a layer

Trigger three: from the second half of 2025 Chameleon has also been offered as a cloud service, and the company says some customers have already upgraded and are running in production on that version. At the same time it launched LayerX, a connective layer for AI capabilities and third-party solutions around the medical record. This is one of the few places in the report where Elad clearly signals a path from implementation-heavy work toward a deeper software layer.

The implication cuts both ways. If this move gains traction, it can increase recurring revenue, improve pricing power, and shift part of the group away from direct dependence on people-hours. But disclosure still does not show how much of this is already becoming revenue, under what commercial model, and at what pace. For now it is a strong option, not yet proof.

The structural merger and the option plan say something about 2026

Trigger four: in August 2025 the statutory merger of Elad Data and Elad Yishumim into the parent company was completed. This is a sensible simplifying step after the IPO and can help with control, reporting, and capital allocation. It does not change the thesis by itself, but it fits the broader move from a private structure into a public operating platform.

Trigger five: after the balance-sheet date, on March 22, 2026, the board approved an outline plan for up to 1.75 million employee options. This may be a rational retention tool in a still-competitive market, but it also reminds investors that Elad’s growth still requires hiring and retaining talent, not just owning intellectual property. If fully exercised, the pool would imply potential dilution of about 3.4% versus the current share count.

Efficiency, Profitability and Competition

Growth continues, but the second half changed the tone

Revenue of ILS 535.1 million in 2025 is a strong number, and the 8.2% increase in gross profit to ILS 86.3 million also looks healthy. But the real issue is where profitability stopped following the top line. Gross margin slipped slightly to 16.1% from 16.3%, operating margin fell to 7.0% from 7.5%, and adjusted EBITDA margin excluding IFRS 16 fell to 7.4% from 8.1%.

The second half sharpens the point. Revenue rose 7.6% to ILS 272.3 million, but operating profit fell 11.4% to ILS 20.6 million. This does not look like a group already enjoying AI-driven operating leverage. It looks like a group investing harder than current growth can absorb.

Management itself points to the reasons: public-company costs, heavier recruitment and training, AI infrastructure, business development in the defense sector, and partner management. Those are legitimate items if they are building a higher-quality next stage. The problem is that disclosure still does not show that those investments are already paying back through higher productivity, better margins, or a visible product mix shift.

AI sits in the strategy and the presentation, not yet in the center of the P&L

Elad presents a broad AI story: agents for service and sales, modernization of core systems, Vibe Coding, AI-oriented data architecture, LayerX around Chameleon, and even M&A in focus areas. Strategically, that is coherent. It also fits a broad enterprise and public-sector customer base.

But when you look at the consolidated report, the picture is still clear. IT solutions and services, consulting, and management generated ILS 508.7 million out of the group’s ILS 535.1 million of revenue. That means the AI story still appears mainly inside the service offering, not as a separately material software layer that already moves consolidated economics.

Headcount highlights the same point. The company ended the year with 1,532 employees versus 1,347 in 2024, and 1,418 of them are categorized as software services staff. That is the heart of the business. As long as that remains true, any real improvement in the story has to show up through revenue per employee, better utilization, pricing power, or a faster shift toward recurring revenue. So far, on an end-of-year basis, revenue per employee actually moved lower.

Growth quality is better than in 2024, but it is not frictionless

There is also a very real positive side. The company has no customer above 10% of revenue, and credit exposure is relatively diversified. 58% of revenue comes from government and public-sector customers, with the remaining 42% from business customers. That is a more stable base than an IT house built around one or two anchor accounts.

But it also means the group remains exposed to the world of tenders, public budgets, and large contracts where competition is intense. The company explicitly lists competition not only from Matrix, Hilan, One, Malam-Team, Aman, Yael, and Abra, but also from the Big Four and from firms backed by private-equity capital. So even if AI demand is rising, it does not automatically produce better margins.

Another point worth noticing is the billing pattern. The company says there is a common commercial practice under which government and institutional bodies use their procurement budgets toward the end of the year, so invoicing shifts more heavily into the second half. That does not automatically create a problem, but it does reinforce the idea that H2 is the real test of revenue quality, margin quality, and collections.

Cash Flow, Debt and Capital Structure

The cash bridge has to be read in two layers

This is where discipline matters. Elad finished 2025 with ILS 73.9 million of operating cash flow. That is a strong number, but it should not be read automatically as clean free cash flow or normalized recurring cash generation.

Normalized cash generation: operating cash flow was supported materially by working capital. The decline in receivables and accrued income contributed ILS 35.5 million, and working-capital lines together contributed about ILS 27.7 million to cash flow. At the same time, ILS 9.2 million of development costs were capitalized and reported in investing cash flow, not expensed immediately through the P&L. So part of the cash strength reflects better collections and part reflects the fact that some technology investment moved onto the balance sheet.

All-in cash flexibility: after real cash uses, the picture is still good, but it needs the right framing. The company used ILS 11.8 million for investing activity, roughly ILS 9.5 million for lease-liability repayments, ILS 3.6 million for interest, and ILS 32.5 million to reduce short-term bank debt. It could do all of that because it had both ILS 73.9 million from operations and ILS 78.1 million from the IPO. That is why the cash balance jumped.

The takeaway is simple: cash quality improved, but the strong cash result should not be confused with a frictionless recurring cash engine. 2025 also benefited from favorable working-capital timing and IPO proceeds.

The balance sheet looks very clean, but year-end was an especially good snapshot

At the balance-sheet level, the improvement is hard to dispute. The company moved from ILS 15.3 million of cash to ILS 109.9 million, from ILS 40.9 million of short-term bank debt to ILS 6.0 million, and from an equity ratio of 27.6% to 57.7%. At year-end it almost looks like a net-cash company, with about ILS 103.9 million of net cash versus roughly ILS 25.6 million of net debt at the end of 2024.

But there is an important qualifier. In the credit-line table, the company shows that near the report date utilized bank facilities had already risen back to ILS 31.1 million, versus only ILS 6.0 million at December 31, 2025. That does not cancel the balance-sheet improvement, but it does mean the year closed on a particularly favorable snapshot and that working capital still knows how to pull on short-term lines during the year.

Covenants are distant, and that is an important external signal

The external signal is also strong. In September 2025 the company committed to financial covenants of minimum equity of ILS 100 million, minimum equity-to-assets of 23%, and debt-to-EBITDA not above 3. By year-end it stood at ILS 169.5 million of equity, 58% equity-to-assets, and a debt-to-EBITDA ratio of 0.43.

That means Elad is not a balance-sheet stress story at this stage. The test now sits elsewhere: whether management can use the cleaner balance sheet to improve growth quality, not just to look safer.

There is also no clear red flag in credit quality at this point. Receivables and accrued income fell to ILS 122.5 million from ILS 158.0 million, expected-credit-loss allowance fell to ILS 2.94 million from ILS 3.13 million, and the company emphasizes that most customers are government or institutional entities operating in independent markets. That supports the story. Still, 2025 benefited from earlier payments and stronger collections, so the market will want to see that 2026 cash stays strong without the same help.

Outlook

Finding one: 2025 solved the balance-sheet question faster than it solved the margin question. This is a financially stronger company, not necessarily a more efficient operating company yet.

Finding two: the AI story is clearly present in management language, the presentation, and the strategy, but disclosure still does not show a material quantitative mix shift at group level. Investors who are already pricing a software-like transformation are moving too early.

Finding three: the improvement in profit during 2025 relied materially on the collapse in finance expense, not only on better operations. That tailwind is hard to repeat with the same force.

Finding four: improvement in the “Other” bucket does not come only from new growth. It also reflects the removal of the loss-making Campus activity. That is better quality, but it is not yet proof of a new scalable engine.

Why 2026 is a proof year

The company’s roadmap for the coming year is actually very clear: Agentic AI, AI-oriented data stacks, Vibe Coding to improve productivity, modernization of core systems, LayerX expansion, moving Chameleon customers to cloud architecture and subscription contracts, and completing the next-generation version of the medical record.

This is a good list, but it is not a breakout list. It is a build-out list. In practice, it can be read as two tracks that both need to work:

1. Improve productivity and customer value inside the services engine.
2. Build a deeper health-software layer that can create more recurring revenue and a different economic profile.

If only the first track works, Elad remains a strong IT services platform with decent margins. If the second track also starts to scale, the company profile changes.

What has to happen for the read to improve

The good news is that Elad enters 2026 with ILS 292.3 million of backlog scheduled for recognition during 2026 and another ILS 0.7 million beyond that. That gives the company a comfortable starting point and covers about 54.6% of 2025 revenue already going into the year. The company also says actual cancellations have historically been negligible, even though part of the customer base formally holds termination rights with notice.

But for the market to believe this is becoming more than steady services growth, it will need four signs.

First, recovery in operating margin in the first 2026 reports, or at least a clear stop to the H2 2025 erosion.

Second, proof that Chameleon cloud and LayerX are becoming a measurable commercial path, not only a strategic narrative.

Third, strong cash generation without another similar release from receivables and without unusually helpful finance income on IPO cash.

Fourth, evidence that higher headcount and AI investment are translating into productivity rather than only into broader overhead.

What the market could misread on first pass

The market could get stuck on the clean ILS 27.9 million net-profit number and miss the fact that operating profitability actually weakened. It could also read the improvement in the “Other” bucket as full AI proof, even though part of it simply comes from removing a weak business. On the other hand, it could miss the fact that the company enters 2026 with a much cleaner balance sheet, wide covenant headroom, and a health asset that could become economically deeper if the cloud model gains traction.

In other words, the near-term story will be decided less by whether Elad “does AI” and more by whether AI and cloud already change the economics of the company.

Risks

Labor remains the central bottleneck

The report itself identifies human capital as a major industry risk. That makes sense. With 1,418 software-services employees out of 1,532 total employees, Elad remains a people business. If the technology labor market tightens again, or if the company keeps spending heavily on recruitment and training to support its AI strategy, margin pressure can continue.

Public-sector customers provide stability, but also budget and tender dependence

58% of revenue comes from government and public-sector customers. That lowers credit risk, but increases dependence on state budgets, tenders, contract renewals, and the ability to preserve reasonable pricing on large public work. The company explicitly points to this as a risk, and rightly so.

Cyber, regulation, and healthcare are not side risks

Elad operates deeply inside information systems, including medical data, public systems, and financial-sector clients. The company describes ongoing cybersecurity investment and cyber insurance of up to USD 1.5 million, but the exposure remains high by nature. As Chameleon and AI solutions become more deeply embedded, sensitivity to privacy, accuracy, uptime, and regulation also rises.

The growth strategy is broad, so it also demands broad execution

The company wants to deepen AI, move customers to cloud, keep developing Chameleon, expand partner relationships, and pursue M&A. Each of those is sensible on its own. Together they become a heavy execution agenda. So even though there is no immediate balance-sheet stress, there is definitely execution-load risk.

Dilution is not the main story, but it is part of the picture

The March 2026 option plan looks reasonable as a retention tool, but it also reminds investors that the effort to build teams and retain senior talent will not be free for shareholders.

---

Conclusions

Elad looks healthier, better funded, and better organized today than it did going into 2025. That matters. But the number that should lead the read is not only the rise in net profit. It is the fact that operating profitability weakened just as the company increased investment and leaned harder into the AI and cloud story. That is why the key test of 2026 is not more growth. It is growth quality.

What supports the thesis today is a broad customer base, a strong backlog, a valuable health-tech asset, and a much cleaner balance sheet after the IPO. What blocks a cleaner thesis is that the group is still measured primarily as a labor-heavy services business, with weaker second-half margins and cash flow that benefited from good timing. That is exactly where the market will need proof, not presentation language.

Current thesis: Elad enters 2026 with a stronger balance sheet and solid backlog, but to justify a higher-quality reading it has to show that AI, cloud, and Chameleon are changing service productivity and revenue mix, not just expanding the growth narrative.

What changed versus the old read of the company: capital structure is no longer the main problem. After the IPO, the bottleneck moved from funding to the quality of turning growth investment into margins and cash.

Counter-thesis: the market may be too harsh on 2025 because the company has already solved the balance-sheet problem, holds a broad customer base and strong backlog, and the heavier spending on AI, recruitment, and public-company readiness may be laying the groundwork for better margins later on.

What could change the market’s reading in the short to medium term: early-2026 reports that show margin stabilization, visible commercial evidence of Chameleon cloud upgrades and LayerX monetization, and strong cash generation without the same working-capital help.

Why this matters: if Elad remains mainly an IT services platform, its quality ceiling is very different from that of a company that can turn AI, cloud, and Chameleon into recurring revenue, better pricing power, and higher productivity.

What has to happen over the next 2-4 quarters for the thesis to strengthen, and what would weaken it: the thesis strengthens if margins recover, cash stays healthy, and Chameleon commercialization becomes visible. It weakens if growth continues to depend mostly on headcount growth, margins keep falling, or short-term credit lines again become a consistently large part of the operating picture.