One Technologies: Cash Is at a Peak, but Growth Quality Has Become More Labor-Heavy

Getting to Know the Company

One is no longer just a software house. It is a broad Israeli IT platform with three engines that have very different economics: technology solutions and services, computing infrastructure and communications, and BPO and support centers. In 2025 the group reached ILS 4.65 billion of revenue, ILS 274.0 million of net profit, and nearly 10,000 employees. On the surface this looks like another clean record year. That is only half the story.

What is working right now is clear. The two core engines, technology solutions and services and infrastructure and communications, kept growing organically at a double-digit pace. No customer accounts for more than 10% of revenue, 97% of receivables were still within agreed credit terms, and the company ended the year with ILS 641.9 million of cash and ILS 495.8 million of net cash. Those are the marks of a mature, disciplined, and diversified business.

What can mislead a first reading is that growth quality changed. The acquisition of One Line pushed the BPO segment up by 58.4% and lifted headcount from 7,083 to 9,978, a 40.9% increase, while revenue grew only 16.3%. One did grow, but a rising share of the group now depends on labor-heavy activity, which affects productivity per employee, wage sensitivity, and the group's ability to expand margins.

The active bottleneck today is not liquidity. It is capital allocation quality and mix quality. The company can keep buying businesses, paying dividends, and even repurchasing shares, but the key question is whether each additional shekel of growth will come from areas that improve profitability and cash generation per share, or from areas that mostly add scale, complexity, and labor.

That is why 2026 looks less like a rescue year and more like a proof year. After a record year, the market will need to see whether the AI and data layer, the Strauss Strategy acquisition, the buyback plan, and continued M&A actually improve business quality, or merely make the platform bigger.

Events and Triggers

The first trigger: the One Line acquisition closed in April 2025 for ILS 33.7 million, or ILS 27.4 million net after acquired cash. At the group level this is a small deal, but at the mix level it is meaningful: it pushed the BPO and support-centers segment to ILS 500.6 million of revenue from ILS 316.0 million in 2024, and made that segment far more material inside the consolidated picture.

The second trigger: the fourth quarter looks weaker than it really is. On a reported basis operating profit fell to ILS 90.0 million from ILS 98.1 million and net profit fell to ILS 67.9 million from ILS 73.7 million. But that is because the fourth quarter of 2024 included ILS 20.2 million of other income. Excluding that one-off item, fourth-quarter operating profit actually rose 15.6% to ILS 90.0 million. That is the first read-through the market could easily miss.

The third trigger: in December 2025 the company completed a private placement of 1.93 million shares for net proceeds of ILS 173.3 million. This is a meaningful capital raise for a company that was already profitable and well funded. The implication is not liquidity stress. It is a deliberate choice to expand financial flexibility ahead of growth moves and acquisitions.

The fourth trigger: in March 2026 the board approved a share buyback plan of up to ILS 50 million, to be funded from internal sources, while no shares had yet been repurchased under the plan. The combination of a regular dividend, a private equity raise, and then a buyback authorization tells the market two things at once: management wants flexibility for continued M&A, but it also understands that the debate has shifted toward capital allocation per share, not just group growth.

The fifth trigger: in February 2026 One signed an agreement to acquire 70% of Strauss Strategy Consulting & Systems for ILS 28 million, with contingent payments tied to 2026 through 2028 results. The disclosed numbers show that Strauss generated ILS 39.6 million of revenue and ILS 6.1 million of adjusted operating profit in 2024, and ILS 32.2 million of revenue and ILS 3.7 million of adjusted operating profit in the first nine months of 2025. This is not a deal that changes One on its own, but it does show where the company wants to move: further up the value chain, into technology strategy consulting, data, and AI.

The sixth trigger: the presentation highlights new entries in 2025 into data infrastructure, SailPoint, Atlassian, and monday.com, and in Q1 2026 also a local offshore expansion focused on AI and innovation. That is evidence of a commercially sensible direction, but not yet proof that a new profit engine has already been built. The filings do not disclose revenue or margin separately for the AI layer. At this stage this is still more of a promising option than a proven earnings pocket.

Efficiency, Profitability, and Competition

The core insight here is that One still knows how to grow, but the quality of each incremental shekel of growth became less clean. At the group level reported operating profit rose to ILS 350.0 million, but operating margin fell to 7.5% from 8.2%. Even excluding the 2024 other income, operating profit from ordinary activity rose to ILS 346.7 million from ILS 306.9 million, while the margin still slipped to 7.46% from 7.67%. There is more profit, but not genuine operating leverage.

What really drove growth

The two core segments still performed well. Technology solutions and services grew 12.6% to ILS 2.80 billion, with operating profit up 14.3% to ILS 226.3 million. Infrastructure and communications grew 13.1% to ILS 1.38 billion, but operating profit rose only 6.0% to ILS 103.1 million. The core engines are still providing the base of growth, yet margin expansion is not keeping up.

The BPO and support-centers segment looks strongest in the headline, with revenue up 58.4% to ILS 500.6 million and operating profit up 27.2% to ILS 41.4 million. But this is exactly where the quality shift is hiding: segment margin fell from 10.3% to 8.3%. In plain terms, One bought scale and presence, not a new engine with structurally higher profitability.

Growth became far more labor intensive

This may be the single most important point in the entire report. Headcount jumped from 7,083 to 9,978. Within that, BPO and support centers rose from 1,794 employees to 4,334 employees. That means the segment now accounts for 43.4% of group employees while contributing only 10.8% of revenue. The group is broader, but it is also much more dependent on labor management, hour pricing, operating discipline, and continued hiring and retention.

At the group level, annual revenue per employee fell from roughly ILS 565 thousand in 2024 to roughly ILS 466 thousand in 2025, a decline of about 17.5%. That does not mean the business weakened. It does mean the mix changed materially, so 2025 should not be read as if all growth came from software-heavy or pricing-heavy areas.

Customer diversification is good, supplier dependence is real

On the positive side, One has no customer above 10% of revenue and no customer above 10% of receivables. That meaningfully reduces anchor-customer risk. In addition, 38% of sales come from government and public-sector customers, which provides a relatively stable activity base in weaker macro periods.

On the other side, the group is clearly dependent on major software and hardware vendors, especially Microsoft and IBM. In the solutions and services segment, purchases from Microsoft accounted for 16.95% of consolidated cost of sales in 2025, and IBM added another 2.22%. Across software and infrastructure activity, IBM represented 6.5% of consolidated cost of sales. The supplier agreements are non-exclusive, some require quality and volume targets, and in some cases the company pays the supplier before collecting from the customer. That is not unusual for the sector, but it does mean pricing power is not unlimited.

AI is an expansion layer, not yet a separate earnings pocket

The annual report and the presentation point in the same direction: cloud, data, GenAI, consulting, and implementation. The company explicitly says it sees data, artificial intelligence, and GenAI as a complementary growth engine. But at this stage there is no quantitative disclosure that allows the reader to determine how much of One's profit comes from that layer. The right reading is therefore not "One has become an AI company." It is "One is trying to move its service basket into areas with higher project value." That is a real opportunity, but not yet a proven basis for a new valuation framework.

Cash Flow, Debt, and Capital Structure

The key insight here is that two cash pictures need to be kept in mind at the same time. Recurring cash generation remains very strong. The picture after all actual cash uses is much tighter.

Recurring cash generation remains strong

Cash flow from operations rose to ILS 448.8 million from ILS 382.6 million in 2024, equivalent to about 9.7% of revenue. That is a strong number for a services and technology company of this scale. Collections also look reasonable: receivables rose to ILS 1.019 billion, only 6.4% above 2024 despite 16.3% revenue growth, and 97% of receivables remained within agreed credit terms.

If the question is whether the existing business knows how to generate cash, the answer is yes. The core business is still working.

On the all-in cash picture the story is much less generous

This framing needs to be precise. I am using the all-in cash-use framework here because it is the right lens for balance-sheet flexibility and capital allocation. In 2025 the company generated ILS 448.8 million from operations, but also spent ILS 75.1 million on investing activity, ILS 100.5 million on lease-related cash outflows, ILS 175.6 million on dividends, ILS 42.1 million on long-term debt repayment, and another ILS 16.6 million on minority purchases and contingent consideration in business combinations. After all of that, only about ILS 38.9 million was left.

That is not a weak position, but it is also not a picture of massive organic cash surplus. This is why the sharp rise in year-end cash was not driven only by operations, but also by the net ILS 173.3 million private placement completed in December. The cash balance is strong, but anyone reading it needs to separate cash that was generated from cash that was raised.

The balance sheet is very strong, and covenants are distant

Still, the other side should not be missed. Gross bank debt fell to ILS 155.2 million from ILS 189.6 million. Equity rose to ILS 1.028 billion and the equity-to-assets ratio improved to 38% from 32.6%. Against the main bank covenant set, the company is required to maintain minimum equity of ILS 100 million and a minimum equity-to-assets ratio of 16%, while it ended 2025 at ILS 994 million and 36.7%, respectively. Net-debt coverage is required to stay at 3, while the actual ratio was minus 1.03, meaning the company was in a net-cash position.

Credit lines add another layer. As of the report date, credit and guarantee facilities stood at ILS 723.6 million, utilization at ILS 373.9 million, and positive bank balances at ILS 650.9 million. The combination of unused facilities and cash exceeded ILS 1.0 billion. That is why the company can keep paying dividends, pursue acquisitions, and authorize a buyback without looking financially stretched.

The lease layer is not trivial

The less visible point is leases. Lease liabilities totaled ILS 163.5 million, and total negative lease-related cash flow reached ILS 100.5 million in 2025. This is not a balance-sheet threat, but it is a real cash use that needs to be included when measuring how much cash truly remains after normal activity.

Outlook and What Comes Next

Before going into the details, here are the five non-obvious takeaways from the current read:

2026 looks like a proof year, not a breakout year

Management itself is signaling a clear direction: keep expanding the core, continue M&A, and deepen activity in data, AI, and consulting. On the labor side it also says that the easing in demand for junior and inexperienced technology workers is expected to continue through the first two quarters of the year. Those are relatively supportive operating conditions.

But there is a gap between direction and proof. For the market to accept that the new layer is genuinely value-enhancing, it will need to see at least three things. First, the core engines need to preserve their organic pace without another sharp jump in headcount. Second, BPO expansion cannot keep diluting the consolidated margin. Third, the Strauss acquisition and the broader AI and data push need to show up not only in language and partnerships, but also in identifiable profit pockets.

What must happen over the next 2 to 4 quarters

First, the company needs to show that the fourth quarter of 2025 was not only a normal year-end seasonal peak, but a platform for continuity. Seasonality still works in its favor, especially with government and public-sector customers that tend to use budgets toward year-end, so the coming quarters matter for testing the durability of the run rate.

Second, if the Strauss acquisition closes after regulatory, customer, and bank approvals, the market will want to see whether it adds a consulting layer that can be sold into existing customers and raise project value, or whether it is simply another small company added to the group machine.

Third, the buyback needs to move from signaling to real capital deployment, or management will need to explain why it prefers to keep the cash for better uses elsewhere. Once a buyback is approved after a record year, the market watches not only the statement but the execution.

Fourth, the market will look for proof that the AI layer is producing something beyond strategic language. The report itself describes a gradual move from proof-of-concept projects toward broader enterprise deployment, and rising demand for solutions that connect models, data, and core systems. If that is real, One should begin to see larger projects, deeper implementations, and perhaps even better profit mix inside the solutions segment.

What the market may miss in the near term

There is a small but interesting dissonance here. On one hand, the numbers are strong and the balance sheet is strong. On the other, growth quality became less impressive, so the stock can still trade as if this were simply another stable IT company rather than a business that deserves a cleaner premium. If management can show that balance-sheet strength is translating into per-share improvement rather than only platform expansion, the market reading can change.

That matters because it shows the market is not sitting here with an extreme short view. Short float stood at 1.11% at the end of March, slightly above the sector average of 0.72%, but SIR fell to 1.76 from 8.75 at the end of January. There is some skepticism, but not the kind of crowding that signals deep disbelief.

Risks

The first risk is dependence on public budgets and tenders. Management writes that the government and budgeted bodies accounted for about 31% of revenue in 2025, while the segment-level split shows 38% exposure to the government and public sector. That is a stable base, but also a real dependence on budget conditions, tender renewals, and contract continuity.

The second risk is dependence on global vendors and the pace of technological change. Microsoft and IBM are not ordinary suppliers for One. They are an important part of the profit structure inside the software-related business, and the agreements include targets, service obligations, and conditions under which the suppliers may choose not to renew. At the same time, in AI the risk is not only competition, but also One's ability to keep talent, infrastructure, and know-how in step with a fast-moving market.

The third risk is cyber exposure. Management is unusually detailed here, and that alone signals that this is a real management-level issue. The company describes stronger threats, a 24/7 SOC operation, dedicated insurance, and compliance with demanding standards, but it also states openly that it cannot guarantee full prevention. For a company whose entire business sits on information systems, this is a structural risk, not a footnote.

The fourth risk is legal and regulatory. The statements include a provision of ILS 915 thousand for claims, but claims against the group total ILS 20.4 million, excluding a number of specifically described material proceedings. One of them is an older matter involving Harel Technology Information, where exposure still cannot be reliably estimated. In addition, in January 2026 the company learned of a motion to certify a class action against a subsidiary over a four-hour waiting window for a service provider, in an amount above ILS 2.5 million, and at this stage the company cannot assess the chances.

The fifth risk is macro and FX. The company has dollar exposure mainly in the infrastructure segment because of inventory and equipment purchases in dollars. It says it hedges material specific deals, but margin erosion is still possible if FX conditions move against it. In addition, the company currently benefits from moderating inflation and interest conditions, but it acknowledges that a renewed rise in rates would mainly hit the financing cost of its short-term variable-rate debt.

Conclusions

One ends 2025 as a strong, profitable, and well-funded company. That is the easy part of the read. The harder part is that this year's growth was broader, but also much more labor heavy and slightly less attractive in mix quality. So 2026 is no longer a test of stability. It is a test of capital-allocation quality and growth quality.

Current thesis: One remains a strong Israeli IT platform with a growing core and an unusually powerful balance sheet, but the next step no longer depends on its ability to grow. It depends on its ability to grow well.

What changed versus the earlier read of the business: 2025 strengthened the balance sheet further, but it also shifted the group more clearly toward labor-heavy BPO activity and made it clear that AI is still not a separate profit engine.

Counter-thesis: it is also possible to read the report more optimistically and argue that the company is simply building more service layers, broader diversification, and more cross-sell capacity into existing customers, and that with its cash position and discipline the new mix could still support better profitability in the years ahead.

What the market will measure now: the closing of Strauss, actual execution of the buyback, the organic pace of the core segments, and whether the AI and data layer begins to show up in numbers rather than only in strategic direction.

Why this matters: because One has already proved that it can build a large group. It now needs to prove that this larger group can create cleaner value per employee, per shekel of capital, and per share.

What needs to happen for the thesis to strengthen? The core engines need to remain double-digit growers, the group margin needs to stop eroding, the new initiatives need to produce identifiable profit pockets, and management needs to use the strong balance sheet to improve the economics of the stock itself. What would weaken the thesis? More volume growth without better mix, more capital deployed without visible return, and continued reliance on an AI narrative without financial proof.