Unitronics: Data Center Growth Is Real, but 2026 Will Be Judged Through Cash and US Demand

Getting to Know the Company

Anyone looking at Unitronics and seeing just another small PLC maker is reading the company too shallowly. In practice this is an industrial technology company sitting at a sensitive junction. It still sells programmable logic controllers and related automation products, but more than half of revenue comes from the United States, and the new engine that stands out in the report is data centers, a vertical that was still negligible through 2024 and already crossed the 10% mark in 2025.

What is working now is clear enough. Direct sales rose to 26% of revenue, up from 24% in 2024 and only 13% in 2022. In the US, which is the company’s largest market, direct sales already reached 37% of that market’s sales, up from 16% in 2022. At the same time, data centers moved from a marginal contribution to a visible wedge, with the company already embedding dedicated capabilities for cooling systems, energy-management applications, and advanced communication requirements.

This is also where the active bottleneck sits. In 2025 revenue fell 19.5% to NIS 154.8 million, operating profit fell 40.7% to NIS 26.4 million, and net profit fell 48.1% to NIS 22.7 million. Management attributes the decline mainly to a US slowdown driven by uncertainty around the American tariff plan, and only part of that pressure was offset by the growth in data-center activity. In other words, the new engine is already real, but it is not yet large enough to neutralize broader weakness in demand.

A superficial reader could also get misled by the cash-flow headline. The company did generate NIS 36.8 million of cash flow from operations in 2025, comfortably above net income. But that is only half the story. After development investment, a NIS 29 million dividend, lease principal, and debt service, year-end cash stood at only NIS 5.25 million, and that was only after the company took a NIS 14 million bank loan in July 2025. So 2026 looks less like a breakout year and more like a proof year: Unitronics has to show that US demand stabilizes, that data centers keep scaling, and that cash no longer needs borrowing support.

There is also an early market frame worth stating clearly. The company’s current market value is roughly NIS 309 million, which means this is still a relatively small industrial tech name, not a platform with unlimited access to capital. Short interest is almost nonexistent, but that does not mean the story is clean. It mostly means the market has not yet built an aggressive negative thesis and is waiting for operational proof.

The 2025 economic map looks like this:

Events and Triggers

The first trigger: in October 2025 the control transaction closed, with shares representing about 40.2% of the company sold to Illgin for about NIS 156 million. From that point Illgin and Haim Shani became joint controlling shareholders, and the new shareholder agreement leaves Shani’s remaining shares locked for 48 months and then sets call and put mechanisms on the balance. This is not cosmetic. It means the change in control is not a passing event but a multi-year framework that is meant to shape the company. The upside is obvious: possible access to new markets, product layers and verticals through the AMBER group. The open issue is that the report still does not show a measurable commercial contribution from that relationship.

The second trigger: data centers became a visible growth engine in 2025. Management says the vertical, which had still been negligible until 2024, already exceeded 10% of total sales in 2025. That matters because it signals real penetration into a high-growth end market, not just a pilot program. But the story is not one-directional. In the same filing the company says that toward the end of 2025 lead times for memory components started to lengthen because of stronger data-center demand, and Unitronics therefore increased safety stock for certain parts. So the same vertical that opens a growth path can also create new pressure on supply and working capital.

The third trigger: the American tariff program was already a real drag in 2025. The company says revenue fell mainly because of a US slowdown caused by uncertainty around the tariff plan, starting in the second quarter. According to the note, the initial tariff on Israeli imports into the US was 10%, and from August 7, 2025 it rose to 15%. Unitronics did update selling prices in the US to reduce the direct hit to profitability, but that only solves part of the problem. It does not eliminate customer hesitation when trade rules are shifting.

The fourth trigger: management incentives tell you something about 2026. The December 30, 2025 shareholder meeting approved an update to CEO Amit Harari’s terms, including monthly salary of NIS 70 thousand, a variable bonus of up to seven monthly salaries subject to sales and net-income targets, and an option grant. After the balance-sheet date, on March 23, 2026, the board also approved another 100 thousand options for the CEO, still subject to shareholder approval. That is a sign that the new controllers are structuring incentives for a longer buildout. That could be positive if it aligns with new growth engines, but it also underlines that the next year is viewed internally as a build year, not as a harvest year.

This chart matters because it shows the weakness was not just a late-year event. Revenue and operating profit both eroded fairly steadily through the year, so the data-center wedge has not yet changed the overall direction.

Efficiency, Profitability and Competition

The real story of 2025 was not a product collapse. It was pressure on volume against a cost base that did not fall at the same speed. Revenue dropped by NIS 37.5 million, but cost of sales fell by only NIS 16.0 million. As a result, gross profit fell by NIS 21.5 million and gross margin declined from 51.1% to 49.6%. Management explains that clearly: when sales fall, fixed-cost items, mainly depreciation, amortization and wages, take up a bigger share of revenue.

Price, Volume and Mix

On the volume side, the main hit came from the US. Revenue there fell to NIS 80.4 million from NIS 100.4 million in 2024, a decline of about NIS 20.0 million in a market that already accounts for more than half of the company’s revenue. Europe fell by NIS 5.5 million to NIS 53.9 million, Israel by NIS 3.3 million to NIS 7.3 million, and the rest of the world by NIS 8.7 million to NIS 13.2 million. Weakness was therefore broad, but the United States is what turned the year from soft to problematic.

On the price and mix side, there are some building positives. The company says it updated US pricing to reduce the direct margin hit from tariffs, and data-center activity grew rapidly. On top of that, the shift to 26% direct sales, and 37% direct sales in the US, should improve customer access, control over the commercial interface and the ability to offer fuller solutions. That does not immediately guarantee a better margin, but it does change the quality of the model.

The important point is that this chart does not stand alone. It shows the company moving away from pure distributor-led selling toward a closer end-customer relationship, but it did so while revenue was falling. So the real test is not the higher direct-sales mix by itself. The real test is whether it translates in 2026 into stronger demand stability and better margin defense.

Reported R&D Expense Fell, but Real Investment Actually Rose

This is one of the easiest numbers in the report to misread. In the income statement, net development expense fell from NIS 4.78 million to NIS 4.13 million. That could look like tightening. It is not. Total development spending actually rose to NIS 16.5 million from NIS 15.5 million in 2024, but NIS 12.38 million of that was capitalized into intangible assets. In other words, about 75% of 2025 development spending did not pass through the development-expense line.

That is not automatically an accounting red flag. The company explains that its intangible assets relate mainly to UniStream controller projects, Motion solutions and UniCloud services, and that the amortization of development assets is recorded inside cost of sales. But analytically the distinction matters. Reported earnings look somewhat better because of capitalization, while the real economic burden of the development cycle still requires cash and headcount.

The chart shows the central issue well: the decline in operating profit was sharper than the decline in gross profit because Unitronics kept carrying a relatively heavy layer of development, selling and overhead costs while revenue moved down.

Competition Still Means Competing Against Larger Players

The company itself says its share of the global PLC market is below 1%, and that its main competitors are larger and better resourced, financially, commercially and technologically. That point matters especially now. Unitronics is not competing in a market that only rewards the cheapest product. It competes on speed, development, fit and a fuller solution. That works well when the company can be faster in its niches. It becomes harder when it also needs to fund development, tariffs, safety stock and dividends at the same time.

To the company’s credit, two points stand out. First, it had no customer above 10% of sales in 2025. Second, its distribution network is broad, with about 180 distributors, including around 105 in North America. Those are real commercial assets. They do not change the core point: the company’s edge rests on execution, not on absolute market power.

Cash Flow, Debt and Capital Structure

This is a case where the correct framing is explicitly the all-in cash-flexibility bridge, not the narrower normalized one. The reason is simple: the thesis at Unitronics is not whether the business can produce profit, but how much real cash is left after the actual uses of cash, including development, dividends, leases and debt service.

The Full Cash Picture

At first glance, NIS 36.8 million of cash flow from operations looks strong. And it is stronger than net income of NIS 22.7 million. But once you follow the cash all the way through, the picture changes. Investing activity used NIS 11.3 million, mainly because of NIS 12.4 million invested in intangible assets and NIS 1.3 million in property, plant and equipment, partly offset by NIS 2.47 million of long-term related-party debt collection.

The financing side is sharper still. In 2025 the company paid NIS 29 million of dividends, NIS 2.27 million of lease principal, repaid NIS 3 million of long-term loans, and reduced short-term bank credit by NIS 2.5 million. Only after taking a new NIS 14 million bank loan in July 2025 did it finish the year with NIS 5.25 million of cash, versus NIS 2.68 million a year earlier.

This chart is the core of the cash argument. The business does generate cash, but the cushion left for shareholders after all real commitments is thin. So anyone looking only at EBITDA, net income or even operating cash flow is missing the main constraint: dividend policy and development investment consumed almost all of the flexibility.

Even the Normalized View Still Has Friction

If you still want to isolate the recurring cash-generation power of the existing business before discretionary uses, the picture is less strained. Operating cash flow was supported by NIS 13.3 million of depreciation and amortization and about NIS 2 million of share-based compensation, while working-capital changes were only about NIS 2 million negative. Inventory fell by NIS 2.4 million, suppliers rose by NIS 2.2 million, but receivables rose by NIS 2.8 million and accrued liabilities fell by NIS 4.8 million.

That means the business itself did not break on a cash basis. Still, that is not the most relevant frame here. Unitronics chose in 2025 to keep investing heavily in development while also distributing cash aggressively. So the right way to judge flexibility is the full bridge, not the narrower normalized one.

Short-Dated Debt, Not a Balance-Sheet Crisis

On one hand, this is not a debt story. At year-end the company held NIS 5.25 million of cash, current bank debt of NIS 11.34 million, and lease liabilities of NIS 1.76 million current plus NIS 0.68 million non-current. In other words, there is no heavy debt stack relative to equity of NIS 84.9 million.

On the other hand, the picture should not be prettified. The loan taken in July 2025 carries floating interest in a range from prime to prime minus 0.5%, and it is repaid in five equal quarterly installments, with full repayment due by September 30, 2026. So the debt may not be large, but it is short. When year-end cash is only NIS 5.25 million, even a modest short-dated loan becomes a real checkpoint.

There is another important capital point in equity. Equity fell from NIS 91.1 million to NIS 84.9 million in 2025 not because the business lost money, but because the company paid NIS 29 million of dividends while earning NIS 22.7 million. So cash distribution ran ahead of annual earnings. That is not an immediate danger, but it does change the read on capital discipline.

Outlook

Before moving into 2026, five non-obvious points should be fixed in place:

1. Data centers are no longer just a future story. They already crossed 10% of sales, but they are still not large enough to offset broader US weakness.
2. The development-expense line does not show the full investment burden. Total development spending rose, but more of it was capitalized into the balance sheet.
3. Backlog fell, but it should not be read like a project-company KPI. Management itself explains that customers returned to immediate delivery once supply times normalized.
4. Tariffs hit confidence before they hit margin. Price updates can partly defend profitability, but they do not remove demand hesitation.
5. The change in control created a new option set, but not yet a proven commercial result. There is promise, but still no measurable evidence.

2026 Looks Like a Proof Year

The right label for 2026 is a proof year. Not a reset year, because the business remains profitable and cash generative. Not a breakout year either, because the 2025 numbers actually moved backward. It is a year in which Unitronics has to prove four things at once: that data-center growth keeps expanding, that US demand stabilizes under the new tariff regime, that direct sales really improve customer quality, and that cash no longer depends on new borrowing to support distributions.

The report gives mixed clues. On one hand, backlog stood at NIS 20.25 million on February 24, 2026, up from NIS 17.96 million at year-end 2025. On the other hand, the company itself says backlog is only a partial indicator because most orders in this market are normally placed for immediate delivery. So the really important number in the coming reports is not just backlog. It is whether US revenue stops declining and whether the data-center wedge keeps expanding fast enough to matter.

What the market could miss at first glance is that the modest early-2026 backlog recovery is not enough on its own. In a model like Unitronics, where customers mostly order for immediate delivery, the real trigger is the quality of actual demand, not the size of a long backlog.

What Must Happen Over the Next 2 to 4 Quarters

First, data centers need to remain a true growth engine rather than a one-off jump. The company does not need another one-year doubling, but it does need the vertical to keep taking a bigger share of sales without creating too much new strain in supply or working capital.

Second, the US needs to stop being the factor that drags the entire top line lower. The company already raised selling prices to defend profitability. Now it has to show that customers keep ordering. If not, the company could end up with better pricing on lower volume.

Third, the dividend policy will have to face reality. The company maintains a policy of distributing at least 50% of net profit, but in 2025 it actually paid NIS 29 million, more than annual earnings. If 2026 remains a bridge year commercially, it is hard to see how that pace stays comfortable without pressure on cash.

Fourth, if the change in control has real economic value, it needs to start showing up in something concrete: customer access, a new market, a broader solution set or a new vertical. Without that, the Illgin story remains mainly an ownership story, not yet a performance story.

Risks

The United States Is Both the Engine and the Fault Line

When NIS 80.4 million of revenue comes from the US, any change in tariffs, customer confidence or trade terms hits the thesis directly. The company itself says it still cannot assess the full effect of the new tariff regime and is only taking steps to reduce the direct impact for now. That is careful language, but it also means the issue is not resolved.

The Technology Risk Is Double: Development Success and Commercial Conversion

The company keeps investing in controller families, Motion and UniCloud, and estimates 2026 R&D spending of about NIS 17 million. That is positive because it preserves product relevance. But it is also a two-part risk: it is not enough to develop, the company also needs to commercialize. If those programs do not translate into demand, Unitronics will be left with a high development-asset balance and stretched cash.

Supply Chain Is Starting to Re-Emerge, Precisely Through the New Growth Engine

After the global component shortage ended in 2024, it would have been easy to assume that chapter was over. The end of 2025 says otherwise. The company describes longer lead times for memory components and some price increases because of data-center demand. So even if the vertical continues to grow, it may do so together with higher inventory needs and a higher funding burden.

Aggressive Capital Allocation Could Turn a Strength Into a Yellow Flag

Unitronics remains profitable, with working capital of NIS 32.9 million and equity of NIS 84.9 million. But when a company pays NIS 29 million of dividends in a year when net income is NIS 22.7 million, while also capitalizing NIS 12.38 million of development costs, it is difficult to call the balance sheet conservatively managed. This is not an immediate threat, but it is clearly a risk of reduced flexibility precisely when the company still needs to invest.

FX Is a Background Variable, Not the Core Issue

In 2025 the company recorded NIS 1.36 million of net financing income, mainly from hedging transactions and exchange-rate effects. It also uses forward contracts with maturities of one to twelve months. Still, the sensitivity tables show that a 5% move in the dollar changes equity by only about NIS 232 thousand, and a 5% move in the euro by about NIS 421 thousand. So FX is part of the picture, but it does not currently look like the main risk. US demand, tariffs and development are much more important.

Conclusions

Unitronics did not finish 2025 as a clean growth company, but it also did not finish as a stuck company. There is a new engine that already proved genuine initial penetration in data centers, there is clear progress in direct sales, and there is still a business that knows how to generate cash. What blocks a cleaner reading is the combination of weaker US demand, heavy development investment and a cash-distribution policy that left too thin a cushion.

In the short to medium term the market will focus less on whether Unitronics has a vision and more on whether it can turn that vision into numbers without losing flexibility. In plain terms: does the US stabilize, do data centers keep growing, and does the cash balance finally rebuild in a way that is not dependent on more borrowing support.

Current thesis: Unitronics enters 2026 with a real new growth engine in data centers, but until US demand stabilizes and cash strengthens, this remains a proof year rather than a harvest year.

What changed versus the simple read of the company is that the new engine is no longer theoretical. Data centers already crossed 10% of sales and direct sales keep rising. At the same time, the financial reading also changed: Unitronics looks less conservative in capital allocation because distributions ran ahead of earnings while the cash cushion remained thin.

The counter-thesis: the caution here may prove too heavy. The company still posts gross margin close to 50%, operating cash flow of NIS 36.8 million, modest bank debt, and a new growth engine that could continue expanding. If the US stabilizes, 2025 may ultimately look like a temporary bridge year.

What could change the market read in the short to medium term is not another discussion about technology potential. It is the first report that shows, at the same time, stabilization in the US, continued growth in data centers, and a real easing in cash pressure. That is the triple checkpoint the market will be looking for.

Why this matters: because Unitronics is sitting at exactly the point where a niche industrial technology company can move up a level and become a more relevant automation-solution provider, but only if the new growth layer connects to cash as well as to sales.

Over the next 2 to 4 quarters the company has to show that data centers keep growing, that US revenue stops sliding, and that cash can rebuild without another new loan filling the gap. What would weaken the thesis is a combination of more US softness, renewed inventory pressure, and continued distributions at a pace that does not match the current earning power.