SPEEDVALUE 2025: Growth Jumped, but Cash Quality Still Has Not Settled

Getting to Know the Company

At first glance, SPEEDVALUE looks like a simple growth story. Revenue rose 52% in 2025 to NIS 170.6 million, defense exposure increased, a cyber consulting layer was added, and the group now includes a string of subsidiaries in Israel, Bulgaria, and Romania. That is only part of the picture. The company does not sell a software product with long-dated hard contracts. It mainly sells hours, projects, and IT services, mostly on hourly or monthly pricing. That is why the interesting question is not whether it can win work. It did. The real question is whether it can hold that work through margin, cash flow, and forward visibility.

What clearly works today is demand. In 2025 the group had about 242 active customers, and growth came mainly from a sharp shift toward defense, public-sector, and cyber work. The defense sector alone rose to NIS 77.8 million, or 45.6% of revenue, from NIS 41.4 million a year earlier. The Ministry of Defense became a major customer at NIS 31.3 million, while Rafael remained a major customer at NIS 27.0 million. On top of that, 57.8% of 2025 revenue came from customers with less than two years of tenure, versus just 34.1% in 2024. In other words, the market for the group’s services is there, and the new-customer penetration is real.

But the picture is still not clean. The company grew faster than its economic quality improved. Gross profit rose only 20.6%, far below revenue growth, so gross margin fell to 19.9% from 25.1%. Operating cash flow did turn positive, but only reached NIS 3.9 million. Cash fell to NIS 9.8 million, below the NIS 11.1 million of short-term bank debt. And the company itself makes a critical point: it has no binding backlog, because work orders are mostly budget envelopes that the customer can usually reduce or cancel on short notice of about 30 days.

That is also the 2026 bottleneck. SPEEDVALUE is not currently constrained by a lack of opportunities. It is constrained by the gap between headline growth and growth quality. More than half of revenue comes from relatively new customers, a meaningful part of the expansion sits in defense and public accounts where budget cycles matter, the model leans heavily on freelancers and subcontractors, and cash still does not sit above short-term debt. So 2026 is not a clean breakout year. It is a proof year.

The Economic Map in Brief

What Works and What Is Still Not Clean

Events and Triggers

The first trigger: 2025 was more a year of mix change than a year of clean organic expansion. The technology sector barely grew and stayed around NIS 57.4 million, while defense jumped 87.7% to NIS 77.8 million, medical nearly tripled to NIS 10.6 million, and public and government added NIS 14.4 million from essentially zero. That is not just more revenue. It is a different identity of work.

The second trigger: the list of major customers changed materially. In 2024 Microsoft was still a major customer at NIS 20.6 million. In 2025 it disappeared from the over-10% list, and the Ministry of Defense entered with NIS 31.3 million. Rafael remained a major customer, but its weight fell to 15.8% from 21.5%. So the company did not just grow. It also rotated its concentration points.

The third trigger: the annual picture hides a softer intra-year pace. The first half of 2025 produced NIS 92.5 million of revenue and NIS 2.9 million of operating profit. The second half fell to NIS 78.0 million of revenue and NIS 1.9 million of operating profit. That does not mean the story broke, but it does mean the 52% full-year headline does not fully describe the run rate entering 2026.

The fourth trigger: on July 1, 2025 the company completed the purchase of the remaining minority stake in Mankitak. That helps from a control standpoint because the minority layer is gone, but it also means the company had to simplify ownership in a subsidiary that still posted a NIS 1.49 million pre-tax loss in 2025. This is a control-improving move, not necessarily a move that immediately cleans up the economics.

The fifth trigger: Liacom, acquired in December 2024 for NIS 21 million, produced both contribution and warning in 2025. On one side it recorded NIS 2.06 million of pre-tax profit. On the other, the company recognized a NIS 1.64 million goodwill impairment and also recorded a NIS 1.50 million gain from marking down the contingent consideration, because Liacom’s 2025 and 2026 forecasts were reduced. That is the heart of the story. The acquisition still works operationally, but no longer on the original read.

What the market can easily miss on first read is the gap between “there is a lot of work” and “there is strong visibility.” The company has work, but no binding backlog. Customers are not committed to minimum volumes, and work orders are mainly budget envelopes. In that kind of model, even a very strong year does not automatically produce equally strong visibility for the next one.

Efficiency, Profitability, and Competition

The central point is that SPEEDVALUE’s growth was much stronger than the improvement in its economic quality. Revenue rose 52%, but gross profit rose only 20.6%, so gross margin fell to 19.9% from 25.1%. Operating profit did rise 79.9%, but still reached only NIS 4.8 million, or 2.8% of revenue. That means the company is not stuck on demand. It is stuck on turning demand into margin.

One reason is the cost structure. Development employee cost rose to NIS 78.9 million, while subcontractor cost rose to NIS 55.3 million. Subcontractors reached 40.5% of total cost of sales, up from 36.8% a year earlier. At the same time, the number of freelancers jumped to 141 from 51. That is not automatically a problem. In this model it is also a source of flexibility. But it does mean part of the growth came through a heavier delivery layer rather than through stronger pricing power.

The service mix tells the same story. Server application, cloud, and related work remain the base, but in 2025 a cyber consulting layer of NIS 31.4 million was added, equal to 18.4% of revenue. By contrast, Web and Mobile barely moved and stayed around NIS 31.5 million. So growth did not come from broad-based improvement across all engines. It came from specific engines and from activities that were added or expanded.

Who Actually Holds Up the Profit

The company reports as one segment, but the subsidiary table gives a sharp clue on the real economics. Code Value ended 2025 with NIS 7.16 million of pre-tax profit, and Wall Dan with NIS 3.31 million. Develop Soft and Code Agile were positive but small. On the other hand, Cloudax lost NIS 1.69 million and Mankitak lost NIS 1.49 million. Liacom earned NIS 2.06 million before tax, but simultaneously triggered goodwill impairment at group level.

What matters is that the older core still carries most of the profit, while parts of the acquired layer have not yet shown that the acquisition economics are closed. That does not have to be negative for the long term. But it does mean the market should be careful not to read NIS 170.6 million of revenue as if the whole group already operates at the same level of quality.

From a competition standpoint, the company itself describes a market with low entry barriers in some areas and competition from players such as IBM, HP, TCS, Accenture, Matrix, One, Malam Team, Hilan, and Ness. So a year of very strong revenue growth combined with gross-margin compression is not noise. It is a reminder that scale alone still has not created immunity.

Cash Flow, Debt, and Capital Structure

When reading SPEEDVALUE through cash flow, the right framework here is all-in cash flexibility. That is the correct test because the thesis question is not how much EBITDA the company can show on paper, but how much cash is left after the real cash uses. On that test, 2025 still does not look clean.

The company generated NIS 3.9 million of operating cash flow in 2025. That is a sharp improvement from negative NIS 3.4 million in 2024, but it is still a very small number relative to revenue. Then the uses arrive: NIS 3.0 million of contingent consideration payments, NIS 3.6 million of lease repayments, NIS 7.9 million of loan repayments, NIS 1.2 million of interest, and another NIS 0.27 million of CAPEX. In other words, before new borrowing, the all-in cash picture of the year is still clearly negative.

If the year is rebuilt in an all-in cash framework, the read gets even sharper: after NIS 3.9 million of operating cash flow, the company paid more than NIS 12 million of CAPEX, contingent consideration, lease cash, interest, and debt principal. It covered part of that through NIS 5.38 million of new borrowing, but even after that cash still fell. So it is still too early to say growth is funding itself.

The Credit Structure

At the end of 2025 the company had NIS 11.06 million of short-term loans and just NIS 56 thousand of long-term loans. It also carried NIS 5.29 million of discounted lease liabilities. The effective interest rate on short-term loans stood at 6.12%. On August 1, 2025 the company renewed credit lines totaling NIS 14 million through January 31, 2027, and by the time the financial statements were signed it was using NIS 7 million of that framework.

That renewal helps. It removes immediate funding risk. But it also comes with a price: the company remains secured by a floating first-ranking charge on all assets, including subsidiary shares, while Code Value and Wall Dan provide unlimited guarantees. So flexibility remains available as long as the banks stay comfortable, not because the balance sheet is already free of pressure.

Covenants and the Real Margin of Safety

On paper, the company is not close to a covenant event right now. Equity attributable to shareholders reached NIS 49.3 million, or more than 46% of the balance sheet, far above the 25% threshold. The company also says it complied with all financial covenants throughout the period. But that does not mean there is no constraint. The practical constraint is that NIS 9.8 million of cash against NIS 11.1 million of short-term bank debt means the year still closed without a comfortable liquidity surplus.

The Positive and Negative Side of the Same Move

Extending and renewing the bank lines is a two-sided move. On the positive side, the company bought time and breathing room for a phase in which acquisitions still have not fully converted into cash. On the negative side, it also means the constructive read on 2025 still depends partly on the bank staying in the story. That is not a critical weakness. But it is also not the picture of a company that has already shifted into calm self-funding mode.

Outlook

Before looking at 2026, four points need to be aligned because the annual headline does not resolve them:

1. The company has no binding backlog, so even after NIS 170.6 million of revenue the visibility level remains only moderate.
2. 57.8% of revenue comes from customers younger than two years, which means a large part of growth is still too young to be treated as seasoned base business.
3. The second half of 2025 was weaker than the first, so the run rate entering 2026 is softer than the full-year headline suggests.
4. Liacom already required a goodwill impairment, and Cloudax’s valuation still relies on a first year of 23.4% negative growth followed by very sharp rebound assumptions. That means not every acquisition layer is already sitting on fully stable ground.

Management outlines three clear directions for 2026: expanding the offshore activity through Bulgaria and Romania, expanding the full outsourcing model and entering geographies such as England, and extracting more synergies between the acquired companies in order to offer end-to-end solutions. It also says no additional funding should be required for ongoing operations over the next year, while still noting that bank and non-bank funding sources are being reviewed from time to time in line with strategy.

That is why the coming year looks like a proof year, not a breakout year. Management is effectively saying: we have demand, we have offshore capacity, we have a broader solution set, and we do not have to raise capital immediately to keep operating. But it does not give a clean quantitative target on growth, margin, or cash. When the message is qualitative rather than quantitative, the market asks for proof in the numbers themselves.

What Has to Happen Over the Next 2-4 Quarters

What can change the market read fairly quickly is a combination of two things: operating cash flow staying positive without another one-time sharp release in receivables, and quarters in which revenue holds up despite the lack of binding backlog. On the other hand, if cash falls again while short credit use rises, the market will read 2025 less as a growth story and more as a soft funding-pressure story.

Risks

No binding backlog means visibility is lower than the headline suggests

This is probably the most important issue a reader can miss. Customers are not committed to minimum volumes, and work orders are mainly budget frameworks. In many cases the customer can reduce or cancel the order on short notice of about 30 days. So NIS 170.6 million of revenue in 2025 does not automatically translate into high visibility for 2026.

Customer concentration and tender exposure

Two major customers explain 34.2% of revenue. In addition, part of the work is awarded through tenders. The company itself states that non-renewal of contracts, failure to win follow-on tenders, or failure to win new tenders could materially hurt revenue. In SPEEDVALUE’s case, that is not theoretical, because the meaningful new customer layer sits more heavily in defense and government.

Human-capital risk

The company operates in a sector where scarcity of skilled people can directly pressure profitability. That is especially visible here because the model relies not only on salaried employees but also on a growing external freelancer layer. If recruiting and retention costs keep rising, or if the company must use even more subcontractors to meet demand, it is difficult to see gross margin expanding quickly.

Funding and interest-rate risk

Short-term loans carry floating rates, and the company itself says inflation and rate trends can hurt both its customers’ activity and its own expense line. At the same time, all company assets are pledged, and part of the group’s flexibility still depends on continued willingness from the banks to provide credit.

Goodwill and valuation-assumption risk

At the end of 2025 the company carried NIS 31.5 million of goodwill and NIS 15.5 million of intangible assets. Liacom already required a NIS 1.64 million impairment. Cloudax did not impair, but its value test still relies on a first year of negative 23.4% growth followed by a fast return to high double-digit growth. That does not mean the numbers are wrong. It does mean the margin for error is not especially wide.

Conclusions

SPEEDVALUE ends 2025 with stronger demand, a broader footprint, and much deeper penetration into defense and public-sector work. That is the support for the thesis. The main blocker is that this layer still has not translated into the same quality of margin, cash flow, and visibility. In the short to medium term, the market will focus less on growth itself and more on whether the company can live with this profile without leaning on short bank lines and without relying on generous assumptions around past acquisitions.

Current thesis: SPEEDVALUE has already shown that it can expand activity and change its mix, but 2026 will decide whether the new growth really turns into durable cash and a revenue base that stays.

What changed versus the earlier understanding of the company: Until recently it was easier to read SPEEDVALUE as an IT services company growing through acquisitions. In 2025 it became clear that the center of gravity moved into defense, public-sector work, and relatively new customers. The right read today is therefore not only acquisition-led growth, but growth through work that is still less locked in.

Counter-thesis: One can argue that the caution here is excessive because the company stayed in compliance with covenants, short-term debt already came down during the year, operating cash flow turned positive again, and the new exposure to the Ministry of Defense and Rafael should actually improve demand quality. If the newer customers mature and the offshore layer stabilizes, margin and cash may simply arrive with a lag.

What could change the market reading in the short to medium term: Better gross margin, cash moving back above short-term debt, and quarters in which revenue remains firm despite the lack of binding backlog would improve the read quickly. Another softer second half, another cash decline, or another acquisition-related surprise would push it the other way.

Why this matters: In an IT services company, it is not enough to show that work is coming in. It needs to stay, it needs to be profitable, and it needs to turn into cash after real uses. That is exactly the test that is still open at SPEEDVALUE.

What must happen over the next 2-4 quarters for the thesis to strengthen, and what would weaken it: The thesis will strengthen if the company shows that even without binding backlog it can hold a broad work base, modestly improve gross margin, and generate more clean cash. It will weaken if the newer customers prove less stable, if short-term debt continues to sit above cash, or if earlier acquisitions keep producing more friction than contribution.