Abra 2025: Scale Improved, but the Backlog and Funding Test Is Still Open

Understanding the Company

Abra is not a classic software product company, and it is not a small body shop selling hours. It is an IT and integration platform built around three layers: infrastructure and cloud, core enterprise systems such as ERP and CRM, and software development. In 2025 that model widened further through three acquisitions, so the right way to read the year is not just "revenue was up," but whether the group is starting to look like a larger, more efficient platform with a broader base for complex projects.

What is clearly working now is the operating scale. Revenue rose 25% to NIS 554.0 million, fourth quarter revenue jumped 39% to NIS 161.8 million, and fourth quarter operating profit rose 61% to NIS 9.9 million. At the same time, selling and overhead costs grew much more slowly than revenue, the backlog for 2026 reached NIS 413.0 million, and by the time the materials were published the company had credit lines of NIS 147.8 million versus only NIS 47.8 million at year-end 2025. There is real evidence here that the platform is broader and more flexible than it was a year ago.

But this is still not a clean story. Gross margin fell from 25.6% to 23.1%, and in the fourth quarter it fell from 24.0% to 22.3%. Reported net profit rose only 6% to NIS 15.8 million, while adjusted net profit reached NIS 25.7 million. That gap matters. It tells investors that acquisitions are already helping revenue and scale, but they are also creating a layer of amortization, put options, and deferred deal costs that still weighs on the bottom line and on the capital structure.

A superficial reading may treat Abra as if it has already become a pure AI story. That would be a mistake. This is still a labor-intensive services company, with 1,107 employees and contractors at year-end, including 967 software professionals, and revenue of roughly NIS 0.5 million per head across the group. Its economics are still driven by the combination of labor, projects, licenses, cloud consumption, working capital, and funding. AI may expand the opportunity set, but in 2025 the core story is still integration, service breadth, and acquisition absorption.

There is also a practical screen investors should apply early. The market cap is around NIS 428 million, but turnover on the latest trading day was only NIS 18.3 thousand. That says nothing about business quality, but it does say something about actionability. At the same time, short interest is almost nonexistent: short float is just 0.08% versus a sector average of 0.72%, and SIR is 0.64 versus a sector average of 1.339. The market is not building an aggressive short case here, but it is also not offering the kind of liquidity that produces a fast and efficient market read.

The quick economic map looks like this:

Events and Triggers

The first trigger: 2025 was clearly an acquisition year. Methoda was added in June, CRG in August, and Medistat in August. Those three deals expanded technological breadth, opened adjacent niches, and pushed Abra further away from the model of a generic software house and closer to an integrated services platform with stronger specialization.

The second trigger: the fourth quarter is the first period where investors can see the group after the new businesses sit more fully inside the reported numbers. It is not a perfect quarter, but it does suggest that integration has not broken the operating machine. If anything, it shows that the platform can absorb more volume and still improve operating leverage.

The third trigger: in January 2026 the company moved from examining a convertible bond issue to publishing a first draft trust deed. This is still not an actual issuance, and no shelf offering report was published. But the move from principle to a detailed draft does show that Abra is looking to upgrade its funding toolkit rather than rely only on banks and short-dated facilities.

The three acquisitions that changed the year

This table tells two stories at once. On the positive side, Abra did buy capabilities that widen the platform in meaningful ways: Atlassian services, defense and cyber, and a niche clinical data operation. On the other side, a large share of the purchase price was booked as goodwill, and in two of the three targets Abra did not buy 100% of the asset. That means value may be created at the platform level, but part of the economic bill and part of the risk still sit ahead through put options, deferred consideration, and minority interests.

New funding is now part of the story

In January 2026 the board authorized management to examine and advance a convertible bond offering. Eight days later the company published a first draft trust deed. Under that draft, the bonds would be unsecured, non-linked, pay a fixed 2% annual coupon, and mature in a single principal payment in February 2030. The draft also included financial covenants of minimum equity of NIS 90 million and an equity-to-assets ratio of 23% over two consecutive quarters, alongside negative pledge language and distribution limits.

That does not mean the company was under stress at year-end 2025. It does suggest that Abra understands the next phase of its acquisition strategy may require a longer and more flexible funding layer. Bank financing is a useful tool for a functioning business. It is not always the best tool for a serial acquirer that wants to keep expanding without continuously adding pressure to the holding structure.

The fourth quarter as the first integration test

Fourth quarter revenue reached NIS 161.8 million, EBITDA excluding IFRS 16 depreciation reached NIS 14.4 million, and operating profit reached NIS 9.9 million. That does not prove the acquisitions have fully succeeded. It does show that the machine did not break. In fact, the company managed to grow selling and overhead costs more slowly than revenue, which suggests that integration has already created scale benefits above the gross margin line.

Efficiency, Profitability, and Competition

Abra's 2025 operating story is not really a story of better unit economics. It is a story of scale. Gross margin weakened, but above that line a clearer operating lever emerged. Investors should therefore be careful with any simple reading of "growth and improvement across the board." Yes, the lines improved. No, that did not happen because every incremental shekel of revenue became inherently cleaner.

The gross line weakened visibly. Gross profit rose to NIS 128.1 million, but the gross margin fell to 23.1% from 25.6% in 2024. In the fourth quarter the compression was even clearer: 22.3% versus 24.0% a year earlier. One driver is the cost mix. Licenses and cloud services jumped 64% to NIS 136.1 million, far faster than revenue growth. That suggests a heavier mix of licensing, cloud, and service activity with lower gross margins, including the effect of acquired businesses.

Above the gross line the picture improves. Selling and marketing expenses rose 26% to NIS 30.8 million, roughly in line with revenue, but general and administrative expenses rose only 4% to NIS 66.8 million. Put differently, the company did not expand gross margin, but it did absorb growth and acquisitions without expanding the corporate layer at the same pace. That is exactly why operating profit rose 21% to NIS 30.5 million despite weaker gross economics.

The fourth quarter makes the same point even more clearly. Gross margin fell, but selling and overhead as a share of revenue dropped from 18.7% to 16.3%, so operating margin improved from 5.3% to 6.1%. That matters, because it suggests the 2026 engine will depend first on continued absorption of the fixed cost layer, not on a sudden jump in the underlying quality of the gross profit base.

Where the company is competitively strong, and where it is still vulnerable

Abra has two clear strengths. The first is customer diversification. The company states that it has no dependence on a single customer or a small group of customers, and no single product or service accounted for more than 10% of revenue. The second is service breadth. The mix of infrastructure, core systems, and development makes the platform relevant for more complex projects, especially as the market shifts toward AI, data, and integration.

But there is a real counterpoint. Service breadth in an integration business does not automatically mean pricing power. In its capital markets presentation the company argues that AI increases the need for integration, governance, data, and security rather than reducing the need for its services. That may well be true. Still, the 2025 numbers do not yet show a separately visible AI contribution. What investors can see today is mainly a broader services platform. Whether that platform can move up the value chain, rather than simply doing more work inside a similar economic model, remains unproven.

Another important nuance is that customer diversification does not mean the absence of economic concentration. The company explicitly says it still has several large customers and that simultaneous deterioration in those relationships could materially hurt results. So there is no single-customer risk here, but there is certainly dependence on the health of the broader large-account book.

This mix chart sharpens another point: Abra is not a single-theme business. That helps stability, but it also means there is no one segment that can carry the whole story by itself. Defense rose to 14% of fourth quarter revenue and is clearly worth watching. But until that becomes a sustained contribution rather than a quarterly mix move, it should be treated as an interesting wedge, not as the entire thesis.

Cash Flow, Debt, and Capital Structure

The most important point in Abra's capital structure is that this is not a covenant stress story. It is also not a clean "cash rich" story. The real situation sits in the middle: banks clearly see a reasonably strong borrower, which is why some restrictions were removed in the fourth quarter. But the business still does not generate enough clean cash to fund acquisitions, lease obligations, minority buyouts, and full strategic flexibility without thinking seriously about the next funding layer.

Two different cash pictures

To read 2025 properly, investors need to separate two cash frames:

That is exactly why the sentence "operating cash flow rose to NIS 20.7 million" is not enough on its own. It is a better number than in 2024, and it does show improvement. But after lease principal repayment of NIS 11.7 million and reported CAPEX of NIS 4.5 million, the remaining room is still quite narrow. After cash payments for the three acquisitions, business-combination liabilities, and the increase in stake, the all-in picture turns materially negative. That is not an argument against the acquisitions. It is simply the correct way to understand who is funding the growth strategy.

Covenants, facilities, and liquidity

At year-end 2025, cash and cash equivalents stood at NIS 58.4 million, pledged deposits and restricted cash at NIS 0.9 million, and bank debt at NIS 48.8 million. The company reported equity of NIS 366.5 million, an equity-to-assets ratio of 55%, and a current ratio of 1.5. Against its bank covenants the company remains far from pressure: equity of NIS 352.9 million against a floor of NIS 150 million, and an equity-to-assets ratio of 53% against a minimum of 23%, which gradually rises to 25%.

That matters because it means the pressure is strategic rather than covenant-driven. Abra can continue to operate comfortably inside its current structure. The question is whether that structure is also the right one for the next acquisition cycle, or whether it is mainly sufficient to maintain the platform that has already been built.

This is where the story becomes more interesting. In the fourth quarter the company removed requirements to maintain a minimum pledged deposit and a minimum cash balance, and by the time the materials were published it had total credit lines of NIS 147.8 million. But only NIS 7.8 million of that amount is committed. The rest is uncommitted and for one year. That is better flexibility, but it is not the same as durable capital.

The hidden debt is the deal structure

There is another funding layer that is easy to miss. Alongside bank debt, the balance sheet includes NIS 28.7 million of liabilities related to put options for non-controlling interests, plus another NIS 3.4 million of business combination liabilities and contingent consideration. In other words, the 2025 acquisition wave did not only buy new revenue engines. It also pushed part of the economic bill forward through minority interests, put options, and deferred payments.

That is not unusual for an acquirer. But it does mean the real test is not only whether the acquired businesses grow. It is whether they generate enough profit and cash to justify both the goodwill booked and the layer of minority and option obligations that remains behind them.

Why the convertible bond came up now

Against that background, examining a convertible bond makes sense. Under the draft deed, the company was targeting unsecured debt, a fixed 2% coupon, and a 2030 maturity, with a negative pledge and distribution limits. That is not necessarily a distress signal. It is an indication that Abra is trying to turn part of its bank-driven flexibility, much of it short and uncommitted, into a longer and more durable source of funding.

The issuance has not been completed, the final terms are not fixed, and there is no certainty it will happen. So no one should write as if the problem is already solved. But it is just as important not to ignore what management is signaling: the next phase of the story probably needs a different funding architecture.

Outlook

Finding one: 2026 starts with a wider revenue base, but not with a cleaner gross margin base.

Finding two: the 2026 backlog is large, but it is not comparable to the hard backlog of an infrastructure contractor or a developer.

Finding three: funding flexibility is better than it looked at year-end, but a meaningful share of it is still short and uncommitted.

Finding four: AI is a real upside option, but at this point it is still not a stand-alone profit engine.

What the backlog does tell investors

A 2026 backlog of NIS 413.0 million, plus another NIS 3.6 million for 2027 and beyond, is meaningful. Relative to 2025 revenue, that is coverage of roughly 74.5%. The quarterly spread also looks reasonable: NIS 111.2 million for the first quarter, NIS 103.0 million for the second, NIS 98.4 million for the third, and NIS 100.4 million for the fourth. That gives investors much better visibility than they would have from trailing revenue alone.

What the backlog does not tell investors

This is where one of the most important lines in the materials appears. The company says most engagements are framework agreements and monthly or hourly arrangements, and that most can be reduced or terminated with notice. It also says that part of the backlog includes expected revenue over 12 months even when the customer is not contractually committed for that full period, and that the company does not have a centralized and orderly view of the portion of orders that can be canceled.

That does not invalidate the backlog. It does change its quality. In Abra's case, backlog is better understood as a strong indicator of repeat demand and long-term client relationships, rather than as a hard project backlog that is fully locked in. That means 2026 will be judged not only on the headline size of the backlog, but on the pace of conversion into revenue and on whether growth can continue without another round of gross margin erosion or renewed working capital strain.

2026 is a proof year, not a breakout year

The right label for 2026 is a proof year. The company no longer needs to prove that it can buy businesses. It already did that. What it now needs to prove is three other things:

1. That fourth quarter 2025 marked the start of a new operating run-rate rather than a temporarily convenient quarter after consolidation.
2. That the acquired companies can sit inside the group without further dragging down gross margin.
3. That the next funding layer will be durable enough to reduce the trade-off between growth, cash, and strategic flexibility.

There is also an AI test ahead. Abra describes the AI wave as an opportunity rather than a threat. Strategically that is a reasonable argument: an integrator that understands core systems, permissions, data, and business processes may indeed benefit more from a market shift toward embedding AI inside workflows. But the market will want to see that in the numbers. If 2026 starts to show a new layer of repeatable projects that improves pricing power or deepens penetration into existing accounts, that would be a high-quality addition to the thesis. If not, Abra will still look primarily like a better and broader integration platform rather than a structurally better business.

Risks

The first risk is backlog quality. A backlog above NIS 413 million looks strong on paper, but the company itself explains that part of it is based on 12-month expectations even without full contractual commitment, and that it does not have a centralized view of the cancelable portion. Any overly aggressive reading of that number would be a mistake.

The second risk is continued gross margin pressure. If the acquisitions and the heavier mix of licensing, cloud, and lower-margin services continue to weigh on the gross line, the company will need to keep squeezing efficiency out of the corporate layer simply to preserve operating improvement. That lever is real, but it is not unlimited.

The third risk is a funding structure that is not yet fully resolved. Yes, some bank restrictions were removed and the company remains far from covenant pressure. But uncommitted one-year facilities are not the same thing as long-term capital. If the convertible does not happen, or happens on less attractive terms, Abra will remain more dependent on banks and on internally generated cash.

The fourth risk is the unfinished layer inside the deals themselves. Put options, contingent consideration, and non-controlling interests mean the economic price of the 2025 acquisition wave has not been fully settled yet. That is not a footnote. It is part of the financial cost of the expansion strategy.

The fifth risk is human capital. The company itself treats labor as a core asset, and the market for skilled technical staff remains competitive. In a business built around 967 software professionals and a wide mix of functions, execution on hiring and retention affects delivery capacity, sales depth, and project quality directly.

---

Conclusions

Abra exits 2025 as a broader and operationally stronger company than it was when it entered the year. Scale improved, the fourth quarter was good, and the banks now give it more room. The bottleneck has shifted. The question is no longer whether the company can grow. It is whether that growth can turn into cleaner cash generation, a more durable funding structure, and a quality of earnings that relies on more than absorbing overhead.

Current thesis: Abra proved in 2025 that it can scale a platform, but it has not yet proved that the platform can translate into cleaner cash generation and a more durable funding base.

What changed: A year ago the question was whether the company could absorb more acquisitions. After 2025 the question is different: whether that absorption will turn into a more reliable 2026 through backlog conversion, steadier profitability, and a less bank-dependent funding mix.

Counter-thesis: Abra may simply be a mid-sized services roll-up that grew revenue through deals while keeping weaker gross economics, partly soft backlog quality, and continuing reliance on external funding without meaningfully improving the underlying business quality.

What could change the market read in the near term: completing a convertible bond on reasonable terms, preserving fourth quarter operating profitability, and showing that the 2026 backlog really converts into revenue without a renewed rise in working capital pressure.

Why this matters: if Abra executes well through 2026, it can move from being a collection of acquired service businesses into an integration platform with real operating leverage and a workable capital structure. If it does not, it may remain a company that grows revenue faster than it improves the underlying economics.

What must happen over the next 2 to 4 quarters: gross margin needs to stop drifting lower, backlog needs to prove its conversion quality, and the company needs to show that the next funding layer is longer and more durable. What would weaken the thesis is renewed cash pressure, additional deal-related obligations, or signs that AI remains mostly a marketing overlay rather than a source of repeatable, profitable work.