Abra in the first quarter: growth is working, cash has not caught up yet

Abra opened 2026 with a quarter that gives an initial answer to the question left after the 2025 reports: whether backlog and acquisitions are actually turning into revenue. The operating answer is positive, with revenue up 33% to NIS 158.4 million and operating profit rising faster than revenue because selling, marketing, general and administrative expenses fell sharply as a share of sales. But the quarter does not close the quality-of-growth test. Gross margin fell from 24% to 22%, operating cash flow declined to NIS 2.8 million, and cash decreased despite higher net profit. The current read is therefore not that the IT services platform has moved into an easier phase, but that scale is starting to work while working capital, funding and the acquisition layer still test how much of that profit is really accessible to shareholders. The next reports will matter less for the existence of growth and more for three proof points: gross-margin stability, cash that remains after receivables and leases, and acquisition funding that does not rely mainly on short, non-binding credit lines.

Revenue answered one test, growth quality did not

Abra is a software services and integration company: enterprise systems implementation, development, cloud, data, open-source services and a platform that has also been built through acquisitions. This is not a classic software-product company with gross margin that expands by itself. The economic machine combines people, projects and acquisitions, so the right test is not only how much revenue grew, but how much of that growth was organic, how much was bought, and how much cash it consumes on the way.

The quarter looks strong on the revenue test. Revenue was NIS 158.4 million versus NIS 119.4 million in the comparable quarter, and the presentation attributes about 13% of the growth to organic activity. On that basis, less than half of the revenue increase came from the existing business, with the rest mainly from consolidating acquired companies. That is not automatically negative, because acquisitions are part of the model. But it changes the quality of the read: the quarter proves that the 2025 acquisitions are already in the income statement, not that the existing business alone made a step-change.

This connects directly to the prior backlog-quality analysis. The question then was whether the 2026 backlog, part of which was based on framework agreements and expected revenue, would behave like a sufficiently stable revenue base. A NIS 158.4 million quarter gives a good first answer because it is above the average quarterly revenue pace of 2025. Still, the company did not update backlog this quarter and did not split firm commitments, framework work and estimated demand. The volume test improved, but the firmness test remains open.

The AI layer adds a growth story, but it is not yet a separate profit layer. The company presents a sales agent operating at Isrotel, a CooperVision project in planning stages for rollout to additional countries, and a project for a global electronic-components distributor where request handling time fell from 96 hours to 24 hours and handling capacity rose by about 80%. These examples matter because they show real use cases, not only marketing language. At the same time, there is still no revenue or margin split that shows how much AI already changes the company's economics. For now, it is an addition to the services engine rather than proof of new pricing power.

Scale lowered expenses, gross margin did not join

The important point in the quarter is not only revenue growth, but where it reached profit. Cost of revenue rose 36%, faster than revenue, so gross margin fell to 22% from 24% in the comparable quarter. That is a clear warning flag: the company sold more, but each shekel of revenue left less gross profit.

The improvement came from the operating-expense layer. Selling, marketing, general and administrative expenses rose only 14%, while revenue rose 33%. As a result, operating expenses fell from 18.8% of revenue to 16.2%, and operating profit rose to NIS 9.2 million. This is exactly what management wants to see after acquisitions: overhead, systems and infrastructure spread over a larger revenue base.

The implication is that the quarter strengthens the scale argument, but not the pricing-power argument. EBITDA excluding IFRS 16 depreciation rose 42% to NIS 13.7 million and the margin improved to 8.6%, but gross profit did not grow at the same pace. For an IT services company, that difference matters: improvement from overhead absorption can continue for several quarters, but it is not unlimited. For the improvement to be deeper, the company needs to show that the new activity and acquisitions are not only adding volume, but also holding gross margin more steadily.

The presentation adds another clue to that tension. Since the start of 2026, the company recruited 104 new productive employees, and at the end of the quarter it had 125 open positions. That hiring fits a company expecting continued demand, but it also reminds investors that the model requires headcount before revenue and profit stabilize. If the hiring converts quickly into projects and billable work, it will support further scale. If not, the same people layer can bring pressure back to the expense ratio.

Cash flow still shows a working-capital burden

The cash test is weaker than the profit test. Net profit rose to NIS 5.5 million, but operating cash flow was only NIS 2.8 million, down from NIS 7.6 million in the comparable quarter. The main reason is working capital: asset and liability movements consumed NIS 10.7 million, mostly from higher receivables and other current assets, partly offset by an increase in suppliers.

The all-in cash picture tests how much cash remained after the quarter's actual cash uses: operating activity, investments, lease repayments, credit movement, payments to non-controlling interests and transaction payments. On that basis, the company did not generate surplus cash in the quarter. Positive operating cash flow was not enough to prevent a NIS 6.4 million decline in cash and cash equivalents.

That gap is especially important because the balance sheet still looks comfortable. Cash and deposits were NIS 52.4 million, and their surplus over financial debt was NIS 7.1 million. But that picture comes before NIS 72.0 million of lease liabilities and before NIS 32.2 million of liabilities for business combinations, put options and contingent consideration. There is no survival pressure here, but there is a limit to how quickly the company can keep buying activity, pay for past deals and invest in growth without stronger operating cash flow.

Acquisition firepower still depends on funding quality

The positive funding point is that the banks are not applying pressure now. Equity for the covenant was NIS 357.8 million versus a NIS 150 million minimum, the equity-to-balance-sheet ratio was 53% versus a 24% to 25% threshold, and the company is no longer required to hold a pledged cash deposit or a minimum cash balance. That is real progress compared with the restricted-cash structure discussed in the prior funding analysis.

Still, funding quality matters more than the headline "firepower". The presentation shows about NIS 150 million of available credit lines for acquisitions, but also says only about NIS 8 million are signed lines, and some of those are used. The implication is straightforward: the company has bank access and willingness to act, but a large part of its ability to buy more activity still depends on future funding decisions rather than a binding source already in place.

The acquisition layer itself has not disappeared. At the start of January, the company bought another 20% of Abra Services for NIS 550 thousand and increased its holding to 80%. After the balance-sheet date, Abra Cloud allocated 7.5% of its shares to the activity manager, and the company's holding declined to 92.5%. These are not large cash events, but they sharpen the value-capture structure: some activities are still held with partners and managers, and part of the economic consideration is built through options, minorities and staged payments. Higher consolidated profit is therefore not always the same as simpler free cash at the public-company level.

2026 looks like a proof year, not a victory year

The first quarter strengthens the company's business direction. Growth did not stall, the 2025 acquisitions are already affecting the report, and operating expenses are starting to benefit from scale. That is a better starting point than a large backlog that remains only a promise.

The main constraint is the quality of proof. Gross margin needs to stabilize, cash flow needs to show it can absorb receivables, leases and acquisition payments, and the company needs to turn the broad funding framework into more binding financing before another acquisition round. The strongest counter-thesis is that the concern is overstated: Abra is growing, far from covenant pressure, has a net financial cash position and presents AI examples that show new demand for integration services. That is a legitimate argument, but for now the quarter mainly proves volume and overhead efficiency, not yet clean margin, cash and funding structure. 2026 is therefore a proof year: each additional strong quarter will reduce the open questions, while renewed cash-flow pressure or non-binding funding would shift attention back from the growth engine to balance-sheet quality.