Accel After The Acquisition Year: Why EBITDA Did Not Turn Into Cash

Where The EBITDA Got Stuck

The main article argued that Accel ended 2025 with a broader platform, but with cash that still lagged the expansion. This follow-up does not retell the broad story. It isolates the mechanism. The gap was built in three layers working together: the company gives customers more time than it gets from suppliers, acquisitions keep consuming cash after closing, and above both layers sits a new funding bridge made of equity, bank debt, and public debt.

That is also why the right lens here is all-in cash flexibility, not normalized cash generation. The key question is not how much the business could have produced in a clean world without acquisitions and investment, but how much cash was actually left after real cash uses. Through that lens, Accel did not hit an immediate liquidity crisis. The board reviewed the forecast cash flow and said the company had no liquidity problem. But the absence of a liquidity crisis is not proof of cash quality. It only means the gap is being funded for now.

The numbers are fairly blunt. Cash flow from operations moved to negative ILS 5.5 million. Investment in businesses consolidated for the first time was ILS 59.2 million, and another ILS 5.6 million went to repayment of liabilities for acquired businesses. At the same time, cash used for lease transactions rose to ILS 5.2 million. These are not accounting adjustments. They are real cash uses.

Working Capital: Customers Get More Time Than Suppliers

This is the core of the story. Accel grew into a model where customer credit expanded faster than supplier credit. The company says explicitly that it may grant customers payment terms of up to current plus roughly 120 days. Supplier credit, by contrast, is described as up to 90 days from delivery of goods or receipt of services. That sounds technical, but once revenue scales and the services mix grows, the gap turns into a cash drag.

On an annual average basis, customer credit rose to ILS 151.9 million in 2025 from ILS 121.7 million in 2024. Supplier credit also increased, but only to ILS 61.2 million from ILS 49.7 million. The credit gap therefore widened to ILS 90.7 million from ILS 71.9 million a year earlier. That is exactly where EBITDA can look healthier than the cash balance.

The cash flow statement shows how that moved into the reported cash result. Customers absorbed ILS 24.9 million of cash, and suppliers and service providers absorbed another ILS 11.4 million. Together, those two lines pulled roughly ILS 36.3 million out of operating cash flow. Net income was ILS 4.7 million. In other words, even before stepping into acquisitions, leases, and financing, working capital on its own was enough to flip the picture.

The year-end balance also shows that this is not just a generic growth effect. Gross receivables reached ILS 206.5 million, up from ILS 155.4 million a year earlier. Of that, ILS 176.2 million was not yet due, but roughly ILS 30.3 million was already overdue across the 0 to 30, 31 to 60, 61 to 90, and over 90 day buckets. That is roughly 15% of gross receivables. In addition, ILS 13.1 million of receivables was already presented as non-current. That means not all of this balance is going to unwind quickly even if the business remains healthy.

There is another detail that is easy to miss. Receivables did not grow only through open trade balances. Open balances rose to ILS 177.8 million, but accrued income jumped to ILS 23.4 million from just ILS 3.4 million in 2024. In plain English, a larger share of reported revenue was sitting in accrued amounts that had not yet turned into cash. That is not automatically an accounting red flag on its own, but when it sits alongside long customer terms and negative operating cash flow, it becomes part of the cash-quality thesis.

So the problem here is not merely higher sales. It is growth that needs balance-sheet funding. As long as the company carries average customer credit above ILS 150 million against average supplier credit of roughly ILS 61 million, part of the expansion is being financed internally rather than through self-funding operations.

Acquisitions Kept Consuming Cash After Closing

A common mistake in reading 2025 is to treat the acquisitions as a completed event and therefore as something that can simply be normalized away. That is too convenient. From a cash perspective, these deals remain active well after the legal closing date.

In the Synel acquisition, the total consideration was estimated at about ILS 92.7 million. Of that, roughly ILS 69.5 million was paid in cash at closing, around ILS 22.9 million was booked as deferred consideration, and contingent consideration was initially estimated at about ILS 0.3 million. That already tells you the deal was not finished on day one. It left a layer of seller obligations on the balance sheet.

By March 2026, the story had not closed. It had been rewritten. The company reported that the nominal deferred balance of roughly ILS 20.3 million was replaced with revised deferred and contingent consideration terms. In other words, even a few months after year-end, the seller payment stack from Synel had not disappeared. It had only changed form.

The year-end balance sheet showed that clearly. Current liabilities for acquired companies and businesses stood at ILS 15.8 million, and non-current liabilities for acquired companies and businesses stood at ILS 8.1 million. In addition, contingent consideration was ILS 209 thousand in the short term and ILS 6.8 million in the long term. So even before the next transaction, year-end 2025 was already carrying roughly ILS 30.9 million of acquisition-linked liabilities and contingencies.

The cash flow statement confirms that reading. In 2025, ILS 59.2 million went out under investment in businesses consolidated for the first time, and another ILS 5.6 million went to repayment of liabilities related to acquired businesses. So even before Starlight and Nextwave, acquisitions had already consumed more than ILS 64 million through direct cash use and follow-on payments.

Leases add another layer that can disappear in a superficial read. Lease liabilities rose to ILS 35.1 million at year-end from just ILS 10.1 million a year earlier. Of that total, ILS 32.0 million related to office leases. Cash used for lease transactions rose to ILS 5.2 million from ILS 3.3 million in 2024. That matters for two reasons. First, it is another real cash use that belongs inside the all-in cash picture. Second, the right-of-use asset roll-forward shows that ILS 35.9 million of additions came from businesses consolidated for the first time. So a large part of the new lease burden is not quiet organic expansion. It is another consequence of the acquisition year.

The Funding Stack: A Bridge, Not Proof

That brings us to the third layer. If working capital is pulling cash and acquisitions are consuming cash, something has to fund the gap. In 2025 that was not the business itself. Cash flow from financing was positive ILS 78.7 million. Inside that figure were ILS 42.9 million from issuance of equity and warrants, ILS 5.3 million from warrant exercises, ILS 31.3 million from bank loans, and another ILS 15.3 million net from bank and other credit. Against that, the company paid ILS 7.1 million of bank loan repayments, ILS 5.2 million for lease transactions, and ILS 4.0 million of dividends to non-controlling interests.

This is exactly where liquidity and quality need to be separated. Management says there is no liquidity problem, and year-end covenant headroom does not look tight. In Solutions, tangible equity stood at about 58% of the tangible balance sheet against a 20% minimum, and the ratio of short-term financial debt to operating working capital was about 28% against an 85% ceiling. In Data Tech, tangible equity was roughly ILS 17.1 million, or around 30% of the balance sheet, and short-term credit was 52% of working-capital needs against an 80% ceiling. So the story today is not about a covenant break. It is about a business that still has not shown it can fund its own expansion.

The September 2025 bank loan taken for Synel makes that point more clearly. The company borrowed ILS 30 million for six years at a prime-linked annual rate of roughly 6%, with a one-year grace period in which only interest is paid. At the same time, it pledged 70% of the Synel shares it had acquired and agreed not to create other liens without bank approval. That is not an emergency structure. But it is a structure that ties the financing directly to the acquired asset and to the transaction that created the need for cash.

Then, after year-end, the company added a public-market layer on top. The issuance sequence matters at least as much as the final number:

That sequence ties directly back to the continuation thesis. The bond was not raised to improve a theoretical capital structure. The company says explicitly that it used the proceeds from the bond-and-warrant issue to finance the Starlight and Nextwave acquisition. So the public funding layer came in to keep the acquisition chain moving. It does not prove that the earlier acquisitions had already turned into cash.

The contractual liquidity table adds another layer of depth. As of December 31, 2025, contractual payments from bank credit stood at ILS 102.7 million including interest, lease obligations at ILS 44.8 million, liabilities for acquired businesses at ILS 25.4 million, and contingent consideration at ILS 10.5 million. Even if part of the short-term bank credit is routinely rolled in practice, the scale is clear: funding bought Accel time. It did not erase the bill.

What Needs To Happen Next

The next test for Accel will not be whether it can raise money again. It already proved that. The next test is whether it can reduce the need for the bridge.

The first thing to watch is a genuine working-capital reversal. It is not enough for recurring revenue to be higher. Customer balances have to stop swallowing cash at this pace, and the gap between customer credit and supplier credit has to stop widening. That improvement can come through collections, commercial terms, or mix, but without it the discussion around revenue quality will remain incomplete.

The second thing to watch is a real decline in the acquisition-liability layer. If acquisition balances and contingent consideration only keep changing form, and the next deal gets funded before the previous one has unwound, then the platform may be larger but the cash base is still stretched. That matters even more after the roughly ILS 20.3 million nominal Synel seller balance moved into a new structure, and after Starlight and Nextwave added a ILS 49.2 million closing payment plus up to ILS 40.3 million of contingent consideration.

The third thing to watch is whether the January 2026 bond turns out to be a one-off financing layer for a specific transaction, or a sign that Accel's capital structure is entering a new phase in which every growth step needs public debt, warrants, and discount. The market can live with that, but only if cash starts to follow.

The bottom line: Accel is not stuck because it lacks activity. It is stuck because in 2025 and early 2026 every engine that made the group bigger also demanded cash, credit, and financing before paying it back. That is why the 2026 question is not how much EBITDA the group can present, but how much of it can make it through customer credit, acquisition payments, and the funding stack and survive as real cash in the end.