Apollo Power: The big backlog, the partial schedule, and the 2026 conversion test

The Gap Inside The Backlog Number

The main article argued that Apollo's 2026 question is no longer demand. It is conversion. This follow-up isolates the sharpest version of that issue: the backlog headline is large, but the dated revenue schedule attached to it is much smaller.

At the end of 2025 Apollo reported backlog of about NIS 65 million, and near the publication date it said backlog was already around NIS 70 million. Against that headline, the dated table in the annual report only lays out NIS 32.7 million: NIS 17.7 million for 2026, NIS 7.5 million for 2027, and NIS 7.5 million for 2028. Even on the conservative year-end basis alone, about NIS 32.3 million is not mapped in that table. If one reads the table wording as tied to the later report date, the undisclosed timing gap is wider still.

That matters more than it first appears. Only about 27% of year-end backlog is explicitly allocated to 2026. In other words, the big commercial number is already there, but most of the visibility for the next 12 months is not. For a company at Apollo's stage, that is not a disclosure detail. It is the line between backlog that signals commercial interest and backlog that already starts to describe a revenue year.

Why VW Is Both The Explanation And The Timing Problem

The company itself makes this easier to decode. It explicitly says the jump from about NIS 6.3 million of backlog at the end of 2024 to the new figure was driven by the VW order, worth EUR 14.5 million over eight years. At the same time, the VW note says the formal restart approval came in August 2025 and serial production is only planned to begin in November 2026.

That is the critical point in the entire backlog discussion. The same project that explains the jump in the headline number is also a project whose serial contribution starts relatively late. So backlog size and 2026 conversion capacity are not the same thing. They point in different directions: the headline grows now, while the portion that should convert into revenue over the next year remains more limited.

The dated schedule reinforces that reading. Of the NIS 17.7 million assigned to 2026, NIS 16.9 million sits in the first three quarters, while only NIS 0.8 million sits in the fourth quarter. If VW serial production starts in November, the reasonable read is that 2026 still depends mainly on shorter-cycle orders and nearer-term execution rather than on a broad VW ramp already this year. That is an inference from combining the disclosures, not a sentence management writes verbatim, but it is hard to read the timing any other way.

The more interesting nuance is that VW already helped 2025, but in a very different way. Apollo recognized EUR 1.0 million, about NIS 3.8 million, in 2025 revenue for development work performed before the project change. In the directors' report, it also says the increase in receivables was driven mainly by a December invoice to Volkswagen of roughly the same size. In addition, the remaining EUR 0.15 million is expected, according to the company, to be paid in the coming months. Put differently, the near-term support that VW already provided is still support from development work, compensation and project reset economics. It is not proof that serial manufacturing is already behind the company.

What Already Gives 2026 Some Concrete Support

This is where the two approved immediate reports become important. They do not solve the whole gap, but they do show that the next year is not resting only on VW.

On January 4, 2026 SolarPaint received a Ministry of Defense order worth about NIS 1.67 million, with delivery planned by the end of the first quarter of 2026 and payment due within 60 days of delivery. On March 13, 2026 Lippert placed an order for solar fabrics and Panda panels worth about USD 1.1 million, about NIS 3.5 million, for delivery through the end of 2026. The same report also disclosed an additional Volta Solar order worth about NIS 0.5 million, excluding VAT, for immediate delivery. Together, those three pieces amount to roughly NIS 5.7 million.

The meaning cuts both ways. On the positive side, these orders strengthen backlog quality. They come with size, delivery timing and in one case even payment timing, so 2026 is not resting solely on a multi-year program that starts serial production late in the year. On the cautionary side, even if all of this converts exactly as disclosed, it still covers only about one-third of the NIS 17.7 million the company assigns to 2026 in the dated table. So the immediate reports reinforce the visible layer of backlog, but they do not close the transparency gap.

There is another reason those reports matter. They show what is actually supposed to carry 2026 in the near term. These are not eight-year mega-contracts. They are shorter-cycle orders with much nearer delivery windows. If Apollo wants to prove conversion this year, it cannot wait for VW. It has to close the shorter and clearer orders first.

The 2026 Conversion Test

That creates a fairly simple, almost accounting-style test. If the backlog table is meant to function as a real guide, 2026 has to start looking like a year of order closure, not just a year of narrative. Four checkpoints matter most:

• The first and second quarters need to show that the Ministry of Defense order moved from announcement to delivery, invoicing and cash collection.
• By the end of the first half, the table points to NIS 12.6 million of expected revenue recognition. If the realized pace is materially lower, the market will need more than a backlog headline as an explanation.
• The Lippert and Volta Solar orders need to move from disclosed orders into visible execution in the financial statements.
• In VW, what matters now is no longer another one-off development contribution. It is evidence that the project is progressing toward November 2026 without another spec reset or timing slip.

The most sensitive part is that the larger the backlog headline becomes, the higher the cost of disappointment becomes too. If Apollo keeps showing a high backlog number while leaving most of it without a fuller time map, readers will struggle to tell whether they are looking at a future revenue base or mostly at a long tail of slowly moving projects. That is exactly why a large backlog is not automatically a high-quality backlog. Here, backlog quality is measured by timing clarity, not just by size.

Bottom Line

Apollo's backlog does not look too small. The problem is almost the opposite: it looks too large relative to the level of timing detail attached to it. The company provided a strong headline figure, explained the source of the jump, and added shorter-cycle orders that support 2026. What it still has not provided is a full timing map that ties those pieces into a clear revenue path.

That is why 2026 remains a two-layer conversion test. First, the shorter-dated and more clearly signed portion of backlog has to turn into revenue and cash on time. Then VW has to move from development and compensation economics into a serial path that can be measured quarter by quarter. Until that happens, the backlog headline is impressive, but the timing quality behind it is still only partial.