Apollo Power: What the GPM EPC pipeline is actually worth

The main article argued that Apollo’s real 2026 test is not whether demand exists, but whether demand can turn into recognized revenue, executable backlog, and cash. The GPM thread deserves its own continuation because it sits exactly on that fault line: the headline is large enough to reshape the way the company is read, but the path from headline to revenue is still much longer than a casual reader might assume.

The bottom line already: the GPM agreement is a serious commercial option, not signed backlog of $53 million. Apollo got preferred-supplier status for feasibility work, development, and EPC, but the move from each site into execution remains with GPM, at its sole discretion, and GPM can also terminate the agreement on 90 days’ notice. That does not make the thread weak. It does mean the market should measure it through actual NTPs, not through the full headline number.

What was actually signed

On January 14, 2026, Apollo Power America signed a binding agreement with GPM for the evaluation, development, and rollout of solar solutions across various sites in the US operated or controlled by GPM. Apollo is to serve as GPM’s preferred supplier for development services, including feasibility surveys, and at least 300 sites are expected to be reviewed. Only if GPM chooses to move ahead at a specific site, at its sole discretion, will Apollo provide EPC services for design, engineering, construction, installation, grid connection, and commissioning, directly or through subcontractors, on a turnkey basis and at a specified price per installed watt subject to adjustments.

That already defines the economic boundaries of the deal. Apollo is not getting ownership of a long-lived asset, and it is not sharing in the tax benefits. GPM is entitled to the full tax benefits arising from ownership and operation of the systems. So if the agreement scales, Apollo participates mainly as a development and EPC contractor. That can still be highly meaningful, but it is contractor economics, not owner economics.

What matters here is that the agreement gives Apollo access to a very large possible pipeline, while almost every economically meaningful milestone remains conditional. The key word is not “binding.” It is “if.” If GPM chooses a site. If an NTP is issued. If the adjusted price per watt remains attractive. If execution stays on track. That is why the right way to read this is not “Apollo signed a $53 million project,” but “Apollo opened the door to a chain of projects that still has to convert site by site.”

The path from site evaluation to revenue is longer than the headline

The encouraging part is that this thread did not start from zero. Following the memorandum of understanding between the parties, a first solar-equipment order of about $1.3 million was placed on December 5, 2025. Apollo also delivered feasibility surveys for several sites and said it expected to receive NTPs for seven initial sites. That matters because it shows the relationship has already moved beyond generic dialogue into actual commercial work.

But it matters just as much to be precise about what has not happened yet. The first order was for solar equipment, not the full EPC stream. The NTPs themselves were described as expected, not as already issued. Under the binding agreement, the company estimates that EPC NTPs will be issued in stages over roughly 20 months for sites totaling about $53 million, or about NIS 167 million. That is a completely different order of magnitude, but it is still a chain of future decisions rather than one executed project already in hand.

That chart is the core of the story. It shows why the headline is easy to get excited about, but also why discipline matters. There is already a real commercial proof point at $1.3 million. There is an $8 million cumulative EPC order threshold that activates the warrant package. And there is a $53 million estimated NTP opportunity. Those are three different milestones, with very different levels of certainty.

Another detail the market can easily miss sits in Apollo’s revenue-recognition policy. Service revenue is recognized over time, over the period in which the customer receives and consumes the benefits. Execution and construction revenue is also recognized over time according to milestones, because the work creates or improves an asset controlled by the customer. So even once the first NTPs arrive, this is not a one-step move from contract headline into the income statement. The path runs from agreement, to NTP, to execution, and only then into recognized revenue.

That is where the base case has to be separated from the option value. In 2025 Apollo reported NIS 19.673 million of revenue, of which NIS 10.751 million came from projects recognized over time. Order backlog stood at about NIS 65 million at year-end and about NIS 70 million near the report date. Against that, the estimated GPM NTP opportunity stands at roughly NIS 167 million. This is large enough to change the scale of the project business, but until actual NTPs appear it cannot be treated as if it were already part of the base.

That chart explains why the market quickly reads GPM as a thesis of its own. If it matures, it is materially larger than Apollo’s 2025 revenue base and larger than the backlog reported near publication. But it also explains why the gap should not be declared closed too early. A very large headline sitting on a still-small operating base is exactly where an investor needs to separate option value from operating base.

The warrants open relatively early, not at the end of the road

Three days after the agreement disclosure, the company reported a non-material, non-extraordinary private placement of unlisted warrants to GPM Investments, LLC, which the company said was wholly owned by Arko Holdings, a Nasdaq-listed company. The package covers up to 2,166,667 shares.

The structure is smart from an incentive standpoint, but it changes how the economics should be read. The warrants are exercisable until December 31, 2027, and only if by December 31, 2026 GPM places cumulative EPC orders of at least $8 million. If that threshold is not reached, the warrants expire. If it is reached, the entire package opens.

That is the key point. An $8 million threshold is only about 15% of the estimated $53 million NTP opportunity. In other words, the full warrant package does not wait for most of the pipeline to convert, or even half of it. It opens at a relatively early proof point. For Apollo, that aligns the counterparty with the company precisely when the relationship has to move from evaluation into real orders. For public shareholders, it means potential dilution can arrive before the market has full evidence that the headline opportunity is actually converting.

The exercise-price formula adds another layer. Each warrant is exercisable at the higher of NIS 6.00 per share, or the average of NIS 6.00 and the share closing price on the date the minimum order threshold is reached. On the day before the private-placement report, the share closed at 505 agorot, so the minimum exercise price was 95 agorot above the market price, or 119% of it. This was not a deep in-the-money giveaway. It was an incentive package that started above spot.

If the $8 million threshold is reached and the whole package is exercised at the minimum NIS 6.00 price, Apollo would receive about NIS 13 million and issue up to 2,166,667 shares. The filing states that this would equal about 2.68% of the issued and paid-up share capital after allocation, or about 2.32% on a fully diluted basis, assuming exercise at NIS 6.00. So the warrants are not extreme dilution, but they do signal that the market should look not only at whether orders arrive, but also at the stage at which the first orders begin to alter the capital structure.

So what is this pipeline worth today

More than zero, and less than the headline. That is not evasive. It is exactly what the evidence supports.

On the positive side, several points should not be dismissed. The agreement is binding, not just an MOU. The size of the funnel, at least 300 sites, is large enough to make GPM a potentially real growth engine. There is already an initial equipment order of about $1.3 million. Apollo has already completed feasibility work for several sites and said it expected NTPs for seven initial sites. If that path keeps moving, Apollo can expand its project-revenue base quickly relative to the NIS 10.751 million of project revenue it recognized in 2025.

On the other side, every economically decisive step is still open. No NTP has been disclosed as a completed fact. The price per watt is subject to adjustments, so even if the headline dollar amount sounds large, the actual economics will depend on final pricing and execution quality. GPM keeps both ownership of the systems and the full tax benefits, which means Apollo’s upside comes from construction and services rather than from holding the assets. And GPM retains a 90-day termination right, so even after signing the customer still holds a meaningful exit lever.

That leads to the practical read. The GPM thread is not just “another large customer.” It is a commercialization test for Apollo in a broader US EPC lane. If conversion happens, it can change the scale of the company. If it stalls at the level of feasibility surveys, initial equipment orders, or a small number of sites, the market is left with the same issue flagged in the main article: a lot of commercial interest, but still too little proven revenue at the pace and quality needed to justify the headline.

What will decide the next read

First trigger: actual NTPs. Not more language about expectation, but clear disclosure that sites have moved into execution. Until that happens, the $53 million remains an opportunity set rather than an operating base.

Second trigger: the pace of accumulation toward the $8 million threshold. That threshold does not only activate the warrants. It also provides the first real proof that the customer has moved from evaluation into actual EPC ordering.

Third trigger: economic quality. Since the pricing is per watt and adjustable, the market will need evidence that GPM conversion comes at economics that preserve a reasonable margin rather than through commercial concessions that add volume but weaken quality.

Fourth trigger: recognition speed. Even after orders start to arrive, Apollo will still have to show that they move through milestones quickly enough to flow into reported revenue, rather than remaining stuck at the contractual stage.

Conclusion

The GPM agreement is one of the more interesting commercial options Apollo has added in recent years, mainly because it is large enough to move the needle if it matures. But the real test of this agreement is not the signing itself. It is a measurable move from feasibility work into actual NTPs, and from early orders into the $8 million threshold that signals the customer has truly decided to execute.

So the correct read today is disciplined: GPM is a thread with real strategic and commercial value, but it is still not a base-case anchor on which Apollo’s story can rest. The market will get its answer not from more statements about potential, but from sites that move into execution, from the pace of EPC orders, and from proof that the path produces revenue of sufficient quality rather than just a large headline.