PAMS: Why 42% of Sales to Customer A Still Does Not Capture the Full Dependency
Customer A ended 2025 at 42% of sales, but more than half of that relationship was recognized in the US service-manufacturing model where the customer supplies the raw materials. At PAMS, the real dependence is broader than the revenue line: it sits in the US footprint, the leased production line, and the only customer framework agreement the company has.
42% Measures Revenue, Not Dependency
The main article argued that PAMS finished 2025 with strong margins, a comfortable balance sheet, and new capacity on the way, but with customer concentration still sitting in the background. This follow-up isolates the part that the 42% figure slightly hides. At PAMS, Customer A is not just a large buyer. It is also the counterparty that supplies raw materials to part of the operation, the customer around which the US production base is structurally built, and the only customer with a framework agreement.
That remains the right starting point even after the customer’s internal reorganization. During 2025 the customer completed a split into several entities, and after that split it is referred to as Solstis. The company says it does not expect a material effect on the relationship format. In plain terms, the name changed. The dependency did not.
That is why 42% of sales is a real number, but not a full map of dependence. It describes how much revenue was recognized against Customer A in 2025. It does not describe how much of the US plant, the long-term agreement, and the operating logic of the business still sit on that same relationship.
More Than Half the Relationship Is Booked on a Lower-Revenue Basis
The critical detail sits inside the US operating structure. In both the business description and the segment note, the company explains that in the US plant Customer A supplies the raw materials and PAMS provides manufacturing services. That means the top line does not capture the full product value. It captures only the cash consideration paid for the manufacturing service. This is not a minor accounting footnote. It is the reason why a simple comparison between Customer A at 42% of sales and Customer D at 18% is not a comparison between transactions recorded on the same basis.
Note 4 gives the key datapoint: about 56% of sales to Customer A in 2025 were carried out through the US plant, meaning through the structure in which the customer transfers the production materials and revenue is recorded only in the amount of the cash consideration. In reported revenue terms that is about 28.8 million dollars out of Customer A’s 51.43 million dollars. The remaining roughly 22.6 million dollars was mainly produced through the Israel plant, where PAMS buys the raw materials itself and sells a finished product.
That is the core point. More than half of the relationship with Customer A in 2025 was recorded through a structure that mechanically makes reported revenue smaller relative to the production volume involved. So 42% of sales does not mean Customer A represents only 42% of the operational dependence. It means that this is the layer that reaches the revenue line after the US service-manufacturing structure has already stripped out the raw-material value.
That distinction also matters for profitability. The company says that profitability in the US manufacturing activity is lower than in the Israel plant for the same production volume. In other words, more than half of the relationship with Customer A in 2025 ran through a layer that management itself describes as both lighter in reported revenue and weaker in profitability.
This Is Dependence in the Plant, the Inputs, and the Footprint
There is another layer that is easy to miss. In the US activity, Customer A is not only the buyer of the output. It is also the party that supplies the raw materials needed for the manufacturing process. From PAMS’s perspective, that reduces exposure to third-party raw-material suppliers in that layer. From a business-dependence perspective, it concentrates even more power in the same counterparty.
The same logic appears in the plant structure. The company states that as of the report date the US site is used fundamentally for production for Customer A, while the Israel site serves a variety of customers and can also produce for Customer A when the customer asks for shipments to regions that are not usually the United States. So this customer does not sit only inside the major-customer table. It also sits inside the division of labor between the two production sites.
| Dependency Layer | What 2025 Shows | Why 42% Is Not Enough |
|---|---|---|
| Reported revenue | Customer A at 51.43 million dollars, about 42% of sales | This is only the accounting-recognition layer |
| Recognition basis | About 56% of Customer A sales were booked through the US plant in a manufacturing-service model | In that layer revenue is recorded without the raw-material value supplied by the customer |
| Plant footprint | The US plant is fundamentally used for production for Customer A | This is dependence in production infrastructure, not only in sales |
| Inputs | The customer supplies the raw materials for US production | The same counterparty sits on both the demand side and the input side of that activity |
| Commercial structure | This is the only customer with a framework agreement | Most other customers work through ad hoc orders |
The Agreement Deepens the Tie, but Does Not Create Hard Visibility
This is where it becomes clear why 42% does not tell the whole story, but also why the dependence is not fully protected. Customer A is described as a roughly 20-year customer, and the company says explicitly that it has no framework agreements with customers other than Customer A. The agreement signed in 2019 replaced the earlier arrangement and sets a 10-year term, with the customer holding the right to extend for another 5 years and then automatic renewal unless early termination notice is given. At the same time, the company leases a production line from the customer, transferred to the US subsidiary at the end of 2019, for a 15-year period that can be shortened by the customer under agreed conditions.
This is already dependence embedded in assets and contracts, not only in turnover. But the important qualification comes immediately after that. The agreement does not state minimum order quantities for the orders submitted from time to time. It does contain a minimum mechanism derived from the scope of the customer’s product sales in North America, but it still does not turn the relationship into a hard contract in which PAMS’s work volume is fully locked in advance. The company also says explicitly that because it is not involved in producing the customer’s final products or selling them to end customers, it cannot detail the relevant operating data of that customer or forecast the order volumes it will receive from it.
That creates a two-sided conclusion. On one hand, PAMS benefits from a deep, long-standing, and unique commercial relationship that does not exist with its other customers. On the other hand, that depth does not remove the fact that order flow still depends on the customer’s own sales pace and operating decisions. That is exactly why 42% is an incomplete metric: too small to capture the depth of the integration, and too flat to capture the uncertainty around it.
What Has to Be Measured Next
Anyone trying to read PAMS correctly in 2026 cannot stop at Customer A’s share of revenue. Four more precise questions matter. First, does the US plant begin serving a broader layer of customers, or does it remain in practice an asset built around Solstis. Second, does the US service-manufacturing share inside Customer A sales start falling, so that more of the relationship is booked as full-product sales from Israel. Third, does the customer’s reorganization really remain neutral at the order-flow level. Fourth, can PAMS preserve this deep relationship without allowing the customer to accumulate even more pricing and operating leverage over the company.
The conclusion of this follow-up is straightforward. 42% is the starting point, not the endpoint. At PAMS, Customer A is bigger than the headline revenue number because it sits simultaneously in revenue, in the raw materials of the US activity, in the US production base, in a leased production line, and in the only agreement that is not built merely on ad hoc orders. As long as that structure remains in place, the real question is not only how much PAMS sells to Customer A. It is how hard it would be to separate the US activity from that relationship.
What the market may miss on first read: the move to Solstis did not change the level of dependence. It mainly made it easier to see. The customer can look smaller when measured only through reported revenue, but on the operating-structure test it still sits on one of the central axes of the company.