Gan Shmuel: What the Kibbutz, the Land and the Service Web Really Cost

The Cost Layer the Headline P&L Hides

The main article argued that 2025 was first and foremost a price-reset year and a Thailand proof year. This follow-up isolates a different layer: the kibbutz and related entities do not sit only in the ownership structure. They run through the land, labor, management, services, water, electricity and part of the raw-material chain. A reader who looks only at operating profit sees a business that earned less. A reader who works through the notes sees what kind of operating base that profit is actually sitting on.

The sharpest number is in the related-party table. Sales to controlling interests in 2025 were only $51 thousand, while expenses to controlling interests reached $17.92 million, up from $13.42 million in 2024. Roughly 81.6% of that charge sits in cost of sales, another $1.203 million in selling expenses and $2.095 million in G&A. That means this is not a governance footnote. It is a cost layer that runs through the company’s core expense engine.

Even before looking at the summary table, the individual agreements already tell a heavy story. The main-site lease, manpower, officer-service arrangements, fruit procurement from Kibbutz Gan Shmuel, general services, catering, additional leases and back-to-back utility transactions added up to at least $16.5 million before VAT in 2025. That does not mean these costs are unnecessary, illegitimate, or not on market terms. It does mean Gan Shmuel’s economic base is built on a denser and less flexible agreement web than a surface reading of the annual report would suggest.

Land: Rent with an Automatic Upward Ratchet

Land is not just a historical legacy of a kibbutz-owned industrial business. It is a cost mechanism with built-in upward steps. Under the 2019 lease, the company rents about 90.6 dunams that include the plant, logistics center, cold storage, storage yards, employee facilities and other buildings. Monthly rent started at NIS 599,058 plus VAT, linked to CPI, and it also rises by 3% every year on top of indexation. By the end of 2025, monthly rent had already reached NIS 842,721 plus VAT.

There is a brake, but not a day-to-day flexibility valve. After five years an external appraiser is supposed to reassess market rent, and in August 2024 the audit committee was updated that rent would remain unchanged. In other words, just before the weaker 2025 year, the land-cost layer was not reset down. It stayed on the same path, with CPI linkage and an annual upward ratchet. On that basis the company paid about $2.806 million in 2025 for the main site, a 13.3% increase.

That is not a technical detail. When the business cycle gets less friendly, this cost does not move with volumes the way raw materials or commissions do. It keeps climbing. On top of rent, the company also carries taxes, licensing, permits and insurance tied to the site, so land is not only rent. It is a full operating-cost layer.

The smaller support areas are not especially flexible either. The lease for the microbiology lab and the computer-lab offices was extended on the same terms for another five years, through July 2029, and the 2025 charge was $66 thousand. The number is small, but the message is clear: even support space is not sitting on short contracts that can be cut quickly.

Labor and Management: The Kibbutz Sits in Payroll and in the Executive Layer

The heaviest layer is manpower. The agreement was extended in January 2025 through the end of October 2028, and the kibbutz can supply up to 120 assigned members. An increase of up to 15 more needs only audit-committee approval; above that, fuller corporate approvals are required. The daily rate at the report date was NIS 856 plus VAT, CPI linked every month, and it is also subject to an annual check against the average wage in the economy. If the average wage rises faster than CPI, the rate catches up upward.

The important point is that the effective labor cost is not only the daily tariff. The agreement comes with a wider benefit shell: 12 to 26 paid vacation days, sick pay from day one, NIS 1,400 of annual clothing allowance for each assigned member, meals, holiday gifts, relocation reimbursements, phone, laptop, professional dues, education support, and up to 30 vehicles at a use-value level of up to NIS 6,000 a month, including a possible tax gross-up with compensation-committee approval. In 2025 the company paid the kibbutz $6.857 million for this manpower layer, up 20.8%.

The easiest point to miss is that the link between the kibbutz and the cost base does not stop at base pay. Once the agreement was extended, it also included entitlement to a profit-participation bonus derived from the annual dividend the company distributes, capped at NIS 1.5 million per year. In March 2026 the company approved such a payment of NIS 313 thousand for 2025. That means dividend policy is not only a capital-allocation decision toward shareholders. In parts of the service architecture it also bleeds back into labor cost.

The same logic runs through the executive layer. Officer services are supplied through the kibbutz, and the company states that at the report date these roles included the CEO, operations manager, chief scientist and HR manager. Here too the tariff is NIS 856 a workday plus VAT, under the same indexation mechanism, and the arrangement includes extra benefits, a car at a use-value level of up to NIS 10,000 a month, an extra annual payment equal to one monthly compensation unit for each assigned officer, and a dividend-linked profit-participation formula. The 2025 payment for officer services was about $405 thousand, and in March 2026 the company approved a 2025 profit-participation payment of NIS 17,579.

The implication is that the kibbutz sits not only on the production line but also in the decision-making layer. That is not automatically a problem. It is, however, a good reason to read Gan Shmuel’s cost model as one in which labor, management and capital are more tightly intertwined than in a standard industrial company.

Services and Electricity: Operating Convenience at the Price of Dependency

General services sounds like a footnote. At Gan Shmuel it is a meaningful part of the machine. The new agreement for 2025 and 2026 carries a fixed monthly payment that stood at NIS 56,638 at the end of 2025, and it gives the company a very broad operating shell: access to kibbutz facilities, security, laundry, architecture and planning, municipal services, vehicles and tractors, part of the garage, access to the diesel station, maintenance of access roads, parking, rights to use the electricity network and its maintenance, water-network maintenance up to the company gates, postal services, clinic access and emergency services. Beyond that, the company can order additional services, including logistics, procurement, bookkeeping, customs clearing, metalwork, electrical work, garage services, printing and gardening. The agreement has a $600 thousand cap, but that cap excludes the back-to-back transactions.

On the ground, the company paid about $538 thousand in 2025 for general services, versus $220 thousand in 2024. Catering, now covered by a separate agreement with a $700 thousand cap, added another $198 thousand, versus $118 thousand in 2024. In other words, the services-and-food shell rose from $338 thousand to $736 thousand. That is no longer a negligible cost. It is a full operating wrapper.

The heavier item is the back-to-back utility layer. In 2025 the company paid $2.775 million for diesel, water and electricity supplied through the kibbutz. The company says something important here: its ability to obtain electricity and water from third parties is limited. That one sentence explains why this cost deserves to be read differently, not as just another routine related-party line, but as an infrastructure expense where bargaining power and replacement freedom are only partial.

Then the post-balance-sheet event sharpens the picture. In March 2026 the company approved a 10-year electricity agreement between Ganir and Gat Electricity Distribution, effective January 1, 2026, at the tariffs published by the electricity regulator plus VAT, including installation, connection, meter reading, operation and technical handling. This matters because it shows electricity is not leaving this relationship web. If anything, it is moving into a more explicit long-term format. It is still too early to quantify electricity on a stand-alone 2026 basis, but it is already clear that the company is not entering the new year with a disconnection thesis. It is entering it with a more formal version of the same dependency.

Fruit Procurement: Not an Open Pipe, But Still a Structural Link

This is the place where it is important not to overstate the case. Fruit procurement from Kibbutz Gan Shmuel, Kibbutz Gat and Beit Nir is a related-party arrangement, but it is also the cleanest part of the web. The agreements run for 20 years, with a 12-month termination right for the company. The price paid to the kibbutzim cannot exceed the price paid to other suppliers at the nearest relevant supply date, and organic citrus fruit cannot exceed the non-organic price by more than 30%. In addition, because of a historical commitment from the 1993 prospectus, fruit purchased from Gan Shmuel cannot exceed 10% of all fruit received at the company’s plants.

That distinction matters because it prevents an exaggerated reading in which the entire raw-material chain is trapped inside the same control circle. That is not what the filing shows. In 2025 the company paid Gan Shmuel $2.859 million for fruit and Gat another $549 thousand, while Beit Nir recorded no payments. So this is still a meaningful procurement layer, but one that is bounded both in price and in volume. It is the cleaner part of the relationship structure.

Still, even this cleaner part says something important about how to read the company. Raw materials, operating services, rent, management and electricity are not all bought in separate open markets. Part of them sit inside the same web of kibbutz, related companies and controlled entities. So even when the company says the terms are market terms, the investor still has to ask not only what the price is, but how much room the company really has if it wants to switch supplier, cut a service or build a different cost structure.

What This Changes in the 2026 Read

The right 2026 read on Gan Shmuel cannot stop at concentrate prices, the Primor distribution transition, or the first commercial proof from Thailand. All of those matter. But they will sit on a cost base that became much more visible in 2025: rent with a fixed upward ratchet, labor indexed to both CPI and wage levels, bonuses that are also linked to dividends, a broader service shell that got more expensive, and electricity and water where supplier substitution is openly limited.

The important point is not to overshoot in the other direction. There is no proof here that the company is paying above market. The main-site lease has an external appraiser mechanism, fruit procurement has market-price clauses, several agreements have caps, and trade payables to controlling interests were only $1.481 million at the end of 2025, down from $1.935 million a year earlier. So the issue here looks less like hidden financing and more like an embedded cost architecture.

That is the core read. In a peak year it is easy to treat this shell as background noise. In a reset year it comes back to the center. When expenses to controlling interests approach 72% of 2025 operating profit, which stood at $25.03 million, the investor does not need to prove that the kibbutz is the problem. The investor only needs to understand that Gan Shmuel’s operating flexibility is lower than the headline operating-profit line suggests.