Triple and IPM: How Much of Alma's Value Is Cash, and How Much Is Still Optionality
Alma marks its Triple stake at NIS 226.8 million, but only NIS 31.6 million actually reached the parent in 2025, and most of that came from refinancing rather than a normal dividend. From here, the remaining value depends mainly on IPM's bilateral-sales ramp, its ability to absorb a lower tariff, and whether the data-center path turns paper value into accessible value.
What This Follow-up Is Isolating
The main article established that Alma's real test still sits at the parent-company layer. This follow-up isolates Triple because that is where three things sit at once: Alma's largest marked holding value, the unusually strong 2025 cash receipt, and the easiest place to confuse model value with cash already in hand.
The first line is simple enough: Alma owns 18.14% of Triple, and the stake is marked at NIS 226.8 million as of December 31, 2025. But this is not one homogeneous type of value. On a look-through basis, roughly NIS 185.9 million of that mark comes from Alma's indirect share of IPM, roughly NIS 55.3 million comes from Triple's land, only about NIS 1.8 million comes from other net current assets and restricted balances, and about NIS 16.3 million is offset by Triple-level debt and liabilities. The headline mark is therefore mixing together potential cash, operating value that is still assumption-sensitive, and land optionality.
| Value layer | Alma look-through value | What supports it today | What still separates it from accessible cash |
|---|---|---|---|
| Indirect share in IPM | About NIS 185.9 million | An operating power plant, high profitability, historical DSCR of about 1.9 | The value still depends on sales mix, tariff, financing, and upstream distributions |
| Land at Triple | About NIS 55.3 million | A NIS 305.0 million land appraisal and NIS 9.0 million of rental income in 2025 | The land is financed, covenant-bound, and the data-center path is not yet cash |
| Other net assets | About NIS 1.8 million | Cash, receivables, and restricted balances | Immaterial relative to the two main layers |
| Debt and liabilities at Triple | About NIS 16.3 million on a look-through basis | Land-level borrowing and current liabilities | A standing friction between asset value and accessible shareholder value |
What Is Already Close to Cash
If the goal is to see what has actually made it up to Alma, the right place to start is the cash received in practice. In 2025 Alma received NIS 31.6 million from Triple, but only NIS 5.3 million of that was a dividend. The remaining NIS 26.3 million came from shareholder-loan and capital-note repayments. Put differently, about 83% of the cash Alma saw from Triple in 2025 was not a normal recurring dividend, but the harvest from a financing event.
Management's own 2026 bridge already resets the base lower. The forecast receipt from Triple falls to NIS 16.0 million, and this time it is entirely dividend. That does not automatically signal operational weakness, but it does mean 2025 cannot be annualized as if it reflected a normal distribution pace.
This is not just a parent-level reading. Triple's own cash flow statement shows what sat behind the jump: in 2025 it released NIS 79.7 million from restricted reserves, raised NIS 831.9 million of long-term bank debt, repaid NIS 901.6 million of bank principal, paid NIS 86.7 million of bank interest, repaid NIS 37.7 million to shareholders and related parties, and paid NIS 29.1 million of dividends to shareholders. The NIS 31.6 million that reached Alma was therefore the product of refinancing and reserve release, not simply of the year's operating earnings.
There is one more small but important detail. In the IPM valuation model, cash flows distributable to owners do not include the rent that Triple receives for the use of the land. That means even inside Triple there are two different value layers: owner-level cash generation at IPM on one side, and a separate land-rent layer on the other. Together they create value, but they do not behave like the same kind of cash.
What IPM Still Has to Prove
The reason not to dismiss IPM's value is that the operating business is real and strong. In 2025 Triple posted NIS 936.4 million of revenue, NIS 282.8 million of EBITDA, and NIS 63.9 million of net income. Cash flow from operations reached NIS 278.7 million. Financing is not currently a distress story either: the subsidiary's historical DSCR for 2025 stood at about 1.9, above the minimum 1.05 threshold, and the group also met the minimum loan-life coverage ratio of 1.08.
But this is still value that depends on what has to happen next, not only on what has already happened. According to the valuation report, sales mix through the first half of 2025 was 15% bilateral and 85% system-manager sales. In practice, from July 1, 2025 through December 31, 2025, 36% of IPM's capacity was allocated to bilateral sales. From January 1, 2026 that allocation rises to 41%, and from April 1, 2026 it is supposed to rise to 51%. Even so, the valuation model assumes that by the second half of 2026, roughly 75% of output sold will be bilateral.
That is the core issue. The valuation itself says the profitability improvement does not come from a better tariff mechanism, but from a higher bilateral mix and more volume sold bilaterally. At the same time, the weighted production tariff fell to 28.90 agorot per kWh with the December 2025 tariff decision that took effect at the start of 2026, down from 29.39 agorot previously. So for the valuation uplift to hold, IPM has to move enough volume into bilateral contracts fast enough to offset a lower weighted tariff.
This sensitivity is not theoretical. According to the valuer's sensitivity table, a 5% drop in IPM's value reduces Alma's marked value in Triple from NIS 226.8 million to NIS 217.5 million, and a 10% drop takes it down to NIS 208.2 million. The discount-rate table makes the same point differently: at an 8.9% rate, IPM is valued at NIS 1.22 billion, at 9.4% it falls to NIS 1.166 billion, and at 8.4% it rises to NIS 1.279 billion.
| IPM value sensitivity | Alma's marked value in Triple |
|---|---|
| Down 10% | NIS 208.2 million |
| Down 5% | NIS 217.5 million |
| Base case | NIS 226.8 million |
| Up 5% | NIS 236.1 million |
| Up 10% | NIS 245.3 million |
There is also a simple way to see the friction on the way up the chain. Alma's indirect economic share in IPM is about 15.24%. In the IPM valuation model, distributable owner cash flow for 2026 is NIS 140.2 million. On a look-through basis, that is equivalent to about NIS 21.4 million for Alma before Triple-level debt, land financing, and actual distribution policy. Alma itself forecasts only NIS 16.0 million from Triple in 2026. That gap is precisely the difference between economic value at the project layer and cash that has already made the full trip to the listed parent.
The Data-Center Path Adds Real Optionality, but It Does Not Shorten the Cash Path
Triple's land is not background noise. It was appraised at NIS 305.0 million as of December 31, 2025, and in the same year Triple recorded NIS 9.0 million of land-rental income. At the same time, the land is financed with a bank loan of NIS 89.8 million. The company remains within the land-loan covenants, with annual rent of NIS 9.046 million, a debt-to-net-operating-income ratio of about 9.9, and a land-value-to-loan ratio of 35%. This is a much better asset than idle land, but it is still not the same thing as an open cash drawer.
On the vacant northern portion of the property, about 30 dunams that are no longer used by the power plant, Triple is advancing a data-center route. As of the report, this is framed as a project of up to 20 MW IT. The presentation gives the wider picture: in December 2024 Triple submitted a permit application for two data centers totaling 40 MW, in November 2025 the application was approved subject to conditions, after the reporting date the company received excavation and shoring approval for one of the two lots, and the estimated timeline for the first 20 MW site points to construction starting in 2026 and commercial operation in 2028.
That is already more than a conceptual slide, but it is still not accessible cash. As of the presentation date, not all conditions required for the building permit had yet been met. Triple's own financial statements add another layer of friction: after the balance-sheet date, court motions were filed regarding the cancellation of, or extension of, the stay on board decisions around the data-center project, and the company says it cannot currently assess the chances of that process.
The right way to read the land layer is therefore in two levels. The first level is a financed income-producing asset that already supports value and produces rent. The second level is a real option to improve land use, but one that still depends on conditions being met, timelines being achieved, and governance friction being resolved. Anyone treating the full NIS 55.3 million look-through land component as immediate cash is skipping the entire distance between conditional permits and shareholder distributions.
Bottom Line
Triple is both Alma's biggest value engine and the place where it is easiest to confuse valuation with liquidity. The part that has already proven itself in cash was the 2025 refinancing harvest that sent NIS 31.6 million up to Alma, but most of that was not a normal dividend. From here, the bulk of value runs through IPM, meaning through the assumption that the bilateral-sales ramp can keep advancing even after the tariff reset, and through land that clearly has value and development potential, but still does not sit in the parent's cash account.
Put more bluntly, Alma does not need Triple to be worth less, it needs the market to stop treating all of that value as if it were already cash. The NIS 16 million expected from Triple in 2026 is the layer of value currently closest to Alma. Above that sits real operating value, but it is still sensitive. Above that sits the data-center option on the land, which is interesting and may become important, but still does not belong in the same drawer as near-term accessible cash.
Disclosure: Deep TASE analyses are general informational, research, and commentary content only. They do not constitute investment advice, investment marketing, a recommendation, or an offer to buy, sell, or hold any security, and are not tailored to any reader's personal circumstances.
The author, site owner, or related parties may hold, buy, sell, or otherwise trade securities or financial instruments related to the companies discussed, before or after publication, without prior notice and without any obligation to update the analysis. Publication of an analysis should not be read as a statement that any position does or does not exist.
The analysis may contain errors, omissions, or information that changes after publication. Readers should review official filings and primary sources before making decisions.