OPC Energy: How Much of the U.S. EBITDA Is Actually Accessible After Cash Sweeps and Project Debt
OPC's U.S. step-up is real at the plant level, but much smaller at the parent-cash level. In 2025 the operating U.S. plants produced NIS 1,098 million of EBITDA, but only NIS 334 million of cash after project debt service, before the minority layer inside CPV.
What This Follow-Up Isolates
The main article already established that the U.S. business is now carrying much more of OPC's story. This follow-up isolates the question that the consolidated headline can blur: how much of the U.S. plant-level EBITDA actually survives the trip upward. At the asset level, 2025 was strong. At the parent-cash level, the picture is much tighter.
Across the five core plants plus Three Rivers, the U.S. operating assets generated NIS 1,098 million of EBITDA in 2025, NIS 819 million of FFO, and NIS 334 million of cash after project debt service. That is already a sharp compression: less than one-third of EBITDA remained after debt service at the project level. But even NIS 334 million is not an automatic dividend. Cash sweep mechanisms direct all or part of excess cash toward accelerated principal repayment and restrict distributions. Beyond that, OPC itself indirectly owned only about 70.69% of CPV. On a simple look-through basis, before parent overhead and additional cash uses, OPC's economic share of that number is roughly NIS 236 million.
That does not mean the assets are weak. Quite the opposite. Maryland, Shore, Valley, Fairview and Towantic are producing meaningful EBITDA. The point is different: if the U.S. story is read only through EBITDA, it is being read through the most optimistic layer in the chain. What actually moves up depends first on each plant's financing structure, and only then on plant scale or power prices.
Where the Cash Gets Trapped
The gap is not uniform. It changes sharply from plant to plant, which means the U.S. EBITDA story has to be read asset by asset, not just segment by segment. Towantic is a relatively clean cash converter, but at a 26% ownership stake it is too small to carry the whole story. Maryland is the largest EBITDA engine, but a large part of the cash still gets absorbed before it can move up. Shore and Valley are the clearest examples of why high EBITDA does not guarantee accessible cash. Fairview looks excellent in the annual number, but it contains a financing distortion.
| Plant | EBITDA | FFO | Cash after project debt service | Cash sweep | Net debt | What it means |
|---|---|---|---|---|---|---|
| Fairview | 171 | 140 | 271 | 32% | 527 | Conversion looks exceptional, but the figure is inflated by a refinancing-driven distribution. |
| Towantic | 117 | 99 | 105 | 8% | 152 | The least restrictive structure in the group, which is why cash conversion is relatively high. |
| Maryland | 336 | 293 | 135 | 53% | 560 | The biggest EBITDA engine, but only part of the cash remains after the debt layer. |
| Shore | 220 | 78 | (223) | 96% | 819 | The plant moved close to full control, but financing still swallowed the cash in 2025. |
| Valley | 201 | 167 | 16 | 100% | 358 | Strong operating asset, but almost no accessible cash until the post-balance-sheet refinancing. |
The pattern is straightforward. Towantic operates with an 8% cash sweep and therefore keeps NIS 105 million out of NIS 117 million of EBITDA. Maryland generates NIS 336 million of EBITDA, but ends up with only NIS 135 million after project debt service because a 53% sweep still absorbs a meaningful share of the surplus. Shore and Valley are in a different category entirely: 96% and 100% cash sweeps, respectively, are not technical footnotes. They are the difference between strong EBITDA and cash that is barely accessible.
Three Distortions That Can Mislead the 2025 Read
Fairview looks like the best converter, but this is not a normal operating run-rate
Fairview contributed NIS 171 million of EBITDA in 2025, NIS 140 million of FFO, and NIS 271 million of cash after debt service. On the surface, that makes it the star asset in the group. That is only a partial read. Part of the number came from an increase in loan principal under an amended financing agreement, after which cash was distributed to the partners. CPV's share of that distribution was about NIS 179 million. In other words, a large part of the annual figure came from refinancing rather than from the plant's recurring operating run-rate. The same distortion is visible in the fourth quarter, where Fairview posted NIS 207 million of cash after debt service while the financing event was still embedded in the number.
Shore shows why a higher ownership stake does not automatically solve accessibility
Shore already looks like a core asset. CPV's ownership rose to 89% by year-end 2025, and after the report date it reached 100%. But the 2025 cash translation was very weak: NIS 220 million of EBITDA became NIS 78 million of FFO and then negative NIS 223 million after project debt service. Two opposite reading mistakes are possible here. One is to read the EBITDA as if it were accessible cash. The other is to read the negative NIS 223 million as if it were the steady-state cash yield. The annual number includes a partial debt repayment as part of the refinancing completed in February 2025. That is why the fourth quarter alone was already back to a positive NIS 20 million after debt service. Even so, the forgiving read still does not remove the central issue: a 96% cash sweep and NIS 819 million of net debt leave the asset far from a free-distribution model.
Valley proves why the financing agreement matters more than the EBITDA headline
Valley generated NIS 201 million of EBITDA in 2025 and NIS 167 million of FFO, but only NIS 16 million remained after project debt service. In the fourth quarter, the number was actually negative NIS 1 million. This is not an operating problem. It is a financing-structure problem: a 100% cash sweep, a 9.8% weighted average rate, and a May 2026 final maturity as of year-end. That is why the most important event happened after the report date. In February 2026, Valley completed a refinancing that cut the interest margin to 2.75% and replaced the 100% cash sweep with a leverage-based mechanism of 75%, 50% and 25%. In the same transaction, about $100 million was used to repay shareholder loans and distribute dividends, with CPV's share at about $50 million. This is the clearest piece of evidence that cash accessibility can change materially once the right asset is refinanced.
What Actually Matters Going Forward
The number that should lead the U.S. read from here is not plant-level EBITDA, but cash after project debt service, preferably also after the ownership layer. Maryland is the easiest asset to overread. NIS 336 million of EBITDA looks like a huge engine, but only NIS 135 million remains after debt service. That is still a meaningful number, and probably the closest thing in the portfolio to a large-scale, repeatable distribution engine, but it is far below the operating headline. Towantic sits at the other end: low cash sweep, relatively modest net debt, and clean cash conversion, but CPV owns only 26%, so the absolute contribution is limited.
What the market should test now is whether 2026 starts to loosen this bottleneck. There are three clear checkpoints. First, whether Valley shows repeatable distributions rather than just a one-time release after refinancing. Second, whether Shore moves from very weak cash conversion toward quarters in which FFO actually converges with positive free cash. Third, whether the higher ownership in Maryland translates not only into more EBITDA, but also into more upstreamable cash. If those three things happen together, the read on the U.S. business improves materially. If they do not, a meaningful part of the value will remain trapped at the asset level rather than at OPC.
Bottom Line
This is the core of the follow-up: OPC's U.S. EBITDA is a starting point, not an endpoint. In 2025 the operating plants generated NIS 1,098 million of EBITDA, but only NIS 334 million remained after project debt service, and on a simple look-through basis OPC's economic share of that figure is roughly NIS 236 million, before parent overhead and further capital-allocation uses.
That is why the U.S. step-up should now be read less as a story about capacity and power prices, and more as a story about cash release, plant by plant and refinancing by refinancing. Maryland and Towantic show that the cash can come through. Shore and Valley show how easily it can get trapped. Fairview is the reminder that part of 2025's conversion came through financing rather than through recurring operations. As long as that gap stays wide, anyone reading only the EBITDA line is reading a stronger story than the one that actually moves upward.
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.