yes 2025: How Much of the Turnaround Is Real, and How Much Still Sits in the Model
yes's enterprise value jumped from NIS 86 million to NIS 762 million because 2025 operating improvement was real, but most of the value still comes from forecast EBITDA, a lower discount rate, and a large terminal value. That makes 2026 the proof year between a genuine business recovery and value that still runs ahead of cash.
What Exactly Is Being Tested Here
The main article argued that yes is no longer just a weak-value pocket inside the group, but that the accounting headline had moved faster than the cash reality. This follow-up isolates only that question: how much of yes's enterprise value jump from NIS 86 million to NIS 762 million is already visible in 2025 operating evidence, and how much still rests on forecasts, terminal value, and a lower discount rate.
The short answer is fairly sharp. The turnaround has become real, but it is not fully proven yet. In 2025, yes already delivered a clear operating improvement: revenue rose to NIS 1.30 billion, adjusted EBITDA rose to NIS 218 million from NIS 173 million in 2024, adjusted net loss narrowed to NIS 42 million from NIS 72 million, and free cash flow almost reached breakeven at negative NIS 1 million. Those are not the numbers of a business that is still stuck.
But most of the valuation jump still does not come from cash that has already landed. According to the valuation report, NIS 622 million of the increase came from a change in forecast EBITDA, another NIS 142 million came from a reduction in the discount rate from 11.0% to 9.25%, and only NIS 187 million of the total enterprise value comes from the present value of 2026 to 2030 cash flows. NIS 575 million, roughly three quarters of the value, sits in the discounted terminal value. That is no longer only a question of whether yes improved. It is a question of how much of that improvement is already proven in cash, and how much still asks the reader to trust the model.
What Has Already Moved From Model to Reality
The encouraging part is that the improvement does not need to be invented. It is already there. yes ended 2025 with 565 thousand TV subscribers, versus 562 thousand a year earlier. IP subscribers reached 492 thousand at year end and roughly 496 thousand near the report date, or about 88% of total TV subscribers. Fiber subscribers reached 118 thousand at year end and roughly 128 thousand near the report date. At the same time, yes ARPU rose to NIS 192, and in the fourth quarter it had already reached NIS 200.
That matters because this is not just a raw-volume story. It is a mix story. More IP, more TV plus fiber bundles, lower dependence on the legacy satellite model, and more revenue layers that are not measured only through traditional TV ARPU. The group presentation explicitly says the fourth-quarter ARPU increase was driven by more TV plus fiber subscribers and by revenue from the Partner transaction. So yes, something real happened operationally in 2025. This is not only an Excel re-rating.
The point that has to be held at the same time is the gap between accounting and operating economics. In the annual report's segment table, yes shows 2025 EBITDA of NIS 467 million and net profit of NIS 485 million. In the presentation, which explicitly defines yes figures as pro forma, adjusted EBITDA is NIS 218 million and adjusted net loss is NIS 42 million. That is not a technical contradiction. It is the center of the thesis. The operating improvement is real, but the accounting profit moved ahead of the economic profit.
| 2025 metric | Reported / segment basis | Adjusted / pro forma basis | Why the gap matters |
|---|---|---|---|
| Revenue | 1,300 | 1,300 | There is no material debate at the revenue line |
| EBITDA | 467 | 218 | The accounting layer still makes the profitability headline much bigger |
| Net profit | 485 | (42) | The accounting turnaround came well before an adjusted earnings turnaround |
| Free cash flow | (1) | (1) | Cash is the line that leaves the least room to hide |
There is also an important counterpoint. A cynical reading of the whole valuation would be a mistake. In EY's appendix comparing actual 2025 performance with the prior valuation model, on the valuation work's own basis, revenue came in NIS 48 million above the old forecast, EBITDA came in NIS 47 million above it, and free cash flow was NIS 69 million better than the prior forecast, even though it was still negative NIS 5 million. In other words, the model did not re-rate out of thin air. First there was a real operating beat. Then came the sharp valuation update.
Where the Valuation Jump Actually Came From
This is where business improvement needs to be separated from what the model is willing to pay for it. The move from NIS 86 million to NIS 762 million did not come directly from cash built in 2025. According to EY's bridge appendix, NIS 622 million of the increase came from a change in forecast EBITDA, minus NIS 88 million of other effects, plus another NIS 142 million from the lower discount rate.
That bridge matters. EY attributes most of the EBITDA-forecast change to the contribution from the Partner agreement and the advertising agreement, to updated wholesale internet-cost assumptions, and to higher subscriber levels after the 2025 actual outcome. But the same appendix also states explicitly that there is a reduction in EBITDA from lower TV ARPU. That is an important nuance. The model is not built on a revival of old TV pricing power. It is built on broader yes economics, where Partner, advertising, ISP, and bundles compensate for erosion in the legacy TV product.
That is exactly where a lazy definition of "improvement" becomes dangerous. Yes, there is improvement here. But at the valuation level, the jump depends less on dramatic subscriber growth and more on the assumption that the new shape of yes can hold a rising overall yes ARPU even while TV ARPU declines, and can do so over time.
That is also why the discount-rate effect is not a footnote. The reduction from 11.0% to 9.25% added NIS 142 million of value. In the sensitivity table, assuming the same 1.0% terminal growth rate, enterprise value falls to NIS 674 million if the discount rate were 10.25% instead of 9.25%. That is still far above NIS 86 million, but it is a good reminder that the new value is still sensitive to the cost of capital, not only to operating execution.
That sensitivity already spills into the group reading. The presentation explicitly says the drop in net debt to adjusted EBITDA to 1.4 also reflected higher EBITDA because of yes's valuation, and that excluding that valuation effect the ratio was similar to 2024. That matters: even outside yes itself, the model is already shaping how the group looks before the cash proof is complete.
Where the Model Still Asks for Belief
The best way to see how much of the value still sits in the model is to open the 2026 to 2030 forecast itself. On one hand, it is no longer the forecast of a broken business. Revenue rises from NIS 1.30 billion in 2025 to NIS 1.409 billion in 2026 and to NIS 1.518 billion in 2030. EBITDA, on the valuation basis, rises from NIS 188 million in 2025 to NIS 261 million in 2026 and to NIS 280 million in 2030.
On the other hand, the way the model gets there tells a much more delicate story. TV subscribers are not expected to explode. After 565 thousand subscribers in 2025, the forecast is 564 thousand in 2026 and 2027, then 559 thousand from 2028 through 2030. TV ARPU is not expected to rise either. It falls from NIS 163 in 2026 to NIS 151 in 2030. What does rise is overall yes ARPU, from NIS 204 in 2026 to NIS 225 in 2030.
That is one of the key insights of this continuation. A NIS 762 million valuation is not built on a TV-volume breakout. It is built on the assumption that IP migration, fiber bundles, the Partner agreement, advertising, and other revenue layers will create a better yes economics even without meaningful subscriber growth. That is possible. It is also still a model.
And the model does not tell an immediate cash story either. In 2026, net cash flow in the valuation work remains negative, at negative NIS 14 million. It turns positive only from 2027 onward. That means even under the updated forecast, 2026 is still a bridge year. This also aligns with yes's own risk wording, which says the company is expected to move to operating and cash profitability beginning in 2026, yet may still accumulate operating and cash losses in some periods in later years. That is not the language of a fully locked turnaround. It is the language of proof still required.
What 2026 Has to Prove
If the story needs to be reduced to one line, it is this: yes made an important move in 2025 from a business fighting deterioration to one that is starting to rebuild better economics. But the valuation jump moved ahead of the cash jump.
That is also the test for 2026. For more of the NIS 762 million to stop looking like model value and start looking like business value, three things need to happen:
- Revenue from Partner, advertising, and bundles needs to keep showing up inside clean adjusted EBITDA, not only inside one strong quarter.
- The move to IP needs to remove costs faster than TV ARPU erodes.
- Free cash flow after leases needs to turn positive and stay there, not just brush against zero.
Until that happens, the right reading is neither to dismiss the turnaround nor to fully embrace the 2025 valuation without reservation. The right reading is that yes already looks much better. But a large part of the value assigned to it at December 2025 still waits to become cash that is hard to argue with.
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.