Bezeq follow-up: yes is profitable, but the ARPU test has only started

The previous yes analysis framed the key test as whether the jump in the segment's value would begin to be backed by cash and profit, not only by a model. The first quarter of 2026 gives an encouraging first answer, but not a complete one. yes ended the quarter with NIS 343 million in revenue, NIS 56 million in adjusted EBITDA, NIS 4 million in adjusted net profit and NIS 21 million in free cash flow. For Bezeq, this is no longer only a story about an asset that stopped weighing on the group, but the start of a real operating contribution. Still, the number that calls for caution is ARPU: total yes ARPU rose to NIS 202, while TV ARPU fell to NIS 165. The current read is therefore positive but conditional. yes is proving that it can earn more around the TV subscriber, but it has not yet proved a clean recovery in classic TV pricing. The next two quarters need to show whether Partner, fiber bundles and the IP migration create a repeatable run rate, or whether a strong first quarter mostly arrived before the proof was complete.

Profit has reached the filing, but it is still small

The newest point at yes is not only that revenue grew 7.5% year over year. Revenue reached NIS 343 million, the strongest quarter since the first quarter of 2019, and adjusted EBITDA rose 14% to NIS 56 million. After several quarters of adjusted net losses, yes showed adjusted net profit of NIS 4 million, while the subsidiary-level financial data show NIS 3 million of profit for the period.

That is an important transition point. At the end of 2025, the yes debate still sat on the gap between a NIS 762 million enterprise value and free cash flow that was barely positive. Now there is a quarter in which adjusted profit is above zero and free cash flow is positive. This does not fully validate the value, but it changes the evidence: yes is no longer asking the reader to believe only in a multi-year forecast.

The caveat is the size of the proof. Adjusted EBITDA of NIS 56 million is very close to the stronger quarters of 2025, and does not mark a new step-up versus the fourth quarter. Adjusted net profit is positive, but still small enough that a modest change in expenses, Partner-related revenue or investment timing could move the story back into the gray zone. This quarter marks progress, not the end of the test.

Total ARPU rose, while TV pricing eroded

The most important point in the quarter is the split between two ARPU measures. yes TV ARPU is calculated without revenue from the Partner agreement, advertising and ISP services. Total yes ARPU includes those layers, so it is a better number for understanding the new yes, but a less clean number for testing pricing power in TV itself.

That is where the gap sits: total ARPU rose from NIS 189 in the corresponding quarter to NIS 202 in the first quarter of 2026, while TV ARPU fell from NIS 174 to NIS 165. TV subscribers stood at 565 thousand, unchanged from the fourth quarter and up 4 thousand year over year. IP subscribers reached 499 thousand at quarter-end and about 503 thousand near the reporting date, around 89% of TV subscribers. Fiber subscribers reached 132 thousand at quarter-end and about 137 thousand near the reporting date, about 24% of subscribers.

The implication is not negative, but it is more precise. yes is not returning to a model in which the classic TV subscriber pays more. It is trying to turn the TV subscriber into an anchor around which it sells Bezeq fiber, IP services, Partner-related revenue and advertising. That can be a better model than the old one, but it needs a different proof: not just higher ARPU, but a clearer breakdown of what actually holds the increase.

Positive free cash flow still needs repeatability

For the yes proof test, free cash flow means the company's metric of operating cash flow minus net investments and lease payments. It is a test of operating cash inside yes, not an all-in view of cash uses at the Bezeq group after dividends, debt and buybacks.

Cash flow was positive in the first quarter, but not unequivocally strong. Operating cash flow rose to NIS 88 million, compared with NIS 75 million in the corresponding quarter. Net investments rose to NIS 61 million, compared with NIS 36 million, and lease payments stayed at NIS 6 million. The result is NIS 21 million of free cash flow, positive but below the NIS 33 million recorded in the corresponding quarter.

That detail prevents an overly quick conclusion that yes has already solved the cash question. The quarter does show that the business can report profit and positive free cash flow at the same time, but it also reminds readers that investments and CAPEX payment timing can move cash flow from quarter to quarter. If free cash flow remains positive even in quarters with heavier investments and continued TV ARPU erosion, the argument that yes has moved from model value to business value will be much stronger.

The next test is repeatability, not another ARPU headline

The current judgment is that yes has passed another stage in its 2026 proof test. Positive adjusted profit, adjusted EBITDA and positive free cash flow improve the quality of the story versus the end of 2025. But the main test is not whether total ARPU rises by another shekel or two. It is whether Partner-related revenue, advertising, TV-and-fiber bundles and the IP migration keep producing EBITDA and cash while classic TV pricing erodes.

The counter-thesis deserves respect: perhaps this is exactly the new model, and there is no reason to miss old TV ARPU if yes can grow total revenue per customer. For that argument to win, the coming quarters need to show three things together: total ARPU stable around NIS 200 or higher, continued growth in fiber and IP subscribers, and positive free cash flow that does not depend on a convenient investment quarter. Until then, the first quarter is a strong positive signal, but still only the start of the ARPU test.