Open in interactive viewer → charts, metric popovers & call review

freenet's Term Sheet Ends the Waiting Game—Now the Bill Comes Due

A year of MNO brinksmanship becomes a Telefonica term sheet, a cheap waipu buy-in, and a guidance mix that still leans on the fine print.
FNTN.DE · Earnings Call · 2026-11-12

freenet's Term Sheet Ends the Waiting Game—Now the Bill Comes Due

For four straight quarters, freenet's earnings calls had a hole in the middle of them: one mobile network operator agreement had gone bad, and management could not say how bad it would get. In February, the CFO called the worst case EUR 50 million. By August, the CEO was still saying talks were constructive and a positive outcome was possible. On the Q3 2026 call, that hole is finally filled—with a term sheet with Telefonica Deutschland. The MNO agreement overhang is no longer an open-ended threat. It is a negotiated document waiting to become a contract.

But the call is not a clean all-clear. The same agreement is still why the reported numbers look soft. CFO Ingo Arnold pinned the gross-profit and EBITDA decline on a “negative effect of EUR 32 million up to now.” — Ingo Arnold · 2026-11-12 Guidance was confirmed, but with a split that tells you where the risk sits: “in EBITDA, we will be at the lower end of the guidance. And in the free cash flow, we will see the upper end of the guidance.” — Ingo Arnold · 2026-11-12 In other words, the existential issue is resolved, but the P&L has not yet been made whole.

The overhang breaks—but the invoice arrives first

The central change is not subtle. In February, the company was openly discussing legal action and a EUR 50 million worst case. By August, management was still in “we are still optimistic that we believe we have a fair chance that there will be -- in the end, there will be a positive outcome for freenet.” — Robin John Harries, CEO · 2026-08-13 Now the CEO says the commercial issues are settled: “No, I would say there are no open points. So the term sheet is very detailed. We have been negotiating hard for many months, as you know, almost 1 year negotiations.” — Robin John Harries, CEO · 2026-11-12 That matters because freenet's model depends on wholesale access to all three German networks. A bad MNO deal is not a margin problem; it is a distribution problem. The term sheet resolves that. But management explicitly says the positive effect is not in guidance until the contract is signed. So investors get the de-risking without the earnings uplift—yet.

Cash generation remains the ballast. Cash generation is still the company's key strength, with cash conversion above 60%. The CFO expects working capital of EUR 40 million to EUR 50 million, taxes below EUR 40 million, capex around EUR 45 million, and lease payments of EUR 60 million. That is why the equity story can tolerate a low-end EBITDA print: the free cash flow is still at the upper end. The term sheet may be the strategic unlock, but cash is what funds the buyback and the waipu buy-in.

waipu.tv: a bargain, or a warning?

The second strategic move is the buy-in of the remaining 25% of waipu.tv for EUR 76 million initial, implying a EUR 300 million enterprise value, plus up to EUR 50 million performance-based. It is funded from the CECONOMY stake sale. Analyst Polo Tang immediately asked the right question: less than 3x 2028 EV/EBITDA if midterm guidance is met—why so cheap? Is there risk to guidance? The CEO's answer was confident but not numerical.

So I think one of our strengths here at freenet is that we are a strong dealmakers. So for us, obvious, it's a good deal. And we were strong shareholders before that year with 75%.

Robin John Harries, CEO · 2026-11-12

That can be read two ways. The benign read: freenet already owned 75%, the minority holder needed liquidity, and freenet bought full control at a price that is accretive if the 2028 plan holds. The skeptical read: a roughly 3x multiple on 2028 numbers implies the seller—or the market—does not fully underwrite the IPTV plan. The truth is probably somewhere in between. waipu added 39,000 net adds in Q3 and is #2 in the German IPTV market, but advertising has softened since mid-year. The CEO conceded: “we saw it was a little bit softer... weaker macroeconomics has affected revenues. But ad impressions remain broadly stable year-over-year.” — Robin John Harries, CEO · 2026-11-12 Stable ad impressions are good; ad pricing is the swing factor. The buy-in simplifies ownership and gives freenet full growth participation. It also makes waipu easier to value—and easier to judge.

The mobile machine keeps grinding—and the AI answer is a rerun

Mobile is the quiet engine. Organic postpaid net adds were 41,000 in Q3; postpaid service revenue rose 1.1% to EUR 480 million; ARPU was stable sequentially; front-book prices edged up again. The value over volume strategy is doing what it says. Competition is described as disciplined. The old worry—that German mobile would turn into a price war—is not the live issue.

The live issue is agentic AI. Analyst Sofija Rakicevic asked what gives freenet confidence it can defend its position if AI makes switching easier. CEO Robin Harries answered that freenet is direct, offers all networks, and has multiple brands: “every smart AI agent will consider freenet and our brands.” — Robin John Harries, CEO · 2026-11-12 It is a reasonable argument. It is also almost exactly the argument he gave in February, when he said “for us, AI is an opportunity, it's not a threat... Freenet is no comparison website. Freenet is no intermediary. We are direct.” — Robin John Harries, CEO · 2026-02-26 The question has changed from “will MNOs disintermediate you?” to “will AI agents disintermediate you?” The answer has not changed. That is either consistency or a blind spot. The company has AI-driven upselling and an AI-first approach, but no quantified defense. For now, the market is being asked to accept that the multi-brand, all-network shop is agent-proof.

What freenet stops saying

The most telling thing about the quarter is what has disappeared from the language. The worst case is gone. Legal action is gone. The old MNO shortfall debate is gone. In its place: term sheet, final contract, and a share buyback of up to EUR 75 million. The balance sheet and CECONOMY proceeds are now doing double duty—funding the waipu buy-in and returning capital. That is a more confident company than the one that spent 2025 apologizing for a bad contract.

But one old pressure remains: decrease in ARPU. Management says ARPU was stable sequentially and customer-base growth more than offsets it. True—but it means mobile service revenue growth is still volume-led. The back-book drag from high-ARPU legacy customers was supposed to fade into the second half of 2027. Until then, freenet is running to stand still on price.

So what changed? The existential MNO risk moved from “worst case” to “term sheet.” waipu ownership moved from 75% to 100% at a price that implies the minority seller was not demanding a growth multiple. Capital returns moved from possible to announced. What did not change: the mobile business is still a disciplined, slow-growth cash machine; the AI threat is still answered with the same direct-to-consumer, multi-brand shield; and the EBITDA guide is still at the low end because the Telefonica fix costs money before it makes money.

That is the trade. freenet has bought itself certainty and a cleaner story. It has not yet bought itself an upgrade cycle. The next catalyst is not another quarter of net adds—it is the signature on the Telefonica contract, and the first set of numbers that show the positive effect rather than the EUR 32 million negative one.