Telecom Italia: The Tender Is Over, and the Memory Bill Just Arrived
A quarter that was designed to be boring
Telecom Italia's third quarter was, by its own CEO's admission, an exercise in saying nothing new: group revenue of EUR 10.4 billion (+2.8%; +3.7% stripping out the MVNO drag), EBITDA after lease of EUR 2.8 billion (+2.3%, +6.0% ex-MVNO), capex at 12.4% of sales, equity free cash flow after lease of EUR 0.8 billion, net debt after lease of EUR 7.2 billion and leverage below 1.9x. Guidance confirmed. That is the arithmetic. The interesting part is what surrounds it — a company that spent 2025 and the first half of 2026 fighting a takeover and defending its own share price, and is now trying to convince the market it has become an operator again.
Poste is an accelerator, not a substitute
Poste Italiane formally assumed direction and coordination of the group on September 30. Pietro Labriola spent a striking amount of the prepared remarks pre-empting the obvious worry — that the new controlling shareholder would rewrite the strategy — insisting that “this is not a change in direction for TIM, but rather an acceleration of the industrial path we have already built” — Pietro Labriola · 2026-11-11. The tangible payoff is already visible: both Moody's and S&P have restored investment grade, which Labriola framed as the culmination of a five-year deleveraging and which should show up as a lower cost of funding before it shows up anywhere else.
The keyword record shows the shift cleanly. In the summer quarter the call was consumed by process words — fairness opinion, public tender, offer document, Poste share price, terminal value arguments. Those have dropped out of the top of TIM's keyword list almost entirely. What replaced them is product language: the five value-creation levers of a household "customer platform," a premium connectivity tier, unified customer intelligence, brand architecture. A company that spent two quarters defending a valuation is now arguing about ARPU.
What management stopped talking about
The quietest number in the release is the MVNO one. The contract migration cost EUR 84 million of revenue over nine months, but the trend is the point: domestic revenue growth went from -0.9% in Q1 to +1.2% in Q2 to +3.8% in Q3, and the MVNO transition is explicitly fading. That mechanical headwind, which management had flagged repeatedly through the first half, is now doing less and less work as an explanation — which cuts both ways. It removes an excuse, and it moves the burden of proof onto the underlying business: domestic EBITDA after lease was down 1.2% reported but up 5.6% ex-MVNO, and the fourth quarter now has to deliver roughly EUR 615 million of domestic EBITDA after lease to hit the year.
Labriola has been pre-positioning this for a while. On the July call he warned that “our third quarter is more challenging than the fourth quarter... we can be weaker in the third and stronger in the fourth” — Pietro Labriola, Chief Executive Officer (CEO) · 2026-07-30, and today he leaned on the same three Q4 drivers — easier OpEx comparatives, TIM Enterprise's year-end seasonality, and the cost transformation programme. TIM Enterprise is genuinely the bright spot: EUR 2.5 billion of revenue over nine months, +6.8%, accelerating to +9.3% in Q3, with cloud up 20% and now 46% of service revenue, and backlog guided from EUR 4.0 billion to north of EUR 4.5 billion.
The memory bill
The most candid moment of the call came from a Bank of America question about capex inflation. Labriola waved off general inflation — labour is contracted for three years — and named a different adversary entirely.
I'm less worried about inflation and more worried about the increase of the memory. So when we talk of the CPE, today, the increase of the cost is much more related to the increase of the cost of memory than on the inflation. This is the same also on the cloud side and on the server side.
This is a company-specific worry riding a genuinely global wave. Across the market's curated keyword set, memory cost and memory pricing have been climbing as themes on other calls, and TIM is now a member of that club — with the twist that for TIM the exposure sits in customer premises equipment, routers and servers rather than in a chipmaker's gross margin. Management has responded with procurement pre-buying to build equipment stock, an admission that this is a hedging exercise rather than a solution.
Here is the part that makes the memory story more than a cost line. Because smartphones are getting dramatically more expensive — Labriola cited devices at EUR 2,000–3,000 — the financing solution TIM built with Santander has turned device retail from a loss-leader into a margin contributor. As Andrea Rossini put it, “we went from having negative margin on the sale of devices to positive margin on the sale of device” — Andrea Rossini · 2026-11-11. That is also why total revenue outpaced service revenue so visibly in the quarter. The device business has become, improbably, both a growth and a retention lever for a telco that used to give handsets away.
The fight nobody wants to have
FiberCop and INWIT remain the unresolved plumbing. On FiberCop, Labriola says the two sides have agreed to try to settle operational matters — delivery and assurance quality, which he again blamed for weak FTTH net adds — by year-end, and to discuss a price list that he insists was fixed in the network sale agreement and must be honoured. The frustration is not new: in July he said flatly “I'm unable to explain to you why there's a deterioration. This is a question that must be asked to FiberCop” — Pietro Labriola, Chief Executive Officer (CEO) · 2026-07-30. On INWIT, a formal letter went out on Friday, and the message was transactional rather than bellicose: “I don't want to fight just for the fighting. It's clear that the best end game should be to proceed with INWIT” — Pietro Labriola · 2026-11-11 — with a plan to go it alone if nothing lands by March or early April.
Access technology is the strategic reframe underneath all of it. Having sold the last mile and kept the backbone, TIM now frames itself as able to arbitrage between fibre, wireless access and satellite depending on economics and geography — a genuinely different posture from the vertically integrated TIM of four years ago, and one that quietly explains the FWA push.
Agentic AI, defensively
The newest keyword in TIM's own vocabulary is Agentic AI, but the framing is the mirror image of the market's enthusiasm. Asked whether autonomous shopping agents could commoditise Italian mobile tariffs, Labriola was dismissive, noting that Italian prices are already so low that “we have the lowest price in the world, not in Europe, in the world” — Pietro Labriola, Chief Executive Officer (CEO) · 2026-07-30 — a line he has used for three quarters running, and the reason he argues an agent-driven comparison war is a German problem, not an Italian one. For TIM, personal AI agents are a cost tool for customer care and network operations, not a threat. Whether that confidence survives the next eighteen months of agentic commerce is the open question the call did not answer.
Strip out the noise and the investment case has simplified: a deleveraged, now investment-grade national operator with a supportive industrial shareholder, a growing enterprise cloud franchise, a consumer business whose revenue trend has turned, and one large unresolved supplier dispute. The tender drama is gone. What is left is execution — and a memory bill.