TMT M&A predictions 2026

Alessandro Ravagnolo
Welcome everybody. Hello. So welcome to listen to this podcast about the TMT M&A predictions for 2026 hosted by Analysys Mason. Today you're here with Alessandro Ravagnolo, myself, Partner in the London office. I'm joined by Charles Murray, also a Partner in the London office.

So this is not the first year we're doing M&A predictions for the TMT space. Actually, this is something, kind of a tradition in the team, right? We started back in 2018. So, it's many, many years now. We started from a small article in the room, writing ideas. It was a much smaller team. And now it has evolved into something much more structured. And this is the first year we're doing a podcast.

So, you can listen to us, you can watch us, unfortunately. So, it's the first time, it's a bit of an experiment, so please bear with us. But I also gave you a clue. You can also look back at our previous M&A predictions, check them out, see where we were right, where we missed the target, we might have been wrong. That's all. We're never wrong, obviously, never. But there's always enough to debate. There is always a good excuse in any case.

So, this is the first year of the podcast that we have exciting guests that you will probably want to hear from more than you want to hear from us. But just a quick look back at 2025 for Analysys Mason and for the transaction team. 2025 was a very exciting year. It was a year of growth, learning. We had a lot of exciting events, exciting opportunities. We had fresh new go-to-market strategy. We welcomed new exciting colleagues. You're going to hear from them, from some of them, shortly. And we also met great new clients on the way with whom we developed new business models.

So, we plan to really continue these exciting things also happening for 2026. Just to give you some idea of numbers, the numbers have gone up over time exponentially and now we are supporting in 2025 over 210 deals in the TMT space, ranging from more traditional telecom digital infrastructure for us to new exciting areas like satellite, space, software, and tech-enabled services. That was really, really great. And it was with great pride that we saw the team expanding, growing, maturing globally and across different regions.

Without further ado, let's move to the first prediction. What do we have first, Charles?

Charles Murray
Well, talk about our new colleagues: we’re delighted to welcome Simon Fischer, Managing Partner in our new Munich office. Simon, over to you, please.

Simon Fischer
Thanks for having me here on this podcast for the first time. New Partner to the club, new kid on the block, so to say. So, it's really, really nice. Just as a quick insight, I've been doing mostly work in the IT services and software space for the past decade. So, bringing maybe some news and some insights into the Analysys Mason ecosystem that we haven't had the chance to speak about in the past much, right? So, I'm very happy to be here. And yeah, after so many years in ICT and seeing AI emerging in the last years, I mean, it's obviously the most quoted buzzword of all times. My prediction for the following year is that after 2025 has been a year that's been really about integration of AI systems in IT firms and ICT and MSPs in particular, I think in 2026 we'll see the harvesting of those efforts and of those initiatives and we'll see valuations changing for ICT firms that are really good in using AI and those really poor at not delivering on the promises. So that's my prediction that there's a widening gap in that space.

Alessandro Ravagnolo
So we're talking about AI, right? Which is great. We cannot avoid talking about AI now in 2026. To what extent you see really AI being a real implementation in ICT service companies?

Alessandro Ravagnolo
How much action is marketing effectively, right?

Simon Fischer
So you see, you see certainly a lot of work that's been done there. This is more than marketing. You see also them being able to bundle that in terms of service offering and what they kind of deliver to the customers in terms of products and sales. But very few have really been good in delivering this internally, right? So the typical thing that ICT firms are bad about their own IT is true about AI, right? And therefore the kind client-zero mentality that you would expect an IT firm to have on AI, that's certainly not taking place yet.

Alessandro Ravagnolo
Great. any more questions?

Charles Murray
No, I'm just very excited by the potential of AI. There's a lot of hard yards, I think, to go with it in terms of implementation, but certainly more nimble companies are going to benefit from implementing AI. Just working out cleverly where to implement it, where to focus initially, rather than try and do it everywhere. I think it's focus and effort and driving it to actually succeed is really important. So there's a lot about implementation.

Alessandro Ravagnolo
And I guess Simon, the key questions for you as advisors on deals is how to call out who did it? How do you actually, how do you call out who did it well? And how do you spot actually who didn't do it correctly or actually doesn't do it for real? It's just really an information teaser, right? Yeah. for investors.

Simon Fischer
So what's marketing and what's not, right? So, you know, at the end of the day, if I look at firms like that, and, you know, we've been doing a lot of due diligence in this space. I think one of the things is you find evidence of really good AI implementation in the organisation itself, in the sense that, you know, they have a specific roles designed to that. And we're not talking about AI officers, right? Everybody can call themselves an AI officer, but we're talking about data stewards, people that really take care of the ins and outs of those systems. And you would want to find those people close to the CEO office, right? It's a CEO agenda item and you would want to see that being really, really close to the top management priority list. Secondly, what you find that ICT firms need to partner with vendors, otherwise they're not able to deliver properly. So that's something you would want to see them building partnerships with strong vendors in the space. And lastly, you would want to see results in terms of efficiency gains and ticket resolution timing and all that, effectively making them better firms. So I think a couple of items to check on the DD list.

Alessandro Ravagnolo
Sure, I'm sure they're on your DD list, straight up, thank you. If you have any question about AI implementation in the ICT service space, look no further, just contact Simon and he will be there for you.


Alessandro Ravagnolo
Great, and it's now time to move to prediction number two from our colleague Sylvain Loizeau in the Paris office. Hi Sylvain, welcome.

Sylvain Loizeau
Hi everyone. Glad to be here. Yeah, I'm Sylvain, Principal in the Paris office of Analysys Mason, and my prediction this year is that NeoCloud, or what we call GPU as a service, these are cloud players that are really focused on compute power provided by GPUs. These players are growing quickly and are going to lead a new wave of investment in data centre infrastructure in 2026.

Alessandro Ravagnolo
Yeah, a really interesting topic. We've seen a lot of noise about NeoCloud, particularly in the US. Do you think it is very focused in the US or do you think it's a global opportunity?

Sylvain Loizeau
Yes, that's a very good point. So for now, it's definitely driven by the US. It's driven by hyperscalers who are doing a lot of the AI deployments right now and are building these huge campuses we hear about on the news. It's also driven by the NeoClouds, as I mentioned, because for now it's mostly in the US that they've been growing a lot. You have players like CoreWeave who are spending billions building huge facilities, hundreds of megawatts. We're talking the type of capacity that could power mid-sized cities in Western Europe just for powering GPU racks. And CoreWeave is definitely one of the bigger players.

And what's also interesting is that in the US, these NeoCloud players are mostly bankrolled by GPU providers like Nvidia, who are trying to create alternative ways to go to market. And hyperscalers themselves, they are so willing to get as much capacity as possible that they're buying capacity from these players. So that's in the US, at least it was the last two to three years, but there's a lot of growth now in Europe we're seeing, especially from new players like Nebius, Fluidstack, Scaleway in France, for instance. There's a lot of growth. So what used to be a US-only opportunity up until, I'd say, last year is definitely going to grow a lot in Western Europe this year.

Alessandro Ravagnolo
So you mentioned growth many times, right? So this is effectively a new vertical and new business model. You could argue whether actually a lot of these are crypto campuses, crypto mining campuses being repurposed. So a lot of greenfield deployment effectively. Greenfield deployment means you need the power, right? Is power the key for everything here? Do you have access to power, the key to success?

Sylvain Loizeau
Yeah, so that's a very good point. So definitely, you know, the amounts of capacity we're talking about, several hundreds of megawatts, is definitely putting a lot of pressure on the grids. And that's one of the reasons you're seeing a lot of new geographies emerge as deployment regions for these kinds of AI cloud or NeoClouds. The reason being, you know, now time to power in mature markets like FLAP-D is so long, sometimes it can be five, six, seven years depending on the geography, that it's easier to go to what you could call tier-two countries or tier-two cities to get access to power in a quicker way.

You're seeing a lot of growth in the Nordics. Spain, Portugal also getting a lot of traction. France as well, because there's a lot of spare nuclear power, which is not renewable, but it's low carbon intensity. New geographies emerging. And the other shift we're seeing as well is a change a little bit in the infrastructure that's being deployed. As you mentioned, it's much bigger, but it's also a bit different. It's a bit more modular because the technology is still a bit uncertain.

It's definitely liquid cooling. Probably everybody's heard about that by now, that AI requires liquid-cooled facilities. Reinforced floors, that's something people think a bit less about because these AI racks are putting pressure of several tonnes per square metre, which legacy facilities cannot handle. And another thing that's a bit interesting we're starting to see in some recent deployments is a bit less focus on the reliability of the infrastructure. You know, Tier III, everybody's heard about Tier III, it used to be the gold standard for any cloud deployments. We're seeing some players look at, let's say, tier one or tier two type capabilities that come with, you know, quicker time to market and a lot less capex for deployment.

Alessandro Ravagnolo
I think it's fair to say that whenever we have a big boom from a new industry, new applications, everyone wants a piece of the new pie that is being baked in the market. And how many times do we receive an email, a call from somebody who has a plant or a land plot and says, how can I participate?

I have power, and therefore I'm sitting on plenty, on a lot of money basically, right? On a lot of money because I can just sell it to somebody building a data centre and power AI. Is that the case? We're going to talk more about banks later, by the way, about land banks. But what's your view? Are land banks a good investment model?

Sylvain Loizeau
Well, the main thing is we're seeing a lot of players that do not know how to do data centres, as you mentioned, and say, OK, I'm sitting on a pile of gold because I have a lot of power that's available right now. Let's just say from experience, you also need to know what you're doing and how you're going to sell it. Because the key question right now is getting power, sure, but there are many players who can get access to power. It's how you're going to sell it and who you're going to sell it to, and with what type of reliability you're going to get from your tenant and how you're going to finance it. You know, having access to power and land is just a tiny piece of the puzzle. It's a requirement, but it's not sufficient.

Alessandro Ravagnolo
Great, great. Thank you, Sylvain, right? Thank you.

Alessandro Ravagnolo
Charles, who's next in our guest list? I'm delighted to welcome Richard Morgan in the London office. Over to you, and what's your prediction?

Richard Morgan
Thanks Charles. So my prediction for this year is that we will see a reality check in some of the valuations around speculative data centre land banks. This is despite the fact that there is a huge boom in demand for data centres. On the back of this, we've seen a lot of players piling into this land grab across Europe, looking to acquire land, applying for power and, in some cases, incurring quite significant amounts of capex in developing that land bank.

And this is both data centre operators, real estate companies, energy groups. And while we do see a lot of demand coming, in some areas this pipeline of land bank could far exceed what there's actually going to be demand for. So just to put that in context, Europe today has around 10 gigawatts of operational data centre power. We're expecting that to somewhere between double and triple over the next five years until 2030.

So we're looking at going from 10 gigawatts up to around 30 gigawatts. But if you look at Italy alone, there are 42 gigawatts of grid connection requests just relating to data centres. So it illustrates the level of pipeline scale relative to what we're actually expecting in terms of deliverable demand in Europe.

Alessandro Ravagnolo
That's a really significant amount of sort of additional demand beyond what the supply chain can actually deliver. Do you think that makes it difficult for investors who have not gained exposure to get exposure to data centres?

Richard Morgan
It certainly plays into the consideration. I think we see a lot of early-stage data centre opportunities where they have secured some land, they have made an application for power. And I think it's quite difficult now for these investors to unpick this secured land bank from what is a viable opportunity that can be transferred into data centre revenues.

So I think not all land banks are equal. And it's certainly going to be a big difference between an established data centre platform, which is developing its land bank in parallel to be able to market new sites to its customers, versus a completely greenfield opportunity where they don't necessarily have that track record in commercialising data centres. So it really needs to be assessed on a case-by-case basis, but it is certainly a challenge for investors today.

Alessandro Ravagnolo
It's fascinating the fact that we just finished talking together with Sylvain about GPU as a service and the prediction is all about growth, growth, growth, right? In 2026 and beyond. So a lot of power is needed, right? And therefore we need those land banks to deploy these data centres. At the same time, what your prediction is, is just we need to be cautious, right? Because there is a lot of potential supply.

What are your recommendations for investors? So if somebody comes today, picks up the phone and says, I have powercubes, I have this land plot which is powered, how do you suggest them to look at it?

Richard Morgan
Yeah, absolutely. So you need to look at the specifics of the management team and their ability to take that from a piece of powered land to a successful commercialised data centre. So typically you need to get all the building blocks in place at the same time, which are land, power, financing, and crucially those customer contracts.

So if it's a player that doesn't have a track record of selling to data centre customers, there may be partnership models, joint venture models that can help with that commercialisation. And it also depends where this land bank is. It's still quite difficult to secure big plots of powered land in certain supply-constrained locations like Frankfurt, Amsterdam, London, Dublin. So there can be quite clear opportunities if you can secure the right land and get an attractive place in the power queue in some of those locations.

But if you have a plot of land in a remote part of the Nordics, for example, that can be quite easy to come by. So we see very large pipelines in certain geographies. And I think there it's really obsessing their ability to commercialise that. So then it's coming down to what will these potential data centre customers look for in selecting which land bank holder they decide to go with.

So some of the key factors here are the ability to deliver capacity on time. So I think this is why these land banks are necessary. You need to be able to show you've got a site you can develop. You need to show confidence on the timing that power can be delivered to that site. But they also want a partner who can manage the increasingly complex environments needed for developing data centres, particularly as we move in towards AI and higher-density liquid-cooled environments.

So then looking at a partner which can show operational credibility, track record in delivering data centres, those are going to be key factors in some of these customers deciding which data centre platforms they go with. So I think from an investor's point of view, a land bank is an important part of the picture, but it's always important to understand how able that particular partner will be to commercialise that site and secure customer contracts.

Alessandro Ravagnolo
Thank you, Richard. I think 2026 is going to be another year of stellar data centre deals. I think there's going to be different tax, different directions taken, but yeah, we're not going to see a shortfall in DC deals this year. Okay, maybe. So thank you, Richard.

Alessandro Ravagnolo and Charles Murray (shared audio)

So maybe changing tack slightly and into fibre and…

Over to you, maybe Ale to talk about your prediction. So my prediction is very simple: 2026 will be a key year for FTTH, and the prediction is refi or retreat. Actually, it was not the real name at the beginning, but our marketing team decided that refi or die was not appropriate for a podcast.

But here I am, please Abi do not edit it out. So refi or die, 2026 a crucial year for FTTH. So why do we expect that? Why do we expect 2026 to be a pivotal year? Yeah, particularly why are so many FTTH players in trouble? In trouble, yes. Well, so if you think about it, the FTTH boom started a few years ago now, right? Five, six years ago, we saw a lot of capex being put into the market, into greenfield deployment, in the access network. Before that, it was more on the core, on the backhaul.

And we get to the point where a lot of the regional financing deals come up to maturity. A lot of them are actually more project finance structures and need to move to a different type of deal. And these basically create trigger events, as that means that the borrower, so these FTTH infra cos, need to access the market to gain access to fresh funding.

If you put these together with the fact that there is a big disparity in performance between FTTH players, either because of under-delivery on the commercial side, so basically take-up, or on the capex side, or difficulties in gaining the right scale to have high margins.

If you put all of this together, it means that 2026 will be a year where access capital will require deeper scrutiny of the performance of the company and the future outlook of the company. So it's going to be a year where we're going to have a natural pause, a natural pause where we can look back at the performance and then look forward and reset our expectations on how FTTH businesses will perform going forward.

And we need to make sure that this performance, this new business plan, is married with the right pocket of capital in order to continue funding this vital part of the digital infrastructure that we need every day.

Yeah, absolutely. Core piece of infrastructure that everybody's needing. What do you think is going to happen if the refinancing processes do not go smoothly? A provocative question there for you. Tricky questions, right? Well, it depends on which point of view you're looking at it, right? If you are the equity investor behind the platform, the main implication is you're no longer in the driving seat.

Somebody else will decide the future of your investment. It may mean that lenders will intervene a bit more, be more active. There might be, we expect, some of the equity investors to either be wiped out or to reduce in terms of value, therefore not get the returns they were initially expecting.

Assets could be sold out, therefore driving consolidation, especially in fragmented markets like the UK or Germany. One interesting thing that we can expect is also new players to come in. We're talking about tactical opportunity credit funds who can sneak in, buy debt on the cheap and use that debt as leverage to become more active in the management of the assets and take over control.

Clearly, the only way to avoid all of this is to be in the driving seat, as I mentioned before, which means that you need to drive and deliver a smooth and efficient refinancing process, be in control over your destiny. Yeah, I do think that 2026 will definitely be the year where winners and losers will become much clearer in the fibre market.

And yeah, there'll be some successful refis and some painful ones, I think, this year. Yes, agreed. It's going to be a good mix. Similar to what some colleagues mentioned before in other sectors, right? There's going to be a big differentiation between the winners and the losers.

And it's not going to be about having the right or wrong business models. Sometimes it's just a matter of who got there first with scale, or you're playing in the right market, or you're not playing in the right markets. There are many factors at play. So there's not a single recipe for success here. That's the problem. Yeah, absolutely.

So I appreciate we're sort of getting towards the end. So one last prediction to go for myself this is, and that's all about four to three consolidation in the mobile industry. So very much a hot topic, should we say?

And one which has been long discussed and long in the making. Now, obviously, we've seen four to three before in a number of markets in Italy and in Spain. Whilst those were very much positioned as transformative for the market at the time, they did have the unfortunate side effect through the remedies of letting in new entrants into those markets. So we have Iliad arriving in Italy.

Which you may know a few things about. Which means that this was not really a four to three consolidation effectively. And now we're seeing a not dissimilar journey for the Spanish market with the arrival of Digi, which is doing a remarkably good job on customer acquisition and infrastructure deployment. So yeah, that's certainly changing the market there quite considerably. So we have seen it before, but not necessarily with the success that it may have had.

However, that has potentially changed with the fact that in the UK, the Vodafone/Three merger finally got clearance in 2025. And because of that, with no significant other remedies, meaning there will be no Iliad or Digi entering the UK market. And from that, you can see other large European operators looking at that as a precedent and thinking, well, isn't it about time that we had a consolidation in our markets.

I think there is a bit of an elephant in the room that we need to address. The year 2026 started with a big announcement, the RAN share deal between TIM and Vodafone in Italy. So you talk about consolidation, the sense of a merger, right? But what about RAN sharing? Is it still a hot topic? I think RAN sharing can play a part.

In some ways, they are complicated deals that take quite a long time to execute, but they certainly give you the infrastructure synergies that are long required and potentially spectrum synergies as well. Don't underestimate the cost of spectrum in many markets. I think RAN sharing certainly plays a part to play and maybe that's the way that Germany will consolidate, potentially a 1&1 and Telefonica or even a Vodafone RAN share could be possible there.

Rather than out and out merger. But I think sort of coming back to the sort of 2026 prediction, I think there's two markets where we're likely to see action. We've obviously seen the original offer for SFR being rejected, but we're pretty confident that a revised offer is coming back on the table. And I think just the pressure that Iliad is under will make that offer at some point be accepted. So I think...

You know, as a prediction that France may be the next market where we will see a four to three merger. And I think you mentioned Italy, I think that there's a good roadmap in Italy too. I think the WindTre and Three option there in Italy will be the most obvious one. I think merger with Three with other operators like Vodafone may be just a bit harder to do, particularly as Fastweb and Vodafone are still being integrated. So from that perspective, we are also predicting that in 2026 we could see an announcement in Italy as well.

Spain is the other market where there's clear opportunity for four to three now that Digi is becoming so active and so aggressive. And Zegona is the obvious seller there, we think. However, exactly how that shapes up given Zegona's high share price means 2026 is probably unlikely, maybe in a few years yet, but that will certainly be discussed all throughout the year. So yeah, a very interesting time, think in the sort of core telco world from a consolidation perspective. And maybe this is the sort of inflection point that the market really needs to start driving value creation for shareholders.

A piece of information that I want to add that you correctly mentioned that RAN share deals are extremely complicated and both me and you probably worked on quite a few RAN share deals that were never surfaced to the press and never do. But the TIM-Vodafone RAN share actually was first announced in 2019. And it's interesting that it's back on the table, never implemented of course. And now it's January 2026 and it's back on the table. So this is how hard and unpredictable this market is sometimes.

One question to close the loop is what does it mean, of course, for lenders and investors in digital infrastructures or better into tower cos? What does it mean for them?

Yeah, I mean when there's a RAN share in the market, I think there's an opportunity to sort of demonstrate as a tower co in particular your understanding of client needs and your ability to flex where required while still maintaining the revenue line that you need to support the investment thesis. So it's all about the tower cos that listen, the tower cos that think, and the tower cos that provide a win-win solution for all players are the ones that are likely to win out when it comes to consolidation.

Because ultimately consolidation is all about trying to drive a certain amount of cost out of the business. So I think there's a lot of learnings from past mergers about how to go about it, and particularly how not to go about it. So I think from that perspective, I think just on the lenders as well, lending into the core operators, it's certainly an interesting time. I think there's a lot of opportunity for these consolidations to produce a potentially new capital stacks, which will be interesting and a lot of money to be spent in terms of consolidation. So in the UK, there's £11 billion to be spent. From that perspective, I think it's a good opportunity for lenders.

So this was prediction number six, the last one for these years. Are we going to be right? Are we going to be wrong? Well, let's discuss it during the years.

I think 2026 is going to be a fantastic year. We're certainly looking forward to it. We saw an uptake in activity in the second half of 2025. And we're only seeing interest and activity straight from the get-go in 2026.

A reasonably good looking pipeline going forward. So I think it's going to be certainly an interesting year. Thank you all for listening and for joining us today. You can continue subscribing to our podcast. Also follow us. We continue producing a lot of free insight and content for all investors, right? Corporate investors, equity investors, lenders. There is content and insights for you all. If you have any questions, you want to launch any debate on any of our predictions, either because you like them or you don't like them, even more interesting at that point, right? Please do reach out and let's discuss it, right? Let's get to know each other. Goodbye.

Note: This transcript was generated using AI and has been lightly edited. Minor errors may remain.