BCE Inc. (BCE) Earnings Call Transcript & Summary

Welcome BCE and BCE CEO, Mirko Bibic, to our conference. Thank you so much for joining, and it's a lot that you made the trip down. Really appreciate it.

My pleasure. Thank you, Matt.

So I thought I'd begin maybe focusing on like kind of your AI initiatives and announcements and Ziply before getting into kind of like the business update and what's going on in wireless and back-to-school and broadband. because I think there's a lot of -- it's very topical, I think, at the moment. So maybe to begin and set the stage because I think the backdrop to a lot of the AI announcements is really the sovereign AI push, which is occurring and why that's important and what it means for the country and for BCE?

Okay. 2 very good questions. So maybe start at what it means for the country at the macro level, maybe thematically. And kind of like you said that the theme -- it's the macro themes that then kind of inform the investment thesis, certainly for us in terms of where we're making investments and obviously, for investors looking to -- at AI opportunities. So at the macro level, if you think of kind of the geopolitical environment globally, you've got countries certainly like Canada who are -- and we're not the only ones, but you've got kind of the national security considerations. You've got the diversification of the economy, economic growth. Those all come together and AI is a significant component of that, where we are very, very strong in AI in Canada and less strong in production and deployment, but certainly in the science of it. It is a massive opportunity for economic growth everywhere. Canada is certainly alive to that issue, and we are on the cusp of -- AI is basically -- with AI, we're on the cusp of kind of massive transformation that we haven't seen since the advent of the Internet. So if you look at it from that perspective, it's a very big opportunity, particularly so when we go back to the geopolitical issue where sovereignty becomes an issue. And what is sovereignty from an AI perspective? It's the idea that the data resides in the country entirely. So it doesn't traverse any other border. It just stays in the country where the data is stored and where the technology is housed are in data centers that are in Canada, owned and operated by Canadian entities who are subject to Canadian laws from beginning to end. And beyond that -- sovereignty goes beyond that. And the technology that's used to power AI is actually also operated and controlled by Canadians. So when you look at it from that perspective, therefore, that macro theme, it's a very large opportunity in Canada, and we should probably -- I think you're going to ask me kind of where we play there. But I'll stop here with one more sentence and let you ask me your specific question on where we're going to play there. If you think about what Bell's core business has been for 145 years, it's been connecting people to each other, connecting companies to their customers and connecting everyone in Canada to technology. And what we're doing in the AI space is connecting Canadians to AI technology. So I'll stop there, and I'm sure you want to...

No, that's a good overview. And so I think you've made some announcements like specifically more on the inference side of AI and on some announcements on size, general size of data center build that would accompany it over a period of time. And so what is it that Bell is doing on the inference and with the data centers? And who are you partnering with to make that happen.

It's a great question. Let me unpack that quickly, but in its entirety. So I said what I said -- my last sentence in the last question was said for a reason. We're connecting Canadians to AI technology. So what that means? We're going to play in the full stack of AI, and let me very quickly go through them. The first layer of that stack is network. You're connecting data centers with fiber. That is our wheelhouse, right? So no mystery there. The next layer of the stack is data centers, purpose-built AI data centers, not standard colo, purpose-built AI data centers. And there, if you've got the land, the power, the cooling, you've got an advantage. And the critical advantage in this space right now is time to power. And we have all of those. And you see we've launched a site in British Columbia in June, and we've announced more. And that is fairly, frankly, it's just -- it's putting up a data center and you have access to the low-cost hydropower. So it's quite capital light. So that's the second layer. The third layer there is compute infrastructure. As you can see, we have not owned and operated, purchased the compute, we've chosen to partner with those who do compute. So in the case of Groq, inferencing. We've done another -- we just announced another one with BUZZ High Performance Computing. So we're partnering in a capital-light way in order to provide the compute infrastructure to our customers. Above that, you've got the software layer. So again, think of LLMs, not going out and designing and building LLMs. We're partnering, for example, with Cohere. We have a massive partnership with Cohere. And then the last layer on that stack in my mind, is all the AI technology advisory services, and we have a managed service provider called Ateko, where we're able to provide those very sophisticated AI advisory services to our customers. So that's part of our business. The top layer on AI advisory is part of our business. The bottom, the very first layer, fiber, that's our core wheelhouse, the data centers, we can do in a capital-light and everything in between, we're partnering.

So can we maybe focus on one element of that right now, which is the compute. Because if I look at Groq and they have an example of another sovereign AI deal that they've done, Saudi Arabia is an example, which I don't know if that is a good analogy or not, but it's the one that I kind of come across. And in that instance, the -- somebody is a partner that purchases the compute from Groq, and then there is a revenue share that pays the partner back for buying the compute. And then once there's a return threshold that's met, revenue gets split among the partner and Groq, splits differently. So Groq makes their money on the back end. So should we assume that there is -- that Groq is following that same model? There'll be somebody will be announced as a partner that's buying the compute that will then be subject to this kind of revenue share split? Or is the deal a different construct?

That is a model that I also understand Groq has deployed elsewhere. That is not the model that Bell and Groq have deployed in the first facility in British Columbia. In the first facility in British Columbia, it's Groq's own compute that they've taken the entire space. They put up the racks. It's their compute. We -- and they pay us for access to the facility on a long-term basis. Now there are other models like the partnership model, which you've actually described a partnership model where somebody, it could be the person, it could be the entity that actually owns the data center will buy the inferencing chips from Groq and put them in -- install the racks in the facility and then there's a partnership split between Groq and the partner. That is not the model we've employed or deployed in our first data center. Look, we take a big step back. We're playing in the layers of the stack where we are very good at or it's our core business or lanes that we have developed like managed service provider with Ateko. The other layers, we have chosen to do it in a capital-light partnership way. And if you take a step back, what we're going to be building with Bell AI Fabric is to be the backbone of the AI ecosystem in Canada, which has tremendous opportunity. There's an addressable market there that's in the billions, that's growing in the double digits and Bell AI Fabric will ride that wave, and there's significant revenue opportunity in the years to come that's in the hundreds of millions of dollars, and we're doing it in a capital-light way. I think we've stated publicly recently that you're looking at kind of in that $200 million to $300 million range of investment over the next 2.5 to 3 years, and we're going to be very disciplined in operating an overall BCE business at a 14.5% capital intensity, and that factors in all the investments we plan to make in Bell AI Fabric. So there's growth -- high growth potential. And when you're talking about sovereign AI in particular, given the answer to the first question and the kind of geopolitical and economic growth themes there, one significant element that stands us apart from most others or many others is the notion of trust. And the Bell brand is one of the most trusted brands in the country in any sector of the economy. And when you combine that with our core assets that it's hard to duplicate, particularly in the enterprise space where most large, frankly, 95 of the 100 top Canadian companies are with Bell, that's a big advantage that nobody else has.

Great. Okay. And maybe just to finish off, you mentioned Ateko, which is sort of a grouping of businesses, I guess, really within Bell's business wireline services that sort of can help kind of modernize companies, updating legacy systems and so forth. That is where kind of this Bell AI Fabric will fit in with the connection to the business customers buying AI services through Bell and there's a connectivity layer there as well. I know that there's a plan within Bell Business Markets to grow that bucket of revenue. Is that what is identified as like growing to up to $1 billion in the next -- I don't think there's an exact time frame. But is this the linkage? Is this the engine that will get that growth within that Bell Business Markets, this AI initiative with -- through the sales of services Ateko offers?

Maybe the best way to answer it is -- let me answer it this way. So the Bell Business Markets, our enterprise business has been significantly underappreciated, and we're starting to surface it more, highlight it more because the potential in that space is quite large. And it comes through a lot of hard work over the years and a lot of really focused strategic thinking. And there's a tremendous amount of potential there and growth potential. And if you think we have a unique set of assets that can't be duplicated in the enterprise space in telecom. Most other large kind of mega-cap traditionally legacy telcos have struggled, including ourselves in the enterprise space, technology disruption and competitive disruption. So it's been kind of a constant decline. We are very unique in the sense that we -- in the core business, the declines are modest, starting to stabilize and inching up to growth and you add the collection of interrelated businesses to the core wireless and wireline business in enterprise, and there's now growth. And so what is that? So it starts first with our fiber assets, which we have the scale that nobody else has. Our brand, the trusted brand, our deep, deep enterprise relationships that span 145 years. Right there, there's a natural advantage. So you have 5G, fiber, the relationships, the brand. And now you add 3 related high-growth, net new growth. This is not displacing anything. This is net new. And the 3 are managed service provider, the MSS -- the systems integrator managed service provider business of Ateko, which is -- which helps companies on their modern -- technology modernization journey just like we are undergoing ourselves. So we have operator expertise in the space. We bring that expertise to bear for our enterprise clients with a particular focus on the hyperscalers ServiceNow, Salesforce and AI integration. We have cybersecurity, the cybersecurity business. And of course, when you think of security, you do think of Bell, core network security, but also cybersecurity, they go together. And then there's Bell AI Fabric. So all those 3 technology solutions businesses, AI Fabric, Ateko and cybersecurity work together. One pulls through the other, Fiber pulls them through or they pull Fiber through. So it all works as a unified strategic whole. And then to your question on the exact kind of expectations of growth. So on the -- just the Ateko business is what we said would grow from -- around $200 million, $250 million in revenue to $1 billion by 2030. The other 2, cybersecurity and AI Fabric are over and above that.

Okay. That's great. That's a good overview. So maybe shifting gears now. I mean we could stay on AI, I think, for a lot longer. But shifting gears to the other kind of exciting change that is happening at BCE is your Ziply deal is closed. I think you've -- the debt has been kind of redeemed, you're sort of moving full steam ahead here. And so maybe just by the end of the year, you'll have 1.5 million passings. There's going to be another 0.5 million that will be built -- or sorry, yes, another 0.5 million that will be built to be 2 million in the footprint. And then you have the partnership, which will build another 1 million in the footprint and then 5 outside in low-cost areas that you've identified. Can you kind of put some context in the pace of the build that you're planning? I know it's early days for the partnership to be able to -- but you must have some targets that maybe it's time to share or maybe just high level would be helpful.

Yes. I think the Ziply Fiber story is a great story. So we -- a couple of high-level points. We closed earlier than anticipated, which is a credit to the Ziply Fiber and Bell teams working together. That's one. The performance since -- of Ziply Fiber, subscriber growth, revenue growth, EBITDA growth has been better than we expected it would be when we made the initial acquisition announcement in November. It's a particular credit to the Ziply Fiber team. And the third high-level point I'd like to make is this is really, really important. The Ziply fiber management team that has built Ziply fiber from scratch essentially back in 2020 they are with us. They are now part of the BCE family. And I think -- I mean, again, that's great news because they're an outstanding team, but I think it's also a credit to how we're working together as a team. And that's important to keep the operational momentum going. So what we're trying -- what we're going to accomplish is we think that by 2028, we'll have above 3 million homes fiber lines built and largely in the 4 core states in the Pacific Northwest with an ambition to get to 8 million, as you pointed out. So I think those are -- gives you a couple of milestones there on the fiber passings.

Okay. And then there's the -- I mean, there seems to be a foot race to build fiber that's only accelerating in the U.S. How comfortable are you about being able to be first to fiber in the areas where you want to build? Like my understanding is you've identified more than enough homes to be able to make that happen. But just if you can get some comments on how you feel your ambitions match up with what's happening in the market recently?

Yes. So again, let's go back to the previous -- one of the previous points I made. One of -- I believe that -- well, I know because I've been told that one of the reasons that the management team has stayed with us and wants to execute the plan is they're energized about what BCE brings to the table and what we brought to the table is the ability for Ziply Fiber to accelerate what was in their initial build plan to get to that 3 million-ish homes in the 4 core states, so accelerate that and expand beyond that. And we wouldn't -- they wouldn't be excited about that if we all didn't collectively believe that it's doable. And what we did do with the PSP partnership is we unlocked certainly the funding component for that. So what do you need? You need to identify areas that don't have fiber, and there are a lot of those areas in the United States. Fiber availability is only 50% of the country, so there's tons of opportunity. You need to have the team to do it. We have the team to do it. And of course, you need to have the funding capacity to do it. And with the PSP partnership, we've unlocked that. So those are the 3 critical ingredients. And we have -- there's Ziply Fiber backbone that extends beyond the Pacific Northwest, which gives us an opportunity to kind of build along those areas. And again, there's a tremendous amount of potential given how little fiber availability there is in the U.S. compared to Canada, and BCE is an expert fiber operator as well.

So in the U.S., the other big theme is convergence, which you have 2 camps. It's not as though it's a decided argument. The AT&Ts and Verizons seem to be pushing the notion that there's a mutually beneficial relationship between offering fiber and wireless. Then you have on the other side of it, T-Mobile, which says, you know what, people aren't buying based on wanting one bill. They buy the best service. And so fiber obviously lines up as the best service and can win. So just where do you fall on that?

I think both camps are right. And here, I'll tell you why. In Canada, clearly, convergence is the predominant theme. And there's a reason for that. You've got 4 very competitive players in Canada and all 4 operate their own wireline and wireless networks to larger and lesser degrees, but they do. And the amount -- the extent of overlap between their owned wireless networks and their owned wireline networks is pretty extensive. So you have a different market structure, but you do see the results of convergence, lower churn, higher lifetime value, et cetera, et cetera. So from that kind of thematic point of view, that camp is correct, and we see it in Canada and Canada, I certainly endorse that point of view. In the U.S., I think both camps are also right there. You -- if you have extensive fiber networks of your own and you have an extensive wireless network of your own, of course, you're going to try to find ways to put those 2 together. But the other camp is also correct that ultimately, the customer just wants better connectivity. And connectivity is critical to every single household and customer will always choose better. And fiber is better. And why -- if that is true, like it's unassailable because you see companies like Ziply Fiber and there are many others who kind of started from scratch and are taking significant share away from the non-fiber players because customer wants better. Will there be a time where the market will evolve to a degree where that convergence becomes the predominant theme, perhaps. And when that time comes, we will be ready. And I think BCE is also uniquely positioned from that point of view because we are also an expert wireless operator. We know what we're doing. We have extensive global relationships in the wireless space. So we can bring that expertise to bear to Ziply Fiber when the time comes. But you can see it in the results, the absence of a wireless offering in Ziply Fiber has not slowed them down. In fact, I think the absence of a wireless offering has perhaps enhanced the early success of Ziply Fiber because Ziply is able to offer such a compelling, simple package to consumers. There's 3, 4 Internet packages, pick the one you want. It's easy and off you go, there's no confusion. The pricing is simple. And so the customer service and the NPS scores are through the roof and you see the penetration.