BCE (TSX: BCE) has been in the news for its plan to build artificial intelligence (AI) for Canada. In the run-up to its rebranding from telco to techno, first announced in 2024, a lot of work has happened in this space. BCE lowered its capital spending on building Canadian fibre infrastructure and shifted its focus to building Bell AI fabric.
Source: Getty Images
A little about Bell AI fabric
The work is ongoing on multiple fronts to bring Bell Canada’s vision for the AI fabric to reality. This fabric is a full-stack AI platform, and at its core is BCE’s nationwide fibre network. This network will be stacked up with data centre infrastructure, software, cloud capabilities, advanced professional integration services, and an ecosystem of Canadian technology partners.
Think of Bell’s fibre infrastructure getting a massive tech upgrade along the lines of Tony Stark preparing his Iron Man suit. This process is capital-intensive, needs a lot of collaboration, and pulls together multiple strings, from technology know-how to infrastructure building to capital.
The long-term growth opportunity for BCE
Bell has the advantage of a base fibre network and experience of building communications infrastructure on a large scale. It is partnering with universities and tech companies to collaborate on building the AI fabric. The government is also supporting AI infrastructure under the nation-building activity.
Bell has partnered with the Université de Sherbrooke on quantum technologies, post‑quantum cybersecurity, and sustainable data centre infrastructure. It has also ventured into direct-to-device satellite service by integrating AST SpaceMobile’s space-based cellular broadband network with its terrestrial network.
This journey of BCE from telco to techno looks promising and could generate higher returns in the long term. To give you a glimpse of the kind of returns BCE is expecting, its $1.3 billion investment in a Saskatchewan AI data centre is expected to generate $500 million in annual revenue and $400 million in earnings before interest, taxes, depreciation, and amortization (EBITDA) by the end of 2027.
The risks that could slow BCE’s upside
Bell AI fabric opportunity looks promising, opening new revenue streams of cloud networking, high-performance computing, and satellite communications. But this comes at the cost of its already established fibre infrastructure in Canada, which is seeing a churn in subscriptions amidst aggressive price competition. The $500 million revenue from the AI data centre makes up for only 2% of BCE’s annual revenue of $24.5 billion. This conversion of revenue from traditional wireless and wireline services to AI and cloud services could see a pullback before a strong jump.
Moreover, the Bell AI fabric infrastructure will increase capital expenditure needs, which means the first few years will see high depreciation costs. It means the earnings per share (EPS) growth might be negative, just as it is for most tech companies.
The biggest risk is the $42 billion debt on its balance sheet, which is 3.8 times its EBITDA. BCE has been disposing of its conventional, non-core businesses to repay debt and acquire new businesses with higher returns and faster growth. It has agreed to dispose of its land mobile radio network services business for $675 million and use the proceeds to reduce the leverage ratio to 3.5 times by the end of 2027.
The fine balance between risk and growth
BCE is currently walking on a thin rope. On one side, there is AI and cloud computing growth, and on the other side, the risk of high leverage and revenue shift from telco to techno. At this point, valuation is a challenge as the company undergoes a turnaround.
If you are looking to invest in BCE, be prepared for three to five years of stagnant growth or decline, as EPS could take a hit amidst the transition. If your motivation to hold the stock is the 5.6% dividend yield, you could consider buying, but do not rely on this dividend for your retirement. Remember, tech companies are not good dividend payers due to their need for constant infrastructure upgrades and investment in research and development.
However, you could hold off buying BCE until its AI fabric starts contributing meaningfully to the revenue. At least then you would see the telco-to-techno conversion altering the fundamentals and the stock valuation.
