BCE (BCE) Q3 2026 Earnings Call Transcript

BCE reported Q3 operating revenue of C$6.182 billion, up 2.2% year over year, and adjusted EBITDA of C$2.786 billion, up 0.9%; adjusted EPS fell 6.3% to C$0.74, while net earnings dropped 84.6%, largely reflecting a C$5.2 billion prior-year gain on the MLSE stake sale. Free cash flow declined 14.9% to C$854 million as capital expenditures rose 49.3% to C$1.33 billion, driven by AI data-center and U.S. fiber investment; BCE reaffirmed its 2026 guidance and C$5 billion CapEx outlook. BCE agreed to sell a 49.9% stake in TowerCo for C$2.45 billion and maintained its target of 3.5x net debt leverage by end-2027; management said Bell AI Fabric’s first data hall is expected online in H1 2027.

Analysis

The key tension is capital intensity versus proof of returns: BCE is redirecting capital from a mature Canadian telecom base into AI capacity and U.S. fiber just as free cash flow is pressured and leverage remains elevated. The tower transaction improves funding flexibility, but BCE retains control and consolidates TowerCo, so it is a balance-sheet and capital-recycling event—not a clean EBITDA deconsolidation or proof that future AI spend is self-funding.

Near term (days to 1–3 months), the risk is that higher Q4 investment and weaker cash conversion overshadow modest operating progress. Wireless pricing is improving at the new-customer level, but that does not yet establish blended ARPU recovery; government additions and regulatory fee changes complicate the read-through. Quebecor’s reported share of industry net additions is a competitive watch item, not enough on this call alone to justify a pair trade.

Over 6–18 months, AI Fabric offers asymmetric upside if contracted capacity is delivered on schedule and customer payments are durable. But management’s EBITDA potential is a company claim; verify contract protections, customer concentration/credit, remaining build cost and power availability before capitalizing it. The nonbinding expansion MOU is not yet incremental contracted earnings. Community opposition, permitting, utility constraints, or customer financing stress could delay returns. Contrarian point: low core telecom capex may flatter near-term cash discipline while leaving less room for network investment if competition or regulation worsens.

Key Decisions for Investors

• No aggressive BCE add at the open: treat the quarter as execution progress, not yet evidence of a durable FCF inflection. Existing holders should size exposure against leverage and the upcoming investment cycle rather than headline AI EBITDA potential.
• Set a catalyst-based BCE entry alert for evidence of the first AI data hall entering service in H1 2027; before adding, confirm commissioning timing, contracted customer payments, remaining capex and power delivery. A delay, contract change, or materially higher build cost falsifies the near-term AI upside case.
• Track Q4 free cash flow, net debt leverage and blended wireless service revenue/ARPU separately from new-activation rates. Failure to sustain pricing momentum or progress toward the 3.5x leverage objective would weaken the equity thesis; improving cash conversion and on-time AI delivery would support reassessment.
• Do not infer a QBR.A short or BCE/QBR.A pair from net-add share alone. Revisit only with comparable evidence on churn, ARPU, promotional intensity and profitability; subscriber share gains bought through discounts may not translate into superior returns.