Telus to build AI data centers in Vancouver and Kamloops, Globe says

Telus plans three AI-focused data center facilities in Vancouver and Kamloops that will require more than 150 megawatts of electricity by 2032. The Kamloops site and a repurposed Vancouver Mount Pleasant facility are expected to come online later this year, while a new downtown Vancouver development near BC Place is slated for 2029. The expansion signals meaningful infrastructure investment to support AI workloads, though the article provides no financial terms or direct earnings impact.

Analysis

This is less a near-term revenue story than a multi-year re-rating catalyst: the market tends to underwrite telecoms on stale asset intensity assumptions, but AI hosting shifts the mix toward power-constrained, higher-utility-use infrastructure with longer contracted duration. If Telus can secure capacity and long-dated tenants, the incremental value is not just data-center EBITDA but a higher-quality cash flow profile that can compress its cost of capital and support multiple expansion relative to domestic telco peers.

The key second-order effect is electricity access, not compute demand. A 150MW buildout meaningfully enlarges Telus’s load footprint, which can create optionality if British Columbia prioritizes grid-connected digital infrastructure, but also introduces execution risk if interconnection, permitting, or community opposition slows delivery. Competitively, this can pressure smaller regional colo operators and enterprise IT landlords that lack the balance sheet to pre-fund power-constrained capacity; the scarce asset is now behind-the-meter or utility-approved megawatts, not rack space.

Near term, the stock can rerate on headline enthusiasm, but the fundamental catalyst is staged over 12-36 months as lease-up and power milestones de-risk. The contrarian issue is that AI/data-center optimism often gets capitalized before economics are proven; if Telus’s cost of capital rises or pre-leasing disappoints, the market could flip from “infrastructure growth” back to “capital allocation concern.” That makes this more attractive as a measured exposure on pullbacks than as a chase after a one-day pop.

The broader winners are power equipment, grid, and cooling supply-chain names that benefit from every incremental MW installed, while the losers are less-differentiated colo providers with weaker access to utilities. The cleanest trading expression is to own the optionality on execution while hedging telco beta, because the upside is driven by multiple expansion on a strategic asset story, but the downside is classic capex skepticism if returns lag.

Key Decisions for Investors

• Long TU on pullbacks over the next 1-3 months, targeting a 10-15% re-rating if management demonstrates power access and pre-lease visibility; stop if the market starts pricing in equity dilution or sub-8% project returns.
• Pair trade: long TU / short a basket of mature Canadian telcos over 3-6 months to isolate the data-center optionality and avoid being paid only for the telecom factor.
• Buy 12-18 month TU call spreads to capture the multi-year re-rating while limiting downside if permitting or interconnection slips; best entry is after any post-headline fade.
• Add exposure to grid/power-enablement names over the next 6-12 months as the second-order beneficiaries; the trade is on MW scarcity, not on Telus-specific execution.
• If TU rallies hard on the announcement, trim into strength rather than chase: the market will likely move ahead of realized cash flow, and the risk/reward worsens until lease-up milestones are confirmed.