Barlow’s Research Roundup: Canadian banks adopting AI much faster than digital currencies

RBC says Canadian banks will likely be slow adopters of AI, stablecoins and tokenization, though AI could still drive incremental cost savings and revenue opportunities over time. The note also flags cybersecurity, strategic and talent risks, while highlighting bank-specific AI targets such as BMO's over $1B of PPPT from AI by 2030, RBC's $700M-$1B incremental enterprise value by 2027, and TD's $1B AI value target. Separately, BMO sees commercial real estate still under pressure in multifamily and industrial, while retail is steady and office is gradually improving; CIBC pointed to record LNG exports from Canada's west coast.

Analysis

The near-term winner here is not the banks’ revenue line; it is their cost base and control stack. Even if monetization from tokenized deposits or stablecoin rails takes years, AI adoption can quietly compress operating expenses and reduce fraud-loss severity before it ever shows up in top-line narratives. That favors the largest incumbents with scale, data breadth, and capital to absorb model risk, while smaller FIs and fintech intermediaries are more exposed to being squeezed on payments, custody, and workflow automation.

The second-order risk is that cyber defense and cyber offense both improve at the same time. That creates a paradox where breach frequency may rise in the short run, but the institutions with better datasets and security budgets can actually widen the gap versus peers through lower incident costs and better customer retention. In that setup, the market may underappreciate that AI spend is less about growth optionality than about preserving franchise durability in a slower-adoption regulatory regime.

On the bank-specific setup, the bigger issue is dispersion: a bank with clearer AI execution credibility should deserve a valuation premium, while a bank with more operational complexity and less visible monetization gets trapped in the “AI promise, no proof” bucket. The consensus may be too focused on future enterprise value targets and not enough on which franchises can convert AI into tangible PPPT improvement within 12-24 months. That makes the setup more of a relative-value story than a sector-wide re-rating.

The macro backdrop from commercial real estate and LNG matters because it shapes credit quality and funding pressure, not just headline sector sentiment. Higher office and retail resilience can offset some CRE concerns, but the multifamily softness and industrial weakness still argue for caution on credit-sensitive lenders, especially if rate cuts stall or vacancy trends worsen into 2026. Record LNG exports are a modest positive for Canadian energy-linked cash flows and Western Canada activity, but the spread improvement may be too small to materially change broad bank earnings unless it triggers a durable capex cycle.

Key Decisions for Investors

• Long TD / short CM on a 3-6 month horizon: TD has more visible AI monetization credibility and a cleaner relative-risk narrative; CM looks more exposed to slower conversion of AI spend into near-term operating leverage. Target ~5-8% relative outperformance, stop if CM shows a material acceleration in fee income or efficiency ratio improvement.
• Buy TD Jan-2027 calls or call spreads to express the market not fully pricing medium-term AI value realization. Use strikes modestly above spot to capture a 12-18 month re-rating if cost saves and workflow automation become measurable in quarterly disclosure.
• Long BMO common into any pullback tied to generic “AI spend skepticism” headlines. The setup is a quiet margin-improvement trade, with upside over 6-12 months if AI-driven efficiency can offset CRE-related provisioning drift; risk is that disclosure remains too opaque for the market to pay up.
• Avoid chasing the entire Canadian bank basket as a thematic long. Prefer pair trades and single-name selection; the market is likely to reward execution dispersion rather than assign a broad AI multiple to the group.
• For more defensive positioning, consider a small short or underweight in Canadian bank names with the most CRE-sensitive loan books versus TD/BMO, as 2026 CRE weakness is a slower-burn catalyst that could surface through provisions after the AI excitement fades.