Amber Kanwar’s Weekly Setup: What investors should watch for in Canadian bank earnings

Canadian banks are set to report this week with the sector at record highs and trading at the highest premium to U.S. peers in 23 years, but Middle East conflict, housing softness and valuation risk remain key headwinds. Canadian Q1 GDP is expected to rise 1.5% year over year, while Citi says tighter financial conditions effectively amount to two rate hikes and the BoC may still cut to 1.75% later this year. The article also previews earnings for Salesforce, Costco and Dell, highlighting AI demand, retail margin pressure and a potential 46% sequential jump in Dell AI server revenue.

Analysis

The bank setup is less about “good earnings” and more about dispersion: the domestic asset-quality story is already broadly understood, so the market is paying for fee resilience, trading income, and balance-sheet optionality. That makes the most levered beneficiaries the firms with the cleanest mix of capital markets, wealth, and U.S. exposure, while lenders with meaningful EM footprints face a second-order squeeze from higher imported inflation, weaker credit demand, and translation risk. The valuation premium matters because when an entire sector is priced for perfection, even modest guidance conservatism can trigger multiple compression rather than just a stock-specific de-rate.

Canada macro is the other hidden short. If bond yields have effectively tightened financial conditions by the equivalent of two policy moves, then the market is already doing part of the central bank’s job; that reduces the bar for a dovish pivot if growth prints soft and inflation continues to cool. The trade implication is that rate-sensitive assets can rally even without an explicit cut, because the first derivative of policy expectations is what matters over the next 4-8 weeks.

In software, CRM is the clearest AI “prove it” event: the market no longer rewards narrative, only monetization velocity. The danger is not just slower adoption, but customer substitution behavior where AI copilots compress seat expansion and reduce upgrade urgency across the software stack. That is a medium-term margin story for incumbents and a near-term multiple story for the group, especially if investors continue rotating toward semis and AI infrastructure.

DELL is the contrarian expression of that rotation: the near-term setup is still favorable, but the stock is increasingly exposed to demand pulled forward by memory inflation and procurement panic. If component pricing stays hot, gross margin may look fine for one more print, but the second-half setup becomes harder if customers already locked in capacity. COST sits at the opposite end of the spectrum — less upside if execution is merely solid, but far lower earnings variance; that makes it a relative haven if discretionary weakness broadens into the lower-income consumer.

Key Decisions for Investors

• Long RY / BMO into earnings, hold 2-5 trading days: favor the banks with cleaner U.S. exposure and stronger fee mix; target 4-6% upside on a beat, but trim quickly if management flags valuation or credit normalization risk.
• Short BNS and CM against long RY in a paired trade, 1-3 weeks: express the view that international EM exposure will be penalized if geopolitical stress keeps inflation and credit costs elevated; aim for relative underperformance rather than outright beta.
• Buy short-dated downside protection on CRM into earnings (weekly or 1-month puts): the risk/reward is attractive because the market is paying for AI monetization proof while the conversion from pilot to paid usage remains the key failure point.
• Stay long DELL but hedge with call spreads instead of outright stock, 2-6 weeks: participate in the blowout AI server quarter while limiting damage if memory costs and pulled-forward demand cause a post-print giveback.
• Initiate a tactical long position in COST on any post-earnings dip, 1-4 weeks: use weakness as a quality-to-defensive rotation trade; downside should be contained unless consumer trade-down is materially worse than expected.