RBC bank analyst previews earnings reports, presents top picks

Daily roundup of research and analysis from The Globe and Mail’s market strategist Scott Barlow

RBC bank analyst Darko Mihelic previewed earnings reports for the sector and presented top picks,

“We adjust our Q3/25 estimates modestly (little tweaks to capital markets and other individual bank adjustments) and expect approximately 6-per-cent year-over-year growth and a healthy 6-per-cent quarter-over-quarter increase in core EPS on average for the large Canadian banks in our coverage, mainly because we do not expect the performing reserve builds in Q2/25 to repeat. We model muted loan growth while NII/NIMs are still supported by tractors and we expect capital markets and wealth businesses to perform well as equity markets were helpful. Credit quality appears to remain stable; although we still see some soft spots in GDP growth, unemployment, and housing market trends, we don’t expect to see “issues” in Q3/25 results. We believe CM, NA and TD can report results stronger than the most recent consensus view”

The analyst has outperform weightings on Bank of Montreal (BMO-T), Canadian Imperial Bank of Commerce (CM-T) and EQB Inc. (EQB-T).

***

BMO chief economist Doug Porter does not see AI as a productivity boost equal to the internet,

“Last week’s Q2 report on U.S. productivity didn’t get the attention it deserved—there is, after all, a lot going on these days. Perhaps that’s because it wasn’t that notable, with productivity rebounding at a 2.4-per-cent annualized rate. in Q2 after a 1.8-per-cent drop the prior quarter. That combo left productivity up a so-so 1.3 per cent from a year ago. Ho hum? But that’s just the point. Productivity has not been doing anything out the ordinary in the past couple years, even with the massive investments in AI. Some would suggest that it is far too early to pass judgement, and there is some merit to that view. But note that it’s now been nearly 3 years since ChatGPT burst onto the scene and many, many billions of investment in the space since. It isn’t exactly comparable, but the internet first became widely used in 1995 (just ask AI, it will tell you), and U.S. productivity was ripping with 3-4-per-cent gains per year by 1998 and 1999. (The best stretch of productivity gains in the past 50 years was from 1997 to 2003.) Allowing that it is still early, but the evidence so far suggests AI will contribute to productivity, but it’s not going to rival the widespread adoption of the Net in the late 1990s.

***

Wells Fargo strategist Austin Pickle has downgraded commodities from overweight to neutral but there’s a lot of detail under the surface,

“While our forecast for a modest reacceleration in U.S. economic conditions in 2026 should still be a tailwind for commodity performance, we see better opportunities in our other favored asset groups. We suggest taking profits in Commodities, bringing their allocations back to strategic weight, and rotating into fixed income until markets present more compelling opportunities. Both Energy and Industrial Metals commodity sectors should benefit from a modest economic rebound in the latter part of 2026, lower short-term interest rates, and economic stimulus in China. However, OPEC+’s l accelerated production quota reduction is a considerable headwind to oil prices, as well as the return potential of the Energy sector. Meanwhile, Industrial Metals should benefit from sluggish global mined copper production as demand grows. Additionally, the recent sharp collapse in copper prices — as tariffs announced proved to be less severe than anticipated — provided an attractive tactical opportunity to overweight the Industrial Metals sector, in our view. As a result, we downgraded Energy from favorable to neutral while simultaneously upgrading Industrial Metals from neutral to favorable”

***

Bluesky post of the day:

Diversion: “AI automatically designs optimal drug candidates for cancer-targeting mutations” – Phys.org