Cibc Bancorp USA 13F Portfolio

The biggest adds spell out what Cibc wants to own for the next phase of this cycle: the picks-and-shovels of AI, scale consumer franchises, and a layer of passive equity exposure as risk management.

On the AI infrastructure side, the manager is aggressively upgrading its exposure from headline chips to the memory and wafer-capital stack:

• Micron (MU) is up 96.5% in shares, a $490.3M add, at a gain of 108.8% versus average cost. That is buying into strength, not bottom-fishing – a loud call that AI demand will live in high-bandwidth memory and storage.
• Applied Materials (AMAT) almost doubles in shares, up 96.0% and $181.7M, complementing existing stakes in ASML and LRCX. This is a vote for sustained wafer fab capex rather than a short, hypey AI cycle.
• AMD gets a 12.1% share boost and an $80.9M add while already sitting on a 123.9% gain vs cost, confirming belief that it is graduating from challenger to core GPU/accelerator supplier.

On the consumer side, Cibc is willing to lean into regulatory moats and scale:

• DraftKings (DKNG) is the single largest dollar add at $633.5M, with shares up 205.0%. At only a 3.2% gain vs cost, they are effectively underwriting a long runway for state-by-state legalization and operating leverage.
• Flutter (FLUT), despite being down 27.4% vs Cibc’s cost basis, sees a 68.6% share increase and a $522.7M add. That looks like a classic average-down in a structural winner move across global online wagering.
• Walmart (WMT) shares jump 69.3% with a $142.9M add – modestly profitable vs cost – signaling preference for volume retail and grocery over more cyclical or fashion-led retailers.

The other big winner is passive equity. VOO’s stake rises 43.8% ($304.6M), VEA is up 23.1% ($111.3M), and small increments go into VO, VB, IWF, and IVV. That mix of US large-cap, small/mid, and developed ex-US looks like intentional “index armor” to keep overall factor exposure sane while the single-stock bets get more idiosyncratic.

Funding for these bets comes primarily from the prior AI darlings and some high-beta cyclicals. Cibc is not abandoning the winners; it is trimming them back from oversized, crowded status into core positions.

On the mega-cap tech side, the pattern is consistent:

• Nvidia (NVDA) is cut 10.1% in shares, or about $461.1M of stock, despite still being the top holding at 5.05%. Alphabet (GOOGL/GOOG) is trimmed across both share classes (down 14.5% and 10.7%), a combined $546.1M reduction.
• Apple (AAPL) and Microsoft (MSFT) see similar surgical cuts, -8.7% and -10.2% in shares, taking out roughly $674.5M between them. All four names are comfortably in the green vs cost, so this is straightforward gain-harvesting, not a thesis break.
• Meta (META) is also reduced 10.5%, even though Cibc is slightly underwater on its cost. That suggests a position-size decision: social/ads are now less central than AI infrastructure and ETFs.

Outside the big tech cluster, they are clearly lightening beta and cyclicals to pay for higher-conviction themes:

• Amazon (AMZN) is cut 14.0% ($390.3M), a classic e‑commerce cyclicals trim as they prefer more stable retail like Walmart.
• NextEra Energy (NEE) is pared by 30.2% ($164.3M) around flat vs cost, a notable step back from quasi-growth utilities.
• AstraZeneca (AZN) is slashed 17.9% ($148.5M) despite being down 16.8% vs cost, an unambiguous downgrade of conviction in that pharma story.
• Substantial reductions in JPMorgan (JPM, -12.6%), Blackstone (BX, -20.7%), and Howmet Aerospace (HWM, -21.8%) confirm that financials and industrial beta are being used as funding sources, not growth engines, this quarter.

Sector-wise, Cibc remains a tech-centric allocator with a more diversified chassis. Technology dips only marginally from 54.54% to 53.77% of the book, even after heavy trims to NVDA, GOOGL, AAPL, and MSFT – because that capital is being recycled into MU, AMD, AMAT, LRCX, TSM, and the like.

What really moves is the structure of non-tech exposure. The unclassified bucket – essentially ETFs and structured index products – climbs from 9.65% to 11.17%, through aggressive adds to VOO, VEA, and steady lifts in VO, VB, IWF, IVV, and ATMP. This is how you run a high-conviction AI/memory book without letting factor and style drifts run wild.

Consumer discretionary edges up from 9.52% to 10.35%, but the character changes: less Amazon, Home Depot, and TJX; more DraftKings, Flutter, and Walmart. That is a rotation from cyclical or housing-sensitive demand into recurring digital wagering and staple-heavy retail.

Health care, industrials, finance, energy, and materials all bleed a bit of weight at the margin. Cuts in UNH, AZN, ABT, RTX, HWM, JPM, BX, XOM, and LIN fund the AI-plumbing and ETF upgrades. The message is clear: non-tech sectors are still present for diversification, but they are not where the alpha budget is being spent this quarter.

Pulling the threads together, Cibc is expressing a view that the AI upcycle is real, durable, and most attractively monetized in memory, tooling, and capital equipment rather than only in the most visible megacaps. Doubling Micron and Applied Materials, topping up AMD and TSM, and maintaining a still-massive NVDA position says they see structural, not cyclical, demand for compute and bandwidth.

At the same time, they are clearly uncomfortable with being hostage to a single-factor tech melt-up. Expanding VOO, VEA, VO, VB, IWF, and IVV gives them broad-market ballast, while reweighting toward Walmart and diversified digital wagering platforms gives them consumer cashflows that don’t rely on a perfect macro or ad cycle.

Going forward, watch for two things: whether they keep migrating from AI front-ends (search, social, consumer hardware) into the less glamorous, higher-operating-leverage plumbing; and whether the ETF sleeve continues to grow as they dial up single-stock dispersion. If both trends continue, Cibc’s playbook will look like a barbelled AI strategy: concentrated bets where they believe the economics are underappreciated, wrapped inside an increasingly robust index shell.