Toronto Dominion Bank 13F Portfolio

The biggest add is unambiguous: eBay explodes to 6.29% of the portfolio, with shares up 297.8% and an estimated additional $2.83B deployed. That is not a trade; that is a franchise-level judgment that a cash-generative, unfashionable marketplace with only a 6.3% gain versus average cost still offers attractive upside.

On the growth side, the AI stack remains the spine of the book, but the emphasis is shifting from pure winners to the broader supply chain:

• Nvidia and Apple both see meaningful top-ups, with Nvidia up 14.7% and Apple up 16.9%, reinforcing them as high-conviction compounders where the fund is sitting on triple-digit percentage gains versus cost.
• Intel is the classic contrarian add: shares up 111.8% this quarter, lifting the stake to 2.0% of the book and adding an estimated $635.3M despite only a 70.4% gain versus average buy. They are clearly willing to underwrite a long, messy turnaround in foundry and PC/datacenter share.
• Marvell gets a 539.2% increase and another roughly $207.5M, a big statement that networking, custom silicon, and storage are key bottlenecks in AI deployment, not just GPUs.

The manager also leans into AI monetizers: Meta is up 39.8%, and Shopify gets a 44.3% boost, while software-adjacent plays like Oracle and CoreWeave see sizable increases despite Oracle being slightly under water on cost. This signals a willingness to average up into platforms that can tax AI workloads over time.

Away from tech, the consumer book is being rebuilt around durable U.S. names:

• Lowe’s and Starbucks each see triple-digit percentage increases in shares (+211.0% and +505.3%, respectively), with roughly $164.3M and $183.9M of fresh capital, and Walmart gets a solid 19.5% add. The fund is deliberately concentrating into scaled U.S. consumer brands rather than chasing higher-beta discretionary fliers.

Under the radar, the bank is also nudging up its exposure to North American hard infrastructure via Canadian Pacific Kansas City, Canadian National, Nutrien, Caterpillar, and an increased Brookfield stake. That cluster looks like a long-term bet on trade flows, freight pricing power, and asset-backed cash flows that can survive different macro regimes.

The funding sources are just as revealing as the buys. The manager is aggressively cashing in on long-held, high-gain positions in Canadian financials and energy, effectively de-emphasizing the domestic macro cycle.

The steepest trims by dollars cluster in three buckets:

• Canadian banks and insurers: Royal Bank of Canada is cut by 30.3% (roughly $711.5M out), Canadian Imperial by 28.3% (about $659.1M), Bank of Nova Scotia by 28.9%, Manulife by 56.9%, Sun Life by 48.4%, and even JPMorgan gets trimmed. These are not panic exits — gains versus average buy are all healthy, often well above 100% — but classic profit-taking in cyclical, rate-sensitive balance-sheet stories.
• Energy and pipelines: Canadian Natural Resources is slashed by 58.8% with an estimated $771.0M taken off the table, Suncor by 47.9%, Enbridge by 42.9%, and TC Energy by 59.8%. Together they represent a conscious step back from commodity and midstream risk after a strong run, redirecting capital into secular growth.
• High-flyer AI names at rich marks: AMD is cut by 45.1%, Micron by 27.9%, and Applied Materials by 24.4%, despite enormous gains versus average cost (Micron is up more than 500% vs cost, AMD nearly 300%, AMAT over 300%). Alphabet’s A and C shares are also reduced, down 29.2% and 11.4%, and Palantir is trimmed by 21.5%.

Read together, this is not an anti-tech move; it’s an internal rotation within tech. The fund is selling partial winners in names where the market has already re-rated the story and pushing that capital into perceived underappreciated or earlier-stage beneficiaries like Intel, Marvell, CoreWeave, Oracle, and the rails. They are also quietly letting some U.S. healthcare exposure bleed down, with UnitedHealth off 41.3% and Eli Lilly down 26.2%, signaling less appetite for crowded GLP-1 and managed-care narratives at current valuations.

At the sector level, the quarter is a decisive reweighting toward secular growth and away from rate- and commodity-sensitive cyclicals. Technology marches from 43.64% to 46.92% of disclosed exposure, while Finance drops from 25.2% to 18.74% and Energy from 8.36% to 4.25%.

The tech push is broad-based: semis (Nvidia, Intel, Broadcom, Marvell, Lam, Qualcomm, Micron, AMD, Applied Materials) are still the core, but software/platform plays like Meta, Shopify, Oracle, CoreWeave, and Palantir now form a meaningful layer of optionality. The manager is clearly positioning for a multi-year AI capex cycle and the follow-through into advertising, cloud, and enterprise workloads.

Real estate looks inflated at 12.55% in the screener data because eBay and Visa are misclassified; economically, this is really an e-commerce plus payments bet rather than a property call. Finance’s decline is almost entirely about trimming Canadian commercial banks and life insurers; Toronto-Dominion itself is actually up 4.5%, underscoring that the de-risking is at the portfolio level, not an existential call on the parent entity.

Energy and Utilities exposures fall sharply via cuts to Canadian Natural, Suncor, Enbridge, TC Energy and others, while Industrials creep up through rails and Caterpillar. Consumer Discretionary edges higher, but the composition matters more than the size: the book is moving towards scale retailers and restaurant chains with pricing power rather than speculative consumer tech.

Health Care’s weight falls as UnitedHealth and Lilly are trimmed, with some of that "defensive growth" capital effectively recycled into gold via Agnico Eagle, where the stake is up 41.8%. That modest Basic Materials move reads like an inflation and tail-risk hedge sitting alongside a much more aggressive growth core.

Put simply, this is a book built for a world where AI capex, cloud workloads, and U.S. consumption carry more weight than interest-rate volatility in Canada or the next OPEC headline. The manager is leaning into platforms that either sell the chips, own the data, or sit at the checkout of global commerce.

The internal rotation within tech — from partially monetized winners like AMD, Micron, and Alphabet into Intel, Marvell, CoreWeave, Oracle, and still-growing positions in Nvidia, Apple, Meta, and Shopify — suggests they see the AI cycle as broadening rather than peaking. They are willing to accept higher idiosyncratic risk in earlier-stage or turnaround names to keep upside convexity.

On the macro side, the sharp cuts in Canadian banks, insurers, and energy point to lower confidence that traditional value and yield plays will outperform from here. Instead, the ballast of the portfolio is shifting toward U.S. consumer staples/discretionary leaders (Walmart, Lowe’s, Starbucks) and North American rails and industrials, which should benefit if nominal growth and trade remain resilient.

The eBay move is the clearest outlier — a single-stock decision to make a somewhat neglected e-commerce asset a core 6.29% holding at only a modest premium to cost. That looks like a classic value-in-a-tech-wrapper thesis: strong cash flow, buybacks, and operational leverage if sentiment normalizes.

If this quarter is any guide, expect future filings to show continued pruning of crowded winners and domestic cyclicals, with incremental dollars flowing into the less-traveled parts of the AI supply chain, asset-heavy infrastructure tied to North American trade, and durable consumer franchises that can compound through different rate regimes.