Fil Ltd rotates from Canadian banks into AI software and defensives

The “biggest buys” table makes one thing obvious: Fil is done treating health care as an afterthought. The new AstraZeneca position at 0.71% and the 157.1% ramp in Zoetis (now 0.50%) are a decisive build-out of a pharma and animal health pillar, even as both sit below cost on a gain_vs_avg_buy_pct basis. They are willing to average into drawdowns here, which is a clear sign of structural rather than tactical intent.

On the growth side, they’re doubling down on AI and enterprise software, but with a bias toward infrastructure and durable franchises over pure hype. Microsoft, Texas Instruments, Nvidia, Oracle, Intuit, Adobe, Salesforce, Marvell, and PDD all see higher share counts, with Oracle up 63.6% and Intuit up 64.5% despite negative gain_vs_avg_buy_pct marks. This is the classic active move: buy your own pain, not your winners.

Real economy leverage is coming through the rails and select industrial tech. Canadian Pacific Kansas City is up 26.0% and Canadian National 7.0%, while Teledyne is up 18.0%. Together with a 38.2% add to Suncor and an 11.0% add to Cenovus, they’re building a backbone of hard-asset, cash-flow engines to balance the software and pharma risk.

Funding for all of this clearly comes from dismantling an overbuilt Canadian financial and yield complex. The biggest single dollar trim is Bank of Nova Scotia, cut by 45.2% (down roughly $516.5M). Toronto-Dominion is down 13.8%, Royal Bank of Canada down 4.9%, Canadian Imperial down 8.0%, and even JPMorgan and Progressive are modestly reduced. They’re not abandoning banks, but they’re acknowledging that a 20%+ financials weight made less sense with better growth opportunities on the table.

Energy income vehicles and traditional “Canadian widows‑and‑orphans” names also get taken down. Enbridge is slashed 28.7% and TotalEnergies 17.4%, even as they pour that capital into Suncor and Cenovus. Within gold, Barrick takes a heavy 18.0% cut, while Agnico Eagle, Franco-Nevada and Kinross see only modest trims — a subtle upgrade toward higher-quality duration exposure in the precious metals sleeve.

The other notable source of cash is domestic telecom. Rogers is down 16.5% and BCE nudged lower, both sitting at negative gain_vs_avg_buy_pct. Fil is effectively admitting these were dead money safety trades and rotating their defensive risk budget into pharma, animal health, and higher-quality utilities instead.

At the sector level, the book is rotating along three clear axes. First, Technology climbs to 33.45% while Finance falls to 18.19% and Real Estate (including mis‑classified services like Mastercard and Accenture) slips from 4.08% to 3.55%. They are exchanging rate‑sensitive yield for platform economics and recurring software revenue.

Second, health care goes from rounding error (0.39%) to a real allocation at 2.34%, driven entirely by AstraZeneca and the aggressive Zoetis add. That is a deliberate new leg of the barbell opposite their still‑large tech exposure.

Third, within cyclicals they’re reshaping, not growing, risk. Energy headline weight is essentially flat at 7.16%, but mix shifts away from Enbridge and TotalEnergies toward Suncor and Cenovus. Industrials rise from 7.26% to 7.85% on rails and Teledyne, while Utilities and Telecommunications edge down as Fortis, Emera, TC Energy, Rogers, and BCE are modestly trimmed. The result is a portfolio less dependent on domestic yield and more tied to global demand for computation, molecules, and freight.

This is not a cautious repositioning; it’s a conviction reset after a -5.9% quarter that leans into the losers Fil wants to own for the next decade. The book now reads like a barbell between AI and software platforms on one side, and hard-asset plus health-care defensives on the other. Mega‑cap US tech, semis, and enterprise software are the growth engine; rails, upstream energy, gold, and pharma are the ballast.

If this is right, Fil is implicitly betting that the market has over‑penalized quality growth on duration fears while underpaying for long‑lived infrastructure and IP. The willingness to add materially to names with negative gain_vs_avg_buy_pct — Intuit, Adobe, Salesforce, Oracle, Zoetis, AstraZeneca — shows they see mispricing, not thesis breaks.

Going forward, expect them to keep bleeding weight out of Canadian yield proxies and diversified financials to fund more focused bets in software, semis, health care, and industrial tech. The portfolio is still diversified, with top‑10 concentration at 20.6%, but the direction of travel is clear: fewer passengers, more drivers of secular growth and real-asset cash flow.