Invesco Ltd. rotates from mega-cap AI to health care and hard assets

The biggest adds cluster around three themes: AI plumbing, large-cap pharma (especially obesity and oncology), and resilient consumption. The allocation data and the top-dollar buys make this clear.

• Walmart: A 74.0% position increase and a $2.30B capital add vault Walmart to 0.83% of the book. That is a statement that scale brick-and-mortar plus e-commerce is the consumer exposure Invesco wants, instead of chasing more cyclically sensitive online names.
• Micron and Applied Materials: Micron is up 23.4% in shares with a $771.4M add; Applied Materials is up 21.6% with $589.1M added. This is a deliberate push into memory and equipment – the scarce inputs and tools behind AI data centers – rather than just the GPUs themselves.
• Caterpillar: The 27.0% share increase and $579.5M add show a clear preference for hard-asset industrial exposure tied to infrastructure and industrial capex, not just software.
• Johnson & Johnson, Eli Lilly, and AstraZeneca: JNJ (+18.2%, +$567.4M), LLY (+14.7%, +$408.0M), and a new $1.99B position in AstraZeneca collectively build a sizable large-cap pharma leg. This points to a multi-year bet on pipelines in oncology, immunology, and obesity/diabetes rather than smaller, binary biotech risk.
• Intuit: A 31.4% increase and $468.1M add into a name currently modestly underwater (gain_vs_avg_buy_pct at -7.2%) suggests they see recent weakness as an entry point, not a thesis break, for mission-critical software.

Taken together, this is Invesco doubling down on the infrastructure of AI and health innovation while pairing it with upgraded, durable consumer demand.

On the sell side, the pattern is straightforward: realize gains in the most extended mega-cap winners and in some financial and payment franchises, then redeploy into the new legs of the thesis.

• Mega-cap platforms: Meta is the single largest trim by dollars, with shares down 14.3% and about $1.49B taken off the table. Microsoft follows with a 6.5% cut and a $1.09B dollar reduction. Amazon is down 4.1% (-$512.7M), and Broadcom is trimmed 3.9% (-$415.4M). These are all huge gainers (Meta up 214.1% vs cost; AVGO up 453.8%), so the motivation looks like classic profit-taking rather than a repudiation of AI.
• Financials and payments: JPMorgan is reduced 14.5% (-$617.9M), Wells Fargo 8.1% (-$225.4M), and Visa 14.4% (-$481.4M). That combination reads like a funding basket: quality franchises that have rerated well (JPM up 186.9% vs cost) but don’t sit at the center of Invesco’s forward growth map.
• Other trims: Modest reductions in Chevron (-8.5%, -$302.5M), Cisco (-5.1%, -$237.0M), and a small Tesla cut suggest a view that traditional energy, legacy networking, and volatile autos are supporting actors, not core engines.

Importantly, they barely touch Nvidia and Apple (both trimmed less than 1%), signaling they still see these as core AI and ecosystem anchors, even as they diversify around them.

Sector data confirms this is a rotation within growth, not away from it. Technology nudges down only slightly from 57.43% to 56.24%, but inside that, dollars are moving from platform front-ends (Meta, Microsoft, Amazon) into semiconductors and equipment (Micron, Applied Materials, Lam Research, AMD, Texas Instruments, KLA, Analog Devices).

Health care is where the real net build is happening, rising from 5.68% to 6.95%. The combination of adds in JNJ, LLY, MRK, PM, and the new AstraZeneca line is a clear attempt to create a second diversified growth engine with lower correlation to pure tech.

Consumer exposure is being upgraded, not expanded recklessly. Consumer discretionary edges up from 11.31% to 11.77%, driven by Walmart and steady stakes in Costco and Netflix, while consumer staples rise from 2.51% to 2.57% on increased Coca-Cola and CVS despite a tiny Pepsi trim. Finance slips from 6.08% to 5.65%, Real Estate (really payments via Visa and Mastercard) falls from 2.62% to 2.35%, and Telecommunications (Cisco) edges down, all consistent with these sectors being used as cash registers rather than core themes.

Put together, this quarter sketches a manager that still believes in AI and growth, but wants to be paid for taking infrastructure and clinical risk instead of just headline multiple risk. The heavy adds in memory, equipment, and industrials show a conviction that AI and reshoring will require enormous capex in fabs, tools, and physical infrastructure.

The simultaneous build-out in large-cap pharmaceuticals and obesity/oncology franchises suggests Invesco wants a second, long-duration S-curve that doesn’t trade tick-for-tick with cloud and software. That’s supplemented with higher-quality consumer demand through Walmart, beverages, and drug retail.

Expect future quarters to rhyme with this one: trims of over-earning, fully rerated financials and platform names, and incremental adds to AI plumbing, industrial capacity, and big, diversified pharma pipelines. If they’re wrong, it will be because the capex and health-care payoff is slower than expected; if they’re right, the portfolio will be levered to the less glamorous, more durable parts of the next growth cycle.