State Street 13F Portfolio

The biggest conviction shift this quarter is simple: State Street wants exposure not just to AI chips, but to the tools and fabrics that make them possible. KLA was the standout move — the stake jumped 922.1%, lifting the position to $19.0B and 0.56% of the book, adding an estimated $17.1B in value. That’s a straight bet that process control and yield management are strategic choke points in the AI build-out.

Marvell was the other major upshift. The position was lifted 79.3%, with an estimated $5.5B added, taking it to $12.5B. That’s a call on networking, custom silicon, and accelerators that sit one level behind Nvidia but are critical to AI data center throughput.

At the top of the stack, State Street still refuses to fade its winners. Adds to Alphabet’s GOOGL line (+5.6%, roughly $4.5B), Apple (+2.1%, about $3.7B), Microsoft (+2.9%, about $3.3B), Nvidia (+1.6%, roughly $3.2B), Broadcom (+2.9%, about $2.1B), and Meta (+4.0%, about $2.0B) show continued willingness to pay up for scaled software, GPUs, and custom ASICs.

Below the headline names, the fund kept layering into the semis and tooling complex: Lam Research, Applied Materials, Texas Instruments, Intel, and Palo Alto Networks all saw incremental increases. The pattern is consistent — build a diversified AI picks-and-shovels sleeve rather than swing on a single hero name.

If the buys say “AI everywhere,” the trims say “we’ll fund it from yesterday’s cycle winners.” Micron was the largest reduction by dollars, with shares cut 4.1% and an estimated $2.5B pulled out even though the position still sits at $57.8B and 1.71% of the book. With a gain vs cost north of 2000%, this looks like disciplined profit-taking to finance higher-conviction names like KLA and Marvell rather than a structural rejection of memory.

Energy is clearly being leaned on as a funding source. Exxon was trimmed 3.3% (about $1.0B out), and Chevron 3.1% (around $0.8B out), nudging Energy down to 2.89% from 3.07%. These are meaningful cuts for a book this large, and they telegraph that the secular return on capital in oil is less attractive than in semis.

Cyclicals and capital-intensive financials also contributed cash. Caterpillar was cut 1.4% (roughly $0.5B), while Goldman Sachs and Morgan Stanley were reduced 2.2% and 1.8%, respectively. The tiny 0.5% trim to AMD, despite big gains, looks more like a micro rebalance than a thesis break — especially given the simultaneous enthusiasm for KLA and Marvell.

Importantly, there are no dramatic liquidations in the top-50 list; the sales are controlled skims. This is a reallocation within risk buckets, not a defensive scramble for safety.

The sector chart shows a subtle move that’s big in signal: technology creeps from 58.71% to 59.54% of the book despite huge mark-to-market gains, meaning State Street is adding net dollars on top of price appreciation. That incremental capital is not going to defensive software; it’s going into semiconductors, equipment, and AI-adjacent infrastructure.

On the other side, classic cyclical and rate-sensitive areas are slowly bleeding share. Finance slips from 6.15% to 5.96% even though the fund nudged up JPMorgan, Bank of America, and Wells Fargo — the trims to Goldman Sachs and Morgan Stanley more than offset those. Industrials edge down from 6.72% to 6.61%, thanks to Caterpillar cuts even as defense names like Lockheed and RTX were topped up.

Energy’s drop from 3.07% to 2.89% is textbook: harvest integrated oils, push capital into higher-growth, higher-multiple tech. Consumer Discretionary, Health Care, and Staples are basically steady — Amazon, Walmart, Costco, Home Depot, Procter & Gamble, Johnson & Johnson, Eli Lilly, and Merck look like long-duration quality rather than tactical bets.

Unclassified holdings Berkshire Hathaway and GE Vernova, plus payments (Visa, Mastercard), round out a core compounding spine that hasn’t been disturbed. The real rotation is intra-growth: from commodity and cyclical earnings power toward AI-driven operating leverage.

Taken together, this quarter says State Street views AI as an enduring capital cycle, not a fad to trade. The firm is willing to hold oversized, massively profitable positions in Nvidia, Apple, Microsoft, Alphabet, and Meta, then keep adding around the edges into KLA, Marvell, Lam Research, Applied Materials, and others to capture the broader ecosystem.

The trims are revealing precisely because they’re measured. Pulling dollars from Micron, energy majors, Caterpillar, and investment banks to fund process control, networking, and cloud platforms implies a clear view on where incremental returns on capital will live: in digital infrastructure and its tollbooths, not in barrels, trucks, or balance sheets.

Defensive and cash-flow names — Berkshire, the mega-banks, health care majors, consumer staples — are being kept as stabilizers rather than growth engines. That ballast gives the fund room to own a structurally higher tech weight without turning the book into a one-factor AI bet.

For anyone reading the 13F as a macro signal, the message is blunt. State Street is not preparing for a near-term mean reversion away from tech; it is positioning for AI and data center capex to define the next cycle, and is aligning its capital accordingly.