New York State Common Retirement Fund 13F Portfolio

The biggest single statement is the blowout add in KLA. The fund lifted KLA by +840.2%, taking it to $448.4M and making it one of the largest incremental capital commitments in the book. That is not a tweak; it is the move of a manager convinced that process control and yield management will be structural chokepoints in the AI and advanced-node buildout.

Alongside KLA, they doubled down on another electronic components name, SanDisk, with a +103.7% increase to $343.6M. Together, those adds say they want more exposure to the memory and storage backbone that actually feeds AI workloads, not just the headline GPU vendor.

Further down the list, Marvell is quietly increased by 6.5% to $293.0M, reinforcing a bet on data-center networking and custom silicon as second-derivative AI winners. Outside of semis, the adds in Walmart and Coca-Cola — $577.0M and $385.0M respectively after low-single-digit increases — look like deliberate ballast: durable cash-flow franchises to sit opposite an even-more-concentrated tech book. A small add to Palantir rounds out the picture: the fund still wants AI software upside, but it’s spreading the bet beyond the Magnificent Seven.

Almost every trim at the top of the book is a classic winner’s cut. Nvidia, still a 6.27% position at $5.0B, is reduced by -5.1%. Apple, Microsoft, Alphabet’s A and C shares, Amazon and Meta all see low- to mid-single-digit percentage reductions. This is systematic profit-taking in names where the fund is sitting on enormous gains — Nvidia up 11349.3% versus its average cost, Apple up 5474.4%, Alphabet A up 1566.5%.

Micron’s -9.3% cut is more aggressive, even though it remains a $1.32B position with a 3228.4% gain versus cost. That suggests a view that DRAM/NAND cyclicality has run ahead of near-term fundamentals compared with the more oligopolistic process-equipment vendors they are adding.

Away from tech, the fund is clearly funding new ideas by leaning on incumbents in sectors they view as ex-growth or late-cycle. Merck is slashed by -17.2% (down to $373.9M), while UnitedHealth, Bank of America, Wells Fargo, Goldman and Morgan Stanley all take high-single to mid-teens share reductions. The message: they still respect these franchises but see better incremental risk/reward in AI and chip capex than in US banks and big pharma at this point in the macro.

On paper, sector weights barely budge; in substance, they move a lot. Technology edges up from 61.92% to 62.59% of the disclosed book, but that masks a meaningful swap within tech: out of mega-cap platform exposure, into semiconductors and equipment like KLA, Marvell and SanDisk.

Health care drifts down from 7.78% to 7.45% as they pull back from Eli Lilly, Johnson & Johnson, AbbVie, Merck and UnitedHealth. This is not a wholesale exit, but the direction is clear: less dependence on pharma pipelines and US managed care, more on AI-driven demand for compute and electronics.

Financials fall from 5.56% to 5.25% as banks (JPMorgan, Bank of America, Wells Fargo) and brokers (Goldman, Morgan Stanley) are trimmed, reducing regulatory and credit-cycle exposure. Energy ticks down from 1.90% to 1.83% with cuts in Exxon and Chevron, while industrials nudge up from 4.71% to 4.77% on a small add to Caterpillar despite trims in Tesla and RTX. Consumer looks barbelled: consumer discretionary is flat at 10.71%, but inside that they rotate slightly toward Walmart and away from higher-beta names like Costco and Home Depot, while consumer staples (Coca-Cola) rises from 0.81% to 0.86%.

Put together, the quarter reads like a high-conviction view that AI infrastructure is early in a multi-year capex supercycle and that owning the tools, memory and connectivity around GPUs will pay better than chasing the most crowded winners.

The fund is not scared of concentration; it is redefining it. Top holdings still cluster in big tech, but incremental dollars are migrating from the broad AI/mega-cap trade into more idiosyncratic, manufacturing-heavy parts of the value chain. That sort of move tends to come from a manager that believes the theme is durable, not peaking.

At the same time, adds to Walmart and Coca-Cola, plus modest support for industrials like Caterpillar, suggest they are not blind to macro or valuation risk. If AI and semis wobble, the portfolio now has more ballast in staples and blue-chip retail; if the cycle persists, KLA, Marvell, SanDisk and Palantir give them geared exposure. For a large public pension that just printed 11.0% in the quarter and a 3-year annualized 24.43%, this is not “de-risking” — it is a conscious shift to own the less obvious, but arguably more durable, legs of the AI trade.