Geode Capital Management Llc

The biggest adds by dollars are not speculative side bets; they are reaffirmations of the core AI and cloud stack. Geode adds to Apple at 5.8% of the book, NVIDIA at 6.5%, Microsoft at 4.33%, Amazon at 3.01%, and Alphabet’s A and C shares, all at substantial gains versus cost, signalling they see this cohort as secular compounders, not crowded momentum.

Beneath the headline platforms, the more telling move is into the AI plumbing. Applied Materials is up +17.8% in share count with an estimated $1.13B added, while Broadcom, Micron, AMD, Lam Research, and KLA all see notable increases and are sitting on multi-hundred-percent gains versus Geode’s average buy. That is a vote that the capex cycle for AI and high-bandwidth compute has a long runway.

Geode is also quietly reinforcing the software layer around this hardware: incremental builds in Palantir, Oracle, IBM, and both Alphabet share classes tie the thesis together. The message is consistent — own the full AI value chain from chips, to tools, to the hyperscale platforms that monetize it.

Selling is surgical, not thematic capitulation, but the pattern is clear: Geode is gently harvesting from slower, cash-heavy sectors to feed its AI and growth complex. The only explicit top-50 trims are Exxon Mobil at -0.5% in shares and Merck at -0.8%, but both moves come after healthy gains off low average costs.

Within health care, Eli Lilly, Johnson & Johnson, AbbVie, UnitedHealth, and Amgen are all being added to rather than cut, which makes the Merck trim stand out as a simple position-size clean-up in a crowded pharma sleeve. Energy tells a different story; Chevron is barely touched while Exxon is trimmed, hinting at a preference for maintaining sector exposure but pruning single-name risk where the easy re-rating has already occurred.

The absence of big new positions or wholesale exits in the top-50 also matters. Capital is being recycled at the margin, not reallocated into new themes — a sign Geode believes its existing winners still offer better risk/reward than any fresh idea they could underwrite this quarter.

Sector data confirms what the name-level moves already suggest: technology drifts up from 56.88% to 56.95% of the disclosed book, even off a weak performance quarter. That incremental move, on a $1.61T 13F base, represents real dollars being re-committed to a theme Geode already dominated.

Consumer exposure is also nudging higher, but very selectively. Amazon, Walmart, Costco, Netflix, Procter & Gamble, Home Depot, and McDonald’s are all increased, tilting the consumer sleeve toward scale platforms and category killers rather than smaller, more cyclical bets.

By contrast, energy slips from 3.09% to 2.99%, and health care edges down from 8.13% to 8.11%, even though multiple health names are being topped up. Finance, telecom, and staples weights are effectively stable — banks like JPMorgan, Bank of America, Wells Fargo, and Goldman Sachs are being added to, but more as yield-and-beta ballast than as a new growth frontier.

Taken together, the quarter reads as a high-conviction continuation, not a reset. Geode is betting that the AI and hyperscale cloud supercycle is still early, and that the right response to volatility is to size up the winners across the entire stack. From NVIDIA and Broadcom to Applied Materials and Lam Research, they are building around the idea that compute intensity and chip capex will keep rising.

The incremental adds in mega-cap consumer platforms and dominant retailers point to a parallel thesis: in a slower, more uncertain macro tape, market structure still favors global scale, network effects, and balance-sheet strength. Names like Walmart, Costco, Netflix, and McDonald’s fit that bill.

On the risk side, the muted trims in energy and Merck suggest Geode is not abandoning defensiveness, just squeezing those pockets to keep funding growth. If this playbook works, expect further rotation within tech — toward enablers and tooling — rather than a wholesale sector shift. If it doesn’t, the ballast in banks, staples, telecom, and big pharma should cushion, but the real P&L will still be driven by whether the AI spend curve keeps steepening.