Schroder Investment Management Group 13F Portfolio

The biggest adds table reads like a deliberate pivot toward defensible cash machines rather than a chase for the next AI ticker. The most striking move is the 306.9% increase in UnitedHealth, taking it to $1.06B and 0.75% of the book — a huge dollar add of about $800.4M in a single quarter. For a diversified manager, that’s not tinkering; it’s a statement that managed care is a core earnings engine, not a sidecar.

Health care more broadly is clearly in favor. Eli Lilly was boosted by 9.2% (about $129.7M), and Stryker by 53.9% (about $233.8M). Schroder is concentrating around high-ROIC, pricing-power names that sit on demographic and innovation tailwinds, rather than lower-multiple but structurally challenged pharma.

On the “real economy” side, Deere was increased by 138.9%, a roughly $323.9M add, and Visa by 10.5%, or about $248.4M. Deere gives them exposure to capital spending and food production, while Visa is essentially a tax on global nominal GDP and e‑commerce.

The willingness to lean into cyclical and bruised growth is also notable:

• Sea was lifted 22.0% (about $127.9M) despite being underwater relative to cost.
• H World Group jumped 64.2% (about $243.6M), a strong vote of confidence in Chinese outbound and domestic travel.

Even in tech, adds like Apple (+4.1%) and Cadence (+13.1%, about $107.8M) show a preference for mature, cash-rich platforms and mission‑critical software over speculative AI names. This is a portfolio that still wants growth — but growth with cashflow, and growth it understands.

The funding sources for this shift are just as revealing as the destinations. The single biggest trim is Arm, cut by 51.0% and roughly $691.8M, even though the position is sitting on gains of 142.8% versus Schroder’s average cost. That looks like disciplined profit-taking in a richly valued story, not a call on Arm’s technology.

Across the AI supply chain and infrastructure, there is a consistent pattern of shaving winners:

• Alphabet’s main line (GOOGL) was cut 3.1% (about $248.1M), despite a gain vs cost of 281.8%.
• Taiwan Semiconductor was trimmed 3.5% (about $99.6M), with a 280.9% gain.
• Vertiv, a data-center capacity play, was reduced 21.8% (about $151.1M).
• Palo Alto Networks and Emerson were cut 20.6% and 22.1% (about $166.9M and $182.5M), respectively.

These are not factor-level de‑risks of tech, because the core AI levered names — NVIDIA (only -0.3%), Broadcom (-0.3%), AMD (+7.9%), Micron (+9.3%) — remain large and, in some cases, are still growing. Instead, Schroder appears to be pruning the more fully discounted AI-adjacent and IT-infrastructure plays to finance step‑ups in health care and real-economy names.

There is also a quiet cleanup in expensive or lower‑conviction industrial and tools positions: Thermo Fisher was cut 11.8% (about $101.9M), GE Vernova by 11.3% (about $93.0M), and Medtronic by 10.7% (about $71.7M). The trims cluster in areas where valuations have run or where secular visibility is weaker, which fits a house style of using rallies to exit the second tier while keeping the crown jewels.

For all the stock‑level activity, the sector bars show a measured rotation, not a wholesale reshaping. Technology eased from 62.49% to 60.44% — still dominant, but this is the first sign that Schroder is capping pure tech beta after a monster 3‑year run.

The beneficiaries are precisely the sectors that can cushion AI volatility while still compounding:

• Health care moved from 6.19% to 7.38%, driven by big adds in UnitedHealth, Lilly, Stryker and a smaller lift in AbbVie.
• Industrials crept up from 3.40% to 3.59%, largely Deere offsetting cuts to Thermo Fisher and GE‑related names.
• Consumer Discretionary climbed from 8.08% to 8.64%, with incremental risk into Sea, Tapestry, H World, TJX and Amazon.

Finance stayed flat at 6.65%, but beneath the surface there was a modest upgrade in quality and cyclicality: Visa, Mastercard and American Express all grew, while Morgan Stanley edged down. Real “Real Estate” in the dataset is actually payment networks and MercadoLibre, and that sleeve ticked up slightly, underlining the theme of owning transaction rails.

The unclassified ETF bucket (VOO, IVV) was nudged up in share terms but down slightly in weight, acting as a liquidity and beta buffer rather than a core conviction area. Overall, sector rotation confirms the stock‑picking narrative: this is not a move away from tech, but a redistribution from AI plumbing and infrastructure into healthcare and real‑world cashflows that should hold up if AI sentiment cools.

Put together, the quarter sketches a manager that still wants to participate fully in AI‑driven upside, but no longer wants its portfolio outcomes to hinge exclusively on how the market prices AI infrastructure. NVIDIA at 6.53%, Microsoft at 4.15%, Alphabet across two lines, Broadcom, AMD and Micron remain a massive concentrated bet that compute will keep compounding.

The difference now is the ballast: a much larger UnitedHealth and Lilly, plus a beefed‑up Stryker position, create a health‑care earnings spine that has little correlation to GPU cycle timing. Deere, Visa, Mastercard and American Express add exposure to nominal GDP, capital spending, and consumer transactions — macro‑sensitive, but anchored in dominant competitive positions.

Schroder is also willing to be early and wrong for a while in select growth names. The 22.0% increase in Sea and the 64.2% jump in H World, both with modest or negative gains vs cost, show a view that Southeast Asian platforms and China travel have more runway than the market is currently pricing.

If this pattern continues next quarter, expect further trimming of high‑multiple AI adjacencies and tools in favor of scaled, cash‑rich franchises in healthcare, payments, and real‑asset‑linked industrials. The bet is that AI will keep driving the indices — but that the best way to monetize it, from here, is to own the highest‑quality nodes in the stack and surround them with businesses that can compound even if AI sentiment pauses.