Jupiter Topco 13F Portfolio

The biggest adds table is basically a roster of this cycle’s AI oligopolists. Jupiter is not trying to be cute on timing; they are writing very large tickets into the most consensus names in global markets, and accepting that crowding risk in exchange for operating leverage to AI demand.

• NVDA (~$16.2B, 6.58%) is the statement piece: the fund is explicitly underwriting NVIDIA as the high‑margin nexus of AI compute. With the position already about 23.3% above their $182.63 average cost, they are willing to buy the name mid‑re‑rating, not waiting for panic.
• GOOG (~$10.6B, 4.33%) and GOOGL (0.43%) suggest a deliberate overweight to Alphabet as the core beneficiary of model deployment at scale, not just advertising. The parallel share classes show size‑before‑precision sizing.
• AVGO ($7.40B) and MSFT ($6.93B) extend the same thesis: Broadcom as the glue in data‑center and networking stacks, Microsoft as the most vertically integrated AI software and cloud platform.
• AMZN ($5.91B) and AAPL ($5.38B) round out the megacap barbell; Amazon for AWS and logistics rails, Apple for device‑side compute and ecosystem monetization.

Below the mega‑caps, Jupiter scales into the tooling layer: LRCX, MU, KLAC, TSM, AMAT, ON, STX and ETN are essentially a bet that capex into foundry capacity, power, and industrial electrification will follow the model‑training wave. This is the picks‑and‑shovels complement to the platform bets.

Finally, the top adds include LLY and a cluster of high‑value therapeutics names (ARGX, AZN) alongside RVMD, PRAX, BBIO, PCVX, MDGL. Jupiter clearly wants exposure to both the obesity/GLP‑1 and oncology/genetic‑medicine arcs, accepting trial and reimbursement risk as the price of long‑dated optionality.

The 13F only shows top‑50 current positions, and every one of them is tagged as new; there are no disclosed trims. That doesn’t mean Jupiter did nothing on the sell side — it means whatever they exited or cut fell either outside the top‑50 or entirely out of the 13F universe.

The cleaner way to read this quarter is as a funding reveal: everything not present is a de‑emphasis. There is no old‑economy energy, no broad utilities, no regional banks, no REITs, and essentially no value cyclicals. For a $245.4B 13F book with just 28.0% in the top ten, that absence is loud.

Even within what’s present, you can see what they refuse to size aggressively. Financials are token — JPM at 0.52%, LPLA at 0.48%, PGR at 0.38% — enough to keep some credit‑cycle and rate‑sensitive exposure, but clearly not a core bet. Industrials like HWM, DHR, TDY, RVTY are sized as growth‑adjacent beneficiaries of AI, aerospace, and lab‑automation rather than traditional GDP plays.

The small consumer positions (ARMK, APG, SPOT, plus AMZN) likewise tell you they don’t see the U.S. consumer as the primary driver of the thesis. They want the secular compounding stories where end‑demand is compute, data, bandwidth, and molecules — not mall traffic.

The sector bar chart shows a top‑down book that is aggressively non‑balanced by design. Technology, correctly understood as semis, semicap, cloud, software, and networking, sits at 63.18% — a venture‑style concentration in public equities.

Inside that tech block, though, the risk is more layered than it looks. There’s a clear split between core platforms (NVDA, GOOG/GOOGL, MSFT, AMZN, AAPL, META) and the capital‑equipment and component vendors needed to sustain AI workloads (AVGO, LRCX, TSM, MU, KLAC, AMAT, ON, STX, ETN, FLEX, APH, ANET, LITE, DDOG, ORCL, DASH, GE). The portfolio is effectively a long AI capex cycle expressed across silicon, power, connectivity, and observability.

Health care at 15.75% is the second anchor, but it is internally barbelled. On one side you have mega‑pharma and managed care (LLY, AZN, JNJ, ABBV, UNH); on the other, a volatile basket of clinical‑stage and early commercial biotechs (ARGX, RVMD, PRAX, BBIO, PCVX, MDGL). That mix gives them both defensive cash flows and idiosyncratic upside.

The remaining sectors read like risk dampeners and thematic extensions rather than co‑equals. Industrials at 4.13% (HWM, DHR, TDY, RVTY) catch aerospace, test, measurement, and lab instrumentation demand. Finance at 2.54% and the mis‑classified payments names MA and V at 2.61% function as high‑ROIC cash machines under NASDAQ’s odd “Real Estate” label. Consumer Discretionary at 6.65% and Telecom at 2.04% mostly host digital‑first models levered to cloud and mobile time‑spent.

Taken together, Jupiter Topco is telling you it wants to be paid for owning the infrastructure of the AI and data‑driven economy, not for outguessing the cycle. That means living with factor risk: momentum, growth, and crowding, offset only partially by healthcare and a sprinkling of financials.

The AI core — NVDA plus hyperscalers and semicap — suggests they expect AI capex to remain elevated and relatively insensitive to near‑term macro wobbles. The gains versus their average cost across names like NVDA, MSFT, AMZN, AMD, DDOG, STX, and ANET show they were willing to build size into strength and have been rewarded so far, but also that future drawdowns will bite hard.

The biotech sleeve is the real torque. RVMD, PRAX, BBIO, PCVX, MDGL and peers can move portfolio P&L far more than their individual sub‑1% weights suggest, because clinical and regulatory events are binary and uncorrelated with the macro tape. Jupiter appears comfortable with that, linking it to a broader thesis that the next decade’s alpha will accrue where compute and biology intersect.

For allocators, the signal is unambiguous: this is not a closet index, nor a balanced core fund. It is a concentrated expression of belief that AI, data‑center infrastructure, and high‑value therapeutics will dominate equity wealth creation. If that worldview holds, this 13F is the blueprint; if it cracks, this is exactly the kind of book that will feel it fastest.