Write-Up: Ask the Managers

SpaceX is the Archetype of the Big Winner Strategy

SpaceX already accounts for over 15% of the portfolio, and the author believes it exemplifies asymmetric returns. Slater's original quote is: "Returns are driven by a small number of big winners. Our job is to try to maximise the return from those big winners for our shareholders." The position is not immediately sellable; lock-up periods are lifted in tranches, allowing the fund to gradually reduce the stake to balance risk and return. Burns notes that SpaceX's core value lies in drastically reducing the cost to reach orbit: Falcon 9 has already cut launch costs per kilogram from $18,000–19,000 to around $2,700, while Starship aims to bring it down to $100–200. This low cost has enabled Starlink, which now generates annualized revenue of $17 billion, up nearly 70% year-over-year, with an operating margin close to 40%. Burns describes SpaceX as "in effect, a near monopoly on access to the rest of the universe." Although the valuation is high, the author argues that the multiple businesses within Starlink's scale are sufficient to support long-term returns.

AI Infrastructure Creates "Royalty-Like" Investments

Burns favors chip-layer companies — NVIDIA, TSMC, ASML, and SK hynix — arguing they sit at bottlenecks in the supply chain and can collect a royalty on AI development. He calls this "a royalty on the development of AI." The author judges that regardless of which model or application wins, these companies will benefit. For mature enterprises, technology alone is not enough; founder leadership, corporate culture, talent, and proprietary data are key. Burns cites Shopify for its early AI focus and Mercado Libre for leveraging transaction and financial data to expand into advertising and credit. Slater also emphasizes that the scale of AI investment must coexist with risk, but he challenges the traditional definition of risk: "Risk is not volatility. Volatility is inevitable. Risk is much more about permanent loss of capital."

AI Demand Far Exceeds Current Build Capacity

Slater believes AI computing demand will outstrip current infrastructure supply, and while the path is volatile, the direction is clear. High-frequency users are consuming more tokens, and the gap with median users is widening. Programming has already clearly generated economic value, and agentic models will further consume compute power. Google, Amazon, and Meta can fund their investments from their own cash flows, while other companies rely on debt, equity, or supplier credit. Slater argues that as long as technology continues to improve, compute demand will "outstrip what can be built today." Burns observes that over the past 18–24 months, corporate attitudes toward AI have shifted from experimentation to clear expectations — improving products, accelerating revenue, or reducing costs — with some already considering restructuring their organizations around AI models. He emphasizes that this is not just about headcount but "having different people in different roles." Slater adds that today's models are "the worst form of AI we'll ever have" and will continue to improve.

Healthcare, Brands, and Quantum Computing as Diversification Growth Levers

In healthcare, the report favors Moderna and Enveda, but remains cautious; Ferrari and Hermès provide scarcity-value hedges; PsiQuantum is a long-term bet on quantum advantage.

• Moderna: After the decline in COVID vaccine demand, the next drivers are flu vaccines and cancer treatments. Positive Phase 3 signals for melanoma are encouraging.
• Enveda: Uses AI to search for useful compounds in nature. The author reveals a selective approach in this area, as healthcare technology adoption is slow, but the potential returns justify the investment.
• Ferrari and Hermès: Burns notes they combine low volume growth, pricing power, consistently improving margins, and mature brands, offering diversification from a tech-heavy portfolio. "Having things that have scarcity value and that are physical could have even greater enduring value over the next 10 or 20 years."
• PsiQuantum: Quantum computing is still early, and the fund holds a small position. Slater borrows Hannah Fry's analogy, comparing classical computers to "a box full of accountants" and quantum computers to "a box full of toddlers," whose advantage lies in simulating atoms and cells — tasks that become too complex for classical computers as scale and precision increase.

Investment Implications

The article clearly outlines Scottish Mortgage's investment philosophy: betting on a few "big winners," accepting high valuations and volatility in exchange for long-term asymmetric returns. A bias in the institutional perspective is that the author holds positions in SpaceX (15% weighting), Starlink, and other private assets, so the discussion of their valuations and market prospects naturally carries a bullish tilt. Readers should be aware that the report does not fully address regulatory risks, technological bottlenecks, or the possibility of competitors (e.g., Chinese commercial space companies).

Position Moves