Baillie Gifford says emerging-market buyers face pricier AI chip sourcing

The consensus view holds that growing AI chip demand offers a broad tailwind for emerging markets, allowing procurement teams to continue sourcing low-cost components. This perspective overlooks a critical shift: the expanding need for specialized manufacturing capabilities. The procurement landscape is rapidly evolving from simple component acquisition to strategic vertical integration.

That matters because the letter’s most useful signal is not simply that AI demand helped emerging-market performance. It is that AI-related semiconductor demand is being treated as a broad emerging-market tailwind without a public account of who captures the margin when chips, packaging, power systems, and supplier access become more specialized.

The portfolio signal is clear; the procurement path is not Baillie Gifford reports “a measured end to 2025 following strong year-to-date performance, driven by AI-related semiconductor demand and a resurgence in commodities,” according to the source summary of the investor letter. That sentence carries the whole story available in the public packet: AI chips and commodities are grouped as performance drivers, but the source does not specify whether emerging-market gains came from foundry exposure, design exposure, assembly exposure, memory exposure, materials exposure, or downstream buyers spending more on imported components.

For executives, that omission is not academic. A chief procurement officer does not buy “AI-related semiconductor demand”; she negotiates supply allocations, lead times, design commitments, and alternative sources. A chief financial officer does not model a theme; he models whether specialized inputs move from a spot purchase to a multi-year supplier dependency. The letter gives investors a macro driver, but it does not give operators a procurement map.

The rejected read is that emerging markets can ride demand unchanged The easy read is that emerging markets keep benefiting from AI chip consumption through their existing roles: components, assembly, commodity inputs, downstream integration, and local deployment of imported systems. That view is attractive because it lets procurement teams treat AI hardware demand as a volume problem. If more chips are needed, buyers search harder, diversify suppliers, and absorb some price pressure.

The counter-read is that AI semiconductor demand does not behave like generic electronics demand once the scarce input is specialized manufacturing capacity rather than the finished component alone.

If the constraint shifts toward advanced process access, packaging relationships, design-specific optimization, or privileged supplier allocation, the buyer that merely imports off-the-shelf parts is no longer buying at the center of the margin pool. The margin begins moving toward companies and markets that can coordinate design, manufacturing, packaging, and customer qualification more tightly.

The letter’s missing baseline makes the performance claim hard to price The letter’s reported phrase “strong year-to-date performance” raises the first diligence question: measured against what baseline? The public packet does not provide the index comparison, sector attribution, valuation starting point, or hardware exposure needed to separate AI semiconductor demand from the commodities rebound Baillie Gifford also cites.

Without that split, an executive cannot tell whether the market is rewarding durable AI manufacturing positioning or simply a cyclical basket of chip-adjacent and commodity-sensitive names.

That is the limitation of the available evidence. This is not a peer-reviewed research claim, a regulator filing, or a company-level disclosure; it is an investor letter summarized in a single-source packet.

It supports the narrow statement that Baillie Gifford identified AI-related semiconductor demand as one driver of emerging-market performance, but it does not prove that emerging markets are already building vertically integrated AI chip capacity or that procurement margins have already shifted.

The work moves from purchasing to supplier architecture

If the thesis proves right, the work affected inside emerging-market companies is not just engineering. It is procurement, finance, legal, and operations. Buyers that once competed on landed cost for standardized components will need people who can evaluate supplier concentration, qualify manufacturing partners earlier in the design cycle, and negotiate access to specialized capacity before product demand is certain.

That changes the labor mix around AI infrastructure. Procurement teams become more technical; legal teams spend more time on supply commitments and cross-border restrictions; finance teams model capacity commitments that look less like ordinary purchasing and more like strategic exposure.

The under-noticed middle is the systems integrator, electronics manufacturer, or industrial buyer that is large enough to need AI-class hardware but too small to command preferred allocation from the most constrained suppliers.

The skeptic’s case is that this is still just a fund letter The obvious objection is that one investor letter cannot carry a claim about a procurement-margin shift. The source does not provide company-level capital plans, customs data, government incentives, supplier contracts, or customer backlogs. It also pairs semiconductor demand with commodities, which means the reported emerging-market performance could be explained by a broader rebound rather than a structural change in AI hardware sourcing.

That objection should constrain the conclusion. The best reading is not that vertical integration is inevitable, or that every emerging market should build domestic chip capacity.

The defensible reading is narrower: when a portfolio manager cites AI-related semiconductor demand as an emerging-market driver but leaves value capture unspecified, operators should ask whether their own exposure sits in high-margin control points or in low-margin procurement channels that become more expensive as specialization rises.

The falsifiable test is capacity behavior, not AI rhetoric The thesis to test is simple: within 24 months, increasing AI-driven semiconductor demand will shift emerging-market procurement margins from low-cost, off-the-shelf components to vertically integrated, specialized manufacturing capabilities. It would be wrong if Q3 2026 company reports show no meaningful increase in advanced manufacturing or AI-specific chip investment, if Q4 2026 customs data show sustained reliance on imported advanced AI semiconductors without matching domestic production effort, or if Q1 2027 government economic reports show no new incentives or investments tied to vertical integration.

Until those signals appear, the Baillie Gifford letter should be treated as a useful marker, not proof. Its omission is the story: AI-related semiconductor demand may be lifting emerging-market performance, but the public source does not say whether the value is accruing to buyers of components, makers of components, or the narrower set of companies able to secure specialized manufacturing control.

For executives planning AI infrastructure, that is the difference between a procurement cycle and a margin-structure shift.