The Siren of Novgorod

Has globalisation ended?

Goods, capital, technology and production still move across borders, it is accounted for 46.3 per cent of world trade in the latest WTO assessment [1], close to its recent peak. Global value chains remain embedded in the world economy, what is ending is the assumption that companies can organise themselves primarily around cost, scale and efficiency while treating geopolitics as an external risk.

The global economy is now moving towards negotiated access. Governments are using tariffs, subsidies, procurement rules, export controls, investment conditions and security partnerships to determine who may participate and on what terms. We will understand some changes in the operating system of global business below.

America: local investment as political insurance

Apple’s five-year agreement to buy $30 billion of chips from Broadcom is a targeted commitment in one of the highest-value and most politically sensitive parts of its production system. The agreement is expected to support the production of 15 billion chips in the United States and a $1.5 billion investment in Broadcom’s Colorado facility. It also sits within Apple’s wider plan to spend $600 billion in America after President Donald Trump threatened tariffs on iPhones unless the company moved more manufacturing out of Asia. [2]

The most revealing detail is that Apple has not attempted to recreate its entire Asian production ecosystem in the United States. Such a move would be expensive, disruptive and unrealistic. Instead, Apple is localising selected high-value components while preserving the wider global network on which its business still depends.

By placing visible investment in the United States, Apple can demonstrate support for American manufacturing, strengthen its relationship with the government and reduce its exposure to future tariff threats, without abandoning the efficiencies of its international supply chain. The wider investment pattern supports this interpretation: foreign investment into the United States reached $232 billion in 2025, its highest level since 2021, and manufacturing accounted for more than half of the inflows. [3]

The commercial value of geography is therefore changing. A factory in America offers more than proximity to US consumers. It may provide tariff protection, access to subsidies and procurement, political goodwill and some insulation from restrictions imposed on imports. This move creates a political location premium, meaning the value of an industrial asset increasingly depends on the political protection associated with its location.

Yet localisation is not automatically efficient: companies investing in US manufacturing face shortages of skilled labour, pressure on electricity networks, ageing infrastructure and tariffs on imported machinery needed to build new plants. Major industrial investments can take six to eight years before production begins, while tariff and subsidy policies may change several times during that period.

The contradiction is clear: Washington may attract more factories while simultaneously making them more expensive to build and more dependent on political discretion.

Europe: turning public spending into geopolitical power

Europe is pursuing a similar objective through a different instrument: public demand.

The proposed Industrial Accelerator Act would introduce European-content requirements in strategic sectors including clean technology, electric vehicles, steel, aluminium and cement [4]. A separate procurement overhaul would allow public authorities to favour EU companies and suppliers from eligible partner countries. Public procurement represents roughly 15 per cent of EU GDP, so if used strategically, it becomes one of the most powerful industrial-policy tools available to European governments. [5]

So now, the doctrine is seeking suppliers who can strengthen European economic security. The concern is valid. Europe imports around 94% of the solar modules and cells it uses from China, alongside roughly half of its batteries [6]. The proposed strategy therefore seeks to rebuild domestic industrial capacity and reduce dependence on external suppliers in sectors considered strategically important.

“Buy European” is more than about nationality. Access to European demand may increasingly depend on whether a company creates jobs, production, research, technology and taxable value inside Europe. European authorities may also be starting to ask a foreign manufacturer whether it transfers useful knowledge, strengthens local capabilities and reduces reliance on a geopolitical competitor. So, it will not only be about price and technology competitiveness now. For Chinese companies, this policy could make exporting finished products into Europe less attractive than establishing joint ventures, employing European workers, conducting local research and sharing more of the production process.

However, local-content rules may preserve factories and technical capabilities, but they may also raise prices, reduce supplier choice and slow investment [6]. This is particularly important for clean energy because Chinese scale has helped reduce the cost of solar panels, batteries and other transition technologies. Reducing that dependence may strengthen Europe’s long-term industrial position while increasing the near-term cost of decarbonisation.

Europe can reserve demand for European firms, but can those firms become competitive enough to survive without permanent protection? Procurement preferences unfortunately cannot indefinitely compensate for expensive energy, slow permitting, inadequate finance or shortages of skilled workers.

China: from disruption to displacement

China’s rare-earth restrictions reveal a more forceful mechanism of conditional globalisation.

China controls more than 90% of global production capacity for yttrium oxide, a material used in electronics, advanced ceramics and semiconductor-related manufacturing. Following Beijing’s export restrictions, Japanese manufacturers faced reduced access while Chinese companies moved to replace their Japanese suppliers with domestic alternatives [7]. The resulting price gap was extraordinary. In July 2026, yttrium oxide was quoted at $7.88 per kilogram in China and $1,175 in Europe. Shares in several Chinese producers of zirconia- and yttrium-based products rose sharply as investors anticipated that the restrictions would improve their competitive position [7].

However, by restricting an upstream input, China can raise costs for foreign downstream producers while preserving cheaper access for domestic manufacturers. Those Chinese companies can then expand into more advanced products previously supplied by foreign competitors. Smart move, right? Simply, this is the logic:

• A foreign company becomes dependent on a Chinese input;
• China restricts that input;
• the foreign company’s costs rise or production becomes uncertain;
• Chinese competitors retain more favourable access;
• buyers begin sourcing the finished product from China instead.

So, the potential long-term plan from that “shortage” is competitive displacement.

Western governments often respond by announcing new mines, but mining is only one stage of the system. China’s advantage also lies in separation, refining, materials science, magnet production, specialist manufacturing, infrastructure and accumulated industrial knowledge. Even where North America possesses the underlying mineral deposits, reducing Chinese dominance in processing and advanced components will require long investment horizons and substantial government support [8]. The strategic prize may belong to those controlling the difficult middle area of the value chain: processing, specialised materials, recycling, magnets, ceramics and components.

Technology access now follows alliances

The US decision to relax export controls on advanced chips, drones and other technologies for the UAE shows that access to strategic technology is becoming conditional on political alignment.

Approved Emirati entities would be able to receive licence-free access to advanced AI chips and servers, unmanned systems, selected satellites and spacecraft, and dual-use technologies used in oil, gas and nuclear industries. Washington linked the decision to the UAE’s status as a major defence partner and its support for US national-security objectives [9]. This is big because the United States had previously restricted advanced technology exports to the UAE over concerns that sensitive capabilities could leak to China.

The policy shows a tension in logic: Washington wants to prevent strategic rivals from accessing American technology, but it also wants American chips, cloud systems and AI platforms to dominate global infrastructure [10]. So it seems they want to create licensed integration. Countries and companies may receive different levels of access depending on security relationships, export-control compliance, ownership structures, data governance, exposure to Chinese technology and willingness to operate within American systems, which will create a hierarchy in the global market.

For countries and companies, this will go far beyond semiconductor procurement because access to advanced computing will shape cloud capacity, AI development, biotechnology, defence production and future productivity growth. Therefore, export-control compliance can no longer be left to a specialist legal team after an investment decision has been made, because it is becoming part of corporate and state strategy.

Canada and the limits of diversification

Canada provides the clearest warning that geopolitical diversification is easier to announce than to implement.

Under pressure from Washington, Ottawa has promoted both deeper North American integration and stronger trade with Europe, Asia and the Gulf. Its proposed “Fortress North America” framework reflects the continued importance of the USMCA (United States-Mexico-Canada Agreement) and the reality that Canadian industries, especially automotive manufacturing, remain connected to the United States. [11]

The Gordie Howe International Bridge illustrates both the strength and vulnerability of that integration. The Detroit–Windsor corridor handles more than four million truck crossings, and almost $70 billion in annual trade, yet the bridge’s opening was delayed while operating and revenue arrangements were renegotiated [12]. At the same time, Canada has sought capital beyond the United States, but a C$70 billion Emirati investment commitment remained undeployed because Canada lacked enough projects ready to receive it [13]. Regulatory approvals, legal arrangements, consultations and financing questions continued to hold back the pipeline. Canada’s decision to select Germany’s TKMS (ThyssenKrupp Marine Systems) for its largest defence procurement represents a more tangible form of diversification. The submarine project was presented as a way to strengthen Canada’s industrial base and connect Canadian companies with European supply chains [14].

Diplomatic commitments and investment announcements are not enough. Diversification requires mature enough infrastructure, finance, competitive firms, regulatory capacity and projects to absorb capital. Maybe, Canada cannot quickly replace the gravitational pull of the United States. Europe cannot instantly replace Chinese clean-technology supply chains. The US cannot rebuild decades of lost industrial skills simply by announcing tariffs.

The macroeconomic cost

This “conditional” trading system may be more resilient in selected areas, but will almost certainly be less efficient overall.

Duplicated factories require more capital. Parallel regional supply chains sacrifice economies of scale. Higher inventories tie up working capital. Separate technology ecosystems increase development and compliance costs. Local-content rules may force companies to purchase more expensive inputs. Export controls can interrupt production even when global supplies remain physically available. The OECD estimates that broad attempts to relocalise supply chains could reduce global trade by more than 18 per cent and global real GDP by over 5 per cent. Its modelling also finds that localisation will increase economic volatility in more than half of the economies examined. [15]

This does not mean governments should ignore strategic dependencies, but self-sufficiency is not the same as resilience. A fully domestic supply chain may still be vulnerable to local disasters, infrastructure failures, labour shortages or political mistakes. Resilience is more likely to come from diversified, transparent and substitutable supply networks than from attempting to produce everything at home.

Globalisation is therefore being rerouted; US imports from China fell 29 per cent in 2025, while Chinese exports were redirected towards Asia, Africa and Latin America [16]. The IMF distinguishes between genuine trade reallocation and trade rerouting. Trade reallocation is where third countries develop production that replaces Chinese exports, and trade rerouting is where goods are largely transshipped through intermediary economies [17]. This means the new world economy is more likely to contain regional production systems, strategic bilateral corridors, intermediary economies and multinational firms operating parallel supply chains under different political rules.

Capital may also become more concentrated; global foreign direct investment rose 6 per cent to $1.6 trillion in 2025, but UNCTAD described the recovery as narrow, fragile and uneven [18]. Large markets with fiscal capacity, infrastructure and strong alliances are best placed to attract duplicated production. Smaller economies risk being bypassed unless they offer something difficult to replace like minerals, energy, logistics, technology or access to a valuable regional market.

So, this is the paradox: governments are trying to reduce national vulnerability, but if every government pursues resilience through duplication, tariffs and exclusion, the collective result may be slower growth, higher prices and greater sensitivity to geopolitical shocks.

What conditional globalisation means for industries

• Semiconductors and AI: US-based production and trusted partners may gain privileged demand and technology access. But high capital costs, power constraints, labour shortages and shifting export controls make political alignment part of computing strategy.
• Automotive and batteries: Production will become more regional. Different rules in the US, Europe and China will require separate supplier networks, increasing engineering, certification and compliance costs.
• Clean energy: Localisation may improve security but raise costs and slow deployment. The challenge is to build domestic capacity without delaying decarbonisation.
• Critical minerals: The greatest value lies in processing, refining, materials, magnets, recycling and components. China’s advantage comes from controlling much of this chain.
• Defence and infrastructure: Procurement is increasingly used to strengthen allied industries. Ports, bridges, pipelines, cables and logistics hubs can also become geopolitical bargaining assets.
• Pharmaceuticals and regulated sectors: Domestic pricing, procurement and regulation are becoming tools of international pressure. Healthcare, technology and telecommunications firms should expect greater overlap between commercial rules and geopolitical coercion.

What companies should do now

Companies should respond selectively rather than attempting expensive and unrealistic full-scale localisation:

• Map political chokepoints: identify where governments or concentrated suppliers can interrupt minerals, technology licences, cloud services, data transfers, transport routes, finance and procurement eligibility.
• Build regional redundancy selectively: focus on the limited number of components, technologies and facilities whose loss would halt production or remove access to a major market.
• Treat government relations as an operating capability: geopolitical, legal, tax, procurement and supply-chain teams should be involved before investments are approved.
• Use scenario-based capital allocation: test major projects against managed interdependence, sharper bloc formation and selective diplomatic stabilisation.
• Avoid performative localisation: a politically attractive factory without energy, infrastructure, skilled labour or long-term demand may become stranded or permanently subsidy-dependent.

---

Globalisation is not disappearing. The factories, ships, capital flows and technologies connecting the world are too deeply embedded to unwind quickly, but the terms of connection are changing.

Governments are increasingly deciding which products may cross borders, which companies may win public contracts, which countries may access advanced technology and where foreign investment must create domestic value. For companies, the question is no longer whether to be global or local. It is how to remain global in a world where every major market increasingly demands some combination of localisation, loyalty and leverage.

The winners will not necessarily be the companies with the cheapest supply chains, but the ones who are able to operate across political blocs without becoming dangerously dependent on any one of them.