China’s AI hardware boom is running into a constraint that is almost the mirror image of weakness elsewhere in the Chinese economy. Demand is not the problem. The problem is that suppliers of optical modules, AI servers, printed circuit boards and related components are struggling to scale fast enough to satisfy orders from cloud providers, data-center builders and AI infrastructure customers.

Capacity constraints and shortages of critical inputs could limit growth through 2026, even as investors continue to treat the sector as one of the strongest parts of China’s equity market. The clearest pressure point is optical communications. AI data centers do not depend only on processors. They also require vast numbers of optical modules, high-speed interconnects, cables, chips, boards and power components to move data between servers and accelerators.

Germany’s gas-storage warning shows that Europe’s energy-security problem is not over; it has merely shifted from emergency winter survival to the harder question of whether markets will refill storage fast enough before the next cold season. Germany could face supply risks next winter even if storage sites are filled to their currently booked level of 76%.

In a severe cold-weather scenario based on 2010 temperatures, there could be a 20 terawatt-hour supply deficit across January, February and March 2027, with more than 35% of gas consumption potentially uncovered on individual days.

The Energy Department’s proposed financing for long-lead nuclear components is an attempt to solve one of the least glamorous but most decisive bottlenecks in America’s nuclear revival: the supply chain. DOE is considering billions of dollars in financing for utilities to secure items such as reactor vessels and steam generators for large AP1000 reactors, components that can take years to procure and whose delay can slow an entire project before construction even begins.

The plan would help several U.S. utilities interested in AP1000 deployment. The idea is important because it changes the sequence of nuclear development. Traditionally, utilities wait until a project is more mature before ordering the largest and most expensive components. But that can leave them trapped in long manufacturing queues once they are ready to build.

The three U.S. LNG cargoes now heading toward Tianjin are a small but politically important test of whether energy can become one of the easier deliverables from this week’s Trump-Xi summit. The vessels Umm Al Hanaya, Al Sailiya and Id’Asah left LNG export plants in Louisiana last week and are expected to arrive in China between June 15 and 20. If any of them discharge there, they would mark the first direct U.S.-to-China LNG delivery since February 2025.

The shipment matters because U.S.-China LNG trade has not disappeared commercially, but it has been politically rerouted. Chinese buyers such as PetroChina and CNOOC still hold long-term contracts with U.S. LNG producers, but over the past year they have largely resold those cargoes into Europe rather than bring them home and pay China’s 25% tariff on U.S. LNG.

India’s record DAP tender shows how the Iran war is moving from energy markets into the core economics of food production. New Delhi is set to import roughly 1.35 million metric tons of diammonium phosphate in a single purchase, equal to about a quarter of its annual DAP imports, after supply disruptions and higher shipping costs tightened the global fertilizer market.

Indian Potash Ltd agreed to buy 765,000 tons for the west coast at $930 per ton on a cost-and-freight basis and 581,500 tons for the east coast at $935 per ton. Argus put the final secured volume at 1.3465 million tons, with letters of intent issued to 14 suppliers.

The rare-earth truce being discussed by Trump and Xi should not be mistaken for a full reopening of Chinese supply. The more important story is that China has learned how to preserve leverage even while appearing to de-escalate.

Although headline rare-earth exports have almost recovered over the past year, shipments of the heavy rare earths yttrium, dysprosium and terbium remain down by around 50% compared with the 12 months before China imposed controls in April 2025. These are not bulk commodities. They move in small volumes, but they are essential to aerospace, defense, semiconductors, electronics and high-performance magnets.

China’s industrial policy has moved beyond the familiar “Made in China 2025” playbook and into something broader, more systematic and harder for advanced economies to counter. Beijing is no longer concentrating only on a few strategic sectors such as electric vehicles, batteries or solar panels. It is expanding state direction across mature industries, foundational supply-chain layers, services and frontier technologies, creating an “industrial policy of everything.”

Up to $650 billion in G7 manufacturing exports could be directly exposed to Chinese market-share gains by 2030 if current trends continue. That is equivalent to roughly 12% of G7 manufacturing exports, with automotive, machinery and chemicals identified as among the most vulnerable sectors.

Britain’s move toward full nationalisation of British Steel marks a major reversal in industrial policy and a recognition that primary steelmaking is now being treated as a strategic asset rather than an ordinary commercial business. Prime Minister Keir Starmer said the government will introduce legislation giving it the option to take British Steel into public ownership, subject to a public-interest test based on national security, critical infrastructure and economic resilience.

The government says the powers are intended to safeguard steelmaking capacity at Scunthorpe and avoid a sudden halt in production while ministers consider the company’s long-term future. The immediate trigger is the failure to find a private buyer on terms the government considered acceptable. British Steel is still owned economically by China’s Jingye, but the UK seized operational control in April 2025 to prevent the shutdown of Scunthorpe’s blast furnaces.