While this has seen many countries reverting to much more explicit industrial policies which were common in the three decades after World War II , which aim to secure supplies of raw materials and energy, rather less political attention is given to the structure and processes of overall value chain systems. Sadly this also comes against the backdrop of the post-WWII ‘rules based system’ breaking down, and being displaced by ‘the weaponization of everything’ and resource nationalism, most notably the supply of raw materials and intermediate goods and supplies.
THE PHYSICAL ECONOMY One simple fact that many are unaware of is that the physical economy of energy and commodities accounts for around 10% of global GDP, ca. $11.0-12.0 Trln. The surprise for many is how this effectively breaks down into four major components: Metals and Mining (ca. $1.2-1.3 Trln), Food Commodities (ca. $1.7-2.2 Trln), Hydrocarbons (ca. $1.6- 2.0 Trln), but Chemicals are by far the largest at ca. $6.5-6.8 Trln. They are all interdependent and as has been seen this year, disruption to the supply of key processing components such as helium, nitrogen or sulphur and other ‘reagents’ can create cascading shocks in downstream sectors. Poignantly the IEA's Global Critical Minerals Outlook 2026 observes that if China were to fully apply its recently expanded rare earth export controls, then around $6.5 Trln of downstream production would be at risk, and bear in mind that the total value of critical minerals production is estimated at just $200 Bln. The latter underlines the point that high tech engineering output often relies on just a few grams of materials such as tungsten or gallium. As was seen in 2025 in the auto sector, if supply of these is constrained and/or disrupted then output has to be cut or even paused completely.
FROM CRITICAL MINERALS TO FINISHED PRODUCTS Equally underappreciated is that all rare earth elements are byproducts, in other words they are not mined individually. Instead they are byproducts of other mining activity, for example the world’s largest rare earth deposit is at China’s Bayan Obo Iron Ore mine. They are also extracted in the processing of aluminium, nickel, uranium, or copper; as well as secondary ‘low value’ byproducts from the extraction of neodymium and praseodymium (NdPr) which are critical to magnet production, or even from phosphate fertilizer waste. The simple point is that they require a lot of complex, time- consuming and often environmentally ‘dirty’ processing to then be turned into what is needed to make the magnets, batteries, catalysts, alloys or even glass that are critical to the energy transition, high tech manufacturing and indeed AI data centres. On the one hand this highlights the real wonders and the complexities of modern chemistry, engineering and technology. On the other it underlines that planning frameworks have to include the complete bill of all materials required to go from raw inputs to finished products, in other words security of supply is as much about critical minerals as the materials needed to process them. Suffice it also to say that substituting a geopolitical rival with an alternative supplier in a ‘friendlier’ jurisdiction where there are governance issues (legal, ethical, environmental, working conditions, etc) does not improve supply chain security, it merely shifts geographical risk, above all in a world plagued by trade tensions and tariffs, and resource nationalism.
SUBSTITUTING A GEOPOLITICAL RIVAL WITH AN ALTERNATIVE SUPPLIER IN A 'FRIENDLIER' JURISDICTION DOES NOT IMPROVE SUPPLY CHAIN SECURITY, IT MERELY SHIFTS GEOGRAPHICAL RISK.
INFRASTRUCTURE: THE NEXT STRATEGIC CHALLENGE While understanding the complexity of these production supply chains is important, there are many other factors that need to be considered. Infrastructure is perhaps the most critical of these. This can for example be seen in the shift to renewable energy, with a rapid expansion of wind and solar production not only confronted by its inherent intermittency in many countries, given a wide divergence between peak production and peak demand periods. By extension it highlights the need for sophisticated BESS (Battery Energy Storage Systems), and a consideration of the challenges of grid distribution in geographical terms, given that power does not travel well (i.e. the potential for a lot of wasted production, per se a potential major inefficiency). That is a particular challenge for AI data centres, which de facto require reliable 24/7 ‘always-on’ power supply, which creates a planning mismatch given that a data centre can be constructed in 1-2 years, but the power supply infrastructure will typically take 5-7 years to construct. In respect of the latter, it also has to be remembered that nuclear power requires large volumes of fresh water for cooling, and as has been seen this summer in Europe (above all Hungary and Romania), water scarcity (in this case due to lower water levels in the Danube) can result in power production being curtailed or even shut down.
24 | ADMISI - The Ghost In The Machine | Q3 Edition 2026
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34
