On the money

Liquid gold. Black Magic. The Lifeblood of civilisation. Crude oil, the world’s foremost fuel since it overtook coal in the 1950s, is widely considered to be one of the most strategically important commodities in the world. With the vast majority of world oil deposits concentrated in a handful of countries, with extraction, refinement and distribution managed by just a few multinational corporations, the concentration of the worlds’ oil, and the global balance of power are inextricably interlinked. Indeed, the security of its’ energy supply has long been a key, if implicit objective of the Collective West’s security policy, having been one of the main factors in the relationship of developed nations with the Middle East (the late Henry Kissinger infamously remarked that oil was “much too important a commodity to be left in the hands of the Arabs” ). Even today, oil and its’ derivatives are a vital component in so many fields – from petroleum and diesel in transportation, to petrochemicals in plastics, even to the bitumen that makes up our roads, modern civilisation was shaped by humanity’s ingenious utilisation of oil.

But oil’s reign cannot last forever. Nations continue to announce new commitments to phase down or phase out their usage of fossil fuels, including oil. Non-petrochemical alternatives are emerging to plug the gaps in the market, from bioplastics to EVs. Whether you welcome the green transition with open arms or you view it as nothing more than an inconvenience, the decline of oil is logically inevitable, if we consider the following:

• As easily accessible oil reserves are depleted, companies must find new deposits of oil, which are harder to reach and more difficult to extract from.
• Greater difficulty of extraction results in higher costs of oil production. This will lead to a combination of rising prices of oil in order to meet demand, and (as a consequence of prices being higher than they otherwise would have been), suppressed demand
• This will lead to a stage where oil can no longer remain competitive against competitors – therefore, the oil markets’ size and influence will begin to diminish. The lifeblood of civilisation will gradually be siphoned out of the veins of the global economy

Of course, technical innovations will also come to the oil industry, which may improve access to some deposits and help slow the tide of falling supply. However, when we compare the already mature field of oil extraction to the fast-developing and innovative field of oil-replacement technologies, it seems clear that at some point ‘peak oil’ will be reached, if not out of our desire to protect the environment, then out of the cast-iron laws of supply and demand.

The Post-oil world

As we contemplate the inevitable decline of oil’s significance in the long term, let’s consider a logical question: what, if anything, will become the “next oil”? One obvious candidate springs to mind: semiconductors. Much like oil before it, semiconductors have grown to become a vital component of virtually every facet of the modern world – anything we use that relies on a computer (which is, in short, anything we use), utilises semiconductors – or specifically, the chips produced from semiconductors. The demand for these, unlike oil, shows no signs of abating – quite the opposite. Advances in fields such as generative AI, autonomous vehicle driving, augmented and virtual reality systems are dependent on their access to vast quantities of processing power – in other words, access to vast quantities of chips. And it’s not just ground-breaking fields that rely on semiconductors, with the consumer electronics sector constantly demanding faster, smaller, more energy-efficient chips to keep up with market expectations.

State control of oil

Firstly, let’s look at how production of oil is managed versus the development and production processes of microchips. The supply of oil is tightly controlled by a variety of supernational organisations and networks. The OPEC+ group, comprising 23 nations that account for roughly 60% of world oil production act as a cartel – during periods of lower demand they co-ordinate production cuts in order to boost oil prices. The sole nation that produces a significant proportion of the world’s oil, and is not a participant in the OPEC+ cartel is the USA. Whilst the effectiveness of OPEC+ as a group in maintaining high oil prices and maximising oil revenues of members is debatable (wide fluctuations in global oil prices since the foundation of OPEC in 1960, often unconnected with OPEC production quotas, show that its’ power over oil prices is far from all-pervasive, often being undermined by discord between members and a failure to fully enforce production quotas).

Owing to a combination of geographical distribution of oil deposits and a re-assertion of national sovereignty over natural resources in the post-colonial era, the vast majority of the worlds’ oil is under the management of state-run companies – 88% of confirmed deposits, to be precise. In contrast, a mere 6% of the world’s confirmed oil is claimed by major private oil companies. This level of state control of oil is crucial to its’ significance as a geopolitical tool – the budgets of several national economies, such as Saudi Arabia, Kuwait and the United Arab Emirates derive almost half their revenues from hydrocarbon profits.

The Diversity of Semiconductors and the role of private enterprise

We’ve established semiconductors are vital to the global economy, and their importance in the foreseeable future is only set to increase. The global supply of advanced semiconductors are also wielded as a foreign policy tool in geopolitical struggles – In recent years, the Biden administration has unveiled a series of export controls regulating the sale of high-end GPUs to China in an effort to check their rapid advancements in computing, amidst a broader deterioration of relations between the two countries. In this sense, chips, much like oil, have become a strategic resource that can be used by nations with an edge in their production in order to further their goals. However, it’s here that the analogy begins to break down. To start with, the semiconductor supply chain is overwhelmingly dominated by private industry. The largest state-run chip manufacturer, the Chinese Semiconductor Manufacturing Industry Corporation (SMIC), generated $7.2 billion last year, which whilst impressive, is barely a tenth of the revenue of the largest private producer, the Taiwanese Semiconductor Manufacturing Corporation (TSMC).

Additionally, the term ‘oil’, generally refers to just a few varieties of crude oil, which can all be refined into various derived substances. Oil is therefore treated as a commodity – we can accurately pinpoint its’ price by looking at futures markets, which five an accurate indication of ‘spot prices’ of reference barrels of oil that have a certain composition, like ‘Brent Crude’ (which indicates the price of oil drilled in the Atlantic basin, and sets prices for over 2/3 of crude oil traded worldwide). In contrast, the umbrella term of ‘semiconductors’ cover a vast array of products which each have specialised production processes, supply chains and uses, from the GPUs that are vital to hosting AI chatbots to the integrated circuits that form the motherboards of our computers. Different grades of microchips appeal to different markets, and competing semiconductor firms utilise a variety of different chip designs and production processes which result in their end products differing and evolving in a way that crude oil simply cannot match – Whilst, say, AMD might release a more advanced GPU with a 20% faster processing speed, it’s impossible for an oil giant like Saudi Aramco to improve the ‘quality’ of their crude oil so that it holds more energy, or is otherwise somehow of more value to consumers.

Oil Embargoes and Chip wars

As a result I believe semiconductors – even the most technically advanced GPUs that seem set to fuel AI-driven progress over the coming years – will neither be a decisive nor an effective geopolitical tool in restricting progress in the way that oil was. Let me explain:

• If one state or group of states choose to wield oil as a diplomatic tool , the changes in price are rapidly felt by suppliers and consumers alike. We see evidence of this if we look at the Oil crises of the 1970s – when Arab nations collectively embargoed oil exports to America, resulting in rapid rises in the price of oil which were immediately felt by consumers
• This is because oil is expendable – its’ value lies in its’ being burned. Therefore, if the supply of oil is restricted, and there are no significant reserves of oil held by non-producers (which when compared to the total demand for oil in the Western world, there aren’t), then buyers of oil will bid up the price of the limited supply that remains.
• Additionally, the use of oil to put pressure on nations is broadly effective against virtually any nation. Whether oil exporters place embargoes on select importers, or vice-versa (as we can see with the use of American sanctions limiting the sales of Iranian oil), every nation on earth is significantly dependent on oil use or production, and as such is sensitive to its’ price.
• In contrast to the binary nature of oil sanctions ( either restricting supply/demand, or not restricting supply/demand), policymakers are able to consider much more nuanced ways to implement controls on semiconductor technology.

One good example of this are the ongoing US export restrictions on advanced chip technologies to China. As a key emerging field, chip technology is something that America and the Western world is keen to retain an edge over China in, especially amidst heating geopolitical competition for dominance in economic, academic, and military fields. However, American policymakers knew that any form of ‘blanket ban’ on exporting chips to China would have been disastrous for all parties involved – China represented over 56% of the global chip market, over 80% of which are supplied by foreign firms – mainly the Taiwanese TSMC and a collection of major American manufacturers such as Qualcomm and Intel. With China making up the lion’s share of their revenues, restricting all semiconductors from being sold to China would have crippled the supply chains of the global tech sector, without hyperbole. Therefore, a more nuanced approach was needed – one that allowed Western companies to continue selling legacy chips to China, whilst limiting the access of Chinese state-controlled entities to more advanced models with applications in AI.

Therefore, in October 2022, the “New Export Controls on Advanced Computing and Semiconductors to China” were released. Whilst not impacting the sale of legacy chips, the regulations heavily restricted China’s access to advanced microchips. Although this did spur Chinese state-owned firms to massively accelerate development efforts of their own domestically produced advanced chips, the American export controls also hindered China’s access to production software to help design advanced chips, and placed restrictions on American citizens working in China which prevented them helping Chinese firms to manufacture semiconductors.

But this appears insufficient to hobble the Chinese chip industry. Revenues of major Western chip companies, such as TSMC and Intel have begun to decline in the wake of reduced markets for their most advanced products. Furthermore, China has begun actively subsidising domestic chip manufacturers, as conscious of its’ vulnerability to foreign sanctions, it seeks to accelerate its’ path to self-sufficiency. in October of 2023, Huawei announced the release of the Mate 60, a phone powered by a new domestically-produced chip. This chip is significant as its’ design utilises a 7nanometre node architecture (with this denoting the size of transistors on a chip – smaller transistors generally result in faster speeds, higher processing powers, and lower energy consumption), a feat hitherto thought infeasible without using advanced chip production machines that were banned from being sold to Chinese firms under US pressure (For reference, the company that produces these particular machines, ASML Holding, is based in the Netherlands, but still complies with US sanctions on exporting equipment to China). If true, this demonstrates the ability of a nation to pursue technical development of chip technologies even when cut off from the international infrastructure. Although there are doubts about the timeframe, there is a consensus that China will be able to continue independent development of its’ own processors, allowing it to remain a major player in the tech world, even if it does lag some 2 or 3 years behind its’ Western counterparts.

Therefore, I draw the following conclusions:

• Despite their superficial similarities with regards to their vital nature in a modern economy, semiconductors are not, and will not become ‘the new oil’ in the foreseeable future.
• This is because, whilst oil is a natural resource – meaning that an oil poor nation can only become oil rich with the (somewhat luck-based) discovery of new oil deposits, semconductors are a versatile manufactured good
• Although it is true that the manufacture of advanced new chips requires significant technical knowledge and advanced development, advanced economies are capable of investing into accelerated development of semiconductors, and increased manufacturing of them.
• This results in the long-term supply of semiconductors, which is managed by several private firms in competition (and some participation from state-managed Chinese firms), being significantly more flexible than the long-term supply of oil, which is largely managed by state-run firms, with hard limits set by mother nature rather than human ambition.
• Owing to the ability of various companies to make technological breakthroughs in semiconductor manufacturing, and continued
• Therefore, even the nuanced approach of restricting access to advanced semiconductors is not likely to benefit major world players today, achieving little more than temporarily stalling the progress of rivals whilst costing billions of dollars worth of future revenues.

Semiconductors, then, are unlikely ever to occupy the same strategic position that oil once held. Oil is a finite, consumable resource bestowed by geology; semiconductors are capital goods created through human ingenuity. An oil-poor nation can only become oil-rich by chance discovery. A chip-poor nation can, with enough investment, knowledge transfer and industrial focus, cultivate its own semiconductor capacity. Supply of oil is ultimately constrained by nature; supply of compute is constrained mainly by our ability to innovate.

This is why the analogy between chips and oil never quite holds. A more fitting comparison is not between oil and GPUs, but between the pumpjack and the GPU. The pumpjack was the enabling technology of its age – the human-made machine that allowed societies to unlock and scale an unprecedented source of energy. GPUs play a similar role today: the critical piece of infrastructure that allows computational power to scale to levels unthinkable only a decade ago. Where the old pumpjack unleashed a tide of oil, the modern GPU unleashes a tide of compute.

Both were transformative, but only one was a natural resource with inbuilt scarcity and immense geopolitical leverage. The other is a manufactured tool – an artefact of human design whose limits are set not by geology but by our own capacity to innovate. And this distinction matters, as semiconductors will define the trajectory of technological progress, but they will not become “the new oil”. Their strategic value will be profound, but never absolute – the key to the expendable resource that is computing power.

Even though it isn’t the ‘next oil’, with tens of billions of dollars of investment pouring into the sector, and dozens of nations aiming to position themselves as silicon superpowers, who knows? The microchip may become an even more important innovation, whilst not technically in the same class as oil. The future is as uncertain as it is exciting.

Thank you so much for reading through! I hope you found this article informative. Obviously, this article was particularly speculative, and goes somewhat against the ‘ flow of conventional wisdom, but I just felt it was an angle worth exploring. As usual, I appreciate any and all feedback, constructive criticism, and advice, so if you have ideas for future posts, or you think I could have done something better when writing this article, don’t hesitate to leave a comment! -Alex