When Things Start to Price Money: From Copper to Compute
Updated: Oct 6
01/10/2026
When Things Start to Price Money
With the global bond sell-off deepening, this seems a reasonable moment to think about the value of money from the other side of the transaction.
We are accustomed to money determining the value of things – are we now entering a period in which things increasingly determine the value of money?
Global debt stands at more than three times annual global GDP and is growing. The debasement argument is well understood – more interesting is what all that money is competing to acquire.
Defence, AI and strategic autonomy may look like different priorities, but increasingly they require many of the same things: energy, minerals, infrastructure, technology, talent and security. Financial capacity remains the most flexible way of acquiring those things when you don't already possess them, but money can be created and mobilised considerably faster than most of the capabilities it is being used to pursue.
For those with sufficient financial capacity, the constraint therefore begins to move from money towards the things money is attempting to acquire. When those things become difficult to replicate or substitute, their value is no longer expressed purely through price.
The return of the bargain
Money increasingly doesn't describe the whole transaction.
China has been particularly effective at constructing bargains over recent decades, while America has more recently begun repackaging its own offering. The new proposition being “give me something I cannot efficiently create, and I will give you something you cannot readily access."
This isn't necessarily a retreat from globalisation - more an adaptation. Nor indeed does resilience necessarily mean independence; it may instead be achieved by spreading dependencies.
Governments have already provided a reasonable map of what they consider strategically important. Critical-minerals lists identify the obvious physical resources along with export controls and industrial policy.
China has gone furthest in demonstrating the leverage this can create, but importantly that leverage does not always come from owning the resource. Processing capacity, infrastructure, technology and scale can matter as much as geology.
Other governments are now spending heavily to bring more of those capabilities within their own spheres of influence.
Financial markets inevitably follow these trends, providing the capital and thus price signals that ‘may’ eventually relieve scarcity. Also creating new ways of gaining exposures: equities, futures, indices, royalties, financing structures and claims on future revenues.
A year ago, I argued that copper was perhaps the clearest market through which to express this collision between financial abundance and physical constraint:
Copper still fits that argument – albeit with this year’s rally maturing/stalling.
The same tension is becoming visible across a much broader collection of resources and capabilities, and increasingly some of them may not be conventional commodities at all.
What comes after critical minerals?
Compute provides an interesting glimpse of what may come next.
Historically, investors wanting exposure to the growth of computing power largely had to own the companies producing the components, building the data centres or selling the services. Compute is now starting to separate economically from the hardware that produces it.
A GPU can increasingly be thought of as a productive asset, electricity as its feedstock, the data centre as its infrastructure.
Rental benchmarks already express output in dollars per GPU-hour, while the soon to launch futures market seeks to make productive capacity hedgeable independently of the hardware itself.
Resources → Energy → Chips → Data centres → Compute
Compute doesn't even need to become a conventional commodity for some commodity-like characteristics to emerge.
Different generations, locations and qualities of compute will have different values, while forward markets can begin to express the relationship between the cost of the inputs and the value of the output.
That could have implications well beyond the compute market itself. Much of the capital value being assigned to AI infrastructure today rests on expected future demand and revenues. If the enormous investment now being made in capacity ultimately drives the price of compute materially lower, the price of the output may begin to tell us something about the value of the assets - and capital - being used to produce it.
The significance may extend beyond the creation of another tradeable market. If compute becomes strategically important in its own right, control of productive capacity potentially creates leverage in much the same way that control of energy, processing or other critical infrastructure already does.
A company deploying critical infrastructure in a geopolitically complicated region need not rely solely on the traditional security umbrella of an allied government. Wider dependencies can be created across political boundaries that give multiple parties an interest in the asset's continued operation.
The deterrence is subtly different: less “if you destroy this, we will punish you”, and more “if you destroy this, you lose access to something you need”.
The above distinction is important, since having financial exposure to an asset does not necessarily give you an interest in preserving it. Indeed, if compute becomes independently tradeable, someone positioned for higher prices might actually benefit from capacity being removed. The more powerful relationship is therefore not ownership itself, but dependence on continued access.
Indispensability creates leverage
This raises an interesting question about where some of that leverage ultimately resides. Twentieth-century dystopian fiction gave us corporations such as Tyrell, Weyland-Yutani and OCP, whose power came not simply from their wealth but from controlling capabilities on which governments and societies had become dependent.
The comparison need not be dystopian. Corporations do not need to replace governments to acquire some of this leverage. They only need to control something that governments, economies or other corporations increasingly cannot do without.
None of this necessarily ends badly since greater interdependence can create incentives for cooperation just as easily as it creates leverage.
All this is to suggest that control of scarce capability is becoming a more important source of economic and geopolitical power, whether that capability sits on a sovereign or corporate balance sheet.
Governments will seek to secure / corporations will seek to control / financial markets will seek to create claims.
All three are ultimately competing for capabilities whose supply can often expand much less readily than the financial resources being deployed to acquire them.
The terms of exchange are changing.
Important Disclaimer
This article is intended for general information purposes only and reflects the market environment at the time of writing. It does not constitute investment advice, a personal recommendation, or an offer to engage in any trading activity. The content does not take into account individual objectives or circumstances and should not be relied upon as the basis for any investment decision. Past performance is not a reliable indicator of future results.
Content may have been created by persons who have, have previously had, or may in the future have personal interests in securities or other financial products referred to therein. All conflicts and potential conflicts relating to our business are managed in accordance with our conflicts of interest policy. For more information, please refer to our Summary of Conflicts of Interest Policy.
For more information and important risk disclosures, please see our Derivative Product Trading Notes and Privacy Policy. AMT Futures Limited is authorised and regulated by the Financial Conduct Authority.
Comments