May the fourth be with you
In economics, the wealth of a nation is built and sustained through production, and that production requires three ingredients: Land, Labour, and Capital. These are the finite building blocks of prosperity. Land provides the raw materials, Labour provides the muscle and the mind to transform them, and Capital represents the tools.
The primary challenge for every society has always been the efficient allocation of, and between, these inherently scarce resources. If you run out of one, growth grinds to a halt. You need all three. That was, it seems, up until recently, when it was proposed a fourth ingredient exists, and today this emerging thesis is inspiring stock market bulls.
That fourth factor of production, according to some economists, is data.
The transformative aspect of this fourth ingredient is that, unlike the physical limitations of land or the finite hours of the human workforce, data is a resource that is functionally unlimited. And importantly, it’s the only factor of production that actually grows more abundant the more we use it.
As a production resource data was often ignored by traditional economic models because, while it existed, it was “dirty”. In other words, it was expensive to extract, difficult to refine, and nearly impossible to use at scale.
In April 1964, the IBM mainframe first brought data processing into the corporate world, but it took almost 60 years for the “Digital Revolution” to evolve beyond merely counting things faster. It wasn’t until the arrival of OpenAI and artificial intelligence that economists can safely say the ‘refinery’ for this new resource was finished.
Notwithstanding its limits and arguments about whether it actually thinks, artificial intelligence (AI) today appears to be an engine that turns raw information into a productive force. One that can boost the efficiency of each of the other ingredients, making workers smarter and capital more effective.
Perhaps without articulating awareness of a fourth resource, enterprise is now, understandably, pouring billions into it, giving life to what was formerly a research paper theory.
In the early days of the computer age, high-tech investments represented only about 20 per cent of what businesses spent on equipment and infrastructure. Today, according to economic researchers like Yardeni Research, that figure has reached a staggering and record-breaking 55 per cent.
While the adoption and embrace of this new way of looking at AI – as a fourth force in production – won’t make the market immune to setbacks, it could very well mean the current stock market rally isn’t just a speculative bubble driven by hype, but a reflection of a structural rebuilding of the global economy.
Companies investing in all layers of AI might unwittingly be betting the ability to process data becomes more valuable than land, labour and capital. If that’s true, demand for the physical hardware required to run the new digital world – specifically, the semiconductors and memory chips that act as the brains and library of AI – is headed way higher.
The logic is that because every interaction with AI generates more data, which in turn requires more memory and processing power to analyse, what’s been created is a self-reinforcing growth loop.
This could be the story currently being reflected in the stock market, particularly in the soaring valuations of semiconductor and memory stocks.
Not everyone subscribes to the theory. Famed short seller Michael Burry last week publicised a new short position through put options on the SOXX semiconductor index expiring in early 2027.
Burry, however, is betting against a broadening theme. For a long time, the gains were concentrated in a handful of massive tech giants. Recently, however, the optimism has spread to the Russell 2000, which tracks smaller, more domestically focused companies. One reading of small and mid-sized technology firms outperforming their larger counterparts suggests the AI trade is trickling down. It’s no longer just about Silicon Valley; it is about a regional manufacturer in the Midwest or a services firm in the South using these tools to find efficiencies that were previously hidden in their data.
The alternative view is one I’ve observed during past bubbles. When the leaders have risen too far, the ‘also-rans’ are bid up amid fear of missing out by those who have indeed missed out. It’s the reason you always see laggards ‘catching up’ to the leaders. After 50 races in which the Ferrari beats the VW Kombi, eventually, people bet the Kombi will have a win.
Nevertheless, despite persistent tensions in the Middle East and the looming shadows of inflation and unsustainable debt in the U.S., the American economy is currently growing at 3.7 per cent. This is a remarkable figure for a developed economy, especially one facing the highest interest rates in decades.
Meanwhile, the U.S. labour market is also resilient, with hiring comfortably exceeding the number of people leaving their roles, and Redbook retail sales, which track the pulse of actual store registers, growing by 7.8 per cent.
Of course, the bulls aren’t without their worries. High levels of optimism can quickly morph into irrational exuberance, and once everyone agrees that the future is bright, there are fewer left to buy and bid up prices. And 2027 harbours geopolitical risk from China and Taiwan – that latter being the global hub for the very semiconductor chips that make the Fourth-Production-Factor theory possible. If that supply chain were to be disrupted, the AI trade would likely face a reckoning.
Ultimately, the bull argument rests on the idea that AI has triggered a historic economic transformation, one that holds that data itself changes the rules of scarcity. If a company can use AI to do more with less – less labour, less physical capital, and less money – then the potential for profit and productivity growth is higher than at any point since the Industrial Revolution.