The relative performance of the world’s major stock markets over the past year can be explained by their exposure to AI. Nations with a large foothold in the “stack” of industries developing AI infrastructure and services are massively outperforming, while those without are lagging by record margins.
The winners include the US and China, thanks above all to their foundational AI models; Taiwan and South Korea, on the strength of their chip manufacturers; and Japan and Israel, on a broad array of AI skills.
The partial winners are secondary suppliers. They include nations that are exporters of circuits, servers, and other AI-related electronic hardware — such as Mexico, Thailand, and Vietnam — or that play a role in the AI stack as both exporters and sizable bases for data centers, such as Malaysia and Singapore.
The losers include much of Europe, with the odd exception (the Netherlands is a major supplier of advanced chips from one big company). Worst off are those countries like India that lack “AI plays” and rely heavily on the industries most exposed to disruption, including IT services.
The AI booster effect continues to help power many economies through one crisis after another, from the tariff war to the Iran oil shock. Expectations for GDP growth have risen by nearly a full percentage point for the AI winners since the start of the year, while falling significantly for the losers.
In countries like the US, Taiwan, and Korea, large gains in advanced manufacturing, the associated surge in profits, and the wealth effect from AI-led stock market gains continue to lift economic growth. In countries like China, Thailand, and Mexico, tech exports are rising rapidly enough to offset weakness in other parts of their economies, including domestic demand.
The debate around the valuation of AI-related companies and whether they have reached the ‘bubble’ territory is ongoing. Recently Palantir CEO raised multiple questions on the pricing policy of AI companies and the service delivery methodology of these entities. While the debate on these matters may and will continue, it goes without saying that the use case for an LLM platform is undisputedly proven. AI will change the way we access information, work, analyse and do every other possible work. This is the reality now, and everything has to adapt and pivot based on this reality. IT service companies worldwide need to reinvent themselves, and the sooner they do, the better off they will be in this new AI world.
Global investors may be focused almost exclusively on AI, but they are not choosing winners at random. The leading AI nations are long-established tech powers with a deep commitment to R&D, spending more than 3 percent of GDP on average — over three times the level of lagging countries. They also invest heavily in technology, with tech spending averaging 3.7 percent of GDP among AI winners, compared with 2.7 percent for partial winners and 1.6 percent for losers.
The debate around the long-term financial viability of “hyper scalers” and their ability to generate revenue sufficient to justify the capital expenditure incurred now will continue in the foreseeable future. This is also relevant, as the pace of innovation has increased significantly, with newer models and chips coming out weekly that are significant improvements over existing models and chips. The speed of upgradation has become extremely critical in this industry. The world also has to adapt to this fast pace of change, and there will be a paradigm shift in how we operate across all walks of life.
In short, it’s an AI-driven world. Of course, this monomania will not last forever. The speculative enthusiasm will fade even as the technological revolution endures and expands in scope. As was the case following the 19th-century railroad boom and the internet craze at the turn of this century, a more balanced global market will eventually re-emerge. But so long as investors continue to see AI as the sole foundation of the next world order, they will keep ranking nations based on their tech prowess.
