How the Middle East conflict threatens the AI boom
Since the start of the year, we have witnessed a marked shift in geographic and sector performance across global equity markets. Between 2018 and early 2025, the American technology sector was the dominant engine, driven by the Magnificent Seven.1 In 2025, that started to change. The strongest returns began to emerge outside the United States: Taiwanese and Korean chipmakers, European industrials.
This rotation took hold against a backdrop of extreme index concentration around artificial intelligence. Consider what "global diversification" actually means today for an investor buying a market-cap weighted ETF tracking the MSCI World. The United States accounts for roughly 70% of the index, up from 49% at the turn of the century. The top ten holdings alone represent around 25% of the total, compared with barely 10% a decade ago. Information technology and tech-adjacent sectors make up over a quarter of the index. In other words, a passive investor buying "the world" is making a concentrated bet on a handful of US technology companies. The valuations of those companies rest on a double assumption: flawless technical execution and a durable monopoly on monetisation.
At TreeTop, we believe in the transformative power of AI. What concerns us is not the technology itself, but the colossal investment required to deploy it at scale, and the valuation multiples the sector currently commands. The top ten names trade at an average of 38 times earnings, compared with 26 times for the index as a whole. These prices leave virtually no margin for error. The smallest stumble could trigger a sharp correction.
The Middle East crisis calls that flawless execution into question. US and Israeli military actions against Iran have introduced a specific kind of risk. Not the vague geopolitical noise investors tend to dismiss. Something more structural. AI and this crisis are connected in two ways.
Energy costs.
Artificial intelligence consumes extraordinary amounts of electricity. Training and running large models demands energy infrastructure that does not scale overnight, and the Gulf states sit at the centre of global energy supply. If the conflict were to disrupt that supply over the medium or long term, even modestly, the economics of scaling AI would become more complex. Energy costs would rise. The margin for error would shrink further still. And the risk is not limited to price: there is also the political possibility of rationing.
Financing the buildout.
The infrastructure required to deploy AI at scale runs into hundreds of billions of dollars: data centres, semiconductor fabrication plants, high-performance computing capacity. To fund these investments, companies are increasingly turning to external capital.
Abu Dhabi's sovereign wealth funds alone manage over $2 trillion. Saudi Arabia's Public Investment Fund manages another $1 trillion. These are not passive spectators of the AI boom. They are anchor investors.
A geopolitical reorientation in the region would not only affect oil prices. It could redirect capital flows. And as inflation and geopolitical uncertainty push interest rates higher, the cost of that external financing rises with them.
Conclusion : Discipline over euphoria
How much of the AI bull market rests on solid fundamentals, and how much on euphoria? Probably some of both. But it would be premature to assume the risks are behind us.
Stretched valuations, historically elevated index concentration and geopolitical disruption do not cancel each other out. They compound. For our clients, this confirms what we have argued since the concentration began: holding a broad, balanced portfolio is not a failure of conviction. It is a form of discipline