The financial world this spring is running a paradox in plain sight. A war centred on Iran is disrupting an estimated 10 to 12 million barrels of oil a day, pushing crude to a four-year high and putting the Strait of Hormuz back into every risk model. And yet: the S&P 500 set a record above 7,270. The Nasdaq-100 hit an all-time high. South Korea’s Kospi pushed past 7,000 for the first time in its history; Taiwan’s Taiex and Japan’s Nikkei printed records of their own. Wall Street, and much of Asia with it, is at peak valuation in the middle of an energy shock. The question is not whether this is strange. It is what is holding it together.
The answer, stated without ornament, is artificial intelligence — and the concentration of that answer is the actual story. Strip the AI complex out and the picture inverts. By one estimate, AI, semiconductor and data-centre companies now account for roughly half the weight of Japan’s Nikkei 225. South Korea’s market, powered by its semiconductor champions, gained some 76 percent in 2025 and has continued climbing; Samsung crossed a trillion dollars in market value. The market’s logic is almost geographic, and one strategist put it memorably: Asia has AI but not energy, Latin America has energy but not AI, and Europe — whose indices remain stuck below their pre-war levels — has neither. Markets are not pricing the global economy. They are pricing exposure to one earnings story.
That concentration is the risk, and VeyrZest readers should sit with it rather than be soothed by the record numbers. A market this dependent on a single narrative is, by definition, fragile to that narrative. The corporate results have so far justified the optimism — first-quarter earnings broadly beat expectations — but investor reaction is already turning selective around AI capital spending and the quality of guidance. The rally is also partly mechanical: research from Goldman Sachs suggests trend-following funds caught short in March were forced to buy back in, manufacturing a multi-billion-dollar squeeze. Some of ‘the market believes’ is really ‘the market had no choice.’
Underneath the equity story sits a macro picture that is genuinely uncomfortable. March consumer inflation ran at 3.3 percent year on year, accelerating, with a gasoline surge driving headline figures sharply higher. Comparisons to 1970s stagflation have re-entered serious commentary, and the rhyme is real: a Middle East oil shock arriving alongside an inflation impulse while growth is already cooling. The differences are real too — the economy is far less oil-intensive than in 1979, and the Federal Reserve will not return to Volcker-era double-digit rates against a federal debt near 100 percent of GDP. But ‘it won’t be the 1970s’ is not the same as ‘it will be fine.’ It means the policy toolkit is narrower, not that the problem is smaller. And the Fed enters this with a leadership transition underway as Jerome Powell’s term as chair ends.
So how should a serious operator read a market like this? Not by shorting euphoria — markets can stay irrational longer than most balance sheets can survive the bet — and not by assuming the records are simply false. The disciplined reading is to understand precisely what you are exposed to. If a portfolio, a business plan or a client’s strategy has been quietly carried upward by the AI rally, the relevant stress test is narrow and specific: what happens to the structure if the AI earnings story merely slows — not collapses, just disappoints? For many, the honest answer reveals far more concentration risk than the diversified-looking surface suggests.
Records amid ruins is not a contradiction the market needs to resolve. It is a description of a market leaning its entire weight on one beam. The beam is, for now, holding — first-quarter earnings kept the optimism grounded in something real. But for those who take the long view, the job this spring is not to celebrate the height. It is to know exactly where the weight is resting, and what else would have to be true for it to hold.
