A Market Built on a Single Theme
One of the defining features of today's equity markets is the degree to which performance has narrowed around a single thesis: the buildout of artificial intelligence infrastructure and the demand for the hardware that supports it. Semiconductors have rallied sharply in the United States, but the same dynamic, in even more concentrated form, has played out across global markets. Nowhere is this more pronounced than in emerging markets, where just three chip companies — Taiwan Semiconductor Manufacturing Company, Samsung, and SK Hynix — now represent roughly 24% of the emerging market index. More striking still, those same three companies account for essentially all of the earnings growth expected from the index this year.
The numbers behind that expectation are remarkable. Samsung and SK Hynix are projected to deliver earnings growth in the range of 300% to 400%, while TSMC is expected to grow earnings by more than 45%. On a forward price-to-earnings basis, emerging markets appear inexpensive — but only because of these extraordinarily high earnings expectations. Strip out the chip giants, or even modestly disappoint against those forecasts, and the apparent value evaporates. This is the core of the concentration risk facing global investors: the cheapness of the index is conditional on a single industry continuing to deliver outsized growth.
The Fragility Behind Hyperscaler Demand
That conditionality is precisely what makes any wobble in the AI demand narrative so consequential. A recent report suggesting that compute demand and AI-related spending might not be living up to elevated expectations was enough to take the wind out of the technology trade. The implications ripple far beyond the United States. If hyperscalers begin to revise their capital expenditure plans, or if questions about return on investment begin to find harder answers, the upstream beneficiaries — the foundries and memory makers in Taiwan and South Korea — would feel it first and most acutely.
For that reason, hyperscaler earnings have become a global event. Markets across continents are watching for any change in capex guidance and for any signal on the economics of AI deployments. Memory chip companies in particular are at an inflection point, as investors weigh whether this cycle truly differs from those of the past.
"This Time Is Different" — and the Cyclicality That Refuses to Die
A growing narrative argues that memory makers may be moving into a structurally different cycle, one where multi-year contracts smooth out the volatility that has historically defined the sector. The hope is that long-term agreements will end the boom-bust pattern that has plagued memory pricing for decades. Some investors have begun positioning for a prolonged boom in memory chips on the strength of this thesis.
But whenever the phrase "this time is different" enters the conversation, it is worth listening more carefully, not less. Memory chip companies remain inherently cyclical. Even with multi-year contracts, demand swings and average selling prices can still move sharply. The shift toward longer contracts may dampen volatility at the margin, but it does not eliminate the underlying cyclicality of the business. Investors counting on a permanently smoother ride may be disappointed.
A Wait-and-See Stance Across Central Banks
While the AI trade dominates equity attention, monetary policy is offering its own complications. Central banks broadly find themselves in a wait-and-see mode, hampered by the fact that they have no more visibility into when the Strait of Hormuz reopens than anyone else. The energy question — how long prices stay elevated and at what level — has become a primary determinant of policy. The longer the closure, the more severe the negative impact on GDP and the greater the upside risk to inflation.
The Bank of Japan recently delivered a hawkish hold, with three members already calling for a hike. After several years of tightening, another increase would not represent a meaningful change in trend. The more significant inflection is at the Bank of England, which had been expected to cut rates before the conflict but is now arguably the most likely to pivot back to hiking. UK inflation never fully moderated, and core CPI in March remained above 3%, leaving the central bank uncomfortably exposed if energy costs push prices higher still.
The Bank of Canada, meanwhile, kept rates unchanged, calling the current setting appropriate. Canada is somewhat insulated by virtue of being an oil exporter — a meaningful contrast with energy-importing economies, which face the brunt of the negative shock.
The Energy Shock and a Stagflationary Tail Risk for Europe
The geography of the pain is straightforward: energy importers, concentrated in Europe and Asia, absorb the worst of the impact, while exporters such as Canada benefit. Negotiations remain stalemated. Iran has offered to reopen the strait if the United States ends its blockade, but accepting that bargain would reduce the leverage Washington holds over concessions on the nuclear program. Reports that the administration is preparing for an extended blockade reinforce the prospect of a drawn-out disruption. At the same time, pressure to end the closure is mounting as midterm elections approach and Iran's own economy faces growing downside risk.
A scenario-based view of these dynamics is instructive. The moderate case envisions a gradual normalization of energy prices through the second quarter. As that quarter ticks by without resolution, however, the probability weight is shifting toward an adverse case in which disruption persists and normalization only arrives as the second half of the year unfolds. The deeper markets slip into that adverse scenario, the greater the risk of a genuinely stagflationary impact, particularly in Europe, where activity data is already deteriorating in contrast to the relative resilience visible in the United States.
Why Europe Is Especially Exposed
Europe's vulnerability is compounded by the fact that it lacks the AI growth story that has been propping up the United States and parts of Asia. There is no European equivalent of TSMC, Samsung, or SK Hynix providing a tailwind powerful enough to offset the energy shock and weakening macro backdrop. The result is a region facing the brunt of higher energy import costs, softening data, and a central bank that may have to tighten into weakness, all without a structural growth narrative to lean on.
What to Watch From Here
The signal-rich items on the calendar fall into two buckets. The first is the AI capex trade: every line of guidance from a hyperscaler, every comment on return on investment, and every disclosure from the memory makers carries outsized weight given the concentration of expected earnings. The second is the central bank cluster, with the European Central Bank and the Bank of England framing how policy will respond to an energy shock whose duration nobody can yet pin down.
The combined picture is one of a global market resting on a narrow earnings base, exposed to a macro shock whose tail is widening, and reliant on a "this time is different" story in a notoriously cyclical industry. None of those legs is broken. But each is worth examining carefully before assuming that the cheapness in headline multiples is as durable as it appears.