Tech Led the First Half of 2026 — But the Real Winners Were Sitting Outside America

There is a version of the first-half-of-2026 story that almost writes itself: technology stocks powered global markets to one of their strongest six-month stretches in years, Wall Street’s benchmark indexes notched their best first halves since the early part of the decade, and the artificial intelligence trade once again proved itself the defining force in global finance. That version of the story is true. But it is also incomplete, and the missing piece is arguably more interesting than the headline itself. When the closing bell rang on the final trading day of June, the technology sector’s biggest gains of 2026 were not made in Cupertino, Redmond, or Santa Clara. They were made in Seoul, Taipei, and across a broader basket of emerging markets that most casual investors rarely think to check.

By the numbers, the gap is not subtle. Among MSCI’s family of sector-specific indexes, the gauge tracking large- and mid-cap technology companies in emerging markets surged more than 90% over the first six months of the year — an extraordinary run by any historical standard. Its European counterpart added a still-impressive 44.8%. The U.S. version of that same index, populated by household names such as Nvidia, Apple, Microsoft, Broadcom, and Micron, rose 19.4%. That is a strong number in almost any other year. In 2026, it was the laggard of the three.

The same pattern showed up elsewhere. The pan-European Stoxx 600 Technology index climbed 23.4% between January and June, outrunning the S&P 500 Information Technology index’s 19.4% gain over the identical stretch. Put simply: wherever you looked at “tech” as a category, the American cohort — long assumed to be the undisputed center of gravity for the industry — came in behind its overseas peers. That is a meaningful shift in a market structure that has, for the better part of a decade, been defined by U.S. dominance.

A Strong Half for Wall Street, Even If It Wasn’t the Strongest

None of this is to say American markets had a disappointing six months. Quite the opposite. The S&P 500 closed the first half up roughly 9.4%, the Nasdaq Composite added about 12.5–12.8%, and the Dow Jones Industrial Average climbed 8.9% to finish above the 52,000 mark for the first time — its best first-half performance since 2021. The second quarter alone was the best quarter for the S&P 500 and Nasdaq since 2020, and the best quarter for the Dow since 2022. Broader still, the MSCI All World index rose roughly 14% for the quarter, its strongest second-quarter showing in six years.

Those are genuinely impressive figures, and they came despite a bruising final stretch. Both the S&P 500 and the Nasdaq pulled back roughly 2–4% in June alone, a reminder that the run higher was neither smooth nor guaranteed. Markets absorbed a spike in energy prices tied to instability in the Middle East, elevated inflation readings, and a growing chorus of questions about whether spending on artificial intelligence infrastructure had begun to outrun the revenue that spending is supposed to eventually generate. That the major U.S. indexes still closed near record or multi-year highs speaks to how resilient the earnings backdrop remained even as the macro noise piled up.

Corporate profits did much of the heavy lifting. S&P 500 earnings per share grew nearly 28% year over year in the most recent reporting cycle, on revenue growth just shy of 12% — a level of operating leverage that helped justify, at least partially, valuations that were already stretched heading into the year. Analysts widely credit a resilient U.S. consumer, still willing to spend despite elevated prices, alongside an AI infrastructure buildout that shows few signs of slowing, as the twin engines behind that earnings strength.

The Real Story: A Leadership Handoff Inside Tech Itself

What makes the first half of 2026 unusual is not simply that non-U.S. tech outperformed U.S. tech. It’s that the internal composition of “winning tech” changed dramatically, and the shift traces almost entirely back to one component of the AI supply chain: memory.

For roughly two years, the market’s AI narrative was dominated by a small handful of U.S. software and hyperscale giants — the so-called Magnificent Seven — whose cloud platforms and consumer ecosystems were seen as the primary beneficiaries of the artificial intelligence buildout. That narrative cracked open in the first half of 2026. Several of the former leaders became some of the year’s most notable laggards. Microsoft is on pace for its worst monthly performance since 2008, down roughly 20% in June alone. Oracle fell about 30%. Meta declined around 15% year-to-date, and Tesla dropped roughly 7%. Even Nvidia, still the world’s largest company by market value, saw its share-price growth trail the broader S&P 500 for stretches of the period — a notable reversal for the stock most closely associated with the AI trade’s early years.

In their place, memory-chip manufacturers became the standout story of the half. Micron and SanDisk, both U.S.-listed, posted year-to-date gains of roughly 248% and 736%, respectively — numbers that would have sounded implausible at the start of the year. But the more striking shift happened overseas. Samsung, SK Hynix, and Taiwan’s TSMC — the trio most responsible for the world’s supply of high-bandwidth memory and advanced chip fabrication — rocketed up the list of the world’s most valuable public companies. Samsung, SK Hynix, and their peers are now reported to rank among the world’s ten to fifteen most valuable listed companies, surpassing long-standing giants such as Berkshire Hathaway and JPMorgan Chase in market capitalization. That surge, concentrated in Seoul and Taipei, is the single biggest reason the MSCI emerging-markets technology index posted its extraordinary 90%-plus gain.

The underlying driver is structural rather than speculative, at least on the surface. AI workloads are increasingly bottlenecked not by processing power alone but by memory — the high-bandwidth chips needed to feed data to AI accelerators fast enough to keep them running efficiently. Estimates circulating among analysts suggest AI-related demand could consume roughly 70% of global high-end DRAM production by the end of 2026, an astonishing concentration of a commodity that, until recently, was viewed as a cyclical, low-margin corner of the semiconductor industry. That scarcity has pushed memory pricing sharply higher — some estimates put potential price increases as high as 70% — and companies positioned to benefit have seen valuations expand accordingly. Semiconductor stocks as a group now make up roughly 20% of the S&P 500’s total market capitalization, the highest share on record and about four times their weighting in 2020.

Why Investors Rotated Away From the Old Guard

The rotation out of the former AI darlings and into chipmakers reflects a market that has grown more discerning about where AI spending actually translates into profit. For much of 2024 and 2025, simply having exposure to artificial intelligence — through cloud infrastructure, enterprise software integrations, or ambitious capital-expenditure plans — was often enough to lift a stock. By early 2026, investors had become considerably more selective, increasingly distinguishing between companies with a demonstrable, near-term earnings connection to AI and those whose exposure remained largely aspirational or capital-intensive without a clear payoff timeline.

That distinction mattered enormously for the hyperscalers. Several of the largest cloud providers have leaned on debt issuance and capital raises to fund the enormous infrastructure buildout AI requires — data centers, power capacity, custom silicon — and markets began treating that financing activity with more skepticism than they had in prior years. Where a capital raise once signaled aggressive growth ambition, by mid-2026 it was just as likely to be read as a sign that internal cash flow could no longer keep pace with the scale of investment being demanded. That reassessment, more than any single earnings miss, appears to explain much of the underperformance among names like Microsoft and Oracle.

Memory-chip makers, by contrast, sit closer to the physical constraint driving the entire AI buildout. Their product isn’t a promise of future efficiency gains; it’s the literal bottleneck determining how fast AI systems can actually be trained and deployed today. In a market growing wary of long-dated capital-expenditure stories, a company selling the scarce input everyone needs right now had an obvious appeal.

Beyond Chips: A Broader Market Than the Headlines Suggest

It would be a mistake to reduce the first half of 2026 purely to a chips-versus-software story. One of the more encouraging developments beneath the surface was a genuine broadening of market leadership beyond the narrow group of mega-cap names that dominated headlines in recent years. Industrial and infrastructure-linked companies posted some of the period’s most impressive gains: GE Aerospace rose roughly 47% over the trailing year, FedEx climbed about 82%, and Caterpillar surged an eye-catching 164%. The Russell 2000, the benchmark for smaller U.S. companies that had spent years trailing their mega-cap counterparts, touched a record high in the final weeks of June.

Sector data tells a similar story of diversification even as technology remained the standout. Among S&P 500 sector ETFs, the State Street Technology Select Sector SPDR (XLK) led the pack by a wide margin, up 28.7% for the half. Energy came in a distant second at roughly 19.8%, buoyed earlier in the year by the spike in oil prices tied to Middle East instability, though those gains cooled considerably by June. Notably, in the weeks following a technology-led pullback in early June, some of the strongest sector performance shifted toward consumer staples, real estate, and healthcare — traditionally defensive corners of the market that tend to attract capital when investors grow cautious about concentration risk in growth stocks.

Industrial fundamentals reinforced the idea that AI-driven demand was translating into activity well beyond the technology sector itself. Companies tied to industrial equipment and infrastructure reported average order backlog growth above 33%, with new order growth near 25% and organic growth exceeding 11%. Power-related businesses — the utilities and equipment makers building out the electrical capacity data centers require — posted sales growth near 35% alongside earnings growth approaching 300%, an extraordinary figure that underscores just how much physical infrastructure investment the AI buildout now demands. Utilities more broadly reported sales growth near 10% and earnings growth of about 6%, evidence that the AI theme has moved well past software and chips into the literal power grid.

The Other Side of the Ledger: What Struggled

Not everything rose with the tide. Bitcoin, one of the most closely watched assets of the past several years, lost roughly a third of its value over the first half, trading at less than half of its October 2025 peak near $125,000. It stood out as the period’s most notable underperformer among major asset classes, a reminder that risk appetite for speculative digital assets did not track the broader equity rally.

The U.S. dollar, by contrast, enjoyed an unexpectedly strong finish to the half, pushing its year-to-date gain to around 2.8% after a period in which many analysts had expected sustained weakness. That strength arrived alongside persistent inflation pressure: the Personal Consumption Expenditures Price Index rose 4.1% in the most recent monthly reading, with core PCE — which strips out volatile food and energy costs — running at 3.4%, a three-year high. Those figures complicate the interest-rate picture heading into the second half, with futures markets now pricing in a meaningfully higher probability of a rate increase rather than the rate cuts many investors had anticipated at the start of the year.

Leadership at the Federal Reserve adds another layer of uncertainty. A new Fed chair with a long public history of criticizing the central bank’s past approach to inflation now holds the responsibility for navigating exactly the kind of sticky price pressure that history warned against. Market watchers broadly expect a communication style that leans toward caution rather than reassurance — an approach designed to keep markets from assuming the Fed will always intervene when conditions deteriorate, even if it means more volatility in the near term.

A Global Mosaic, Not a Single Story

The picture outside the United States adds further texture to the “tech led, but not from America” narrative. In India, for instance, the first half unfolded along almost the opposite lines of the U.S. experience: defense, infrastructure, aviation, and select financial stocks were among the strongest performers, driven by sustained government spending on roads, railways, and manufacturing capacity, along with a steady recovery in travel demand. Indian information-technology stocks, heavily reliant on exports and overseas client spending, faced pressure instead — a reversal of the sector’s usual role as a market leader, and a useful reminder that “technology” as an investment category behaves very differently depending on which end of the global supply chain a company sits on.

That divergence captures something important about 2026 more broadly: this was not a year in which a single theme lifted every market uniformly. It was a year in which very specific, physically grounded supply constraints — memory chips, advanced fabrication capacity, power infrastructure — determined winners and losers with unusual precision, while broader macro themes like inflation, interest-rate policy, and regional instability shaped the edges of the picture without derailing the core trend.

What the Concentration Numbers Reveal

Perhaps the most striking underlying statistic from the half is not a return figure at all, but a concentration figure. The ten largest companies in the S&P 500 now account for nearly 40% of the index’s total value — a level of concentration that continues to raise questions about how durable current market leadership really is. Semiconductor stocks alone have grown to represent roughly a fifth of the index, a fourfold increase from their share just six years ago. Valuation metrics reflect that concentration: broad measures of market pricing suggest U.S. equities are trading near the most expensive levels on record, with only real estate, telecom, and healthcare currently priced below their five-year average multiples.

Whether that concentration represents a genuine structural shift — a world where memory and advanced semiconductor capacity are now as strategically important as oil once was — or a temporary distortion destined to unwind is the central question hanging over the second half of 2026. Optimists point to the underlying fundamentals: capital expenditure by major technology companies continues to climb, earnings estimates keep moving higher rather than lower, and demand for AI infrastructure shows no clear signs of cooling. Skeptics counter that markets have been here before, and that periods of extreme sector concentration — however well justified by present-day earnings — have historically preceded periods of painful rebalancing.

Reading the Divergence Through an Investor’s Lens

For portfolio managers who spent the last several years building U.S.-centric technology exposure, the first half of 2026 delivered an uncomfortable but instructive lesson: geography still matters, even in a supposedly borderless industry like semiconductors and software. A portfolio weighted heavily toward the familiar names — the large-cap U.S. platforms that dominated headlines from 2023 through 2025 — would have captured a respectable but distinctly middling slice of the sector’s total upside this year. A portfolio with meaningful allocation to South Korean, Taiwanese, and broader emerging-markets technology names would have captured something closer to the full story.

That gap has practical implications for how index-tracking and thematic AI funds are likely to be constructed going forward. Many popular “AI exposure” exchange-traded funds launched over the past three years were built with heavy U.S. weighting almost by default, reflecting where the largest, most liquid technology companies happened to be listed. If memory and advanced fabrication genuinely remain the choke point for AI infrastructure growth through the back half of the decade, fund managers and index providers may find themselves under pressure to rebalance toward the companies that actually control that choke point — many of which trade on exchanges in Seoul, Taipei, and elsewhere in Asia rather than New York or Nasdaq.

There’s also a currency and liquidity dimension worth noting. Gains posted in local-currency terms on the KOSPI or the Taiwan Stock Exchange don’t always translate one-for-one into dollar returns for U.S.-based investors, and trading volumes, settlement conventions, and disclosure standards can differ meaningfully from what American investors are accustomed to. None of that diminishes the scale of the outperformance, but it does mean capturing it requires either direct access to those markets or reliance on emerging-markets and international funds that have historically drawn less retail attention than their domestic counterparts.

The Semiconductor Supply Chain as Strategic Infrastructure

One of the more consequential shifts underway in 2026 is conceptual rather than financial: markets are increasingly treating advanced memory and chip-fabrication capacity the way they once treated energy reserves — as a form of strategic infrastructure whose scarcity carries pricing power independent of broader economic cycles. That reframing helps explain why memory producers, historically viewed as commodity businesses subject to boom-and-bust pricing cycles, have instead behaved more like scarce-resource monopolies over the past two quarters.

Supply discipline has played a meaningful role here as well. Years of underinvestment in new memory fabrication capacity, following a prolonged downturn earlier in the decade, left the industry with less room to respond quickly to the sudden surge in AI-driven demand. Building new fabrication facilities takes years, not months, which means even a well-telegraphed demand increase can outrun the industry’s ability to respond, keeping pricing power firmly in the hands of the small number of companies capable of producing at the required scale and precision. That dynamic — inelastic supply meeting a demand curve steepened by the AI buildout — is a large part of why Micron, SanDisk, Samsung, and SK Hynix have been able to post triple-digit percentage gains while broader technology benchmarks posted comparatively modest double-digit returns.

Whether this pricing power proves durable is one of the more consequential unknowns heading into the second half of the year. Semiconductor capital expenditure tends to be reactive: extraordinary profitability invites new capacity investment, and new capacity, once it comes online, has historically compressed the very pricing power that justified the investment in the first place. If AI infrastructure demand continues to grow at anything close to its current pace, that supply response may simply keep pace with demand rather than overshoot it. But if AI spending growth decelerates even modestly while new fabrication capacity continues coming online as planned, the memory sector’s current position at the top of the performance leaderboard could prove considerably more cyclical than its year-to-date numbers suggest.

Looking Ahead to the Second Half

Three questions will likely define how the rest of 2026 plays out. First, whether inflation, currently running well above the Fed’s comfort zone, finally shows signs of genuinely cooling, or whether the central bank is forced into a tightening stance that markets have not fully priced in. Second, whether the geopolitical premium currently embedded in energy prices — tied to ongoing instability in the Middle East — eases or intensifies, given its direct bearing on both inflation expectations and corporate input costs. Third, and perhaps most consequential for equity investors specifically, whether market leadership continues to broaden beyond a handful of chip and infrastructure names, or whether a narrow cohort of technology-adjacent companies continues to determine the direction of global markets almost single-handedly.

If the first half of 2026 offered any clear lesson, it’s that markets have shown a remarkable willingness to look past macroeconomic uncertainty as long as the earnings trajectory keeps improving. Profits, not politics, remained the dominant force shaping investor returns through six turbulent months — but the location of those profits shifted in a way few forecasters anticipated at the start of the year. The technology sector didn’t just lead global markets in the first half of 2026; it redrew the map of who benefits from that leadership, handing some of the period’s largest rewards to Seoul, Taipei, and a broader emerging-markets cohort that spent years operating in the shadow of Silicon Valley. Whether that handoff proves temporary or marks a lasting rebalancing of the global technology hierarchy will be one of the more consequential stories to watch as the second half of the year unfolds.