AI momentum stocks hit record run as global trade narrows

Global momentum stocks pushed deeper into record territory on Monday as investors kept crowding into artificial-intelligence winners, leaving MSCI’s global momentum gauge 17 percentage points ahead of the MSCI All Country World Index since the end of March, according to Bloomberg Markets. From an analyst’s seat, the core question is no longer whether AI can lift a handful of megacap names. It is whether factor leadership itself is being rewritten by a trade that keeps attracting capital even as growth worries, war risk and sticky inflation would normally argue for a broader market reset.

Beneath the headline gain sits a more consequential shift in market structure. Scramnews has already tracked how hyperscaler and corporate AI spending is stretching balance sheets, with US Big Tech set to spend more than $700 billion on AI this year, while CNBC reported estimates that AI infrastructure spending could top $1 trillion within two years. What the momentum print shows is the market-wide spillover: investors are no longer merely paying up for Nvidia-style growth, they are clustering around an AI complex that now shapes factor screens, passive allocations and the definition of defensive exposure.

Skeptics read the same tape differently. Reuters argued in January that a maturing AI rally would eventually push investors toward cheaper parts of the market, while CNBC’s market wrap last week showed equities advancing even as traders were also weighing correction fears tied to inflation, bond yields and the Iran war. That contrast matters because it suggests the current move is being powered less by an all-clear macro signal than by a willingness to keep owning the same winners until something forces a reprice.

Bloomberg Markets quoted Hao Hong as saying the trade may still have room to run, even if the path gets rougher.

“The momentum story will persist for another few months, with significant volatility in between, till it finally sees a climax.”

— Hao Hong, via Bloomberg Markets

His point helps answer one of the analyst community’s central questions: is this move really about broadening earnings participation, or is it still mostly a factor trade? Available evidence still leans toward the second reading. Bloomberg reported separately that AI’s growing hold on benchmarks is trapping active managers, a sign that performance pressure is coming from concentration rather than from a clean expansion in sector breadth. In other words, stock-picking is losing ground to exposure management. Owning the right AI basket is doing more work than having the right opinion on the median company.

Asia becomes the proxy

Across Asia, the same trade is becoming easier to see because local indexes are more explicit about what they own. CNBC reported that Taiwan’s stock market is now worth $4.7tn and South Korea’s $4.4tn, lifting them to the world’s sixth- and eighth-largest equity markets. The move is not a vague regional-growth story. It is a direct expression of hardware concentration, with Taiwan Semiconductor Manufacturing Co. accounting for 40.0 per cent of Taiwan’s market capitalization and Samsung Electronics plus SK Hynix making up 42.2 per cent of the Kospi.

That concentration answers one of the insider view’s most practical questions: how much of the Taiwan and Korea surge is really about national markets, and how much is simply a levered expression of AI hardware demand? On the figures CNBC compiled, a very large share of the answer is a handful of names. Once a benchmark becomes dominated by a chip foundry, a memory champion and the supply chains around them, investors are not making a broad country call so much as buying a listed wrapper around the global AI buildout.

June Chua told CNBC that the regional benchmarks are no longer behaving like diversified national indexes.

“Both indices have effectively become AI and semiconductor proxies.”

— June Chua, via CNBC

Seen through that lens, Asia is less a side effect of the rally than its transmission mechanism. When Washington cloud spending climbs, when chip lead times tighten or when another round of capex guidance resets expectations, the reaction can move through Taipei and Seoul as quickly as it does through Nasdaq futures. That helps explain why the AI trade has stayed resilient even while macro headlines have turned less friendly. Investors hunting for pure exposure can find it more cleanly in markets where the index itself has become the thesis.

Why the trade looks fragile

Still, concentration cuts both ways. Bloomberg’s main report and Reuters’ earlier analysis point to the same vulnerability: once a rally becomes this crowded, the next shock does not need to be catastrophic to do damage. It only needs to force investors to ask whether they are being paid enough to own the same expensive exposure everyone else already owns.

For bubble-watch desks, the obvious breakpoints are inflation, yields and earnings disappointment. A hotter inflation print can lift real yields and compress the valuations of long-duration growth assets. A bond-market repricing can do the same from another direction. An earnings season that validates spending but fails to confirm monetization would be even more damaging, because it would undermine the idea that the AI premium is a bridge to future cash flow rather than a self-reinforcing narrative.

Jun Bei Liu told Bloomberg Markets that investors are already stretching past the macro backdrop.

“It’s risky — we are seeing bubbles forming in a number of areas, and the AI-led basket clearly has ignored all macro development.”

— Jun Bei Liu, via Bloomberg Markets

That warning carries more weight because the rally has not been happening in a placid backdrop. CNBC’s market wrap described traders looking through both geopolitical turmoil and inflation fears. Momentum can survive that tension for a while. What it cannot do indefinitely is convince investors that macro conditions have become irrelevant. Once the market starts distinguishing between AI beneficiaries with real earnings torque and those with only thematic association, factor leadership will look a lot less effortless.

The IPO wave could deepen it

Another force is about to test if the trade broadens or simply becomes more concentrated. Reporting from Bloomberg Markets, CNBC and the Financial Times suggests that SpaceX, OpenAI and Anthropic are pulling forward a new IPO cycle tied directly to AI demand. Semafor reported that the resulting listings could make 2026 the biggest IPO year on record.

If that pipeline materializes, the market’s next concentration problem may not sit inside the existing public indexes at all. It may sit in the scramble to fund, classify and benchmark a new set of giant AI-linked issuers. CNBC reported that SpaceX’s debut alone could crowd out demand for other flotations, while a separate CNBC report on valuation expectations showed how quickly private-market AI narratives are being translated into public-market size assumptions. That is why the current rally looks like a market-structure story rather than a simple tech-stock boom. Each new listing has the potential to deepen the same exposure loop that already dominates public equities.

Wall Street can tolerate a narrow leadership regime longer than skeptics expect, especially when earnings, policy and issuance all seem to point back to the same theme. Durability, though, will depend on breadth. Until gains start spreading beyond the AI complex, record momentum will remain less a verdict on the health of the global equity market than a sign that capital still prefers the clearest, most concentrated way to buy the future.