As SpaceX falls below its IPO price, we ask: Are this year’s mega-cap IPOs worth chasing?

2026 promises to see a reshaping of the US market as a wave of megacap IPOs raise capital. Elon Musk’s SpaceX launched on the Nasdaq in June, while Anthropic and ChatGPT may also come to market in the second half of the year. For active investment managers, these IPOs exaggerate an existing dilemma. To what extent do they choose to participate in AI, and at what price?

Planned IPOs are increasing the dominance of AI

The market was already shifting. The previously dominant hyperscalers (Amazon, Microsoft, Meta, Apple and Alphabet) have lost ground to semiconductor and memory names. After astonishing performance from Nvidia and other chip-focused companies, the IPOs are ultimately likely to see indices become more even more concentrated in a single theme – AI.

That said, the IPOs are not yet a meaningful part of the index. While Nasdaq has bent its rules to include SpaceX, its relatively small float size means it is currently only around 1% of the index despite its $1.8 trillion valuation[1]. MSCI and FTSE Russell also fast-tracked the float under their rules for large IPOs, but its starting weight in their indices is also small. S&P has stuck to its usual rules and won’t include the new listings for at least a year. It also needs to be ‌profitable under US accounting standards and hold a free float of at least 10%.

SpaceX has been a big winner for Scottish Mortgage

Among investment trusts, Scottish Mortgage has been the highest profile supporter of SpaceX, having held it since its early incarnation as a private company. Its initial £150m stake is now worth billions. As is standard ahead of a public listing, existing shareholders are subject to various restrictions, though Baillie Gifford said this would be a “staggered, tiered system rather than a simple 180-day (six month) hard lock-up”. The SpaceX position accounted for over a quarter of the value of the trust at 30 June.

SpaceX on proposterous valuation?

It has been a natural fit with its high growth mandate of Scottish Mortgage, but other groups are more wary. Julian Bishop, manager on the Brunner Investment Trust, says: “The forecasts that are out there for SpaceX are not just stretched, they are preposterous. We’re a sober, free cash flow-centric trust, and therefore it would not be on our radar.” He says financial information about the two other IPOs – Anthropic and OpenAI – is also still relatively scant. As to whether he might be forced to buy if SpaceX becomes a larger part of the index, Bishop is clear that he sees risk in absolute rather than relative terms.

These two trusts show the flexibility of investment trust managers to back their judgement on the new IPOs. For Scottish Mortgage, this is the ability to hold more a significant stake in SpaceX – even if it temporary – whereas for others, it may be the ability not to hold it at all. The same is true for the AI trade more widely, where there is a question around its dominance of markets and how active managers participate.

To beat the index you need to be very different from it

The dominance of handful of companies has been a dilemma for investors who value a high active share approach. James Budden, director of marketing and distribution at Baillie Gifford, says: “We would contend that mirroring or investing around the index is not active investing because your active share is too low to give you an opportunity to outperform. To beat the index you need to be very different from it.”

Does market concentration make life harder for managers?

However, deviating significantly from the index has become more difficult – and not just in the US. John Citron, portfolio manager on the JPMorgan Emerging Markets Growth & Income Trust, says: “It’s become more challenging to have a very high active share when the concentration in the asset class has increased. We have fewer, larger companies dominating the market and the benchmarks we measure ourselves against. Today, that is a challenge for us.”

They like the large technology companies in South Korea and Taiwan – Samsung, SK Hynix and TSMC are all significant positions in the portfolio. For Citron, these are businesses that make products that are integral to anything that involves computation. “You have to have memory, you have to have logic. The names we’re invested in are foundational.” But inevitably it means the trust looks more similar to its index benchmark.

The disrupters of the last decade have become the establishment

Budden agrees: “Unusually the big winners of late have been the main constituents of the index like the Magnificent 7 so it has been very difficult to beat the index. The disrupters of the last decade have become the establishment in index terms.”

the [AI] industry structure doesn’t look great

He says there is a question whether these names will be disrupted over the next decade. There are already signs of this happening. As they take on more debt to fund AI spending, the market has reappraised the valuations of the hyperscalers in the US. Bishop says AI is “clearly a very a hyped area”, adding “it’s obviously becoming more and more competitive all the time. It’s not only OpenAI and Anthropic, Meta is launching very good models, there is Gemini from Google, plus a whole raft of low-cost Chinese competitors that are seeing really, really extraordinary growth at much cheaper prices than these Western models. It strikes us that the industry structure doesn’t look great.”

He believes it is likely to be far more competitive than the internet revolution, where many companies managed to build up monopolistic positions. Even in those companies exposed to the AI infrastructure build-out, there are also concerns that investors are not recognising the potential cyclicality of semiconductors and memory chips.

The index should not be a starting point for any investment trust manager

AI or not, it should be a reminder that the index should not be a starting point for any investment trust manager. Budden says: “If you look at the concentrations in US tech right now with 40% of the S&P 500 and 30% of global indices allocated to tech, just following the index amounts to a huge bet on AI. The “risk” is to the record and reputation of the fund manager who is probably remunerated on an annual basis against an index so sticks to it like glue.”

In the longer-term, the IPOs add another layer of concentration on stock markets that were already concentrated on a single theme. Investment trusts have the luxury of discernment and investors need to ensure that they are using it.

[1] https://finance.yahoo.com/markets/stocks/articles/spacex-shares-face-fresh-pressure-103721256.html