One week / one topic: The final frontier

What happened?

The SpaceX IPO was a massive success, and it matters for investors of all stripes.

Demand for the shares was unprecedented: $75bn were allocated out of requests for $350bn, immediately valuing the company at $2.2tn on day one.

The shares surged 40-50% within days, rapidly pushing the company into the top 6 globally and further validating the overall bullish thesis for AI-adjacent equities.

The success also acted as a clean clearing event for the ‘AI boom’ equity story, effectively leaving the door wide open for other upcoming mega‑AI IPOs from OpenAI and Anthropic.

Arguably, the company’s massive valuation might look weak vs current fundamentals, given annual revenues of $19bn and a net loss of $4.3bn in the first quarter of 2026.

The IPO, however, is a forward bet on AI dominance – reinforced by rapid moves like the $60bn Cursor acquisition.

While execution risks are clear – heavy capex, talent issues, lagging capabilities in AI and the complexity of… ahem… building datacenters in space – strong beliefs can trump all of this in the near term.

Also – if you had to bet on one human to actually pull this off, I think most people’s money would be on Elon.

What’s perhaps underappreciated – however – is how quickly AI unit economics are deteriorating, despite the asymptotic acceleration of AI investment.

The market is still trading the AI buildout and hyperscalers are effectively forced to keep spending due to competitive pressures stemming from the ‘arms race’ dynamic, both versus each other and versus China.

Meanwhile, token prices, compute costs and model efficiency are falling rapidly as open-source and local models keep narrowing the gap. Perhaps it’s not a coincidence that hyperscalers have been underperforming lately?

What to do then with all this?

Join the party? Pretend it doesn’t exist? Go to the beach early?

Our observations

• Fundamentals: Once again, we can directly observe how – at least in the short-to-medium term – it is all about sentiment.

  Investors believe in the ‘AI buildout’ story, and it’s hard to come up with something cooler than datacentres in space…

• Price action: Not only animal spirits are alive and well… investors are hungry for more.

  I am pretty sure that the people happiest with the SpaceX IPO performance – just after existing shareholders, of course – have unvested shares in OpenAI and Anthropic.

• Investor beliefs: The self-fulfilling nature of runaway price moves should be balanced vs massive growth in real, actual earnings. (Korean equities, I am looking at you)

  Meanwhile, volatility and dispersion remain elevated, signaling extended positioning but also continued discernment.

So what?

‘May you live in interesting times’ said someone, likely in an ironic way and wishing someone misfortune.

Opinions about misfortune may vary, but I think few would deny that we are indeed experiencing ‘interesting times’…

In real-time, no one really knows what to do with rapid, structural change like what we are now experiencing. (It’s all obvious in hindsight, of course…)

Meanwhile, what we can – and should – do, then, is stick to the process.

With that in mind, we have recently trimmed our position in Japanese equities – which have been a clear beneficiary of the AI buildout theme.

While we still like the asset class, we can also observe that the bond market is increasingly questioning PM Takaichi’s fiscally expansionary plans.

This might all sound a bit passé in light of the AI revolution – and indeed Japanese AI stocks are rallying hard – but we are still happy to take some of our gains and reduce dependency on one single theme.

Mood music: Dua Lipa – Levitating

By popular demand, here is the One week / One topic playlist

The information provided should not be considered a recommendation to purchase or sell any particular security.