The Average SpaceX Investor Is Now Losing – These Investors Aren't

Jensen Huang reiterated that the U.S. is behind on energy production, reinforcing the AI power/infrastructure trade and highlighting names across nuclear, uranium, grid, and electrification. Butterfly Network (BFLY) surged 56% last Thursday on new Midjourney Medical-related demand visibility, with subscribers reportedly up roughly 50% to 400%+ on the move. The article also warns that SpaceX (SPCX) has already fallen roughly 25% from its IPO peak, leaving many post-debut buyers near breakeven or underwater.

Analysis

The cleanest expression here is not “AI” broadly; it’s the bottleneck trade. When compute growth outruns grid buildout, pricing power shifts to regulated-asset owners, power-service contractors, and fuel suppliers with scarce permitted capacity. That favors names with visible backlog and low execution risk over the more promotional nuclear-development cohort, because the market will pay up first for duration and reliability, then only later for speculative capacity additions.

The secondary effect is that every incremental AI capex announcement raises the value of boring infrastructure. That should keep re-rating pressure on transmission, switchgear, and EPC names even if headline AI multiples compress, because utility interconnection queues and transformer lead times are the real choke points. The contrarian risk is that this basket becomes crowded quickly: once the market treats power as a generic AI proxy, forward returns likely collapse unless earnings revisions outpace multiple expansion.

BFLY’s move is a reminder that asymmetric upside usually comes from a genuine external validation event plus a tight float/technical setup, not from narrative alone. The Midjourney-linked catalyst is potentially meaningful, but the deeper question is whether it becomes repeatable procurement or just a one-off demo headline; if the former, the valuation can still rerate materially, if the latter, the stock can give back most of the pop. For SPCX, the lesson is the opposite: high-quality asset, poor entry point. Post-IPO price discovery remains dominated by flow and scarcity, so early open-market buyers are funding distribution rather than owning a durable edge.

Key Decisions for Investors

• Overweight PWR and HUBB versus the broader AI basket for a 3-6 month trade: these names monetize the bottleneck directly and should see estimate revisions first; use any post-news pullback to add, with 15-20% upside if backlog converts cleanly.
• Hold a basket long CEG/VST/CCJ/BWXT, but size OKLO/SMR materially smaller; the former group has earnings visibility while the latter trades more like optionality. Prefer a 2:1 allocation to the mature names versus the SMR names until commercialization milestones are de-risked.
• Consider a pair trade: long PWR, short a high-multiple AI software or semiconductor proxy if you want to express 'picks-and-shovels over headline AI.' The trade works if power capex remains the binding constraint over the next 2-4 quarters.
• For BFLY, treat strength as a momentum continuation trade only if it holds prior breakout levels on declining volume after the initial spike. Otherwise, fade rally attempts with defined-risk calls or cash-secured puts; expect high volatility and headline-driven retracements over days to weeks.
• Avoid chasing SPCX in the open market until the post-IPO volume settles and price stabilizes for several sessions. The better risk/reward is to wait for a secondary basis or to own suppliers/adjacent beneficiaries rather than the listing itself.