3 Genius Stocks I'm Buying Instead of the SpaceX IPO

The article argues Microsoft, Nebius Group, and Nvidia offer better risk/reward than SpaceX, citing Microsoft’s AI annual recurring revenue up 123% to $37 billion, Azure revenue up 40%, and quarterly revenue up 18% to $82.9 billion. Nebius posted 684% year-over-year revenue growth in Q1, with analysts projecting 550% growth in 2026 and 225% in 2027, while Nvidia is expected to grow 81% in FY2027 and 41% in FY2028 at 31x earnings. The piece is opinionated stock-picking commentary rather than new company-specific news, so the likely market impact is modest.

Analysis

The key market signal is not that these names beat SpaceX; it is that public-market capital is still rewarding two very different scarcity profiles: durable cash-flow compounding (MSFT, NVDA) and extreme operating leverage to AI infrastructure buildout (NBIS). That matters because the second-order winners are likely to be the vendors selling picks-and-shovels into a capex supercycle — networking, power, and data-center infrastructure names — while the losers are any private-market “story stocks” whose valuation still depends on future optionality rather than near-term monetization.

MSFT looks like the cleanest defensive AI exposure because the business can absorb disappointment without a multiple collapse. If AI monetization slows, the downside is cushioned by enterprise software stickiness; if it accelerates, margin expansion has room to surprise. NBIS is the opposite: it is the highest torque, but also the most fragile to any hiccup in financing, customer concentration, or execution on capacity ramp, so the stock can reprice violently on even modest changes in growth assumptions over the next 6-18 months.

NVDA remains the best expression of the current AI stack because demand is still being pulled by hyperscaler capex rather than just hype. The contrarian risk is that consensus may be underestimating how quickly supply normalizes: if lead times compress and competitors narrow the performance gap, earnings growth can decelerate faster than the market expects, even if absolute demand stays strong. In that scenario, NBIS likely outperforms on multiple expansion only if it proves it can translate growth into durable gross profit rather than just revenue.

Net: this is a relative-value AI memo more than a “buy everything” memo. The better setup is to own the assets with either recurring cash generation or structural bottleneck exposure, and avoid the names whose upside depends on a narrative staying pristine for multiple years.

Key Decisions for Investors

• Long MSFT on a 6-12 month horizon as the lowest-volatility AI compounder; target a 15-20% upside with materially lower drawdown risk than higher-beta AI names.
• Long NVDA as the highest-quality growth/value expression; use pullbacks to add over the next 1-3 months, with a 12-18 month upside case still tied to hyperscaler capex durability.
• Speculative long NBIS sized small (e.g., 50-100 bps) for 6-9 months; upside is asymmetric if growth converts to backlog, but downside is sharp if funding or execution slips.
• Pair trade: long NVDA / short a basket of crowded AI beneficiaries with weaker monetization visibility over 3-6 months; the trade benefits if the market starts differentiating between real earnings power and narrative premium.
• Avoid chasing the latest private-market AI optics; if the sector de-rates, the first losses usually show up in the least liquid names before the megacaps, so keep dry powder for dislocations.