SpaceX Slides Again Friday but Still Holds Most of a Big Weekly Gain

SpaceX shares slid ~4% midday to about $137.14 after opening at $141.29, with the drop attributed to IPO unlock supply digestion and profit-taking rather than a new catalyst. The underlying fundamentals remain strong: Q2 revenue was $7.8B (+92% YoY), including an AI segment at $2.6B (+247% YoY) and adjusted EBITDA of $3.5B, while Musk reiterated the $100B ARR target for December. Despite the positive setup, sentiment is choppy—prediction markets show a 94% chance SPCX finishes down today—and investors will watch whether the stock holds the mid-$130s into the close as the pivot versus a potential round trip back to unlock-era prices.

Analysis

The market is starting to distinguish between "AI demand" and "AI monetization." CRWV looks like the cleaner near-term cash conversion story because backlog and current revenue visibility matter more in a higher-rate environment than raw capacity ambition; NBIS is the higher-beta duration asset and will keep working only if the market continues to reward future power buildout more than current economics. That dispersion should ripple into the broader AI infrastructure stack: power, grid, and data-center suppliers should remain bid, but names with the most financing dependency will be punished first if sentiment cools.

SPCX remains a flow trade disguised as a fundamental one. The next 1-2 weeks are about supply absorption, not operating performance: if the post-unlock seller is exhausted, the stock can squeeze rapidly; if it cannot hold the mid-130s, systematic de-risking likely forces a fast retracement. The key falsifier is simple: a sustained move back above recent resistance on heavy volume would tell us the market has stopped caring about unlock mechanics and is re-rating on growth.

Contrarian take: consensus is treating the whole complex as one AI basket, but capital intensity should create dispersion, not correlation. The winners over 1-3 months should be the names with the clearest funding path and the least dilution risk; the losers will be the ones forcing the market to underwrite ever-larger capex without a matching near-term revenue bridge.

Key Decisions for Investors

• Long CRWV / short NBIS as a 1-3 month relative-value pair. Thesis: CRWV has better near-term monetization visibility, while NBIS is more exposed to narrative fatigue and financing-duration risk. Invalidate if NBIS keeps outperforming after the next capacity/customer update.
• Short SPCX only on a failed hold of the mid-130s; use a tight stop above the prior rebound zone. Best expressed as a put spread into the next 1-2 weeks if volume rolls over, because the trade is about unlock digestion rather than fundamentals.
• If staying directional on AI infrastructure, prefer CRWV over NBIS until the market proves it will pay for speculative future capacity at the same multiple as contracted backlog. Reassess after the next earnings/capex update.
• Set a watch alert on SPCX borrow and trading volume: falling borrow cost plus declining volume would confirm the squeeze is ending; rising borrow with flat price would argue the shorts still have ammunition.