$1 Trillion in Market Cap Erased, but This AI Stock Could Soon Be the World's Most Valuable Company (According to Elon Musk)

SpaceX shares remain about 5% below their IPO price after its valuation fell from a $2.7 trillion peak to roughly $1.4 trillion in mid-July, as investors question its AI-led spending strategy. The company reported $12.5 billion of first-half revenue while capital expenditures surged 308% year over year to $28.5 billion, with 86% directed to AI; Evercore ISI estimates capex could reach $360 billion by 2030. Musk's target of $3.5 trillion in 2033 revenue faces substantial execution risk amid intense competition from OpenAI, Anthropic, Microsoft, and Alphabet, and the article argues continued negative free-cash-flow concerns could pressure the stock further.

Analysis

The investable issue is not the addressable-market narrative but the financing loop: sustained build-out without visible utilization converts an equity story into a duration-sensitive capital-allocation story. If SPCX must repeatedly access equity or debt markets before AI revenue is independently disclosed, its valuation multiple should be set by cash burn and funding risk rather than by comparable software or cloud multiples. First diligence item before any position: independently verify the public listing, reported financials, AI segment revenue, customer contracts, and capex commitments; these claims cannot be underwritten from management commentary alone.

A large new AI-capacity buyer is incrementally favorable for memory suppliers MU and SNDK over the next 1-3 quarters if procurement translates into HBM, enterprise SSD, and networking storage orders. The more important second-order risk is downstream: excess third-party GPU/data-center capacity could force aggressive pricing to fill utilization, creating a cloud-margin headwind for MSFT and GOOG over 6-18 months, even as hardware suppliers initially benefit. That risk remains conditional on delivered capacity rather than announced capex.

Consensus bearishness may be early rather than wrong: equity investors often punish the spending phase, then re-rate rapidly when utilization, contracted backlog, and gross-margin evidence arrive. A reversal requires two consecutive quarters of disclosed AI revenue/backlog growth and evidence that incremental capex is producing improving free-cash-flow conversion; absent that, any sentiment rebound is likely a tradable squeeze rather than a durable rerating. Falsify the bearish view if AI bookings support a credible path to positive segment contribution margin without further material external financing.

Key Decisions for Investors

• Avoid initiating a standalone SPCX long until listing status, audited cash-flow data, AI revenue, and customer backlog are independently verified; treat unverified disclosures as a governance/liquidity risk rather than a valuation opportunity.
• For a 1-3 month hardware-demand expression, maintain a watchlist long MU and SNDK basket only after supplier commentary or reported inventory/order data confirms incremental AI-storage demand; exit if HBM/NAND pricing or data-center order visibility weakens.
• If SPCX is confirmed as liquid and optionable, consider a 3-6 month long MU / short SPCX pair after a failed rally into earnings or financing news: the pair isolates near-term component demand from the risk that SPCX's capital intensity drives multiple compression. Cover the short on disclosed contracted backlog and improving free-cash-flow guidance.
• Monitor MSFT and GOOG cloud gross-margin guidance over the next two earnings cycles. Do not short solely on prospective competition, but a material margin-guide reset alongside evidence of external AI-capacity oversupply would justify a tactical underweight versus NVDA.