Here's the 1 Thing I'd Wait to See Before Buying SpaceX Stock.

The key issue is capital allocation, not simply whether AI eventually grows: investors may discount Starlink’s cash generation if it cannot be separated from AI’s funding needs. That creates a conglomerate discount and makes the equity unusually sensitive to financing costs and proof of incremental AI returns. But operating losses alone do not establish equivalent cash burn; depreciation, stock compensation, intercompany charges, and the split between expensed costs and capitalized infrastructure need verification.

Near term, the risk is further multiple compression if results show widening AI losses without evidence of monetization. Over 1–3 months, segment disclosure, capex guidance, and any financing plans matter more than broad AI product announcements. Over 6–18 months, the structural test is whether AI assets produce external revenue or measurable benefits to other businesses before Starlink’s economics are diluted. Higher rates amplify the downside by raising the hurdle rate on long-duration projects; falling rates could support the valuation without resolving execution risk.

Contrarian point: the market may be treating AI investment as pure leakage when it could create valuable infrastructure or product synergies. Conversely, Starlink profitability is not a sufficient backstop if cash flows are not ring-fenced or transparently reported. T-Mobile’s role should not be inferred from the spectrum reference alone; deal terms and any commercial arrangement are needed before assigning it a material earnings impact.