The 2026 IPO pipeline isn’t just about quantity - it’s about the sheer scale of valuations we’re seeing. Let’s examine what’s coming.
The Trillion-Dollar Candidates
SpaceX
Expected: Second half of 2026
Potential valuation: $1-1.5 trillion
- Aiming for IPO in H2 2026 (entire company, not just Starlink spin-off)
- Potentially raising $30 billion - the largest tech IPO ever
- Combines proven revenue (Starlink) with moon-shot ambition (Mars)
OpenAI
Expected: Possibly 2026
Potential valuation: $830B - $1 trillion
- Recent funding discussions suggest valuations approaching $1 trillion
- Has been preparing infrastructure for public offering
- Complex governance structure may require restructuring
Anthropic
Expected: Possibly 2026
Potential valuation: $183-350 billion
- Financial Times reported IPO considerations as early as 2026
- Backed by Google and Amazon with strategic interests
- Growing enterprise revenue provides IPO narrative
The $100B+ Club
Databricks
Status: IPO-ready, possibly early 2026
Valuation: $100-130 billion
- Raised at $100B+ in September 2025
- Reportedly seeking new round at $130B+ valuation
- Strong enterprise revenue growth story
Stripe
Status: Perennial IPO candidate
Valuation: $50-65 billion
- Has been “IPO ready” for years
- May wait for optimal market conditions
- Payments infrastructure story well understood by investors
What These Valuations Mean
Market Absorption Question
Can public markets absorb $100B+ in new tech listings in a single year?
- 2021 peak saw ~$150B in tech IPO proceeds
- SpaceX alone could exceed that
- May require staggered timing
Valuation Reality Check
Not all private valuations will hold:
- 2021 vintage unicorns often raised at peak
- Some will IPO at lower valuations (down-round IPOs)
- Others may pursue M&A instead
Crowding Effects
Mega-IPOs may affect smaller offerings:
- Investor attention and capital concentrated on big names
- Timing matters - avoid launching alongside SpaceX
- May benefit from association (“the AI IPO wave”)
Are these valuations justified, or are we seeing private market froth translate to public markets?
David, these valuations raise serious questions about market absorption and pricing dynamics.
The Market Math Problem
Can Public Markets Absorb This?
Let’s look at the numbers:
2021 Peak (best year in history):
- Total US tech IPO proceeds: ~$150 billion
- Total US IPO market cap at listing: ~$315 billion
2026 Pipeline:
- SpaceX alone could seek $30B at potentially $1.5T
- OpenAI + Anthropic could add another $50B+
- Databricks, Stripe, others add tens of billions more
We’re potentially looking at a year that needs to absorb more than the 2021 peak.
The Institutional Allocation Question
Large institutional investors have allocation targets:
- Tech allocation: typically 15-25% of portfolio
- New IPO allocation: subset of that
- Single position limits: usually 2-5% maximum
A $1.5T SpaceX IPO would be larger than most existing tech companies. How do institutions fit that in their portfolios?
Valuation Justification
SpaceX at $1.5T:
- For comparison: Apple ~$3.2T, Microsoft ~$3.1T, Alphabet ~$2T
- SpaceX would be 4th largest US tech company
- Is Starlink revenue + Mars optionality worth that?
OpenAI at $1T:
- Projected $115B in cumulative losses through 2029
- Revenue growing but so are costs
- Competition intensifying (Anthropic, Google, Meta)
Historical Precedent
When private valuations get this high, IPOs often disappoint:
- Uber IPO: Fell 7.6% day one, down from $120B private valuation
- Lyft IPO: Down 22% within a week
- WeWork: Never made it to market at original valuation
I’m not saying these companies aren’t valuable. I’m saying the private-to-public translation may include haircuts.
Let me examine the technical moats that could justify (or not) these valuations.
Technical Moat Analysis
SpaceX: The Hardware Moat
What makes SpaceX defensible:
- Reusable rocket technology: 10+ years ahead of competitors
- Launch cadence: More launches than entire rest of world combined
- Starlink constellation: First-mover with 5,000+ satellites deployed
- Manufacturing capability: Vertical integration at scale
The technical reality:
SpaceX has compounding advantages. Each launch provides data that improves the next launch. The gap with competitors is widening, not narrowing.
OpenAI: The Data Moat?
What OpenAI claims:
- Largest training datasets
- Most advanced model capabilities (GPT-5, etc.)
- API customer lock-in and integration depth
The technical reality:
- Model capabilities are converging (Claude, Gemini competitive)
- Training data advantages may be temporal
- Open source models catching up faster than expected
- Inference costs remain high with unclear path to profitability
Databricks: The Ecosystem Moat
What makes Databricks defensible:
- Unity Catalog as the metadata layer
- Customer data gravity (once data is there, hard to move)
- Integration with ML workflows
- Enterprise relationships and contracts
The technical reality:
Databricks has strong product-market fit in enterprise data. The moat is real but faces competition from Snowflake and cloud providers.
The Verdict
SpaceX’s technical moat is the most defensible - it’s physical infrastructure plus accumulated know-how that takes years to replicate.
AI company moats are less certain. The technology is evolving so fast that today’s leader can become tomorrow’s second-place.
Fascinating analysis on the mega-cap pipeline. Let me add some data perspective on these astronomical valuations.
The Valuation vs Revenue Reality Check
Looking at the numbers:
- SpaceX at $1.5T would be ~30x their estimated $9B annual revenue - but with 60%+ revenue growth
- OpenAI at $1T on ~$12B ARR projects to 83x revenue multiple
- Databricks at $100B on $2.4B revenue = 41x multiple
For context, public cloud/AI companies trade at 10-25x revenue. These premiums assume continued hypergrowth.
What the Data Shows About AI Valuations
From my analysis of 2021-2025 IPO cohorts:
- Companies with >50% gross margins AND >40% growth retained 85% of IPO valuations after 2 years
- Companies missing either threshold lost 40-60% on average
- AI companies specifically showed higher variance - the winners won bigger, but the losers crashed harder
The Absorption Question
The market absorbed ~$125B in IPO value in 2025. Can it handle $500B-1T in mega-cap listings? Historical data suggests mega-IPOs actually expand market capacity by attracting new investors. The Nvidia effect: more investors got comfortable with AI exposure.
My Prediction Model
Based on current fundamentals:
- SpaceX: Most justifiable premium - monopolistic position, defense contracts, Starlink revenue trajectory
- OpenAI: Highest risk - $115B projected losses through 2029, competitor emergence
- Databricks: Cleanest path - strong enterprise revenue, clear profitability timeline
The market will differentiate. Not all mega-caps are created equal.