I am watching three potential IPOs that could define this market cycle: SpaceX, Anthropic, and OpenAI. Each carries a different path to growth, a different burn rate, and a different risk map. The question is simple and fun to debate: by the end of 2027, which one will have made public investors the most money? I run through the case for each, the traps to avoid, and the scenarios that could crown a winner.
SpaceX has the best business moat on Earth. Rockets, satellites, Starlink, data centers, government contracts. Space is the final frontier of infrastructure.
OpenAI created a category. ChatGPT became one of the fastest adopted products in history, and it’s still the name that everyone knows.
The Office Pool Question
Inside my team, we have a quiet pool on this. We will settle it on December 31, 2027. The rules are clean: assume each company lists and trades for at least a year. The winner is the stock with the best total return from its first closing price to year-end 2027. It is a thought exercise, but it sharpens the analysis. The goal is not to cheer. It is to weigh business models, cash needs, and likely valuation paths with a cool head.
View this post on Instagram
What Matters Most for These IPOs
- Moat strength: Durable advantage that is hard to copy.
- Revenue quality: Recurring, high-margin streams beat one-time sales.
- Capital intensity: Heavy buildouts can delay free cash flow.
- Customer stickiness: Long contracts and switching costs help returns.
- Regulatory risk: Rules can reshape growth and margins.
- Starting valuation: Your entry price still matters more than the story.
SpaceX: Infrastructure With Lift and Lock-In
The SpaceX thesis ties to one word: control. The company builds and launches rockets at scale. It runs Starlink, a global satellite network. It serves national agencies and private firms. It does the hard things that limit rivals. That is a moat in practice, not in prose.
Launch is cyclical and contract-driven. Margins can swing with payload mix and cadence. But the launch also builds the backbone for Starlink. Starlink is where recurring revenue lives. It packages hardware, access, and service. Churn should be lower in areas with limited alternatives. That mix creates pricing power in the right pockets of the map.
Government and defense work add stability. Those customers value reliability and speed. The bid process is long. The contracts are sticky. The revenue is less flashy than consumer uptake, yet it helps smooth cash flows that fund the next build phase.
The push into network infrastructure that touches data centers and global traffic creates another layer. If satellites serve not only homes but also enterprise links and edge sites, the addressable market grows. Paired with a proprietary launch, SpaceX can price the bundle rather than each part.
The risk is capital intensity. Constellations need constant replenishment. Ground gear must scale. Next-gen launch systems demand cash and time. Returns are strong when cadence is high, and utilization is full. They can tighten when the company shifts to new vehicles or must space out launches. An IPO would likely highlight sum-of-parts math: launch, Starlink, and services. The market will try to price each leg differently. That can help or hurt, depending on disclosure and segment margins.
Anthropic: The Clean, Enterprise-First AI Play
Anthropic has a clear story: build safer, more reliable models for business use. The focus is on enterprise demand. Buyers want accuracy, audit trails, and guardrails. They must meet internal risk rules. That fit matters more than being the loudest brand.
The company reported revenue growth in the first quarter, driven by strong enterprise traction. That is the shape you want to see early: paid pilots that turn into platform deals. Unit economics can scale quickly as usage grows on the same base model and tooling. Anthropic also benefits if customers prefer model choice across clouds. That keeps pricing from compressing too quickly under the weight of a single platform.
Risks cluster in two areas: compute costs and distribution. Training and inference spend can swell if pricing to customers lags. Anthropic must maintain high model quality while driving down cost per token. Distribution depends on partners and marketplaces. If cloud partners tilt the field to their own models or bundles, closing big deals gets harder. An IPO prospectus will be picked over for gross margin details and long-term credit terms with computer suppliers.
Still, the “clean story” angle is real. If revenue is concentrated in multi-year enterprise contracts with rising usage floors, public markets will pay for that visibility. Investors like steady expansion revenue. They reward low churn even more.
OpenAI: The Category Creator With Consumer Pull
OpenAI did something rare: it put an AI product in everyone’s hands. ChatGPT became a reference point. It owns the mindshare. That brand has optionality. It can sell to consumers, developers, and large enterprises. It can monetize through subscriptions, API usage, and embedded tools.
A brand alone is not a business plan. But it lowers the go-to-market cost. A known name shortens sales cycles. It helps recruit partners and developers. It makes upsell paths clearer. An IPO reduces the amount of new capital needed to fuel growth. It also supports a premium multiple if margins and retention are robust.
OpenAI’s risks echo the sector’s: compute costs, competitive models, and regulation. It also manages partnerships with powerful platforms. That can help distribution and hurt margins at the same time. The company must show that enterprise seats and developer APIs have strong gross margins and that consumer plans are sticky. If it does, the market will likely price it like a high-growth software platform with an AI kicker.
How Public Markets Might Value Each
Investors use simple tools to judge complex stories. I favor three quick lenses here.
- SpaceX: A sum-of-parts view. Launch is valued based on cash flow and backlog. Starlink is valued on recurring revenue and ARPU trends. Services valued like telecom or network capacity.
- Anthropic: A software-style revenue multiple if gross margins are high and usage is recurring. Discounts apply if computing costs soak up too much of each dollar.
- OpenAI: A hybrid multiple. Consumer subs get a direct-to-consumer multiple. API and enterprise seats get a software multiple with usage-based seasoning.
Starting price matters. If any of these are at frothy levels, even good execution may not bail out the first year’s returns. On the other hand, a disciplined IPO price with strong early beats can lift a stock fast. Watch the lock-up schedule and insider selling. Supply of shares can cap rallies.
Moat Check: Why It Endures—Or Fades
Moats fail when switching becomes easy. In AI, portability is rising. Model choice and orchestration tools let buyers test and swap. That puts pressure on providers to keep quality high and cost low. It also boosts the value of distribution and trust. A company with direct ties to customers and simple pricing wins more often than one with just raw power.
In space, moats fail if rivals match cadence and cut costs. That is hard. Rockets are unforgiving. Scale matters. Vertical integration helps hold lead time and price advantage. Starlink strengthens the moat. The network improves with more users and more satellites. That network effect compounds when the service expands into enterprise links and mobility.
What Could Go Right—or Wrong—By 2027
We can map three clean scenarios that could decide our office pool.
Scenario A: SpaceX wins. Starlink crosses a key scale mark, costs per bit drop, and enterprise wins stack up. Launch cadence stays high. Government work remains steady. The stock gets a premium for recurring revenue and a pipeline of new services. Cash flow improves even with ongoing capex. Investors reward the bundle.
Scenario B: Anthropic wins. Enterprises standardize on Anthropic for compliance and reliability. Gross margins rise as compute deals improve. Pricing power holds due to accuracy and safety features. Expansion revenue produces attractive net retention. The company prints consistent beats and guides higher. The market values it like a top-tier software platform.
Scenario C: OpenAI wins. The consumer brand turns into a durable subscription base. Enterprises adopt deeper seats. Developers build widely on the API. Unit costs fall faster than prices, lifting margins. The company becomes the default interface for daily work. The multiple stays high because churn is low and growth stays brisk.
Each has failure paths, too. SpaceX could face delays on next-gen systems or regulatory bottlenecks. Anthropic could see partner pushback or pricing pressure. OpenAI could face tougher rules, rising costs, or a brand stumble that slows signups. In all cases, a hot IPO price can set a bar that is hard to clear.
What I Will Watch Before Placing My Bet
Before I pick my entry in the pool, I want to see a few disclosures.
- Revenue mix and churn: How much is recurring, and what is net retention?
- Gross margin trend: Are compute and network costs falling faster than prices?
- Contract length: Average term, renewals, and usage floors for enterprise deals.
- Capex cadence: For SpaceX, the timing and scale of constellation refresh and launch upgrades.
- Partner economics: For AI firms, the take rates and credits with cloud providers.
- Regulatory exposure: Pending rules that could change product scope or data use.
With those in hand, the call gets clearer. Strong net retention, rising gross margins, and steady capex plans usually point to outperformance. If I only had a brand to go on, OpenAI would lead. If I only had a moat to go on, SpaceX would lead. If I only had sales cleanliness to go on, Anthropic would lead. But stocks move based on execution relative to expectations. That is why starting price and guidance often trump story alone.
My Working View Today
As of now, my base case is this: if pricing is fair at listing, SpaceX has the best shot to win by 2027. The mix of recurring Starlink revenue, government ties, and launch control gives it multiple avenues for growth. It also spreads risk across lines of business. The cash needs are significant, so I would size it carefully. But the path to durable free cash flow looks most visible.
Anthropic is my dark horse. If the company keeps converting pilots to multi-year deals with tight churn, the market may reward it with a strong software multiple. Execution on compute cost is key. The more they improve efficiency without sacrificing accuracy and safety, the better the margin story gets.
OpenAI has brand gravity. If consumer subscriptions stay sticky and enterprise seats expand, it could sprint out of the gate. The risk is that a rich listing price bakes in too much perfection. If costs or rules pinch growth, the stock could lag even as the business does well. That is not a bear case—just a note on how public markets trade.
I am a long-term planner by trade. I like businesses with cash flow visibility, lock-in, and pricing power. That bends me a bit toward infrastructure and recurring networks. Your view may tilt toward software multiples and speed. That is what makes this office pool fun. Which one would you pick?
By the end of 2027, we will have an answer. In the meantime, I will keep tracking margins, contracts, and capex. Stories change. Numbers tell you when they do.
Frequently Asked Questions
Q: What should I review first when a tech IPO files?
Start with revenue mix, gross margin trend, and net retention. Then look at customer concentration, contract length, and any long-term compute or capex commitments disclosed in the risk section.
Q: How do AI companies reduce the impact of high compute costs?
They improve model efficiency, secure favorable cloud credits or custom hardware deals, and pass usage-based pricing to customers. Better tooling and caching can also lower per-request expense.
Q: Why might SpaceX’s recurring revenue matter more than launch income?
Recurring service revenue can be steadier and have a higher margin. It smooths cash flows between launch cycles and may command a stronger valuation multiple from public investors.
