The Space Economy: As Launch Costs Plunge, the Real Business Begins

Industry Research No. 9 — Stepping away from the AI main thread to look at a sector I've written about before, but one that demands even more patience: space. It has the most compelling narrative, but it's also the easiest place to pay an irrational price for the romance of 'stars and seas.'

011. The Single Underlying Variable of the Space Economy: Launch Costs

When people talk about the space economy, concepts fly everywhere—satellite internet, space tourism, asteroid mining, in-orbit manufacturing. But to understand it all, you have to grasp the one variable driving everything: launch cost (the price of sending one kilogram of payload into space).

For sixty years, space was a national game, sealed off from ordinary commerce, for one fundamental reason: it was too expensive. Rockets were expendable; each launch burned an entire vehicle. Sending one kilogram into orbit cost so much that only national budgets could bear it. Under that cost structure, the 'businesses' possible in space were extremely limited—aside from national defense, science, and communications, almost no business model penciled out.

The fundamental change over the past decade is the collapse of launch costs. Reusable rockets (led by SpaceX) turned rockets from 'single-use' into 'reusable,' slashing per-kilogram launch costs by one to two orders of magnitude. This change cannot be overstated—it's not 'space commerce got a little cheaper'; it's 'space became commercially viable for the first time.'

Consider the analogy: the collapse of launch costs for the space economy is like the collapse of cloud computing costs for internet startups. When AWS brought the cost of compute for 'starting an internet company' from millions to hundreds, an explosion of web startups became possible. Similarly, when launch costs collapsed, a wave of 'space businesses' finally penciled out. A collapse in the underlying cost unlocked the entire layer above. This is the single most important lens for understanding the space economy.

022. Cost Collapse Unlocks an Entire Downstream Industry

As launch costs fell, the space economy split from a 'national project' into a layered value chain. Like the compute stack, we can break it down:

Note how strikingly similar this is to the AI compute stack—the greatest value ultimately sits in the application layer (Starlink-style satellite internet is the first proven, massive commercial application in the space economy). And the access layer (rockets) is the 'shovel seller,' the prerequisite for everything.

The collapse of launch costs first made the infrastructure layer possible (only with cheap launches can you afford to launch tens of thousands of satellites into constellations), then triggered the explosion of the application layer (constellations enable global satellite internet and high-frequency Earth observation). The prosperity of the entire chain derives from the downward curve of that base-level cost. If you want to judge the outlook for the space economy, the core question is: will launch costs keep falling, and how fast? Because each percentage point drop unlocks new demand downstream.

033. RKLB's Real Position: Trying to Be 'The Non-SpaceX Option'

When we turn to public markets, there's an inescapable reality—the strongest player in the space economy, SpaceX, is private; ordinary investors cannot buy it. This is a fact that must be recognized before looking at the sector: the dominant player is not in public markets.

Among the pure-play space names that are publicly traded, RKLB (Rocket Lab) gets the most attention, and I've written about it before. Its position can be understood this way—it wants to be 'the non-SpaceX option,' and it is evolving from a 'rocket company' into a 'space systems company.'

Two points are most critical:

First, in launch, it takes a 'small and specialized' approach and then pushes into 'medium.' RKLB started with the small Electron rocket, building a reliable small-satellite launch capability, and became one of the few players with a mature commercial launch record besides SpaceX. Its next step is the medium-lift Neutron rocket—if it succeeds, it will gain access to a much larger market. I've written before about Neutron's delays—let me be clear: rocket development delays are the norm in this industry, not a failure unique to RKLB. SpaceX itself failed countless times early on. In this business, 'delay' is almost certain; the real question is 'can it eventually succeed?' not 'did it hit the deadline?'

Second, and more importantly—it is vertically integrating into 'space systems.' RKLB is not content to simply sell launch services (launch is a low-margin, high-risk activity). It's moving upstream into satellite manufacturing, space components, and space system integration. This is its most critical strategy—moving from the hard business of 'selling launches' toward the higher-value position of 'providing full-stack space infrastructure capabilities.' If launch is 'selling shovels,' RKLB wants to be 'selling not just shovels, but the entire gold-panning kit.' Whether this vertical integration path works is the core metric for judging RKLB's long-term value.

044. My View: Space Is a Real Trend, But Buy It as an 'Option'

How should investors think about space? My judgment boils down to one sentence—space is a genuine, multi-decade trend, but for ordinary investors in public markets, it's better treated as an 'option' than a 'core holding.'

Why an 'option'? Because the space sector has several distinct characteristics that shape its payoff structure:

First, it's high-variance. Rockets explode. Projects get delayed. Technology paths reverse course. The outcome distribution in this industry is not smooth; it's binary—'massive success or catastrophic failure.' Such high-variance assets are naturally suited to 'small positions with huge upside' rather than heavy weighting.

Second, it has limited downside and enormous upside (convexity). If you take a small position in a space name: the worst case is you lose that small amount (limited downside); but if launch costs continue to fall and the space economy really materializes, the upside could be tens of times (enormous upside). This 'downside-capped, upside-uncapped' convex structure is exactly what I described in my Bitcoin piece and what Taleb emphasizes—good bets are convex, and convex bets require a position that is 'small, held for the long term, and can be lost.'

Third, the dominant player (SpaceX) is not in public markets, which weakens the payoff. What you can buy in public markets (RKLB, etc.) are the 'runner-ups,' not the dominant player eating the largest slice. This means public-market space investing inherently misses the fattest part of the pie. This further supports the 'small option' rather than 'heavy core' positioning.

So my attitude toward space is 'long-term bullish on the direction, strictly disciplined on position size.' Treat it as a small, affordable bet in your portfolio—one that can absorb a total loss—to capture extreme upside, rather than a core holding to bet the farm on. Paying a little bit for the romance of the stars (to buy a convex future) is rational; betting the farm on that romance (treating it as certainty) is dangerous. The biggest investment trap in space is precisely that its story is too sexy, making it easy to break the 'small position' discipline.

055. Final Thoughts

The space economy is the most romantic industrial story of our time—it's about humanity leaving Earth, about stars and seas. But precisely because it's so romantic, we need to look at it with cool heads.

The first step toward coolness is to grasp the one unromantic underlying variable: launch cost. All the prosperity of the space economy derives from the collapse of this cost curve—each percentage point drop unlocks more downstream. Understanding space isn't about how cool rockets look; it's about 'how much cheaper can we get something into orbit, and how many businesses that once didn't pencil can now pencil?'

The second step is to face the reality of public markets: the dominant player is private, the available names are lagging, the industry is high-variance, and outcomes are binary. This determines that for ordinary investors, space is a convex option, not a core holding—small position, affordable loss, aiming for extreme upside. Hold that discipline, and you can avoid missing the stars while not getting burned by their romance.

If I leave you with one sentence—

The space economy rests on a single underlying variable: the collapse of launch costs. And the correct posture for ordinary investors is also singular: treat it as a small option with limited downside and enormous upside—pay a little for the romance, but don't bet the house.

Next, we'll look at a sector close to AI but often overlooked—optical communications: the 'blood vessels' connecting millions of GPUs inside AI data centers.

—

Risk Disclosure: This article is an industry chain analysis. Companies mentioned are for analytical purposes only and do not constitute investment advice. Market risk exists; invest with caution.