$FLY me to the moon (yet again)...and right into the Pentagon's budget and beyond
[In for $37k dollarydoos](https://preview.redd.it/kl6o5pxbeobg1.jpg?width=1080&format=pjpg&auto=webp&s=466014fb9f0e58a026a2ae376aed7a9dbd5294fc)
The biggest asymmetric bet in the upcoming Space hype cycle of 2026/7 is staring you in the face. That said, I appreciate reading more than 100 words, or words at all, can be cognitively difficult for many here, but let's give it a go
**Firefly isn't just a launch provider:**
The market has massively overlooked this imo - Firefly is now a space and defense technology company after closing the SciTec acquisition on Nov 5, 2025.
This isn't some random acquisition, this was strategically timed, is a f'ing smart play and adds AI-enabled defense software proven in operations across missile warning/defense, ISR, space domain awareness, and autonomous command & control.
This is mission-proven defense software, big data processing, classified facilities, 475+ employees (yes it adds cost) and $164m in existing annual revs. Their customers are a literal shopping list of the kind who keep coming back for more:
* Space Force
* NGA
* Strategic Command
* Missile Defense Agency
* SDA
* National Reconnaissance Office
* Missile and Space Intelligence Centre
Again, Firefly is explicitly positioning SciTec as the software layer that pairs with Firefly’s hardware across launch, lunar, and in-space vehicles, including missile warning and defense and space domain awareness. To keep it simple - this is an obvious play for pieces of the Golden Dome.
**Eclipse:**
Eclipse (Firefly's medium lift) is designed to put 16,300 kg into LEO, designed for reusability with a first stage intended to return and land.
The under appreciated point imo is who is standing next to Firefly on this.
* Northrop Grumman - this first-stage architecture is being used for Antares 330...it is therefore unlikely to be shit
* A Tier-1 prime like Northrop has every incentive to be conservative - it has 0 incentive to remain tied up with a failing going nowhere program
* If Northrop had material doubts about Firefly’s path to a reliable first stage, it wouldn’t deepen its exposure, meaning they wouldn't have dropped another $50m towards it in May 2025
* The fact that Northrop is still there is as good as any validation of Eclipse’s core architecture for me
Now yes, Miranda is a tap-off engine. Yes, tap-off is less common than staged combustion, and yes people love to call it weird and all sorts of shit, usually people who know nothing about the topic incidentally. But understand the rationale at play here - it is simple, fewer subsystems, very reliable, and manufacturable at scale.
There are 2 massive tells for me here that this will likely be a success:
1. **Qualification** \- testing is underway with 60+ Miranda hot fire tests, including full duration fires...ok it could still fail
2. **Hiring** \- the even bigger tell (see below) 170 ish roles being hired for
When are we launching? 2026 is the target. It already slipped from 2025. My personal expectation is later rather than earlier (meaning Q3/Q4 feels more realistic than “early 2026”)
**Propulsion hiring**
Firefly is going big on propulsion hiring - manufacturing engineering, production quality, supplier quality, test operations, and reliability. These are not roles associated with figuring out whether the shit works. They are roles associated with scaling a design that is already locked, and on that I firmly believe Miranda architecture is now likely frozen. If Miranda were still in trouble or likely to fail, the hiring would be for combustion instability, cycle redesign, advanced thermal modelling - it isn't.
**Systems engineering**
There's a massive emphasis on vehicle-level systems engineering, integration, mission assurance, and program execution roles. For me this implies that Miranda is now treated internally as a known subsystem within a larger vehicle architecture (again, we're past the "will this shit work" stage), not the central uncertainty.
**Blue Ghost + Elytra:**
As with the SciTec acquisition, this is grossly misunderstood
* **Blue Ghost Mission 1** was a proof point that Firefly can execute end-to-end in deep space. Blue Ghost landed on March 2, 2025, delivered 10 NASA instruments, and operated on the surface for roughly one lunar day (14 Earth days). NASA liked what they saw, because they came back for more...
Firefly has a repeatable Commercial Lunar Payload Services cadence already mapped out.
* **Mission 2** is targeted for 2026 to the lunar far side with LuSEE-Night (NASA/DOE/UC Berkeley)
* **Mission 3** is targeted for 2028 at the Gruithuisen Domes with six experiments under NASA’s lunar payload selections
* **Mission 4** is even more explicit on economics and intent - NASA awarded $177M for a south pole delivery
It's a contracted multi-mission pipeline for those who missed it...
**Then we turn to Elytra...**
Elytra is where this whole thesis shifts from “lander missions” to “infrastructure and services.” On Mission 4 Elytra Dark will deploy Blue Ghost into lunar orbit, then remain operational to provide a long-haul communications relay, and after surface operations it will stay in lunar orbit for 5+ years supporting Firefly’s Ocula lunar imaging service.
That is not trivial. Building and operating an orbital vehicle that can execute lunar transfer/orbit insertion, sustain power/thermal/communications over long durations, and provide persistent services is materially harder than building a one-off lander. It also pushes Firefly up the value chain: instead of being paid once for delivery, they can be paid repeatedly for ongoing capability (relay + imagery + domain awareness).
The deeper implication, for those who have missed it, is that NASA doesn’t structure repeat lunar payload awards around a provider it believes can’t deliver, lunar payload is competitive and political. If Elytra performs as advertised, Firefly starts looking like a lunar infrastructure operator, which is a fundamentally different revenue and valuation profile.
**Bringing it together:**
Firefly is trying to become a prime with the trifecta below, Spanish added for effect
* **Uno - Launch and spacecraft** give them access to orbits/trajectories + payload delivery
* **Dos - Elytra/Ocula/mission ops** give them persistent presence + data capture
* **Tres - SciTec** gives them the software + data processing + classified footprint that defense buyers lap up
**Valuation and float:**
* \~$4.0Bn market cap
* \~159.25M shares outstanding
* \~76.05M shares in the free float
* on Feb 3, 2026, \~121.95M shares unlock (note this point, although I think the dip will be contained to 10/15%)
* yours for 1/10th of the valuation of RKLB (currently with good reason, as above, this will change)
**SpaceX IPO:**
Add this into the mix if it happens and we re-rate the entire industry, with moderate execution and nothing more than a finger waving in the air, I think a reasonable impact is:
* **$10B–$15B** market cap if Firefly show credible defense-software growth + program embedment + Eclipse looks real and actually works
* **$15B–$25B** if they start stacking sticky renewals / scope expansions and Eclipse hits meaningful milestones (reuse progress being a big one)
**What could go wrong?**
A shit load - Alpha (Firefly's small launch rocket) has had 6 flights:
* 2 complete failures
* 2 partial failures
* 2 successes
But we don't worry about that.