We Cut Our ASTS Base Value 42% to $43.45 After Two Downgrade Conditions Fired
Key Takeaways
- The deployment test we set in August went backward. No BlueBird launched in the third quarter. BlueBirds 14 to 16 reached Cape Canaveral in September and were still waiting for a SpaceX launch date on the call, the next batch of three rides New Glenn's return to flight, and the target of about 45 satellites in orbit moved to mid-2027. That is the third date for the same milestone in three quarters, and end-user beta testing now starts in early 2027.
- Revenue of $42.3 million missed the most-cited $45.7 million estimate by 7.5%. It rose 34.3% sequentially and fell below every consensus figure we found, which ranged from $44.5 million to $47.9 million. Services revenue tripled to $21.75 million at an 86.8% direct margin, while gateway products sold below their direct cost. The reaffirmed $150 million to $200 million annual range needs a fourth quarter of $61.4 million to $111.4 million, and we lower our 2026 estimate to $165 million from $175 million.
- Capital commitments grew faster than the funding that backs them. The Blacklift rocket program commits about $800 million upfront to Firefly ($400 million in cash and 10 to 14 million shares), another $500 million for engines and engineering, and several hundred million dollars of 2027 capital spending. The Japanese subsidy of ¥148 billion (about $938 million) is real, but AST must also fund about $1.4 billion of the joint venture's preferred shares. Our bridge takes net debt from $1.36 billion at September 30 to about $3.75 billion by November 2027, which requires new financing next year.
- ASTS fell 53.0% from our August reaction close of $71.63 to $33.69. Most of the fall came before the results: the shares were 48.5% lower at the pre-print close and fell a further 8.7% on the results. Two of the four sessions with drops of 10% or more followed SpaceX news. SpaceX agreed to buy nationwide 800 MHz low-band spectrum in October and was reported to be in talks for a mobile carrier in November, and it controls the launch that AST's waiting satellites need. No one on the call named SpaceX as a competitor.
- Rating: Maintaining Hold; conviction falls to 4 from 5. Two of the three Underperform conditions we published in August fired. At the August price they would have taken us to Underperform, because our re-derived $43.45 base value is 39% below $71.63. At $33.69 the base offers 29.0%, but the bear case now sits near the equity's floor, and our probability-weighted $37.24 is only 10.5% above the price.
Results vs. Consensus
AST SpaceMobile reported third-quarter revenue of $42.321 million, up 34.3% from $31.520 million in the second quarter and 2.87 times the $14.739 million of a year ago. Every consensus figure we found sat above the print. The net loss attributable to common stockholders narrowed to $(132.380) million from $(230.909) million, because the second quarter carried the $125.911 million BlueBird 7 charge. The loss per Class A share of $(0.44) was $0.03 wider than the $(0.41) consensus.
Q3 2026 Scorecard
| Metric | Q3 2026 actual | Consensus / guide | Result | Magnitude |
|---|---|---|---|---|
| Revenue | $42.321M | ~$45.7M most-cited estimate; range $44.5M to $47.9M | Miss | (7.5)% vs. $45.73M |
| Loss per Class A share (GAAP) | $(0.44) | $(0.41) | Miss | $0.03 wider |
| Adjusted opex excluding adjusted cost of revenues | $109.386M | $105M to $115M guide | Within guide | Includes a $4.5M one-time R&D tax credit |
| Management capital-expenditure metric | ~$310M | $350M to $425M guide | Below guide | $40M below the low end; launch-payment timing |
| Revenue less cost of revenues | $17.007M | n/a | Up sequentially | 40.2% direct margin vs. 25.2% in Q2 |
| Operating cash flow less cash PP&E | $(464.246)M | n/a | Smaller outflow than Q2 | Q2: $(694.770)M |
Year-Over-Year Comparison
| Metric | Q3 2026 | Q3 2025 | Change |
|---|---|---|---|
| Total revenue | $42.321M | $14.739M | +187.1% |
| Products revenue | $20.571M | $7.746M | +165.6% |
| Services revenue | $21.750M | $6.993M | +211.0% |
| Revenue less cost of revenues | $17.007M | $9.228M | +84.3% |
| Operating loss | $(168.705)M | $(79.676)M | $89.029M wider |
| Net loss attributable to common stockholders | $(132.380)M | $(122.874)M | $9.506M wider |
| Loss per Class A share | $(0.44) | $(0.45) | $0.01 narrower |
| Adjusted opex excluding adjusted cost of revenues | $109.386M | $67.712M | +61.5% |
Sequential Comparison
| Metric | Q3 2026 | Q2 2026 | Change |
|---|---|---|---|
| Total revenue | $42.321M | $31.520M | +34.3% |
| Products revenue | $20.571M | $24.428M | (15.8)% |
| Services revenue | $21.750M | $7.092M | +206.7% |
| Revenue less cost of revenues | $17.007M | $7.953M | +113.8% |
| Operating loss | $(168.705)M | $(297.577)M | $128.872M narrower |
| Operating loss excluding loss on involuntary conversion | $(168.658)M | $(171.666)M | $3.008M narrower |
| Net loss attributable to common stockholders | $(132.380)M | $(230.909)M | $98.529M narrower |
| Loss per Class A share | $(0.44) | $(0.77) | $0.33 narrower |
| Adjusted opex excluding adjusted cost of revenues | $109.386M | $95.882M | +14.1% |
| Management capital-expenditure metric | ~$310M | ~$610.4M | Launch-payment timing |
| Operating cash flow less cash PP&E | $(464.246)M | $(694.770)M | $230.524M smaller outflow |
Third-quarter cash-flow lines are the nine-month figures less the six-month figures: operating cash flow of $(169.265) million and cash purchases of property and equipment of $(294.981) million.
Nine-Month Comparison
| Metric | 9M 2026 | 9M 2025 | Change |
|---|---|---|---|
| Total revenue | $88.575M | $16.613M | +433.2% |
| Net loss attributable to common stockholders | $(554.301)M | $(267.974)M | $286.327M wider |
| Loss per Class A share | $(1.87) | $(1.09) | $0.78 wider |
| Net cash used in operating activities | $(314.477)M | $(136.486)M | $177.991M more cash used |
| Purchase of property and equipment | $(1,154.196)M | $(669.045)M | $485.151M more spent |
Quality of the Miss
Revenue: The miss was concentrated in gateway equipment, where product revenue fell 15.8% sequentially to $20.571 million. Gateway deliveries follow operators' build schedules and are lumpy by nature, but they also follow the constellation: an operator has less reason to rush ground equipment while the satellites it would connect to are waiting at the Cape. Services revenue tripled to $21.750 million on U.S. government milestones, which is the better-quality revenue and the line that most supports the annual range. The quarter came in $5.7 million below our own $48 million estimate. Nothing in it was a commercial subscription dollar: the service has not launched.
Margins: Revenue less cost of revenues rose to $17.007 million, a 40.2% direct margin against 25.2% in the second quarter. The improvement is entirely mix. Services earned $18.887 million at 86.8%, and products lost $1.880 million at a (9.1)% margin. Over nine months, products have earned a 4.3% direct margin, below the 10% we assume for gateways in our forward years. The direct contribution covered about 16% of the $109.386 million adjusted operating cost base. Higher-margin government work is the route to better economics, but it rises and falls with milestone timing.
EPS: The $(0.44) loss per Class A share is cleaner than the second quarter's $(0.77), because the $125.911 million BB7 charge is gone and net other expense was only $(6.890) million. Interest expense of $30.340 million and interest income of $30.110 million now offset almost exactly, and that balance will tilt toward expense as cash is spent on the Firefly payments and fourth-quarter capital spending. The loss allocated to noncontrolling interests, $(39.430) million, again absorbs part of the operating loss for Class A holders. Excluding the charge, the operating loss narrowed by only $3.0 million, because operating costs rose $7.8 million against a $10.8 million revenue gain.
Segment Performance
ASTS reports a single segment. Its two revenue categories have very different economics: products are mostly gateway equipment and software sold to operators, and services are mostly milestones under U.S. government contracts. The commercial service, which carries the equity value, has generated no revenue yet.
Revenue Categories, Q3 2026
| Revenue category | Q3 revenue | Cost of revenues | Direct contribution | Direct margin | Revenue QoQ |
|---|---|---|---|---|---|
| Products (gateways, software) | $20.571M | $22.451M | $(1.880)M | (9.1)% | (15.8)% |
| Services (mostly U.S. government) | $21.750M | $2.863M | $18.887M | 86.8% | +206.7% |
| Total | $42.321M | $25.314M | $17.007M | 40.2% | +34.3% |
Cost of revenues is reported exclusive of depreciation and amortization and includes $1.137 million of stock-based compensation.
Revenue Categories, Nine Months
| Revenue category | 9M 2026 revenue | Cost of revenues | Direct contribution | Direct margin |
|---|---|---|---|---|
| Products | $58.404M | $55.916M | $2.488M | 4.3% |
| Services | $30.171M | $4.614M | $25.557M | 84.7% |
| Total | $88.575M | $60.530M | $28.045M | 31.7% |
Gateway Products: The Ground Network Grows, the Margin Does Not
The presentation counts more than 60 gateways deployed, installed or in planning, against nearly 50 at the August update, so operators continue to build the terrestrial side. Product revenue nevertheless fell to $20.571 million, and the category sold below cost for the quarter. Related-party product revenue, mostly gateway equipment sold to an equity-method investee, was $1.422 million in the quarter and $11.192 million over nine months. Management expects gateway sales above $100 million next year.
"We expect revenue to continue to be driven by gateway deliveries, achievement of contracted milestones for the U. S. Government MNO consulting services and software licensing." — Andrew Johnson, CFO
Assessment: Gateways are enabling infrastructure, not a profit line. We lower our 2027 gateway estimate to $125 million from $150 million and keep the 10% forward direct margin, which is already above the nine-month 4.3%. A weak gateway quarter matters mainly as a signal of operators' pace while the constellation lags.
Government Services: The Quarter's Strongest Line
Services revenue of $21.750 million exceeded the full first-half total of $8.421 million (Q1 plus Q2) by a wide margin. The release adds an approximately $100 million U.S. Space Force contract for space-based sensing, 16 U.S. government awards to date, and a priority rating under the Defense Priorities and Allocations System for a dual-use satellite program. Services carried an 86.8% direct margin.
"This is a small step towards a much larger program of record for AMTI, or Airborne Moving target indication, a capability which is highly strategic for The United States and we believe to be a top priority for the Space Force." — Scott Wisniewski, President
Assessment: The government pillar strengthened again, and it is the only revenue line that is growing on schedule. We still cut our 2027 government estimate to $275 million from $350 million, because milestones under these contracts generally need satellites in orbit and the constellation is later. Management's own floor of "$100 million-plus" already executed for 2027 sits well below our figure.
Commercial Service: Still Zero, Now Framed as Intermittent Activation
No subscription revenue has been recognized. The release promises nationwide intermittent service activation in the United States and four other markets (Canada, Japan, Saudi Arabia and the United Kingdom) before year-end. On the call, that means operators can see the network working when a satellite passes overhead, with end-user beta trials in early 2027 at about 25 satellites. Remaining performance obligations rose to approximately $1.3 billion from $1.2 billion, and the share expected within twelve months rose to about 10.1% from 6.6%, which implies roughly $131 million against roughly $79 million.
Assessment: The year-end milestone is a network test for partners, not a consumer launch, and it should not be read as the start of paid service. We move our base start of paid commercial service to the fourth quarter of 2027 from the second quarter, two quarters after management's mid-2027 target of about 45 satellites.
Key KPIs and the August Commitments
Constellation and Operating KPIs
| KPI | As of the Q3 update (Nov 9) | As of the Q2 update (Aug 10) | Read |
|---|---|---|---|
| BlueBirds in operation | 12 (BB1 to BB6, BB8 to BB13) | BB11 to BB13 launched Aug 5 | No launch since Aug 5 |
| Satellites shipped and waiting | BB14 to BB16 at the Cape since September | n/a | Awaiting a SpaceX date |
| Production horizon | Through BB51; BB28 in final stage | n/a | Manufacturing ahead of launch |
| Target of ~45 in orbit | Mid-2027 | Early 2027 | Delayed again |
| Gateways deployed, installed or planned | 60+ | Nearly 50 | Ground build continues |
| U.S. government awards to date | 16 | n/a | Space Force sensing award added |
| Remaining performance obligations | ~$1.3B; ~10.1% within 12 months | ~$1.2B; ~6.6% within 12 months | Near-term portion rising |
| Cash, equivalents and restricted cash | $3,233.480M | $2,722.834M (June 30) | July convertible received |
| Debt principal | $4,160.969M | $3,022.152M (June 30) | July convertible added |
What Management Committed To in August, and What Happened
| August commitment | Evidence by November 10 | Grade |
|---|---|---|
| BB14 to BB16 shipment and launch, then the next batch | Shipped in September; still awaiting a SpaceX launch date | Missed |
| Progress to ~45 BlueBirds in early 2027 without depending on New Glenn | Target now mid-2027; the next batch flies on New Glenn's return | Missed |
| Non-commercial beta later in 2026; paid service timing | Partner activation by year-end; end-user beta early 2027; no paid-service date | Slipped |
| Q3 capex $350M to $425M | ~$310M on launch-payment timing | Below range |
| Q3 adjusted opex excluding cost $105M to $115M; ~$400M for 2026 | $109.386M; Q4 guide implies $415.1M to $425.1M for the year | Quarter met; year raised |
| Sequential revenue growth; $150M to $200M for 2026 | +34.3%; range reaffirmed with potential upside | Met; below our estimate |
| Conversion of new government awards | ~$100M Space Force award; DPAS rating; 16 awards | Advanced |
| Final J-LEO terms and obligations | Signed; ¥148.0B grant to the joint venture; ~$1.4B AST preferred subscription | Resolved; costlier than the headline |
| Mid-band production; L-band resolution | Next-generation ASIC pulled forward to BB34; L-band still open | Mixed |
| Use of July proceeds; acquisition discipline | Firefly agreements worth about $1.3B plus Blacklift capex | Failed our test |
| Operator economics and the U.S. joint venture | Not disclosed | Open |
The August scorecard was a recovery story: two successful batches and booked launches with two providers. Three months later, the satellites are built but not flying, and the plan for regular access to orbit now runs through a rocket that AST does not expect to launch before late 2028.
Key Topics & Management Commentary
Overall Management Tone: Management was more expansive than at any point in our coverage, presenting a rocket program, a sovereign subsidy and a wider market (sensing, AI inference) as the story of the quarter, while the operating test it accepted in August, a regular launch cadence, produced no launch. Answers on payment timing and financing structure were specific; answers on when the waiting satellites fly were not. The posture moved from August's recovery focus to expansion. No one on the call named SpaceX's spectrum and carrier moves, which weighed on the stock from October. The one question on spectrum drew a general answer about AST's own strategy.
1. Zero Launches in a Quarter Built Around Cadence
In August, management described ten booked launches with two providers and a target cadence of one launch every month or two. The third quarter delivered none. BlueBirds 11 to 13, launched on August 5, completed deployment by September 30, so the operating fleet stands at five Block 1 and seven Block 2 satellites. The next three were shipped to Florida in September and, on November 9, still had no confirmed date from the provider.
"With respect to manufacturing, in September of this year, we shipped BlueBirds 14, 15, and 16 to Cape Canaveral and are currently awaiting their SpaceX launch date confirmation." — Andrew Johnson, CFO
The CEO said the company hoped the date was imminent, and gave no reason for the wait. The batch after that, BlueBirds 17 to 19, ships by the end of November for New Glenn's return to flight in early 2027, and six more (BlueBirds 20 to 25) ship in December and January for launches under the government priority rating in the first quarter. If all of that happens, twelve satellites reach orbit in roughly four months. In the three months since our last note, none did.
Assessment: Bear-1 stays materializing and has worsened. In August we credited the recovery because two batches had flown and the plan no longer depended on a single vehicle. This quarter, the binding constraint was a launch date, not a satellite, and the next batch is again tied to New Glenn.
2. The 45-Satellite Target Moves for the Third Time
About 45 BlueBirds in orbit has been the threshold for continuous service in the most attractive markets throughout our coverage. The date was end-2026 at the first-quarter call, early 2027 in August, and mid-2027 now. The 10-Q keeps the coverage framework: 25 satellites for the most attractive markets, 45 to 60 for continuous service across the United States, Europe and Japan, and about 90 for wider coverage. Launch agreements now cover more than 100 satellites cumulatively across 2027 and 2028.
"That manufacturing progress positions us well to support our current network deployment plan, targeting approximately 45 BlueBird satellites in orbit by the middle of [2027]." — Andrew Johnson, CFO
Reaching about 45 by mid-2027 requires about 33 more satellites in roughly seven months. In the roughly eleven months since December 2025, the company has placed seven working Block 2 satellites in orbit and lost an eighth.
Assessment: We no longer take the deployment calendar at face value. Our base case assumes about two quarters of slippage against the stated plan, which moves paid commercial service to the fourth quarter of 2027. A quarter in which management delivers its first-quarter launches on time would be evidence against that haircut.
3. "Service Activation" Is a Partner Test, Not a Consumer Launch
The headline promise of the release is nationwide intermittent service activation in the United States, Canada, Japan, Saudi Arabia and the United Kingdom before year-end. The answer to a question on the call defined it more narrowly: operators will see gateways working, spectrum available and nationwide availability when a satellite is overhead. Beta use by end customers, run by the operators, comes in early 2027 and works best with about 25 satellites.
"First, service. We are on track for nationwide intermittent service activation in United States and other key markets before year end 2026. This is the moment we have been building towards." — Abel Avellan, Chairman and CEO
Two pieces of evidence support the claim that the network works: the first handover of a live connection between AT&T's terrestrial network and AST's satellites, and the 10-Q's statement that BB6 and BB8 to BB10 were ready for their intended use from September 1. Neither tells us when a paying subscriber appears, or at what price.
Assessment: The beta milestone has moved from summer 2026 (the plan until the first-quarter call) to later in 2026 (August) to early 2027. The year-end activation is useful for integration, but we treat it as a technical milestone. Bull-4's demand evidence remains intact, while its monetization moves later again.
4. Blacklift: Buying Launch Control at a Price Paid Before Any Benefit
AST will build its own reusable 20-ton rocket, Blacklift, designed to carry up to five BlueBirds per launch from late 2028. It will license Firefly Aerospace's Miranda and Vira engines in perpetuity and manufacture them in a new U.S. plant. The license costs about $800 million upfront, $400 million in cash and 10 to 14 million Class A shares set by a forward VWAP. A separate supply and services agreement adds $250 million for the engines of the first three rockets and $250 million of engineering support over three years. Closing is expected in the fourth quarter.
"The Blacklift rocket is the springboard to significantly expand our total addressable market, providing AST SpaceMobile with regular unconstrained access to orbit and providing for most favorable owners’ economics in the industry." — Abel Avellan, Chairman and CEO
"Additionally, we are assessing our forward CapEx and expect to spend several hundred million dollars of incremental CapEx in 2027 associated with [Blacklift]." — Scott Wisniewski, President
The economic case is that launch is "a significant majority" of unit cost, so owning the vehicle should take the $21 million to $23 million average satellite cost "much lower." The engine has a test record (more than 175 hot-fire tests), but the vehicle is new, and the first launch falls after the 45-satellite and continuous-service milestones that decide the next two years.
Assessment: This is the "expansion spending that materially worsens the funding bridge" we named in August. About $900 million of cash commitments, a share issue, and several hundred million dollars of 2027 capital spending go out before the rocket can lower any cost. We include the costs in our base case and give no launch-cost saving before 2030; any saving is bull-case value.
5. J-LEO: A Real Subsidy Inside a Larger Commitment
Rakuten and AST formed Rakuten AST, 51% and 49% owned, to build a Japanese-flagged mid-band constellation, with domestic operations to start by the end of March 2029. Japan's Ministry of Internal Affairs and Communications will grant the joint venture up to ¥148.0 billion (about $938 million) over about three years. The 10-Q adds the other side: AST subscribes for no-coupon preferred shares to fund the remaining cost, an estimated $1.4 billion over about three years. The joint venture buys satellites, launch support and software from AST, including a one-time $75 million software license.
"We expect the funding to be made ahead of investment in spacecraft and launch. With the majority of proceeds received during 2027. In consideration for this subsidy, the joint venture which we will hold 49% of the equity will receive the revenue for 3 targeted exclusive markets in Asia." — Scott Wisniewski, President
Satellite capacity outside Japan joins AST's global mid-band network, and AST shares a low single-digit percentage of qualifying revenue with the joint venture. Management describes the net cash subsidy as about $1 billion. The accounting treatment, and therefore how much of the flow becomes AST revenue, is to be determined.
Assessment: The grant is the best-quality capital AST raised this year, and it de-risks the mid-band roll-out in one market. But it comes with a $1.4 billion preferred commitment, revenue in Japan and three other markets that belongs to the joint venture, and a delivery obligation tied to a constellation that is late elsewhere. We credit $500 million of grant receipts against net debt by November 2027, and no J-LEO revenue in our base case beyond a small international line.
6. Government and Sovereign Demand Keep Building
The government leg added an approximately $100 million Space Force sensing award, a dual-use priority rating that gives access to government launches, and a count of 16 U.S. government awards. Management described a handful of active international sovereign discussions on the J-LEO model and a wider set of applications: radar, secure communications, emergency networks and AI inference in orbit.
"We also have access to further launches for U. S. Government missions alongside our DPAS-rated programs, the Defense Priorities and Allocation System, which helps assure that strategic resources for national security are timely available to meet current national defense and emergency preparedness program requirements." — Scott Wisniewski, President
The priority rating matters most to the deployment question, because it is a route to orbit that does not wait in a commercial queue. Six satellites are planned for those launches in the first quarter.
Assessment: Bull-3 is the pillar that strengthened. Government revenue is the fastest-growing line, the awards are getting larger, and the sovereign model now has a signed example. The limit is the same as in August: most of this work needs the same satellites in orbit.
7. The Revenue Range Holds on a Fourth Quarter That Must Nearly Double
Nine-month revenue is $88.575 million. The $150 million to $200 million range therefore needs $61.425 million to $111.425 million in the fourth quarter, 1.45 to 2.63 times the third quarter. Management reiterated the range, claimed potential upside from recent contract signatures and the J-LEO program, and listed contingencies: satellite deployment for government milestones, gateway sales and service revenue.
"We continue to work on several Q4 revenue opportunities that will ultimately determine where we fall within our guidance range, but have high confidence that revenue acceleration this quarter will result in achievement of our full year 2026 revenue guidance with potential upside." — Andrew Johnson, CFO
The third quarter came in 11.8% below our own estimate, and the range's midpoint now needs a fourth quarter 2.04 times the third.
Assessment: We lower 2026 to $165 million, which needs $76.425 million in the fourth quarter (+80.6% sequentially), inside the range but below its midpoint. A one-time J-LEO license fee could provide the upside management mentioned. We leave it out until the accounting is known.
8. Costs Step Up Again, and Capex Was Lighter Only Because Launches Were Late
Adjusted operating expenses excluding adjusted cost of revenues were $109.386 million, inside the $105 million to $115 million guide. A one-time $4.5 million R&D tax credit helped: without it, the figure was $113.9 million. The fourth-quarter guide of $130 million to $140 million is 23.4% above the third quarter at the midpoint, and it takes the year to $415.1 million to $425.1 million, against the roughly $400 million framework in August, which implied about $114 million for the fourth quarter.
"For the fourth quarter of 2026, we estimate that our adjusted operating expenses, excluding adjusted cost of revenues, will increase to the range of approximately $130 million to $140 million as we continue to absorb the full quarter of the cost of our expanded workforce..." — Andrew Johnson, CFO
"This amount for the quarter was below the guidance of $350 million to $425 million that I provided during our last earnings call, primarily due to the timing-related shift in launch payments." — Andrew Johnson, CFO
Capital spending was about $310 million because launch payments shifted into the fourth quarter, which is guided at $400 million to $500 million. That gives about $1.58 billion to $1.68 billion for the year on the management metric.
Assessment: Bear-3 moves to materializing. Fourth-quarter operating costs of $130 million to $140 million compare with third-quarter revenue of $42.3 million, the cost base now includes the rocket program, and the capital underspend is a timing item caused by the launch delay, not discipline. We raise our 2027 adjusted operating cost estimate to $650 million from $550 million.
9. The Funding Bridge: Funded for the Satellites, Not for Everything Now Planned
Cash, equivalents and restricted cash were $3,233.480 million at September 30, of which $434.569 million was restricted. Debt principal was $4,160.969 million, so net debt against unrestricted cash was $1,362.1 million. The 10-Q keeps the claim that the company is fully funded to manufacture and launch about 90 BlueBirds and says it plans to raise more capital. The new $1.2 billion "quasi-governmental" pipeline is mostly debt.
"We believe that we are fully funded for our costs necessary to manufacture and launch a constellation of approximately 90 BB satellites." — Form 10-Q, quarter ended September 30, 2026
"And then the balance is think of it as first lien or CapEx support type of debt." — Scott Wisniewski, President
The constellation claim does not cover the new uses: Firefly payments of about $900 million in cash, the Blacklift capital program and the J-LEO preferred subscription. Our bridge to November 2027 consumes about $2.4 billion net of $500 million of J-LEO receipts. That leaves about $0.4 billion of unrestricted cash before new financing and takes net debt to about $3.75 billion.
We also correct our own August framework. It assumed $2.0 billion of net debt at the target date, which allowed about $1.1 billion of cash consumption over roughly thirteen months. The third quarter alone consumed $464.2 million of operating cash flow less capital purchases, and the fourth-quarter guide plus the Firefly upfront payment imply about $1 billion more. The August figure understated the burn before the rocket program was announced.
Assessment: Bull-1 moves to at risk. Liquidity covers the next twelve months, as the 10-Q states, but the 2027 build now requires a financing event. Whether it is secured debt, subsidy or common equity at depressed prices decides much of the per-share outcome.
10. The Competitive Question the Call Did Not Address
Two of the stock's four worst sessions between the two prints followed SpaceX news. On October 8, SpaceX agreed to buy a nationwide 800 MHz low-band portfolio (up to 14 MHz of paired spectrum, reported at about $8 billion), on top of its earlier purchase of EchoStar spectrum and its authorization for a much larger direct-to-cell constellation. On November 2, a press report said SpaceX was in talks to buy, or use the network of, EchoStar's Boost Mobile. Amazon's purchase of Globalstar was approved in October. Neither the release, the call nor the 10-Q mentions SpaceX as anything but a launch provider.
"And we think that the best approach and the fastest approach to market is partnering with the global operators, and we will continue to focus on that." — Abel Avellan, Chairman and CEO
"Going forward, we are never going to be the top tick on spectrum buys, of course, but we will be smart and we are following that organic strategy." — Scott Wisniewski, President
The partner-first answer has real force. A rival that owns spectrum and may own a carrier is a threat to the operators, which gives them more reason to back a wholesale partner, and the release says AST expects to work with all major operators in the United States and Europe. The weakness is supply: the satellites at the Cape wait on that rival's rocket.
Assessment: We add Bear-4, competition from a vertically integrated rival, at emerging. It affects our valuation through a lower commercial user base and a lower terminal multiple, and it explains why management wants its own rocket.
11. ASIC Pulled Forward, and a Factory Running Ahead of the Launch Pad
The next-generation ASIC now enters production at BlueBird 34 instead of BlueBird 50. Management said it raises processing bandwidth ten times compared with the in-orbit Block 1 satellites. Production is under way through BlueBird 51 in two Midland factories, with a rate of six ASIC satellites a month targeted for December. BlueBird 11 took longer than its two batch-mates to unfold, and management said the cause requires no design change.
"Every satellite we launch from BlueBird 34 onwards, will be dramatically more capable, 10x. And we expect to hit 6 satellites per month in December." — Abel Avellan, Chairman and CEO
"We needed to literally take it to spin in order to warm it up and open it." — Abel Avellan, Chairman and CEO
Spectrum also advanced: Brazil authorized commercial service on 50 MHz of combined S-band and low-band spectrum, and the target is about 100 MHz of access in the United States and more than 60 MHz globally, combining operators' spectrum with AST's own.
Assessment: Bull-2's technology case is the strongest it has been. The ASIC raises capacity and lowers cost per gigabyte from BlueBird 34, so the delay in launching BlueBirds 14 to 33 matters less for long-term capacity than it does for the date of first revenue. Six satellites a month, however, is twice the pace of the twelve satellites scheduled to launch over the next four months, so inventory, not capability, becomes the constraint.
Guidance & Outlook
| Metric | Prior (August call) | Current (November call) | Change |
|---|---|---|---|
| FY2026 revenue | $150M to $200M | $150M to $200M, "with potential upside" | Maintained |
| Q4 2026 adjusted opex excluding adjusted cost of revenues | ~$114M implied by ~$400M for the year | $130M to $140M | Raised |
| FY2026 adjusted opex excluding adjusted cost of revenues | ~$400M | $415.1M to $425.1M implied | Raised |
| Q3 2026 capital expenditure | $350M to $425M | ~$310M actual | Below; timing |
| Q4 2026 capital expenditure | n/a | $400M to $500M | New |
| ~45 BlueBirds in orbit | Early 2027 | Mid-2027 | Delayed |
| BlueBirds deployed under launch agreements | 10 launches booked with two providers | 100+ satellites cumulatively over 2027 and 2028 | Expanded |
| End-user beta | Later 2026, non-commercial | Early 2027, about 25 satellites | Delayed |
| Nationwide intermittent activation for partners | n/a | Before year-end 2026, five markets | New |
| Revenue in the 12 months after commercial service starts | Approaching $1B | Approaching $1B; "increasingly confident" | Maintained |
| Average capital cost per satellite | $21M to $23M | $21M to $23M; "much lower" with Blacklift | Maintained |
| Blacklift | n/a | First launch late 2028; several hundred million dollars of 2027 capex | New |
| BlueBird production rate | n/a | Six ASIC satellites a month from December | New |
Management's guidance is now organized around milestones it controls (production, ground activation, contract signings) and away from the launch schedule it does not. The financial guide is explicit for the fourth quarter only. For 2027, the only revenue framework is the run rate in the twelve months after commercial service begins, which moves with the service date.
Implied fourth-quarter ramp: The low end of the range needs $61.425 million, a 45% sequential increase. The midpoint needs $86.425 million, 2.04 times the third quarter, and the top end needs $111.425 million. Our $76.425 million estimate assumes $40.0 million of products and $36.425 million of services.
2027: Management agreed that a breakdown of gateways, U.S. government work, international work and commercial service was "generally right," and declined to give a calendar-year number. Its own floors are more than $100 million of gateways and more than $100 million of executed government work. Our $500 million estimate ($125 million gateways, $275 million government, $50 million international and $50 million of paid service) sits at the low end of the $500 million to $1 billion framework raised on the call.
Guidance style: Financial guidance has been reliable within ranges: the revenue range has held all year, and each quarter's operating cost has landed within or just above its guide. The deployment calendar has not been reliable, and the revenue ramp after 2026 depends on it. We give the financial guide credit and haircut the calendar.
Analyst Q&A Highlights
How to Build 2027 Revenue Without a Service Date
The opening question offered a calendar-2027 framework: $100 million to $200 million of gateways, $200 million to $300 million of U.S. government work, $100 million to $200 million of international work, and commercial service as the swing factor, for $500 million to $1 billion in total. Management restated its guidance for the twelve months after commercial service begins, gave floors for gateways and government, and endorsed the building blocks without a calendar-year number.
Q: "As you think about 2027 revenue building blocks, rough numbers, is it fair to think of so like if we break it apart into gateways, US government work, international work, and commercial service, you know, assuming, call it, $100 to $200 million of gateway work, maybe $200 to $300 million of US government work, and then maybe rough numbers $100 to $200 million of international work and swing factor is commercial service to get to kind of the midpoint of $500 million to $1 billion revenue opportunity."
— Colin Canfield, Cantor Fitzgerald
A: "So we stand behind our guidance on the 12 months following commercial service start, right? We see that approaching $1 billion... in terms of looking at calendar year next year, I think over $100 million-plus of gateway sales probably makes sense. But otherwise, we are not going to give obviously forward guidance on revenue, but I think your building blocks are generally right."
— Scott Wisniewski, President
Assessment: The $1 billion figure is tied to a start date that has moved every quarter, so it is not a calendar commitment. The floors management did give, about $100 million each for gateways and government, add up to well under half of the framework's low end. Our $500 million sits at that low end and depends on government growth more than on paid service.
When the Firefly Payments Leave the Balance Sheet
A follow-up asked for the timing of the Firefly cash, share and service payments. The answer was precise: the upfront cash and stock go out within days of closing, the service fee is paid roughly evenly over three years, and the engine fee is paid partly upfront and the rest on delivery. Management presented the share component as a way to preserve cash.
Q: "And then just for my 1 follow-up question, and I do not know, maybe it is best for Andy, but you give rough timelines on when some of these components get paid to Firefly the $400 million of cash."
— Mike Crawford, B. Riley Securities
A: "Upfront cash and consideration in stock will be paid imminently within days. And some of those shares will be held back in order to true up for that VWAP calculation. And then the service agreement will be paid out over 3 years, pretty much pro rata... But importantly, we designed it to have a mix of cash and stock so that the upfront payment was a little lower, and we could preserve capital for other things."
— Scott Wisniewski, President
Assessment: The answer confirms that $400 million of cash leaves in the fourth quarter and that the share issue arrives at a price near the 52-week low. Paying in stock preserves cash but transfers the cost to existing holders. If the share half of the roughly $800 million upfront consideration is worth about $400 million, 10 million shares imply a VWAP of about $40.00 and 14 million about $28.57.
Whether Licensed Engines Shorten a Rocket Program
Another question challenged the engine lineage directly, citing the success rate of the smaller rocket whose engine technology the Miranda builds on, and asked why reusability should work when few companies have proven it. Management argued that licensing a tested engine skips the usual small-rocket stage and that AST's composite manufacturing transfers to rocket structures.
Q: "So the alpha engine, I think it had about a 57% success rate. With that engine. So, you know, first, you know, what gives you confidence in the next generation engines?"
— Michael Funk, Bank of America
A: "So what we have done here is basically bypass all that process, which takes anywhere between 7 and 10 years historically... We had acquired the IP with certain parameters, but we have acquired the IP to manufacture the engines."
— Abel Avellan, Chairman and CEO
Assessment: The answer addresses the engine and not the vehicle. Owning a tested engine design removes one development risk, but the stages, recovery, launch-site upgrades and flight qualification are new, and the late-2028 first launch has no margin shown. We treat the date as a target with high slippage risk.
How Firm the Launch Commitments Are
An investor asked how firm the launch commitments are and how much visibility the company has over twelve months. The answer cited contracted launches, a promising New Glenn return and the government priority rating. It did not mention the satellites waiting at the Cape.
Q: "And then maybe just on the launch side, how firm are the launch commitments that you have received? And how much visibility do you have on that over the next 12 months?"
— Timothy Horan, Oppenheimer
A: "So we feel very good. I mean, we have a lot of contracted launch. Our launch with Blue Origin looks very promising. They are getting back to the pad very soon. And then we also have this DPAS rating that gives us access to a government priority launch."
— Scott Wisniewski, President
Assessment: This was the most important question of the call, and the answer was confidence instead of dates. Contracted launch capacity is not the issue; launch timing is. The answer leans on a vehicle that is still returning to flight and a priority rating that has not yet produced a launch.
Why Only Three Satellites on New Glenn's Return
A short exchange established that the New Glenn launch will carry three BlueBirds, below the vehicle's capacity, as deliberate risk management for the return to flight. Management added that it has satellites ready through BlueBird 28.
Q: "I think you mentioned for the upcoming new Glenn launch, it is 3 satellites, is below the rocket's capacity. Is that due to the availability of satellites? Or is that a risk mitigation as they return to flight?"
— Chris Quilty, Quilty Space
A: "That is a risk mitigation as they return to flight. We obviously we have satellites to launch all the way to satellite 28. In the very early launch from Blue Origin coming back to launch, we do not want to put more than 3 satellites on the on the immediate launch."
— Abel Avellan, Chairman and CEO
Assessment: Limiting exposure after BB7's loss is prudent. It also confirms that inventory is not the constraint: satellites through BB28 are ready while three fly on the next available rocket. Each launch-provider decision now caps deployment more than the factory does.
What the $1.2 Billion of "Non-Dilutive" Financing Is
A two-part question asked what the $1.2 billion quasi-governmental pipeline consists of and how firm it is, and what the Blacklift program costs after 2027. The answer described a small subsidy component, one piece awaiting final approval and a much larger piece about six months out, and said the balance is secured or capital-expenditure debt. It declined guidance for 2028 Blacklift spending.
Q: "First, wanted to see if you could expand on the additional pipeline of $1.2 billion non-dilutive U. S. Quasi government financing. If you could just talk about maybe what that is, how firm it is, and what can the funding be used for."
— Bryan Kraft, Deutsche Bank
A: "The $1.2 billion that we referenced in the pipeline for quasi-governmental financing is very serious... A few small components are subsidies that are active in particular 2 programs, 1 of which we have received and another that is in process. And then the balance is think of it as first lien or CapEx support type of debt."
— Scott Wisniewski, President
Assessment: "Non-dilutive" here mostly means secured debt, which avoids issuing shares but adds to the claims ahead of equity. Our net-debt bridge treats it as debt. The missing 2028 Blacklift figure is a material gap, because 2028 is the year of first launch and peak build.
What "Intermittent Service" Means for Subscribers
An analyst asked whether the year-end trials need 16 satellites and whether they will be broadly available. The answer separated the network activation, delivered to operators, from end-user beta, which the operators run in early 2027 and which works best with about 25 satellites.
Q: "For the planned trials for the intermittent service, prior to year end, Are these trials contingent on 16 active satellites or will the trial still proceed with 13? And also will these trials be broadly available to wireless customers? Or will it be for a very select user pool?"
— Louis De Palma, William Blair
A: "So, the coverage that we are lighting up for our top partners in a few key markets is providing the service to them... So, providing beta to the end user, that is with our customer... So we think that those beta trials are best with about 25 satellites, although there is some flexibility there."
— Scott Wisniewski, President
Assessment: The answer makes clear that no consumer uses the service this year. With 12 BlueBirds operating and the first-quarter launches not yet dated, a 25-satellite beta in early 2027 depends on nearly all of the scheduled first-quarter launches happening on time.
Net Cash From J-LEO After AST's Own Subscription
The sharpest question set the release's "$1 billion of non-dilutive funding" against the 10-Q's $1.4 billion AST subscription for preferred shares, and asked for the net cash benefit and whether the Japanese satellites are inside the roughly 90-satellite plan. Management confirmed the net subsidy figure and did not answer the second part.
Q: "First 1 here, the release frames, the J. Leo contract is $1 billion of non-dilutive funding. But the 10 Q shows ASK [AST] subscribing about $1.4 billion of preferred shares in the JV over 3 years. Wondering if you could help us understand what the net cash benefit to AST would be after that subscription and after the satellite sales and the licensing fee it receives back from AST."
— Michael Filatov, Berenberg
A: "Yeah. So the short answer is, and the disclosure's not perfect, of course, but there is a lot of puts and takes. But net, the net subsidy to AST to build out satellites and both materials and launch is that $1 billion figure."
— Scott Wisniewski, President
Assessment: The net subsidy is real, and "the disclosure's not perfect" is fair. The unanswered question matters more: if the J-LEO satellites are in addition to the roughly 90, the "fully funded" constellation claim covers less of the total build than it appears to. We assume they are partly additional and credit only $500 million of receipts by November 2027.
What They're NOT Saying
- A launch date for BlueBirds 14 to 16: The batch has been at the Cape since September. The release, the call and the 10-Q give no date and no reason for the wait. This is the single fact that would most change the deployment read.
- The competitive response: Starlink does not appear in the release, the transcript or the 10-Q, and SpaceX appears only as a launch provider. After a quarter in which the shares fell most on SpaceX's spectrum and carrier moves, saying nothing is a choice.
- Blacklift's full cost: We have the Firefly payments and "several hundred million dollars" of 2027 capital spending, but no 2028 figure, no total program budget, no breakeven and no first-flight margin. Management deferred the unit-cost effect to "the weeks to come."
- Whether J-LEO satellites are inside the 90: Asked directly, management answered only the net-subsidy half. The answer determines how far the "fully funded" claim stretches.
- An updated funding bridge: The 10-Q repeats "fully funded" for about 90 satellites and, in the same section, says the company plans to raise additional capital. No bridge reconciles the constellation, the rocket and the joint-venture subscription against $2.8 billion of unrestricted cash.
- The backlog headline: The August release listed about $1.30 billion of contracted revenue backlog among its highlights. This release has no backlog figure; only the 10-Q's remaining performance obligation of about $1.3 billion remains, so the total is roughly flat.
- Paid-service pricing and the start date: The $1 billion run rate is defined from a service start the company does not date. Neither retail pricing nor the revenue share for the planned U.S. operator venture was disclosed.
Market Reaction
- Pre-print setup: ASTS closed November 9 at $36.89, up 14.1% that day on the Rakuten AST announcement. Entering the print it was down 49.2% year to date, 27.6% over 30 days and 46.7% over twelve months, against a 52-week closing range of $30.20 to $133.09 and an S&P 500 up 10.3% year to date. From our August reaction close of $71.63, the shares had fallen 48.5%.
- After-hours move: Shares rose roughly 1% after the release and during the call.
- November 10 reaction session: The stock opened at $34.20, a 7.3% gap down, traded between $31.96 and $36.84, and closed at $33.69, down $3.20 or 8.7%. Volume was 25.7 million shares, 1.5 times the 30-day average of 17.6 million. The S&P 500 fell 0.4%.
- Path between the two prints: The shares fell 20.5% from August 11 to October 8, then 10.5% on October 9 on 54.3 million shares as the market absorbed SpaceX's spectrum agreement. They fell another 20.2% through October 16, and 21.1% in the week to November 6, which included a 15.2% drop on November 5 to the 52-week closing low of $30.20. They rose 7.1% on the Blacklift announcement on November 6 and 14.1% on the J-LEO news on November 9.
The J-LEO rally, handed back: The results session reversed most of the 14.1% gain from the morning of November 9. The after-hours reaction was positive, but buyers did not hold that level into the regular session, and the stock traded as much as 13.4% lower before recovering part of the loss. Across the J-LEO announcement and the print, the shares closed 4.2% above the November 6 close of $32.33. The market put a small positive value on the subsidy and the results combined.
Competition weighed on the price before the print: Most of the 53.0% fall from our August note happened before the results. Two of the four sessions with drops of 10% or more followed SpaceX news, on October 9 and November 2, and a broker downgrade on October 13 cited SpaceX's push into the market. The largest drop, 15.2% on November 5, had no company-specific cause in the coverage we found. That changes the question the stock trades on. In August it was whether AST could launch often enough. Now it is whether AST launches before a rival that owns spectrum and the dominant launch vehicle reaches the same operators. The quarter's lack of launches gave no evidence against the second fear.
Costs and dates outweighed the new money: Volume at 1.5 times average shows real repositioning, although a hotter-than-expected inflation report the same morning weighed on long-duration growth stocks generally. The stock-specific content was the fourth-quarter operating cost guide, the capital commitments for the rocket, and a call that gave no date for the satellites waiting at the Cape. The reaction suggests that investors priced the subsidy as offsetting part of the new spending, not as relief for the deployment problem.
Street Perspective
Debate: Does SpaceX's Spectrum Buying Make AST a Casualty or the Carriers' Answer?
Bull view: A rival paying billions for low-band spectrum confirms that direct-to-device service is valuable. A vertically integrated SpaceX threatens the major operators, and that pushes them toward a wholesale partner that leaves them the customer. AST's talk of working with all major U.S. and European operators fits that shift.
Bear view: The rival owns its spectrum, its satellites and the launch vehicle AST's next batch is waiting for, and it may soon own a carrier. It can price retail service aggressively and set the pace of AST's deployment at the same time. The partner-first model does not help if AST's satellites cannot reach orbit on time.
Our take: Both sides are right about different time frames. The strategic position of a neutral wholesale network improved, and over five years that may be worth more than spectrum ownership. For the next twelve months, the launch dependency dominates. We cut our base commercial user assumption by a third and the terminal multiple to 18 times.
Debate: Is Blacklift Vertical Integration or Capital Diversion?
Bull view: Launch is most of the unit cost, external launch is scarce and controlled partly by a competitor, and a licensed engine with a long test record shortens the usual development cycle. Owning the rocket gives AST control of its schedule and lower costs for a constellation of more than 100 satellites, plus sovereign and government opportunities.
Bear view: New rockets rarely meet their first-flight dates, and this one starts after the milestones that decide the next two years. About $1.3 billion goes to Firefly, plus several hundred million dollars of capital spending in 2027 alone, while the company still has no subscription revenue and needs financing next year.
Our take: The strategic logic is sound and the timing is poor. The cash goes out in 2026 and 2027, and the benefit cannot arrive before 2029. We include the costs in full in our base case and treat any launch saving as bull-case value.
Debate: Is There a 2027 Funding Gap?
Bull view: AST holds $3.2 billion of cash, the Japanese grant pays most of its funds in 2027, and a $1.2 billion quasi-governmental pipeline is in progress. The company says it is fully funded for about 90 satellites, and its capital spending can be slowed if needed.
Bear view: Net debt was already $1.36 billion at September 30. Fourth-quarter capital spending, the operating cost guide and the Firefly upfront payment consume about $1 billion in one quarter. The pipeline is mostly secured debt, and the joint venture requires a $1.4 billion preferred subscription.
Our take: The gap is real. Our bridge leaves about $0.4 billion of unrestricted cash by November 2027 before new financing, with net debt near $3.75 billion. The company can probably finance it. The form of financing (subsidy, secured debt or equity near current prices) is a larger driver of per-share value over the next year than the operating results.
Model Update & Valuation
Every input that moved this quarter moves our numbers. We lower 2026 revenue to $165 million, cut 2027 revenue by a third to $500 million, raise the cost base, and rebuild the balance-sheet bridge, including a correction of the net-debt assumption we published in August. The valuation method is unchanged, and the August inputs reproduce the August $75.4 base exactly.
Estimate Changes
| Item | August recap | Now | Reason |
|---|---|---|---|
| FY2026 revenue | $175M | $165M | Q3 $5.7M below our $48M; Q4 $76.425M vs. $80.745M |
| FY2026 adjusted opex excluding direct costs | ~$400M | $420.1M | Q4 guide $130M to $140M |
| FY2026 adjusted EBITDA estimate | ~$(349)M | ~$(359)M | Lower revenue, higher costs |
| CY2027 revenue | $750M | $500M | Gateways $125M; government $275M; international $50M; paid service $50M |
| CY2027 adjusted operating cost | $550M | $650M | Q4 exit rate annualized ($540M) plus Blacklift engineering and operations |
| CY2027 adjusted EBITDA estimate | ~$(55)M | ~$(338)M | Paid service later; cost base higher |
| Paid commercial service start | Q2 2027 | Q4 2027 | Two-quarter haircut to the mid-2027 deployment target |
| Target-date net debt | $2.0B (August 2027) | $3.75B (November 2027) | Burn understated in August; Firefly; Q4 capex; net of $0.5B J-LEO receipts |
| Economic shares | 410M | 415M | 390.11M plus ~12M Firefly shares plus ~12.6M unvested awards |
| FY2030 base revenue | $4.5B | $3.5B | Commercial $1.8B (from $2.7B); government $1.4B (from $1.5B); other $0.3B |
| Base terminal multiple | 20x | 18x | Integrated rival lowers the durability premium |
| Base 12-month value | ~$75.4 | $43.45 | See walk below |
Revenue mechanism: The 2027 paid-service line assumes one quarter of service at about 5.6 million average users and $3 of net monthly AST revenue per user. The FY2030 commercial line assumes 50 million average active users at $3 a month, about 1.7% of the more than three billion subscribers that AST's partners serve. In August we assumed 75 million. The cut combines the slower deployment with a competitor that can sell direct-to-device service under its own brand and spectrum.
Net-Debt Bridge to November 2027
| Item | Amount | Basis |
|---|---|---|
| Net debt at September 30, 2026 | $1,362.1M | Debt principal $4,160.969M less unrestricted cash $2,798.911M |
| Q4 2026 capital expenditure | $450M | Guide midpoint |
| Q4 2026 EBITDA loss | $102M | Our Q4 estimate |
| Firefly upfront cash | $400M | Closing expected in Q4 |
| 2027 capital expenditure, eleven months | $1,375M | $1.5B annual, including Blacklift |
| 2027 EBITDA loss, eleven months | $310M | Our 2027 estimate |
| Firefly services and engine payments | $125M | Pro rata services; part of the engine fee |
| L-band annual payment | $80M | At least $80M a year under the Ligado agreements |
| Net interest | $60M | Interest income falls as cash is spent |
| J-LEO grant receipts | $(500)M | Majority of ~$938M expected in 2027 |
| Net debt at November 2027 | ~$3,764M; $3.75B used | Implies new financing in 2027 |
What the Current Price Requires
The method is the one we used in August: FY2030 EBITDA at a terminal multiple, discounted three years to a target date (now November 2027), plus the present value of FY2028 to FY2030 unlevered free cash flow, less target-date net debt, divided by economic shares. These are our underwriting assumptions, not peer multiples or company guidance.
| November 2027 value scenario | Bear | Base | Bull |
|---|---|---|---|
| FY2030 revenue | $2.5B | $3.5B | $5.5B |
| EBITDA margin | 40% | 55% | 60% |
| FY2030 EBITDA | $1.0B | $1.925B | $3.3B |
| EV / EBITDA multiple | 15x | 18x | 22x |
| Annual discount / years | 18% / 3 | 15% / 3 | 13% / 3 |
| Discounted terminal enterprise value | $9.13B | $22.78B | $50.32B |
| Unlevered FCF, years 1 / 2 / 3 | $(1.6)B / $(0.9)B / $0.1B | $(1.45)B / $(0.2)B / $0.625B | $(0.8)B / $0.4B / $1.4B |
| Present value of interim FCF | $(1.94)B | $(1.00)B | $0.58B |
| Total target enterprise value | $7.19B | $21.78B | $50.89B |
| Net debt deducted | $4.75B | $3.75B | $3.0B |
| Economic shares | 430M | 415M | 415M |
| 12-month value per share | $5.67 | $43.45 | $115.40 |
| Return from $33.69; no dividend | (83.2)% | +29.0% | +242.5% |
| Probability weight | 45% | 40% | 15% |
Totals use unrounded components, so the bull column's rounded parts differ from its total by $0.01 billion.
Base walk from August: Net debt of $3.75 billion and 415 million shares take the August $75.43 to $70.30. Cutting FY2030 revenue to $3.5 billion from $4.5 billion takes it to $52.88, the 18 times multiple to $46.77, and the weaker interim cash flows (a present value of $(1.00) billion against $0.38 billion) to $43.45. The total change is 42.4%, and about $17.42 of the $31.98 comes from the revenue cut.
Base cash flows: Revenue of $1.1 billion, $2.1 billion and $3.5 billion in FY2028 to FY2030 produces EBITDA of $0.05 billion, $0.85 billion and $1.925 billion. Capital spending of $1.45 billion, $0.95 billion and $1.0 billion includes Blacklift's peak build in 2028 and fleet replenishment, and taxes and working capital absorb $0.05 billion, $0.10 billion and $0.30 billion. The bear case adds about $1 billion of net debt and 15 million shares of dilution for further delay; the bull case assumes launches on management's schedule and faster government collections.
Why we weight the scenarios now: In August the bear and bull returns sat roughly symmetrically around a base that offered 5.3%, so the base could stand for the expected value. With net debt heading toward $3.75 billion, the bear case is now a few dollars from the equity's floor, while the bull case still more than triples the price. One scenario can no longer represent the expectation. We give the bear case 45%, because every deployment date in our coverage has slipped, the company needs financing in 2027, and its launch supplier is now its competitor. The base gets 40% and the bull 15%. The weighted value is $37.24, 10.5% above the price. With 40% bear and 45% base, it would be $39.13, or 16.1%.
Sensitivities: A further 20% cut in base FY2030 revenue lowers base value to $32.47. A 15 times multiple gives $34.30, and the August 20 times gives $49.55. Each additional $500 million of net debt costs about $1.20 a share.
Grading the August Triggers
In August we wrote: "An upgrade requires the revised launch path to support credible commercial activation and a revenue or capital-efficiency improvement that lifts expected return beyond the market. A further deployment slip without offsetting government cash generation, a material revenue-guide cut, or expansion spending that materially worsens the funding bridge would favor Underperform." We grade each part as written.
- Upgrade, "credible commercial activation" through the revised launch path, plus a revenue or capital-efficiency improvement: not met. No satellite launched, beta moved to early 2027, and capital efficiency worsened.
- Downgrade, "a further deployment slip without offsetting government cash generation": fired. The 45-satellite target moved to mid-2027 and the quarter had no launches. The Space Force award and the J-LEO grant are funded commitments, but they offset only part of it, and neither is cash generation yet.
- Downgrade, "a material revenue-guide cut": not fired. The range was reaffirmed.
- Downgrade, "expansion spending that materially worsens the funding bridge": fired. The Firefly agreements (about $1.3 billion) and the J-LEO preferred subscription (about $1.4 billion) add more than $2.7 billion of commitments before the grant, plus several hundred million dollars of Blacklift capital spending in 2027, and our bridge now requires 2027 financing.
We honor the two triggers that fired in the numbers, not only in words: base value falls 42.4%, the bear case falls to $5.67, and conviction falls to 4. At the $71.63 price of our August note, these changes would mean Underperform, because the new base is 39.3% below it. In August we did not say how a large price decline would net against these conditions. We say it now: the triggers measure the business, the rating measures expected return, and the shares had already fallen 53.0% by the reaction close. At $33.69 the weighted expected return of about 10.5% is a few points above our 6% to 8% assumption for the S&P 500, a margin well inside the uncertainty of the estimate. That is a Hold.
New triggers, stated before they fire and set outside our base case: We would move toward Outperform if all twelve scheduled BlueBirds (14 to 25) are in orbit by the end of the first quarter of 2027, as management plans and our base case does not assume, and the 2027 funding need is met without new common equity beyond the Firefly shares. We would move to Underperform if BlueBirds 14 to 16 are still on the ground at the fourth-quarter report, if 2026 revenue falls below $150 million, or if new common equity is sold below our $37.24 weighted value to fund 2027. If the launches arrive on schedule but the funding comes from equity below that value, the two offset and the rating stays Hold.
Thesis Scorecard Post-Earnings
| Thesis point | Status | Q3 2026 evidence |
|---|---|---|
| Bull 1: Funding capacity for the full constellation | ON TRACK → AT RISK | $3.2B of cash covers twelve months, but Firefly, Blacklift and the J-LEO subscription extend the uses; net debt rises to ~$3.75B by November 2027 and new financing is needed. |
| Bull 2: Widening spectrum and technology moat | ON TRACK | ASIC pulled forward to BB34; six-a-month production; AT&T handover; Brazil authorization. The relative spectrum position is tracked under Bear 4. |
| Bull 3: Government and defense revenue leg | ON TRACK; stronger | Services revenue tripled; ~$100M Space Force sensing award; priority rating; 16 awards; signed sovereign model in Japan. |
| Bull 4: Commercial demand and commitments | ON TRACK; later | 60+ operators and 60+ gateways; partner activation by year-end; end-user beta moved to early 2027; no paid-service date. |
| Bear 1: Launch cadence at scale | MATERIALIZING; worse | No launch in the quarter; BB14 to BB16 waiting since September; ~45 moved to mid-2027; next batch on New Glenn's return. |
| Bear 2: Valuation embeds continued execution | EMERGING → CONTAINED | Shares 53.0% below the August reaction close and below our cut $43.45 base. |
| Bear 3: Losses and heavy investment | EMERGING → MATERIALIZING | Q4 opex guide $130M to $140M; Firefly commitments of about $1.3B; Blacklift capex in 2027; Q3 outflow of $464.2M. |
| Bear 4 (new): Competition from a vertically integrated rival | EMERGING | SpaceX's 800 MHz purchase, reported carrier talks and direct-to-cell authorization; it also controls the waiting batch's launch. |
Overall: The thesis weakened. Government demand and the technology improved, but the variable the whole case depends on, getting satellites into orbit on a schedule, went backward, and the company answered it with a long-dated, expensive rocket program. Two pillars changed status against us (Bull 1 and Bear 3), one changed in our favor only because the price fell (Bear 2), and a new competitive risk enters the scorecard.
Commitments for the next quarter: We will grade the SpaceX launch of BlueBirds 14 to 16; the shipment of BlueBirds 17 to 19 by the end of November for New Glenn's return in early 2027; BlueBirds 20 to 25 for priority-rated launches in the first quarter; partner activation in five markets before year-end; six ASIC satellites a month from December; fourth-quarter adjusted opex of $130 million to $140 million and capital spending of $400 million to $500 million; 2026 revenue within $150 million to $200 million; closing of the Firefly agreements; J-LEO accounting and receipts; progress on the $1.2 billion financing pipeline; and the Blacklift cost figures promised for the coming weeks.
Action: Maintain Hold, conviction 4. Two of our three downgrade conditions fired, and we have cut our value by 42.4% to reflect them. We have not moved to Underperform, because the shares fell further than our value did. We have not upgraded either, because a lower price does not fix the launch problem. The next four months bring twelve scheduled satellites and a financing decision, and those two events will decide which side of the new triggers the stock lands on.