BlueBird 7 Is Lost and New Glenn Is Grounded: The Reversion Condition Fired, Downgrading ASTS to Hold

Results vs. Consensus

For most of our coverage we have treated the ASTS income statement as a secondary read behind the four forward variables that actually drive the equity: launch cadence, funding, commercial demand, and spectrum. Last quarter, for the first time, the print and the operating story pointed the same way. This quarter they point the same way again, only downward. Revenue of $14.7M missed a thin ~$39M consensus by roughly 62% and fell back sharply from the outsized $54.3M fourth quarter, and the net loss to common widened to $(191.0)M. Management had told us at Q4 not to build off the $54.3M Q4 and to expect a softer first quarter, so the sequential decline was signaled; the depth of the shortfall against the Street line was not.

Q1 2026 Scorecard

Quality of the Miss

• Revenue: the miss is large but partly structural. Q1 revenue is the trough of a lumpy, milestone-driven year, and management had explicitly steered off the $54.3M Q4 as a run-rate. The $14.7M was gateway deliveries to MNO partners plus U.S. government service-milestone achievements across five existing contracts, the same two live vectors as prior quarters, with no commercial-service revenue (that starts in 2H). Still, roughly 62% below a ~$39M Street line is a wide gap, and it lands in the same quarter the launch program stumbled, so the print offers no offset to the operating setback.
• EPS / net loss: the $(0.66) loss to common (and the $(249.6)M pre-NCI loss) is, as every quarter, a poor read on operations, with a large share sitting below operating income in interest, financing costs, and non-cash fair-value marks on the enlarged convertible and derivative stack. But the widening is real and directionally negative: the loss to common more than doubled sequentially and quadrupled year on year, and the $91.2M adjusted opex, while down $4.5M QoQ, reflects a workforce and facility base that keeps scaling ahead of orbit.
• Opex (in line): adjusted opex ex-COGS of $79.8M landed at the top of the $70-80M guide, so on the metric management controls tightly there was no overrun. The sequential composition shifted: engineering services jumped $37.9M and G&A rose $17.4M versus Q4 (workforce, expanded production, spectrum and regulatory legal), offset by lower cost of revenue on the smaller revenue base. Discipline on the guided line held; the issue is the level, which keeps rising into a deployment that has not accelerated.
• Capex (the watch item, complicated): Q1 capex of ~$257M printing below the $350-425M guide is not the reassurance it looks like. Management was explicit that the underspend is launch-payment timing, and it guided Q2 capex up to $575-650M to absorb the shift. Netting the two quarters against the original plan, spend is not down, it is deferred and then front-loaded, into a launch campaign that just lost a satellite. The capex-discipline flag we raised at Q4 has not been cleared; it has moved a quarter to the right and gotten larger.

Grading Last Quarter's Commitments

A recap is a chapter in a running story, so before assessing anything new we grade the checklist management left us at Q4 against this quarter's actuals. This is the quarter that checklist was built to test.

Read: the two commitments that mattered most for the rating, the BB7 launch and the first stacked launch that would prove cadence, both failed. BB7 is lost and the stacked launch has not flown. Against that, the technology commitments were met (ASIC integrated, BB6 throughput record) and the funding and revenue-guide commitments held. But a recap grades the high-order items highest, and last quarter we said in plain language that a material launch slip would revert the rating. The scorecard is the downgrade in miniature: the two launch commitments are red, and they outweigh the green technology checkmarks because launch cadence, not throughput or the ASIC, is the binding constraint on the 2026 plan.

Business-Line Performance

ASTS still does not report conventional P&L segments; revenue flows through the same vectors as prior quarters (U.S. government milestones and commercial gateway hardware, with an MNO-services line and commercial subscriber service still nascent or pre-revenue). We assess those vectors alongside the operating and capital metrics that govern the cash runway and, for a pre-service builder, the constellation KPIs that are the real scorecard, which this quarter are the whole story.

The Launch: BlueBird 7 Lost, New Glenn Grounded

The defining event of the quarter is a satellite that is no longer flying. BlueBird 7 launched in April on Blue Origin's New Glenn, the first flight to reuse a previously flown booster. The booster reuse worked; the upper stage did not. A second-burn underperformance placed BB7 into an orbit too low for the satellite's on-board thrusters to sustain, and the company declared it a loss to be de-orbited. The FAA classified the event a mishap, which grounds New Glenn while Blue Origin investigates and implements corrective actions under FAA oversight. Management was candid about the outcome and framed it as a launch-vehicle problem outside its control.

"We were pretty open on Bluebird 7, the day of. We knew what happened immediately and we were very open on what it is. And at the end of the day, remember, we have 30, 33 satellites in advanced stage of production at the factory. So it was a loss, we're on to the next... they're working through the investigation in [an] upper stage anomaly. [An anomaly] like this is not uncommon early in programs, and we feel optimistic about them getting back to the pad soon." — Scott Wisniewski, President

Assessment: This is the fact that changes the rating. At Q4, BlueBird 6 unfolding on orbit retired the array-deployment risk and earned the upgrade. This quarter reintroduces the risk one layer up the stack: not "can the satellite deploy," but "can it get to orbit on the cadence the plan requires." The loss itself is insured and, at one satellite, survivable. The grounding of New Glenn is the larger problem, because New Glenn's larger fairing and booster-reuse tempo were the load-bearing assumptions under the every-one-to-two-months cadence and the 45-by-year-end target. "We're on to the next" is the right posture for a builder with 20-plus satellites in the factory, but the constraint was never the factory. It was the launch slots, and one of the two primary vehicles is now on the ground indefinitely.

U.S. Government: Still Compounding Quietly

The government leg was the quarter's steadiest performer, and one of the few places the story advanced. The company executed across five existing contracts, advanced milestones under the SDA HALO Europa Track 2 prime and under contracts where Fairwinds is prime (a follow-on to previously demonstrated NTN tactical SATCOM capability), and disclosed three additional awards through prime contractors spanning secure-communications and non-communications use cases. Management continues to frame these as proof points ahead of larger programs of record, with Golden Dome-related budget momentum as a tailwind and a wholly owned government-and-defense subsidiary now structured to serve the customer.

"On the U.S. government side, we continue to grow the pipeline with 3 additional awards through prime contractors. These awards address 3 unique use cases across secure communications and non-communications capabilities, reflecting strong proof points ahead of larger contracts... including those related to Golden Dome." — Scott Wisniewski, President

Assessment: The government pipeline continues to broaden, and the dual-use, single-platform model means each new BlueBird serves both TAMs. That is genuine and on-thesis. The caveat we have raised since Q3 still binds and binds harder this quarter: government revenue scales with the same launch cadence as everything else, so it is not an independent hedge against launch risk. A grounded vehicle slows the government ramp as surely as the commercial one. The leg is a source of durability, not an offset to this quarter's setback.

Commercial: Backlog Holds, Partners Keep Signing

The commercial story neither advanced nor retreated dramatically. Contracted commitments held at over $1.2B, the partner roster grew to nearly 60 MNOs covering more than 3 billion subscribers, and two new agreements landed: TELUS as a second Canadian partner (which also made an equity investment in ASTS) and Axiom Telecom, a pan-African operator across 11 countries. Ground-readiness work continues at pace, with hardware deployed across five continents and integration underway with customer network cores, which management frames as a leading indicator of eventual service activation.

"We announced an agreement with TELUS as our second partner in Canada, who also made an equity investment in ASTS... And in Africa, we are pleased to be partnering with Axiom Telecom, a pan-African operator in 11 different countries... we expect additional MNO agreements to be signed with increasing velocity throughout 2026." — Scott Wisniewski, President

Assessment: The demand pillar remains the least-troubled part of the story. Partners keep signing, a carrier taking an equity stake is a stronger signal than a distribution deal, and the ground-integration work is real, physical evidence of intent. But every commercial milestone downstream of here (beta this summer, activation in 2H) is gated on satellites reaching orbit in numbers, and that is precisely what slipped. The book is a floor and a financing tool, as management conceded last quarter that per-year contracted contribution is only in the low hundreds of millions. It is not a reason the stock re-rates while the constellation stalls.

Operating & Capital Metrics

Assessment: The bottom row is the quarter. Last quarter it flipped from zero to one and that single change carried the upgrade. This quarter it stayed at one while a second satellite was built, launched, and lost, so the metrics above it read differently than the raw QoQ moves suggest. The lower capex is deferral, not discipline; the lower revenue is the seasonal trough, not a signal; the wider loss is mostly below-the-line noise. Strip the noise and the operative facts are simple: spend is stepping up (Q2 capex $575-650M), the satellite count on orbit did not move, and the vehicle that was supposed to move it fastest is grounded.

Constellation & Launch Milestones (the KPIs that actually matter)

Assessment: The green rows are technology (ASIC integrated, 98.9 Mbps demonstrated), and they matter: they confirm the platform does what the thesis needs it to do at the physics level. The red rows are logistics (one on orbit, one lost, primary vehicle grounded, ~45 target now improbable), and for a company whose value is a call option on a constellation, logistics is the binding constraint. The path from one satellite to 45 in roughly seven months, with New Glenn on the ground and the first stacked launch not yet flown, is a very steep ramp that just got steeper. We do not need the throughput proof; we have it. We need satellites on orbit, and this quarter subtracted one rather than adding several.

Key Topics & Management Commentary

Overall Management Tone: Management was composed and on-message about a genuinely bad operational quarter, leading the prepared remarks with manufacturing scale, spectrum, the 98.9 Mbps record, and the "fortress" balance sheet before addressing the launch failure largely in Q&A. The posture on the BB7 loss was matter-of-fact and forward-leaning ("it was a loss, we're on to the next"), and the recurring move, familiar from prior calls, was to answer launch-cadence questions with factory and contracted-capacity assurances rather than committed launch dates. Confidence in the destination was undiminished; specificity on the timeline to get there was, again, conspicuously withheld.

The BlueBird 7 Loss and What It Says About the Cadence

Management's framing of the loss rested on two pillars: the anomaly was in the New Glenn upper stage and therefore outside AST's control, and the deep production backlog (20-plus satellites in assembly) makes any single loss a setback rather than a crisis. Both are true. What management did not offer was a revised path to 45-by-year-end that accounts for New Glenn being grounded, leaning instead on the multi-launcher strategy and the assertion that Blue Origin will return to the pad soon.

"We've designed the rocket... in our business strategy to be launch vehicle agnostic, and we're buyers of launch across the entire heavy launcher footprint. So we were prepared for this years ago with our strategy. We think we selected the right partners." — Scott Wisniewski, President

Assessment: The launch-agnostic architecture is a real design advantage and the single best mitigant to this quarter's event, and the pivot to a Falcon 9 stack for the next launch demonstrates it. But agnosticism does not create launch slots; it lets you re-shuffle among them. With New Glenn, the higher-capacity vehicle, grounded for an undisclosed period, the near-term cadence leans harder on SpaceX, and the 45-by-year-end math needs New Glenn back. Management's confidence that Blue Origin returns "soon" is doing a lot of work in a plan that has no committed date for it.

The 98.9 Mbps Record and the Throughput Case

The morning of the call the company announced a peak downlink of 98.9 Mbps to unmodified, off-the-shelf smartphones over international waters using the in-orbit Block 1 satellites, with management arguing that BB6 (Block 2, already in orbit) and the coming 8/9/10 will nearly double that as spectrum is enabled region by region.

"This morning, we did announce that from the middle of nowhere in the middle of the ocean, on international water, we did achieve big data rates, very close to 100 megabits per second, into [a] standard device without any modification... With the BB6, which is already in orbit, and BB8, 9 and 10, we expect to nearly double that capacity." — Abel Avellan, Chairman & CEO

Assessment: This is a real and important data point, and it is the answer to the "does the broadband thesis work" question that dominated our early coverage. Approaching 100 Mbps to an unmodified handset is genuine broadband, not emergency texting, and it validates the architectural bet. But throughput was never the variable gating the 2026 plan; deployment cadence is. A record data rate on the satellites already up does not put more satellites up. We bank the proof and note it does not offset the launch setback.

The ASIC, AI Spectrum Management, and the Next-Gen Roadmap

Management confirmed the AST-5000 ASIC is complete and integrated into the production line, lifting per-satellite processing bandwidth from roughly 1 GHz on the FPGA satellites to 10 GHz, and separately described AI edge-compute and AI spectrum-management features slated for production batches toward year-end. Management was careful to disaggregate the two: the ASIC governs how much spectrum a satellite can manage, while raw peak data rate is a function of the array and does not depend on the ASIC.

"On the FPGA satellites, we had around 1 gigahertz of spectrum; with the ASIC... we have 10 gigahertz of spectrum. So it's a factor increase on number of gigahertz that can be constantly used per satellite. The big data rates are actually not dependent on the ASIC." — Abel Avellan, Chairman & CEO

Assessment: Integrating the ASIC on schedule is a clean execution win on the technology track, and the 10x processing headroom matters for capacity once traffic scales. The AI spectrum-management concept remains, as we characterized it last quarter, genuine optionality rather than modelable revenue. The through-line of the call holds: the technology roadmap is on time; the launch roadmap is not.

Manufacturing, Stacking, and the Composite Structure

The factory remains the grounded half of the equation. Phased arrays are complete through BlueBird 28, with BB11-BB33 in assembly, and management reiterated the target of six fully assembled satellites per month at roughly 95% vertical integration, including in-house production of the stackable composite structures that enable multi-satellite launches (up to eight on New Glenn, five on Vulcan, three on Falcon 9).

"We own all the IP [for] how this is done... we have over 1,000 people dedicated to build these composite structures across our satellites... All of this to achieve and keep our 6 satellite per month fully assembled every month." — Abel Avellan, Chairman & CEO

Assessment: Production is not the constraint, and management knows it; the six-per-month cadence and completed arrays through BB28 mean there is inventory waiting for rides. That is exactly the point. A factory producing satellites faster than they can be launched, into a market where one of two primary vehicles is grounded, converts a manufacturing strength into stranded working capital until the launch bottleneck clears. The stacking capability is the right lever; it needs a flying vehicle to pull.

The Balance Sheet: Still a Fortress

Cash, cash equivalents, and restricted cash stood at approximately $3.5B as of March 31 (cash and equivalents of $3.03B), inclusive of the February 2026 convertible (2.25% coupon, ten-year, ~$116.30 effective strike). Management reiterated no plans for further convertible debt in 2026 and framed the balance sheet as funding the full 100-plus-satellite constellation plus spectrum deployment, higher-cost-debt reduction, and opportunistic investment.

"Our company has key assets, including IP, manufacturing, partnerships, spectrum and balance sheet cash with approximately $3.5 billion to build and launch over 100 Bluebird satellites to enable global coverage of space mobile service." — Abel Avellan, Chairman & CEO

Assessment: This is why the downgrade stops at Hold rather than Underperform. A ~$3.5B liquidity position against a launch program that just lost a satellite means the setback is a timeline problem, not a solvency problem. The company can absorb a lost satellite, a grounded vehicle, and a step-up in capex without a dilutive raise. The balance sheet remains the strongest pillar in the thesis and the reason the equity is not impaired by a bad quarter, only its risk/reward reset.

Spending Step-Up: Q2 Capex to $575-650M

The CFO guided Q2 capex sharply higher to $575-650M, attributing the increase to launch-contract payments that shifted out of Q1 (which is why Q1 landed below its own $350-425M guide). Netted together, management argued, the two quarters' spend is roughly what was planned, just re-phased.

"We expect our capital expenditures to increase in Q2 of 2026 to a range of $575 million to $650 million, primarily driven by the timing of launch payments... had the launch payments been made in Q1 like we originally planned instead of making them in Q2, our guidance for Q2 capital expenditures would have remained in the same general range as Q1." — Andy Johnson, CFO & Chief Legal Officer

Assessment: The timing explanation is credible and the balance sheet covers it. But two things are worth holding in view: first, the "spending relapse" half of our reversion condition is about the trajectory of burn, and a jump to $575-650M in a single quarter is a step-up even if it is phased rather than incremental; second, those launch payments are being made into a manifest whose primary vehicle is grounded, which is an uncomfortable juxtaposition. We do not read this as indiscipline, but we do read it as burn accelerating into a program that just stumbled, which is not the setup the Outperform thesis assumed.

Regulatory: FCC Commercial Authorization Granted

A clear positive amid the operational setback: the FCC granted authorization to operate the BlueBird constellation commercially in the United States on premium low-band spectrum, in coordination with Verizon, AT&T, and FirstNet. Management framed the grant as validation of both the technology and the ability to operate alongside terrestrial networks.

"We are granted FCC authorization to operate our Bluebird satellite constellation commercially in the United States, enabling direct-to-device connectivity in the U.S. on premium low-band spectrum in coordination with our partners, Verizon, AT&T and FirstNet." — Abel Avellan, Chairman & CEO

Assessment: This is a real regulatory milestone and removes a domestic gating item for commercial service. It is the kind of progress that, in a quarter where a satellite reached orbit, would headline the note. Here it is a genuine positive on the demand/regulatory side that does not touch the binding constraint. Separately, L-Band FCC approval, the gating event for the mid-band constellation, went unmentioned, so the spectrum-approval picture is mixed rather than uniformly improved.

Adjacencies: Government Non-Comms, Radar, and Golden Dome

Management again spent airtime on the dual-use platform's non-communications capabilities (radar, geolocation) for defense customers, arguing the core capability is already built into the satellites being produced and that Golden Dome budget momentum is a multi-year tailwind. It declined, appropriately, to detail classified applications.

"We are basically in all aspects of government usage... from FirstNet, from the classified, all the way to the Golden Dome. So for communications and non-communication capabilities, we see ourselves as a very important asset to our government... we expect a very significant growth in revenue and opportunity." — Abel Avellan, Chairman & CEO

Assessment: The defense optionality is real, dovetails with the signed SDA prime, and is a legitimate part of the long-term case. But it is optionality on a constellation that must first fly, and this quarter the flying got harder. We continue to treat the adjacencies as free options embedded in the platform, valuable to the terminal narrative, immaterial to whether the 2026 deployment plan holds.

Guidance & Outlook

Guidance is best read, as always with ASTS, as a set of operational commitments rather than a P&L model. The accountable number remains the reaffirmed FY2026 revenue guide of $150-200M, which management says is roughly half covered by contracted backlog, with Q1's $14.7M as the trough and a heavy back-half weighting tied to gateway deliveries, government milestones, and the beginning of commercial service in 2H. The most consequential change is not on the revenue line but on capex: the jump to $575-650M in Q2, which the CFO frames as launch-payment timing rather than incremental spend.

Implied 2026 shape: holding the $150-200M revenue guide against a $14.7M Q1 implies roughly $135-185M across the final three quarters, a steep back-half ramp that leans on commercial-service activation, which is itself gated on satellites reaching orbit in numbers. That is the circularity at the heart of the guide: the revenue reaffirmation assumes a launch cadence that this quarter's events have made less certain. We treat the revenue guide as reaffirmed on paper but carrying more downside skew than it did a quarter ago.

Street at: for a pre-service name, the meaningful benchmark is not the consensus revenue line (dispersed and of limited use, as the ~$39M Q1 estimate that missed by 62% illustrates) but management's own targets and, above all, the pace of satellites reaching orbit. The 45-by-year-end target is where the debate now concentrates; independent assessments cited in coverage put realistic 2026 deployment closer to ~28.

Guidance style: management continues to reaffirm annual and terminal targets while the near-term launch specifics slip, a pattern now three quarters deep. The FY revenue guide has been met at the top once (2025), which earns it some credibility; the launch-cadence guidance has slipped every quarter of our coverage and now carries an outright vehicle failure. We read the revenue guide as achievable-but-riskier and the 45-satellite target as the number most likely to be revised.

Analyst Q&A Highlights

What Happened to BlueBird 7 and Whether New Glenn Can Scale

The first and most thesis-relevant exchange went straight at the loss and at whether the vehicle it exposed can still carry the year. Management characterized the failure as an upper-stage anomaly of a type not uncommon early in a launch program, expressed optimism about Blue Origin returning to the pad, and pointed to the deep production backlog as the reason a single loss does not derail the plan.

Q: "Now that you've had a few weeks to digest, can you just walk us through what happened with Bluebird 7 and what gives you comfort this will not repeat going forward and that New Glenn can scale accordingly?"
— Christopher Schoell, UBS

A: "We were pretty open on Bluebird 7, the day of... So it was a loss, we're on to the next. So yes, I would say that we're working closely with Blue. They're working through the investigation in [an] upper stage anomaly... and we feel optimistic about them getting back to the pad soon... they have 2 boosters sitting in their integration facility ready to get into the cycle."
— Scott Wisniewski, President

Assessment: The answer is candid about the loss and appropriately unpanicked, but it substitutes optimism about Blue Origin's return for a revised, vehicle-specific cadence plan. "Getting back to the pad soon" is not a date, and the entire 45-by-year-end arithmetic depends on when New Glenn actually flies again. The most important number on the call, the return-to-flight timeline, was the one management could not provide.

The FAA Investigation Timeline and Service at Launch

An analyst pressed for a timeline on the FAA investigation and, separately, for how many MNOs and covered subscribers would be addressable once ~45 satellites are up. Management said no public timeline exists for the investigation, reiterated that such reviews are commonplace, and pivoted the second question to the breadth of ground-integration efforts across the disclosed country list rather than a hard live-carrier count.

Q: "Is there a timeline associated with the FAA investigation on when you would expect that to be concluded? And then... how many MNOs are you expecting to be live at launch?"
— Scott Searle, Roth Capital Partners

A: "No, there hasn't been a publicly disclosed timeline. But like I said, these sorts of investigations are pretty commonplace... we're really focused on our next launch, obviously, with Falcon 9 and the next 3 Bluebirds. But... we have a multi-launcher strategy."
— Andrew Johnson, CFO & Chief Legal Officer

Assessment: The absence of an FAA timeline is the honest answer and also the crux of the uncertainty: the grounding runs until Blue Origin closes the mishap, and that clock is not AST's to set. The pivot to the Falcon 9 next launch is the right operational response, but it also quietly concedes that New Glenn is not part of the very-near-term plan. On the service question, deflecting a live-carrier count to a population-coverage figure is the same substitution of breadth for specificity we have flagged before.

Contracted Launch Capacity and the Path to 45

The most quantitatively pointed exchange asked whether the company has contracted launch capacity to average roughly one launch per month from June through December, what happens if only one more New Glenn flies this year, and whether the next New Glenn would carry the maximum stack. Management asserted it has contracted capacity to meet the 2026 target via a mix of New Glenn and SpaceX (or equivalent) launches, and said the next New Glenn would carry four satellites as a stacking ramp rather than the full eight.

Q: "Do you have contracted launch capacity to do an average of basically one [launch a] month from June through December, because I think that's what you need to do in order to get close to that 45 [satellite] number? ... On the next New Glenn, can you go right to 8? Or do you have to... do 3 or 4 to make sure it goes smoothly?"
— Bryan Kraft, Deutsche Bank

A: "Yes, we do have contracted launch capacity to meet our target for 2026... a handful of [New Glenn] launches and a handful of SpaceX or equivalent launches, and that's what gets us to the approximately 45... we expect on the next New Glenn, we'll launch 4 satellites. Part of that is kind of ramping into the stacking capability."
— Scott Wisniewski, President

Assessment: The candor that hitting 45 requires roughly monthly launches from June is useful and, frankly, sobering: from a base of one, with New Glenn grounded and the first stacked launch not yet flown, "roughly one a month for seven months" is an aggressive schedule that leaves no room for a second slip. Contracted capacity is necessary but not sufficient; the slots only convert to satellites-on-orbit if the vehicles fly on time. This exchange, more than any other, is why we think the 45 target is the number most likely to move.

ASIC Versus AI Spectrum Management on the Throughput Roadmap

An analyst sought to disentangle which technology drives which capability: whether the AST-5000 ASIC or the AI spectrum-management layer is responsible for pushing peak data rates toward 200 Mbps, and whether the non-communications defense work implies mid-band military radar. Management clarified that peak data rate is array-driven and independent of the ASIC, that the ASIC increases manageable spectrum bandwidth tenfold, and that the near-term defense capability uses low-band and requires no satellite redesign.

Q: "With the AST-5000 ASIC... expected to enable 120 megabit per second peak data speeds, is it the AI spectrum management that gets you up closer to 200 megabits per second by year-end? And... on the non-communications capabilities... would we be correct in assuming we're talking about mid-band military radars?"
— Mike Crawford, B. Riley Securities

A: "The big data rates are actually not dependent on the ASIC... the ASIC basically allowed us to upgrade the amount of bandwidth we can manage... [it] does not require mid-band spectrum. That's something we can do with low-band spectrum, which we've deployed today."
— Abel Avellan, Chairman & CEO

Assessment: A clean technical answer that reinforces the throughput case and clarifies that the defense non-comms capability is available now on existing spectrum rather than waiting on mid-band. It is genuinely reassuring on the technology, and genuinely beside the point on the quarter's actual problem. The roadmap questions all resolve favorably; none of them is the launch cadence.

Stacking Hurdles and the AI Edge-Compute Roadmap

An analyst asked about the authorizations and hurdles required to move up the stacking curve toward the maximum satellites per launch, and separately how the AI edge-compute features planned for next-generation satellites improve performance. Management reiterated that the stackable composite structure is manufactured in-house at rate (up to eight on New Glenn, five on Vulcan, three on Falcon 9) and described the AI layer as on-board edge compute plus dynamic AI spectrum management that allocates power and spectrum across the satellite's field of view.

Q: "I think you talked last quarter about stacking up to 8 satellites per launch. So just wondering about the hurdles [and] authorizations required to get to [that]... and then... you mentioned deploying AI edge computing features [on] the next generation of satellite. [I] want to get a better understanding of how that's going to improve the efficiency or performance."
— Michael Funk, Bank of America

A: "We do have now the technology that we are manufacturing at rate... to be able to stack multiple satellites in a single launch... What we are incorporating to our satellites, which you will start seeing in the production batches towards the end of the year, is the ability to edge compute and load AI capabilities on board... And the AI spectrum management is a system behind all of that basically predict traffic... predict where people are and then allocate that very, very intelligently."
— Abel Avellan, Chairman & CEO

Assessment: The stacking capability is genuinely in hand and is the mechanism that makes a 45-satellite year arithmetically possible, so the manufacturing answer is solid. The AI edge-compute features are a year-end-and-beyond roadmap item, firmly optionality rather than near-term value. As with every technology answer on the call, it resolves favorably and leaves the launch-cadence question, the one that actually gates the year, untouched.

The Amazon-Globalstar Deal and the Competitive Landscape

An analyst asked how Amazon's acquisition of Globalstar and its early entry into direct-to-device change the competitive picture. Management drew the same broadband-versus-emergency-messaging distinction it has drawn since initiation, arguing that genuine broadband requires hundreds of megahertz of spectrum and a large aperture that competitors do not possess, and that it sees no material change to the landscape for the next several years.

Q: "What do you view as the impact of Amazon's acquisition of Globalstar? And do you view any potential partnership opportunities with Amazon as they seem to be very much in the early stages of entering this industry?"
— Louie DiPalma, William Blair

A: "We don't see any real change of the landscape, at least for the next 7 years... nobody is anywhere close to the capability that we have technically to deliver hundreds of megabits directly to a phone from something that is flying at [thousands of km/h] 500 kilometers above you."
— Abel Avellan, Chairman & CEO

Assessment: The broadband-versus-texting distinction is real and, with the 98.9 Mbps demonstration, partly proven rather than asserted, and a well-capitalized entrant like Amazon validates the TAM more than it immediately threatens ASTS's specific broadband niche. The "no change for seven years" framing is characteristically confident; we would note that the deeper-pocketed the entrants, the more launch execution, the thing that stumbled this quarter, becomes the moat that has to hold.

Commissioning Time From Launch to Service Activation

An analyst asked, once ~45 satellites are on orbit, how long commissioning takes before MNO service activates. Management put the current target at 45 days from launch to usable 4G/5G connectivity, with an aspiration to compress toward two weeks in later batches as the process matures, while declining to promise the faster figure for early batches.

Q: "When you do hit 45 satellites on the launch side, what is the commissioning time period we should be thinking about until sort of the activation with service with MNOs?"
— Greg Pendy, Clear Street

A: "The target is 45 days, that's what we are planning with the MNOs. Every time that we launch, in 45 days we should be using either 5G or 4G connectivity through them. But as we keep launching, we plan to reduce that time frame... all the way down to 2 weeks; we don't want to promise that in the early batches."
— Abel Avellan, Chairman & CEO

Assessment: A useful operational data point that also compounds the timeline concern: even after a satellite reaches orbit, there is a ~45-day commissioning lag before it carries service. Layered on top of a grounded primary vehicle and a base of one, the 45-day activation window makes the 2H commercial-service start dependent on launches happening very soon, which is exactly what is now uncertain. The disclosure is candid; its implication is that the service timeline has less slack than the reaffirmed guide suggests.

Whether Government Needs Dedicated Satellites

Continuing a line from last quarter, an analyst probed whether the non-communications defense work would require a satellite redesign (optical cross-links, dedicated payloads) and whether such work would be customer-funded. Management was firm that the core capabilities are already built into the production satellites, with any additional features added on request and not disclosable.

Q: "Since your satellites were specifically designed as a communication platform, does that imply that you're going to have to do [a] large redesign of satellites and adding things like optical cross-links and onboard processing... or if so, is that something that would be customer[-]funded?"
— Chris Quilty, Quilty Space

A: "All the capabilities that they require are already built in, in what we are producing on the line. There will be additions after the request [that we] are not permitted to discuss. But basically, the core capability of what they're using was incorporated many years back."
— Abel Avellan, Chairman & CEO

Assessment: The single-platform answer is economically the right one and consistent with prior quarters: one constellation serving both TAMs without duplicating capital. The flip side we have noted before still holds, and holds harder this quarter: because the government capability rides the same satellites, government revenue scales with the same launch cadence that just stumbled. The dual-use model is a strength; it is not a hedge against launch risk.

What They're NOT Saying

With the launch program the quarter's dominant fact, the omissions cluster tightly around the timeline to recover it. These gaps are where an investor's remaining doubt should concentrate.

1. A New Glenn return-to-flight timeline: the single most important number for the 2026 plan, and management explicitly has none to offer ("no publicly disclosed timeline"). The grounding runs until Blue Origin closes an FAA mishap on a clock AST does not control, yet the 45-by-year-end target, which leans on New Glenn's capacity, was reaffirmed anyway without a bridge that reconciles the two.
2. A revised satellite-count path to year-end: the ~45 target was reiterated, but management did not walk through how many launches on which vehicles, by when, actually get there from a base of one with the primary vehicle grounded. The Kraft exchange forced out "roughly one a month from June," but no formal, updated deployment schedule was provided.
3. The root cause of the BB7 upper-stage anomaly: beyond "upper stage anomaly," management offered no technical detail on what failed or why it will not recur, deferring entirely to Blue Origin's investigation. For the event that defines the quarter, that is thin, and it leaves the recurrence risk unquantifiable from the outside.
4. Quarterly 2026 revenue shape: as last quarter, management gave a full-year range and declined to shape it quarterly, which conveniently removes the near-term accountability checkpoints in a year where the back half now carries even more of the load.
5. L-Band FCC approval progress: the domestic commercial low-band authorization was announced, but L-Band, the gating event for the mid-band constellation slated to begin launching by end-2026, went unmentioned. Silence on a spectrum approval that gates a whole capability layer is notable.
6. Insurance recovery specifics: the BB7 loss is described as insured, but no figure, timing, or reinvestment plan for the recovery was disclosed. For a loss of a ~$21-23M asset, the economics of the insurance offset are a fair thing to have quantified.
7. What "beyond the first 100" capital is being deployed on: with Q2 capex jumping to $575-650M, the specific use of the surplus February-raise capital beyond the base constellation remains open-ended, the same deployment-discipline watch item we opened last quarter.

Market Reaction

• Pre-print setup: ASTS closed at $82.55 on May 11 entering the print, up 13.7% year-to-date but down 13.0% over the trailing 30 days (from $94.90 on April 10) and well off its 52-week closing high of $122.09. Over the trailing twelve months the stock was still up 220.5%. The S&P 500 was +8.3% YTD. The setup was a name that had run enormously over the year but had already given back ground into the print, in part on the April New Glenn failure that was public before the call.
• Reaction session (May 12): shares opened at $73.43 (a -11.0% gap), traded a $69.92 to $79.87 range (-15.3% to -3.2% intraday), and closed at $72.96, down 11.6% (-$9.59) on volume of 32.8M shares versus a 16.8M 30-day average (2.0x). The S&P 500 fell 0.2% the same day, so the move was roughly -11.4% on a relative basis.

The tape did what the fundamentals did: it went down, hard, on heavy volume. A -11.6% close on 2.0x average volume, into a roughly flat market, is an unambiguous negative reaction, and it is the mirror image of last quarter's +6.6% rally on a satellite reaching orbit. The BB7 loss was already public before the call (New Glenn placed it in the wrong orbit in April), so the print itself was not the only catalyst; rather, the call confirmed the read that the loss plus the grounding plus a soft print plus a capex step-up adds up to a materially harder 2026, and the stock priced that in.

The relative move underlines the point. On a day the S&P was roughly flat, ASTS shedding 11.6% is almost entirely idiosyncratic, and the 2.0x volume says it was not a thin, low-conviction drift but a real repricing. The wide intraday range ($70 to $80) is characteristic of a high-beta, heavily-shorted retail favorite, and days like this can round-trip, so we do not anchor the downgrade to the reaction-day tape. We anchor it to the launch failure and the grounding the tape was reacting to; the price action is confirmation, not thesis.

One note of proportion: even after the drop, at $72.96 the stock is up 220% over twelve months and carries a large valuation on a company with $14.7M of quarterly revenue and no commercial service. The de-rating removes some froth, but it does not by itself create a margin of safety on what remains a binary, cadence-dependent outcome.

Street Perspective

The post-print debate reorganized around a single question: how much does one lost satellite and a grounded vehicle actually change the 2026 story? We frame the four live debates below.

Debate: Is the BB7 Loss a Blip or a Cadence-Breaker?

Bull view: launch failures are common early in a vehicle's life, the loss is insured, AST has 20-plus satellites in the factory, and the launch-agnostic architecture lets it re-route to Falcon 9; one lost satellite from a base that was always going to grow to 45-plus is noise, not signal.

Bear view: it is not one satellite, it is a grounded primary vehicle with no return date, in a year that needed roughly monthly launches from June; the loss exposes exactly the third-party-vehicle dependency that was always the real risk, and the 45 target is now improbable.

Our take: closer to the bear, which is why we downgrade. The lost satellite itself is survivable and insured; the grounding of the higher-capacity vehicle, with no timeline, in the specific year the cadence had to inflect, is the material fact. This is the risk we named at Q4, and it materialized worse than a slip.

Debate: Does the ~$3.5B Balance Sheet Make the Setback a Non-Event?

Bull view: with ~$3.5B of liquidity and no need to raise, AST can absorb a lost satellite, a grounded vehicle, and a capex step-up without dilution; a well-funded builder gets to be patient, so a launch slip costs time, not the thesis.

Bear view: money does not buy launch slots or un-ground a rocket; the balance sheet solves the problem AST does not have (funding) and does nothing for the problem it does have (getting satellites to orbit on schedule), while the burn steps up regardless.

Our take: both are right, and together they define the rating. The balance sheet is precisely why this is Hold rather than Underperform, there is no solvency or dilution risk here. But cash does not fix a cadence problem, which is why it is not still Outperform. Funded patience is worth a lot; it is not worth a re-rate while the constellation stalls.

Debate: Is the 45-Satellite 2026 Target Still Credible?

Bull view: the factory is producing six per month, arrays are done through BB28, stacking is proven, the manifest is contracted across multiple vehicles, and Blue Origin has boosters ready; the pieces for a steep H2 ramp are in place once New Glenn returns.

Bear view: from one on orbit, with New Glenn grounded indefinitely and the first stacked launch not yet flown, ~45 in seven months requires near-flawless roughly-monthly execution with no slack; independent reads already put realistic deployment near ~28.

Our take: we think 45 is now the number most likely to be revised. It is not impossible, but it requires a clean return-to-flight on an undisclosed timeline plus a launch roughly every month with zero further failures. That is a distribution skewed to the downside, and prudent underwriting assumes something between the ~28 independent read and the 45 target.

Debate: After an 11.6% Drop, Is the Stock Now Cheap Enough?

Bull view: the de-rating off the $122 high plus a bad-news day has washed out froth, the technology is proven (98.9 Mbps), the FCC granted commercial authorization, and the balance sheet is a fortress; a lower price on a de-risked-on-funding story is a better entry.

Bear view: at $73 the stock is still up 220% over twelve months on $14.7M of quarterly revenue and no service, and it still fully discounts the constellation being built on schedule, which just got less likely, so a lower price on a higher-risk timeline is not a bargain.

Our take: the valuation is still full on any near-term metric, and the equity remains a call option on a constellation whose deployment odds just worsened. The de-rating improves the entry at the margin but does not create a margin of safety against a cadence outcome that is now more uncertain. That balance, cheaper but riskier, nets to Hold.

Debate: Does the Moat Hold as Better-Capitalized Entrants Arrive?

Bull view: the 98.9 Mbps demonstration proves genuine broadband to unmodified phones, a capability no rival has matched; the multi-band owned-plus-partner spectrum position, ~3,900 patent claims, and the record-size array are hard to replicate, and Amazon buying Globalstar validates the TAM more than it threatens ASTS's broadband niche.

Bear view: deeper-pocketed entrants with their own launch capacity and existing constellations can iterate around a first mover whose primary vehicle is grounded; being first to a broadband spec matters less if a better-funded rival reaches "good enough" scale while ASTS is stuck at one satellite on orbit.

Our take: the broadband-versus-texting distinction is real and, with this quarter's throughput record, demonstrated rather than asserted, and the spectrum-plus-patent position remains the durable moat. But the more capital that flows into direct-to-device, the more the moat has to be defended with execution, and execution, specifically launch cadence, is exactly what slipped. The technical lead is intact; the operational lead narrowed this quarter.

Capital Structure & Dilution Mechanics

Because the reported loss keeps moving on financing rather than operations, a brief reconciliation helps square the $(0.66) Q1 loss to common with the operating story.

• The below-the-line drag. As in prior quarters, a large share of the net loss sits below operating income in interest, financing costs, and non-cash fair-value marks on warrant and derivative liabilities tied to the convertible and debt stack. The pre-NCI loss of $(249.6)M versus $(191.0)M to common reflects the Up-C structure allocating roughly $58.6M of the loss to the noncontrolling interest. These marks swing quarter to quarter, so the more-than-doubling of the loss to common sequentially is not an operating deterioration of that magnitude.
• The funding cushion. Cash and equivalents of $3.03B (with restricted cash bringing total liquidity to ~$3.5B) against the February 2.25% ten-year convertible struck at ~$116.30, above the current price, leaves the company funded past the 100-plus-satellite constellation with no further converts planned. There is no near-term refinancing wall and no dilutive-raise catalyst in view.
• The burn trajectory. The operative cash question is not solvency but pace: adjusted opex around $91M a quarter plus capex stepping to $575-650M in Q2 is a heavy burn, and while the balance sheet covers it comfortably, the spend is accelerating into a launch program that just lost a satellite. Coverage-wide burn estimates cited post-print run to roughly $1.6B in 2026, with positive cash flow not expected until 2028.

Assessment: The liability stack is clean, long-dated, and fully funded, and the reported loss remains a poor proxy for value, as it has every quarter. Readers should continue to judge the company on cash, capex, satellites on orbit, and revenue against guide rather than on GAAP EPS. The balance sheet is doing its job; it simply cannot do the job that matters this quarter, which is getting satellites to orbit.

Model Update Needed

Valuation impact: ASTS is not valuable on any near-term cash-flow multiple; it remains a call option on a constellation, and the near-term GAAP loss is a poor proxy for value. What changed this quarter is the probability weighting on the option, and it moved against the bulls. The array-deployment risk retired at Q4 stays retired (BB6 works, 98.9 Mbps demonstrated), but the launch-cadence risk we always flagged as the live variable materialized in the worst near-term form: a lost satellite and a grounded primary vehicle. The balance sheet still funds the whole build, the demand book still holds at $1.2B, and the FCC granted commercial authorization, so the terminal case is intact. But the path to it in 2026 is materially less certain than a quarter ago, and at $73, down 11.6% on the print, the stock still discounts on-schedule execution that just got harder. We move to Hold. We would grow constructive again on a clean Falcon 9 stacked launch in June followed by a New Glenn return-to-flight and a genuine multi-launch cadence; we would move to Underperform on a second launch failure, a slip of the June stacked launch, or a formal cut to the 45-satellite target without a credible recovery plan.

Thesis Scorecard Post-Earnings

We grade the standing thesis, pillar by pillar, against what this quarter's print and business update revealed. The reversion condition set at the Q4 upgrade (a material launch slip or a spending relapse reverts us toward Hold) fired.

Overall: the thesis weakened. Three of four bull pillars are confirmed and the balance sheet, demand, and government legs are intact, but the gating bear point, launch cadence at scale, materialized in its worst near-term form, and the capex/burn watch item escalated. The durability of the business is not in question; the timeline to realize it is, and the timeline is what the price embeds. This is the reversion condition we wrote into the Q4 upgrade, and it fired cleanly.

Action: Downgrade to Hold from Outperform. When we upgraded at Q4 we called it measured and set an explicit exit condition: a material launch slip or a spending relapse would send us back toward Hold. A launch failure that lost BlueBird 7, a grounded primary vehicle with no return date, zero net satellites on orbit, and a capex step-up to $575-650M is more than a slip. We upgraded when our pre-committed trigger fired in the bulls' favor, and discipline requires that we act now that it has fired against them. We stop at Hold rather than Underperform because the ~$3.5B balance sheet removes any solvency or dilution risk and the core technology is proven, so the setback is a matter of timing, not existence. We would return to Outperform on a clean June stacked launch plus a New Glenn return-to-flight and evidence of a genuine multi-launch cadence; we would move to Underperform on a second failure, a slip of the June launch, or a formal cut to the 45-satellite target. For now, the burden of proof is back on the company, and the rating reflects it.