Newsletter · · Ashutosh Agarwal

The Street Finally Put a Number on AST SpaceMobile - The Satellite & Space-Comms Race - Week of September 4, 2026

The Satellite and Space-Comms Race for the week of September 4, 2026: Berenberg initiated AST SpaceMobile at Buy with a $92 base case, a $242 bull case and a $23 bear case, SpaceX clawed back above its IPO price, and MoffettNathanson delivered the cleanest physics-based bear argument yet on satellite phones.

The Satellite & Space-Comms Race

Week of September 4, 2026: The Street Finally Put a Number on AST SpaceMobile

Week of August 28 to September 4, 2026. Sourced entirely from podcasts published in the last seven days. All figures USD unless noted.

Every so often a story stops being a message-board argument and becomes a spreadsheet. This week AST SpaceMobile crossed that line. A big European bank, Berenberg, started covering the stock with a "buy" rating and a $92 target, and, more usefully, showed its full math. That matters because for two years the ASTS bull case has mostly lived in the mouths of retail hosts and true believers. Now there's an institutional model you can argue with, and it lays out a genuinely wide fan of outcomes: a $23 downside if the satellites are late and nobody signs up, and a $242 moonshot if T-Mobile comes to the table and people actually use the thing.

Meanwhile SpaceX, the private giant that hangs over every name here, quietly clawed its way back above its own IPO price, and a sharp analyst from MoffettNathanson made the single best bear case on satellite phones I've heard all year. Rocket Lab, oddly, went silent. Let's get into it.

A note on the week's voices: this was an analyst-and-commentator week, not an operator week. No executive sat for a podcast interview. The best material came from a sell-side research note read aloud, a professional investor guest, and a couple of well-informed industry commentators. Where you see a quote from a CEO below, it was cited inside the Berenberg note, so treat it as second-hand.

TL;DR

  • AST SpaceMobile got its first proper big-bank initiation. Berenberg: Buy, $92 base target, on roughly $655M revenue in 2027 scaling to $4.5B by 2030 at 75%-plus margins. Bull case $242 if T-Mobile signs; bear case $23 if service slips. A spectrum "floor" of about $26 is the reason the downside isn't zero.
  • SpaceX's stock bounced back over its IPO price, from below $135 in late August to roughly $149 and up 6% on Wednesday, testing a breakout. The retail crowd is still wary into year-end, but the technicians say it "found its footing."
  • The real debate crystallized: MoffettNathanson's Julie Zhu made a physics-based case that Starlink will not become a fourth U.S. carrier and that its broadband market is smaller than the hype, the cleanest steel-man of the bear side to date. Rocket Lab, Globalstar, Iridium and EchoStar were all quiet on podcasts this week.

What's new

1. AST SpaceMobile: the Street finally put a number on it

Podcast: AST SpaceMobile Podcast, "Why AST SpaceMobile Could Hit $242 (Read by AI)" (Sept 3), an AI narration reading Berenberg's equity research note verbatim.

This episode is unusual: it's a text-to-speech reading of Berenberg's actual initiation report, word for word. Strip away the robot voice and it's the most complete institutional model on ASTS we've seen.

The core pitch: AST is "the only company to have demonstrated true cellular broadband from space to unmodified smartphones," meaning a normal phone in your pocket, no special hardware, connecting straight to a satellite. (That's what "direct-to-device," or D2D, means throughout this newsletter.) Berenberg's argument is that AST complements the big carriers instead of stealing their customers, which is why AT&T, Verizon, Vodafone and others have signed on, more than 60 mobile-network-operator partnerships covering roughly 3 billion subscribers.

The numbers that move the stock:

  • Revenue ramp: about 45 satellites in orbit enabling continuous service in key markets by mid-2027, driving "revenue of circa $655 million in 2027, scaling to $4.5 billion in 2030 at 75%-plus EBITDA margins." Berenberg stresses this only needs paid uptake from less than 3.5% of partner subscribers, a low bar if the service works.
  • Valuation method: 15x 2030 EBITDA, discounted back at 15% for execution risk, gets you the $92 base case.
  • The fan of outcomes: a blue-sky $242 (238% upside) if T-Mobile joins and take-rates run hot, on $12B revenue and $8.8B EBITDA in 2030. A bear case $23 (64% downside) if service slips to end-2027 with 1% adoption. And critically, a spectrum floor around $26: Berenberg argues AST's 45 MHz of L-band rights alone are worth roughly that even if the business never scales.

That floor is the whole reason this isn't a zero-or-hero lottery ticket. Spectrum, the licensed radio airwaves a signal travels on, has become scarce and expensive, and AST owns or controls a chunk of it (45 MHz of L-band, 60 MHz of S-band) plus shared access to its carrier partners' premium low-band.

On deployment: 13 Bluebird satellites are now in orbit after Bluebirds 11 to 13 flew in early August (six spacecraft launched within 50 days). AST has 10 launches booked with two providers, aiming for one launch every one-to-two months. Blue Origin has fixed the root cause of its late-May failure and targets a return to flight this year, though management says it is "explicitly not counting on New Glenn" in its plan. Beta service with real users is reaffirmed for 2026; over 3,000 of about 5,600 U.S. ground cells are already active.

Why it moves numbers: this is the first time a mainstream bank has published a defensible framework, a price target, and a spectrum-based downside floor. It gives long-only funds permission to own the story and gives everyone a shared scoreboard: hit about 45 satellites and start beta, and the base case firms up; land T-Mobile, and you're chasing the $242.

2. SpaceX climbs back over its own IPO price

Podcasts: Schwab Network, "The Big 3: VZ, NEM, SPCX" (Sept 3, technicians Dan and Rick) and Not Me, But You!, "SpaceX/Tesla Update" (Aug 31, retail host).

Two very different voices, one story: the mood on SpaceX shares turned this week. On the retail side, the "Not Me, But You!" host noted SpaceX "closed yesterday at $140 per share, just about $10 less than the initial price on opening day of the IPO," and said he expects it to drift lower into December, so he's dollar-cost-averaging (buying small slices over time).

By Wednesday the technicians on Schwab Network were looking at a very different chart: SpaceX around $149, up more than 6% on the session, having pushed above its heaviest trading zone (roughly $132 to $142) and now testing a breakout at $150, with $176, the post-IPO peak, as the next big resistance level. As their trader put it, "I think it's found its footing… there is potential here to capture a significant breakout above the 150 level."

Why it moves numbers: SpaceX's private-market price is the sentiment anchor for the entire sector. Last week it had cracked below its June IPO price on four straight down days; this week it recovered the line. For the pre-revenue names (ASTS, GSAT), a firming SpaceX makes risk capital cheaper and the whole D2D theme easier to fund.

3. Starbase Louisiana is real, and it reshapes the launch map

Podcasts: Main Engine Cut Off, "T+339: Starbase Louisiana, the Falcon Fade, and the Next Wave of Launch Vehicles" (Aug 28, Anthony Colangelo) and FYI, "Elon Musk's $1 Trillion Chip Factory" (Sept 2, ARK Invest's Sam and Brett).

SpaceX officially announced a giant new launch and manufacturing complex on the Louisiana coast, west of New Orleans. The plan Colangelo lays out is staggering: access to 125,000 acres (only a fraction to be built out), five launch complexes with two pads each, a stated $100 billion spend, construction starting 2027 and first launch around 2029.

The clever part is geography. From the existing Texas Starbase, SpaceX can't easily fly the "sun-synchronous" polar paths (orbits that keep a satellite in near-constant sunlight) because you'd overfly all of Mexico. The Louisiana coast opens a clean southbound corridor over the narrowest slice of Mexico, which is exactly the trajectory the planned AI and data-center satellites and polar Starlink launches need. It also sits on some of the cheapest natural-gas pipeline capacity in the country, and rockets this size burn methane made from natural gas.

ARK's team framed the scale in plain terms: this ranks among the largest infrastructure projects in history. The U.S. highway system was roughly $700 billion in today's dollars; the International Space Station about $200 billion. To reach Musk's stated goal of a terawatt of orbital compute per year, ARK says he's claiming he'll need $1 trillion of capital in the facility. Their bull logic rests on return on capital: they estimate SpaceX's terrestrial data centers earn a roughly 75% internal rate of return, and that by the "thousandth AI satellite" the returns beat Starlink itself.

Colangelo also flagged the quieter structural shift: the "Falcon fade." As of this week there are no more Falcon 9 Starlink launches off the U.S. East Coast; as Starlink migrates to bigger V3 satellites that only fit on Starship, Falcon 9's launch count will trend down over the next couple of years. He compared incumbent-carrier dismissiveness about Starlink's phone ambitions to the famous Nokia-versus-iPhone moment:

"Somebody kind of putting their head in the sand… it sounded to me a lot like a famous comment when the iPhone was coming out… 'the PC guys aren't just going to walk in and figure this out.' And well, here we are, 20 years later, they figured it out and flipped the industry on its head."

Why it moves numbers: the Louisiana buildout is the physical proof of SpaceX's "become the biggest company in ten industries at once" thesis, and the launch-cadence ramp it enables is what ultimately decides how fast Starlink Direct-to-Cell and orbital compute can scale against AST.

4. Starship Flight 14 slips to mid-September, but with a first

Podcast: Astronomy Daily, "Forty Times Brighter Than the Full Moon" (Aug 28).

Older chatter had Starship's Flight 14 flying around August 28. That's stale: per internal SpaceX and FAA planning documents cited on the show, the real "no earlier than" date is September 15. When it flies, it aims to do something Starship hasn't yet, "a full orbital flight deploying the newer, larger Starlink V3 satellites." The tower-catch attempt on the returning booster is pushed back "a few months"; the ship will splash down in the ocean instead. Separately, Ship 40 became the first Starship upper stage to survive reentry and splashdown intact and was towed home for inspection.

Why it moves numbers: the V3 satellite is the workhorse of Starlink's next phase, with bigger antennas, more capacity, and the platform SpaceX needs for serious Direct-to-Cell. Every slip in Starship pushes that capacity out; the first successful V3 batch is the milestone to watch.

The debate: how big is the direct-to-device prize, really?

This is the argument the whole universe turns on, and this week we got both sides articulated well.

The bull steel-man (Berenberg, via the AST SpaceMobile Podcast): Coverage is becoming a service you can buy. Roughly 30% of U.S. land has no 4G, and carriers own low-band spectrum sitting idle over that emptiness. AST is the only operator that can turn those dead zones into broadband to a normal phone without disintermediating the carrier, so carriers hand over their most strategic asset (spectrum) willingly. The three big U.S. carriers are forming a joint venture for satellite coverage, and two of the three (AT&T, Verizon) are already AST partners and investors. Verizon's Dan Schulman, quoted in the note, drew the line straight at the Starlink model:

"There is no reason that we can see to extend an MVNO to any satellite player."

Translation: carriers want to buy satellite coverage wholesale and resell it themselves, not let a satellite company sell phone service straight to consumers. That's the whole ballgame for AST, whose model is partnership; it's a headwind for Starlink, whose model is to go direct.

The bear steel-man came from Julie Zhu of MoffettNathanson on The Rundown (Aug 30). Zhu's report argues satellite phones are a supplement, not a business that reorders telecom, and she reaches for high-school physics to explain why:

"When you move away in distance, your signal weakens by the square of that distance… you have a satellite that's 220 miles in the air versus these macro towers that are maybe five miles away maximum. So it's just a difficult proposition."

Her conclusion: SpaceX won't become a real fourth U.S. carrier; the likely endgame is a wholesale MVNO deal to fill rural gaps. She's also more bearish than consensus on Starlink broadband: by 2030 to 2031 she thinks the constellation can only serve U.S. households in the third-to-fourth least-dense deciles, because nobody in Manhattan is putting a dish on the roof, which caps the market. And she's skeptical of the femtocell "radios on rooftop dishes" idea SpaceX floated, because it runs into the same permitting headaches as towers and it's a low-margin business: "If you look at Verizon, AT&T… these aren't 80% margin businesses."

Where that leaves us: the two sides actually agree on the physics, satellites can't match tower capacity, but disagree on what that means. The bull says it makes AST's carrier-partner model the only sane way to do this. The bear says it makes the whole D2D prize smaller and lower-margin than the TAM slides suggest, and confines it to a supplemental, rural, wholesale niche. Both can be right at once: AST wins the D2D niche and the niche is smaller than the $9B-by-2035 dream. That's the crux for position sizing.

Stocks in play

AST SpaceMobile (ASTS), the week's main event

  • Bull: First big-bank Buy (Berenberg, $92); only demonstrated broadband-to-phone player; owned spectrum gives a roughly $26 valuation floor; 13 Bluebirds up; beta reaffirmed for 2026; carrier JV "frees up a third and fourth customer"; Japan's JLEO win brings up to about $1B of non-dilutive government money.
  • Bear: It's a pre-revenue stock on an "elevated near-term multiple"; the whole thesis leans on hitting about 45 satellites by early-to-mid 2027, and launch cadence is the binding constraint; Legato L-band approval is contested (GPS-interference objections); the carrier JV concentrates buying power and could cap AST's pricing. Berenberg's own bear case is $23.
  • Next catalyst: Start of beta service; the next Bluebird launch batch; any Grain 800 MHz partner selection; a formal T-Mobile agreement (worth about $29/share on Berenberg's math).

SpaceX / Starlink (private), sentiment anchor and the competitor

  • Bull: Stock recovered above its IPO price and is testing a breakout; Louisiana buildout underway; ARK models roughly 75% data-center IRR and a path to $1T of AI-compute investment; first exclusive nationwide D2D spectrum (about 65 MHz of ex-EchoStar airwaves) plus a filed 15,000-satellite plan.
  • Bear: MoffettNathanson says the phone business is an MVNO niche and broadband TAM is capped; the "Not Me, But You!" host expects the shares lower into December; the femtocell mobile plan is low-margin and permitting-heavy; Starship Flight 14 slipped to mid-September.
  • Next catalyst: Starship Flight 14 (no earlier than Sept 15) and the first Starlink V3 deployment; further trajectory versus the $150 and $176 technical levels.

Rocket Lab (RKLB) drew no dedicated podcast commentary this week: nothing new on the Neutron first-flight date, the $8B Iridium acquisition, or launch economics. After a very loud few weeks (Q2 results, the $397M Space Force award, Neutron slippage, the Iridium deal), that's a notable change of pace. Watch: a confirmed Neutron first-flight date; the Iridium shareholder vote and deal close; any revised valuation once Iridium's cash flows are folded in.

Globalstar (GSAT) appeared only indirectly: MoffettNathanson's density argument and the carrier-JV framing raise the bar for every small-satellite player. Iridium (IRDM) surfaced only as Rocket Lab's acquisition target in prior weeks. EchoStar / Hughes (SATS) is notable mainly because SpaceX's newly exclusive D2D spectrum is the former EchoStar AWS-4 and H-block, a reminder of how much value has already migrated out of this name.

Read-throughs

  • Carrier partners (VZ, T, TMUS). Verizon appeared on Schwab's "Big 3" segment but purely as a chart, not a strategy discussion. The substance lived inside the Berenberg note: the three-carrier JV, Verizon's "no MVNO for any satellite player" line, and the hire of Chris Sambar, the ex-AT&T network chief and AST board member, as T-Mobile's chief enterprise officer (effective by Oct 14). Berenberg reads all of it as tilting toward a carrier-controlled, AST-shaped architecture, while stressing its base case does not assume T-Mobile picks AST.
  • Spectrum and the Grain 800 MHz process. A contrast to last week's "$6 billion spectrum war" noise: the only update came inside Berenberg's note. AST holds a 30-day FCC test authorization (granted Aug 13, running to Sept 12) to trial the exact Grain-held 800 MHz blocks near Midland, Texas and Lanham, Maryland, and "more than 80% of its Block II satellites in production carry this frequency." Berenberg believes Starlink's current satellite design can't form compliant beams in that low band, making AST close to the only technically qualified bidder if Grain runs its selection.
  • Defense and national security. A striking read-through came from Click Here, "Iran: Why send a Tomahawk when Starlink will do?" (Sept 1). The guest argued that "if we were to drop a couple million activated Starlink terminals into Iran, that would cost the American taxpayer less money than a single Tomahawk missile," and pointed to direct-to-cell as a way to bypass a state-controlled internet, noting Congress has weighed legislation to enable it. It's a vivid illustration of the sovereign and national-security demand that both AST (its SHIELD and Halo defense awards) and Starlink are chasing.
  • Launch and component suppliers. Blue Origin's New Glenn came up only as AST's backup launch provider (root cause of the May anomaly resolved; return to flight targeted this year, but not in AST's plan). ULA was quiet. The bigger supplier story is the coming "Falcon fade" and the new crop of launch vehicles Colangelo flagged as under development.
  • SpaceX private-market valuation as sentiment gauge. Covered above: the reclaim of the IPO price is the single most useful sentiment tell for the group this week.

What changed vs last week

  • AST went from message-board to model. Last week's AST coverage was a retail-analyst host narrating the Grain "spectrum war." This week a mainstream bank initiated at Buy with a full framework: $92 base, $242 bull, $23 bear, roughly $26 spectrum floor. That's a real escalation in the quality of the bull case.
  • SpaceX sentiment flipped. Last week the stock had cracked below its roughly $135 IPO price on four straight down days. This week it recovered to about $149 and is testing a breakout, even as the retail crowd stays cautious into December.
  • Starbase Louisiana became concrete. Last week it wasn't a discrete topic; this week it's an officially announced $100B, five-complex site (first launch around 2029), and both MECO and ARK dug into what it means.
  • Starship got a date. Last week: nothing specific on Flight 14. This week: no earlier than Sept 15, first V3 deployment, tower-catch deferred.
  • The D2D-TAM debate got a rigorous bear. Last week the skepticism was scattered. This week MoffettNathanson's Julie Zhu gave it a spine: physics, density deciles, MVNO-only outcome.

The bottom line

The useful thing that happened this week wasn't a price move, it was the arrival of a shared scoreboard. Berenberg's fan of outcomes gives every ASTS holder the same three checkpoints to argue over: satellites in orbit, beta service, and whether T-Mobile ever signs. Set against Julie Zhu's physics, the honest synthesis is that AST probably wins the direct-to-device niche and the niche is probably smaller than the slides say. Meanwhile SpaceX reclaimed its IPO price and committed to a $100 billion coastline. Watch launch cadence, watch the Grain selection, and watch whether Rocket Lab breaks its silence.