Snapshot

Satellite communications is a $105B global services market (SIA, 2025) dominated by a handful of operators selling voice, data, IoT connectivity, and broadband from orbit. The three US-listed companies here occupy different niches: Iridium (IRDM) operates the only LEO constellation with pole-to-pole voice/data/IoT coverage and earns $872M/yr at 57% OIBDA margins; Globalstar (GSAT) runs a 31-satellite L-band network that is 85%-dedicated to Apple's iPhone emergency satellite features and is being acquired by Amazon for $90/share (expected close 2027); Redwire (RDW) does not operate satellites but manufactures satellite components — solar arrays, antennas, sensors, and in-orbit infrastructure — selling to NASA, DoD, and commercial operators at $335M/yr revenue with negative EBITDA. The unit of product differs sharply: IRDM and GSAT sell bandwidth (measured in kbps-to-Mbps narrowband links, not Gbps broadband); RDW sells hardware.

Satellite bandwidth demand is being pulled by two forces: direct-to-device (D2D) smartphone connectivity — Apple's iPhone already uses Globalstar for emergency SOS, and Amazon plans D2D voice/data from 2028 — and IoT/M2M connections growing at 26% CAGR est.. Supply of narrowband L-band/S-band spectrum is physically finite and fully allocated; broadband capacity (Ku/Ka-band) is expanding rapidly via Starlink (~7,000 satellites) and Amazon Kuiper but does not compete directly with IRDM/GSAT's narrowband niche. IRDM trades at $6.9B EV generating ~$305M FCF (4.4% yield) and returning 9.5% of market cap annually via buybacks + dividends; GSAT is an arbitrage on a $90/share Amazon deal closing in 2027 vs. $82.64 today; RDW is a money-losing satellite component manufacturer trading at $4.1B on $335M revenue with $411M backlog.

The product & how money is made

These three companies sell fundamentally different products in the satellite value chain:

Iridium (IRDM) — satellite network operator

Iridium operates 66 active LEO satellites (plus spares) in 6 polar orbital planes at 781 km altitude, providing the only satellite network with true pole-to-pole global coverage. Iridium satellites have inter-satellite cross-links (Ka-band, 10 Mbps each), meaning calls and data can route through space without touching a ground station at every hop. This is architecturally different from Globalstar's "bent-pipe" transponder model, which requires line-of-sight to a ground gateway.

Source: IRDM FY2025 earnings release (Feb 2026); IRDM FY2024 10-K.

Service revenue (74% of total) is recurring subscription income. ARPU ranges from $7.29/month (IoT) to $268/month (broadband). The DoD contract (EMSS) provides ~$108M/yr with high certainty.

Globalstar (GSAT) — Apple's satellite partner (Amazon acquisition pending)

Globalstar operates 31 L-band LEO satellites using a bent-pipe architecture (no inter-satellite links; each satellite relays signals to the nearest ground gateway). Since 2022, Apple has used Globalstar's network for iPhone Emergency SOS via satellite. In November 2024, Apple invested $1.5B total — $1.1B in service prepayments, $400M for a 20% equity stake, and $232M to retire debt — and secured 85% of Globalstar's network capacity. In April 2026, Amazon announced it will acquire Globalstar for $90/share (cash or 0.3210 Amazon shares), expected to close in 2027. Amazon also signed a deal with Apple to continue powering satellite connectivity on iPhone 14+ and Apple Watch Ultra 3.

Source: GSAT FY2025 10-K via stockanalysis.com; Apple partnership per satellitetoday.com (Nov 2024); Amazon acquisition per press.aboutamazon.com (Apr 2026).

Apple dominates Globalstar's economics. The $1.26B in "other long-term liabilities" (up from $14M pre-Apple) represents deferred Apple prepayments. Revenue roughly doubled post-Apple deal. With the Amazon acquisition pending at $90/share vs. ~$82.64 current, this is now primarily a merger-arbitrage situation.

Redwire (RDW) — satellite component manufacturer

Redwire manufactures components and subsystems used by satellite operators, NASA, and defense agencies: roll-out solar arrays, deployable antennas, sensors, in-orbit 3D-printing systems (Archinaut platform), and space station infrastructure. It acquired Edge Autonomy in 2025, adding a Defense Tech segment (unmanned systems, ISR payloads). Customers include NASA (Artemis program), U.S. DoD, ESA, and commercial satellite builders.

Source: RDW FY2025 earnings release (ir.redwirespace.com); Q1-2026 10-Q.

The Edge Autonomy acquisition explains the goodwill jump ($71M to $779M), share dilution (67M to 192M shares), and the worsening cash burn. Revenue guidance for FY2026 is $450M-$500M. Book-to-bill was 1.32x in FY2025 and 1.92x in Q1-2026 (backlog grew to $498M).

Demand

Contracted / locked-in demand

• IRDM: DoD EMSS contract provides ~$108M/yr in government service revenue. contracted ~2.54M billable subscribers (Dec 2025), up 3% YoY, with IoT subscribers growing 6%. Aireon ADS-B hosted payload revenue is under long-term contract. FY2026 guidance: service revenue flat to +2% ($634M-$647M); OIBDA $480M-$490M.
• GSAT: Apple committed 85% of Globalstar's network capacity under a multi-year take-or-pay arrangement. Apple prepaid ~$1.1B in service fees plus $232M for debt retirement. contracted Amazon's $90/share acquisition (announced Apr 2026) — if the deal closes, shareholders get $90/share regardless of operational performance. Closing conditioned on regulatory approval and satellite milestone delivery, expected 2027.
• RDW: Contracted backlog of $498M as of Q1-2026 (record). contracted Includes $1.8B Andromeda IDIQ ceiling for advanced spacecraft (task orders awarded as needed), ELSA solar array program for Moog, Marine Corps Stalker orders, NASA and ESA contracts. Book-to-bill 1.54x LTM.

Demand forecasts

• Satellite IoT (IRDM's growth engine): The satellite IoT market is projected to reach $4.7B by 2030, growing at 26% CAGR (IoT Analytics). est. Iridium's IoT subscriber base (2.0M, +6% YoY) is the largest in the industry. 3GPP NTN standards (Release 16-19) will enable satellite connectivity on standard cellular chipsets, potentially expanding the addressable market from specialized satellite modems to billions of smartphones and IoT devices.
• Direct-to-device (D2D): Apple pioneered D2D with Globalstar in 2022 (emergency SOS). Amazon's acquisition of Globalstar signals intent to build D2D voice, data, and messaging from 2028. Iridium has its own D2D program, "Iridium NTN Direct," accepted into 3GPP standards (Sept 2024), targeting service with a next-gen constellation. Revenue is nascent. est.
• Satellite broadband: 10M+ satellite broadband subscribers globally in 2025, up 62% YoY (SIA). Nearly all growth is Starlink. This does not directly benefit IRDM or GSAT (narrowband operators) but validates the broader satellite demand thesis.
• Space/defense manufacturing (RDW): 4,434 satellites launched in 2025 across 296 commercial launches, up 65% YoY (SIA). Satellite manufacturing revenue was $20.4B globally (SIA, 2025). Government space budgets drove ~80% of demand. est. RDW's 2026 revenue guidance of $450M-$500M implies 34-49% growth.

Supply

Capacity — what exists today

• IRDM: 66 active satellites (Iridium NEXT generation, launched 2017-2019 on SpaceX Falcon 9), plus 9 on-orbit spares and 6 ground spares. Each satellite supports up to 1,100 concurrent phone calls at 2.4 kbps. Iridium Certus provides up to 704 kbps per terminal. Total system throughput is on the order of low single-digit Gbps across the entire constellation — a narrowband network. L-band spectrum: 7.775 MHz exclusive plus 0.95 MHz shared with Globalstar (1,617-1,626 MHz). This spectrum is globally licensed and cannot be expanded — a hard supply constraint.
• GSAT: 31 L-band satellites. Apple uses 85% of capacity. The MDA-built replacement constellation (17 satellites, options for 9 more) will refresh the fleet but not dramatically expand total capacity. Ground station footprint limits coverage to areas within view of gateways (no inter-satellite routing). Total throughput is similarly narrowband (sub-Gbps constellation-wide).
• RDW: Manufacturing capacity, not orbital capacity. $498M backlog implies roughly 12-15 months of production at the $450-$500M guided run-rate. est. Facilities in the US, Luxembourg, and Belgium. The Edge Autonomy acquisition added unmanned systems production lines.

Supply bottleneck

• Spectrum is the binding constraint for IRDM and GSAT. L-band and S-band allocations are fixed by ITU treaty. Unlike broadband (Ku/Ka-band, where SpaceX and Amazon are adding massive capacity), narrowband MSS spectrum cannot be expanded. More satellites on the same spectrum only improves geographic coverage and link reliability, not total system bandwidth. This is a permanent, physics-level supply cap.
• Iridium's next-gen constellation: Iridium is designing a follow-on system with roughly 4x more satellites than the current 66. CEO Matt Desch indicated the cost would be "not greater than the network cost we experienced last time, and probably a bit lower" (the NEXT constellation cost ~$2.9B total including launches). No deployment timeline has been committed. Current satellites were designed for 15+ year lifespans (launched 2017-2019), so replacement is not urgent before the late 2020s to early 2030s.
• Launch cost deflation: Launch costs have fallen from ~$50,000/kg to LEO in 1995 to under $3,500/kg on SpaceX Falcon 9 — a 14x reduction that makes constellation replenishment and expansion far cheaper than the original deployments. est.
• RDW's bottleneck is integration complexity, not factory floor space. Space-grade hardware has long development cycles (12-36 months from contract to delivery), and the company is simultaneously integrating Edge Autonomy while scaling production.

The gap

The demand-supply picture differs fundamentally between broadband satellite (Starlink, Kuiper — Gbps per user, capacity-rich) and narrowband MSS (Iridium, Globalstar — kbps per link, spectrum-constrained). These three tickers sit in the narrowband / hardware niches.

Pricing direction: For narrowband MSS (IRDM), pricing is stable-to-up because spectrum is finite and demand is growing — the signature of a supply-constrained market. For satellite hardware (RDW), revenue is growing fast but the company is not yet profitable, suggesting costs (especially from acquisitions and integration) are absorbing top-line gains.

The biggest unknown is what happens when D2D scales. If Iridium NTN Direct succeeds, it could unlock billions of smartphone endpoints on Iridium's network — but the same finite spectrum must serve all of them, which means either strict bandwidth rationing per device or a much larger constellation. Amazon's acquisition of Globalstar and plan for "thousands of advanced satellites" suggests the tech giants are building their own supply rather than relying on legacy MSS operators.

The players

Source: IRDM FY2025 earnings, Q1-2026 10-Q; GSAT 10-K FY2025 via stockanalysis.com; RDW FY2025 earnings, Q1-2026 10-Q; market data Jun 2, 2026.

Not listed but relevant competitors: Starlink (SpaceX, private) — ~7,000 LEO satellites, broadband, dominant in consumer/enterprise, entering IoT; Amazon Kuiper (private, launching) — acquiring GSAT, planning D2D from 2028; AST SpaceMobile (ASTS, public) — building broadband-speed D2D from LEO; Viasat (VSAT, public) — GEO broadband, $2.0B revenue; SES (Luxembourg-listed) — MEO/GEO broadband.

The price of exposure

IRDM — $49.60/share, $5.24B market cap, $6.92B EV

The P/E of 50x vs. P/FCF of 17x gap exists because Iridium carries ~$212M/yr in depreciation on its $2.9B constellation, which suppresses reported earnings but does not consume cash (the satellites are already paid for). The constellation has 10+ years of remaining useful life. FCF is the better representation of cash generated. Management projects at least $1.5B-$1.8B in cumulative FCF through the end of the decade. Target net leverage: below 2.0x by end of decade (from 3.4x today). Cash taxes expected below $10M/yr through 2027 (NOL shields).

GSAT — $82.64/share, $10.64B market cap (Amazon deal pending at $90/share)

With a $90/share acquisition price and $82.64 current market price, the spread is $7.36/share (8.9%). Expected close in 2027. The spread compensates for: (1) deal-break risk if GSAT fails to hit satellite milestones, (2) regulatory approval uncertainty, and (3) time value of money over ~12 months to close. If the deal breaks, the stock likely re-rates sharply lower — pre-deal the stock traded near $2 (before the Apple partnership transformed the business in late 2024). The downward adjustment clause allows up to $110M reduction ($0.85/share) for missed milestones.

RDW — $20.58/share, $4.07B market cap

Revenue is growing 34-49% (FY2026 guide) with a 1.54x LTM book-to-bill, but EBITDA is negative, FCF burned $201M in FY2025, and the share count nearly tripled from 67M to 198M on the Edge Autonomy acquisition. The $779M in goodwill (54% of total assets) creates impairment risk if growth disappoints. Q1-2026 showed improvement: gross margin rose to 26.6% (from 14.7%), and operating cash burn narrowed sharply to -$6.7M (from -$45M). The $1.8B Andromeda IDIQ ceiling represents long-term potential but is not a guaranteed order — task orders are awarded individually.

Moat & risks

Structural advantages

• IRDM: Only pole-to-pole LEO constellation with inter-satellite cross-links. L-band spectrum is permanently allocated and cannot be replicated. DoD dependency (EMSS contract) creates switching costs. 2.54M subscriber base with embedded IoT devices that are expensive to replace. Constellation replacement cost (~$3B) is a barrier to new entrants in narrowband MSS.
• GSAT: Apple's 85% capacity lock-up is a de facto exclusive. Amazon's acquisition, if it closes, removes GSAT from the competitive landscape entirely.
• RDW: Space-qualified manufacturing certifications (ITAR, NASA flight heritage) take years to obtain. Few competitors can build roll-out solar arrays and deployable antennas at scale. But the moat is narrow — contract-based revenue with no recurring subscription component.

Key risks

• IRDM: Next-gen constellation capex ($2.5B-$3.5B est.) could consume years of FCF or require new debt. Starlink and Amazon Kuiper entering IoT/D2D could erode pricing power in adjacent markets. Regulatory risk if 3GPP NTN standards favor competitors. The 3.4x net leverage leaves limited margin for error if revenue stalls.
• GSAT: Deal-break risk is the dominant concern. If Amazon walks away or milestones are missed, the stock could fall 80%+ from current levels. Even with Apple, standalone GSAT generated only $273M revenue at negative net income.
• RDW: $779M goodwill on $335M revenue — any slowdown in defense/space budgets triggers impairment. Cash runway is limited (~$145M cash, -$201M FCF in FY2025) and may require additional dilutive financing. Edge Autonomy integration is absorbing management attention. Government contract concentration means budget sequestration or program cancellation is existential.

What to deep-dive next

• IRDM next-gen constellation economics: How much will it cost ($2.5B-$3.5B range implied by management), when will it deploy, and how will Iridium fund it? If buildable for ~$3B from operations, the business becomes a perpetual FCF machine. If it requires $5B+ and new equity, the capital structure changes materially.
• IRDM's D2D / NTN Direct timeline: Accepted into 3GPP standards (Sept 2024). What chipset partners? How much incremental revenue per connected smartphone?
• GSAT deal-break analysis: What exactly are the "HIBLEO-4 replacement satellite milestones" that condition the Amazon close? What is the realistic probability of milestone failure? If the deal breaks, what is GSAT worth standalone?
• RDW path to profitability: When does gross margin improvement (14.7% to 26.6% in one year) translate to positive EBITDA? What does the Space segment look like on its own without acquisition noise?
• Read the 10-K filings directly: IRDM 10-K for EMSS renewal terms, next-gen capex estimates, and D2D investment cadence. GSAT 10-K for full Apple contract terms and Amazon merger proxy. RDW 10-K for Edge Autonomy purchase price allocation and Andromeda IDIQ structure.

Sources & confidence

Data freshness: Company financials current through Q1-2026 filings (Mar 31, 2026). Market prices as of Jun 2, 2026. SIA industry data published 2025. Amazon-Globalstar deal announced April 2026, expected close 2027. GSAT individual-quarter financials beyond Q1-2026 were unavailable for direct verification; FY2025 totals are from 10-K filings.