Business

Know the Business

Iridium is a toll-bridge utility for places terrestrial wireless cannot reach: oceans, polar regions, remote land, war zones, and the cockpits and bridges that legally must have a backup channel. The economics are 51% EBITDA margins on $872M of revenue, sitting on a fully-paid-for 66-satellite constellation that does not need replacing for ~6 more years. The market debate is not whether the engine works — it is whether SpaceX, Apple-via-Globalstar, and AST SpaceMobile take the next leg of growth (direct-to-device on smartphones) before Iridium's own NTN Direct service can claim it.

1. How This Business Actually Works

Iridium is best thought of as a satellite tollbooth with a wholesale tax-collection arm. The 66-satellite constellation is a sunk cost that was rebuilt in 2010–2018 ("Iridium NEXT") for roughly $3 billion. Every additional minute of voice, every IoT ping, every aviation safety message routed across that constellation is sold not to end users but to ~520 partners (service providers, VARs, VAMs) who package and resell it. Iridium collects a wholesale access fee per subscriber plus usage, takes the cash, and lets distributors absorb sales, marketing, billing, and churn risk.

The economic engine has three features that matter:

1. The marginal cost of one more subscriber is essentially zero. The constellation costs the same to operate whether it carries 2.5M subscribers or 5M. So every dollar of incremental service revenue drops through at near-100% gross margin. This is why service revenue grew only 3% in 2025 yet operating income grew 18%.

2. The cost base is back-loaded by 15 years. Capex was $300–500M annually from 2010–2018 to launch Iridium NEXT. It collapsed to $40–100M afterwards, and management says it stays there until ~2031. Depreciation, however, runs at $210M/year — that's mostly a non-cash echo of the 2010s spend. The "real" cash earnings power is roughly EBITDA minus maintenance capex, not GAAP net income.

3. The EMSS contract is a cash annuity, but with a hard expiration. The U.S. Department of War pays $110.5M/year, fixed-price, for unlimited DoW airtime. It expires September 2026 with one six-month extension. A renegotiation is the single biggest 2026–27 catalyst — up or down.

The single most underappreciated mechanic: a 5% revenue increase produced an 18% operating-income jump in 2025 because almost all the increment fell to the bottom line. The same lever works in reverse if subscribers shrink.

2. The Playing Field

The peer set is not really one industry — it is four very different businesses lumped together by the word "satellite." Iridium is the only one currently making money at GAAP, EBITDA, and cash-flow levels simultaneously. Competitors are split between cash-burning growth bets (ASTS), turnaround stories (VSAT, SATS, CMTL), and Apple-juiced single-customer plays (GSAT).

The chart tells the story bluntly: Iridium has the best margin profile in the group but trades at a fraction of the valuation multiples awarded to GSAT (Apple optionality), SATS (spectrum hoard), and ASTS (D2D dream). The market is paying for narrative, not for cash. That is either a value opportunity or a signal that the whole "global D2D" wave reroutes around Iridium. Both interpretations are defensible — which is the actual investment debate.

What the peer set reveals about advantage:

• Only Iridium has a paid-for, in-orbit, profitable LEO global network. Globalstar's coverage is regional (no polar, limited oceans). Viasat/Inmarsat is GEO and broadband-shaped. Starlink (private) is broadband-shaped and weather-sensitive Ka-band, not L-band.
• Iridium's competitive vulnerability is precisely in smartphone D2D, where Apple+Globalstar (emergency SOS) and SpaceX+T-Mobile (now armed with EchoStar's S-band spectrum bought in 2025) are spending faster.
• Iridium's response is NTN Direct — a 3GPP-standard service launching from the existing constellation in 2026. If it works, IRDM is the cheapest way to play the D2D theme. If it does not, the 5 peers above will all chip at the edges of Iridium's commercial IoT and personal-tracking franchises.

3. Is This Business Cyclical?

Iridium is not economically cyclical in the usual sense — subscribers do not drop in a recession the way ad spend or auto sales do. Mining trucks, fishing fleets, FEMA, the DoW, and Boeing avionics keep paying the toll. The cyclicality is capital-cycle, not demand-cycle: every ~15 years the entire constellation must be replaced, and that single decision swamps ten years of operating profit.

Where short-cycle volatility does show up:

• Subscriber equipment revenue is the most cyclical line — handset and module sales swing 10–15% year to year on inventory and channel timing (down 11% in 2025).
• Maritime broadband behaves like a tech-replacement cycle: VSAT/LEO companion plans are eating Iridium's standalone broadband ARPU (broadband revenue down 10% in 2025 even as subscribers were nearly flat).
• Government engineering revenue (SDA contract, Golden Dome RFQs) is policy- and budget-driven, not economic-cycle driven, but is meaningfully lumpy quarter to quarter.
• 2018–2019 was a "stress test" — the COVID-era recession barely dented service revenue (+4% in 2020), confirming the mission-critical nature of the customer base.

4. The Metrics That Actually Matter

Forget P/E for this name — GAAP earnings are distorted by depreciation of a constellation that is being charged off faster than it is wearing out. The five numbers that genuinely drive value:

A few of these deserve a sentence of interpretation that the table alone does not give:

• Leverage at 3.7x is a deliberate choice, not distress. With FCF of $300M, debt is paid down slowly while management returns ~$60M/yr in dividends and (when not paused) buys back stock at low share-count prices. The Term Loan covenant kicks in only above 6.25x. There is real headroom.
• The 60% OEBITDA-to-FCF conversion is the number management talks to. It already accounts for capex, interest, and taxes. Watch this line for any deterioration — it is the earliest warning that the harvest phase is ending.
• EMSS revenue is concentrated, but the substitution risk is genuinely low. The DoW operates a dedicated gateway that physically only connects to Iridium's network. Switching cost is measured in years and billions.

5. What I'd Tell a Young Analyst

Three things to remember and one to fight about.

Remember: This is a capital-recovery story, not a growth story, and it should be valued accordingly — on FCF yield against a steady-state capex assumption, not on revenue multiples. The market keeps trying to value it like AST SpaceMobile (growth optionality) and then re-pricing it like Comtech (commoditized telecom). The truth is closer to a regulated utility with a 15-year reinvestment overhang. Build the model around per-subscriber unit economics, the $300M FCF baseline, and the expected timing of the next constellation. Everything else is decoration.

Remember: The moat is real but narrow. It exists in narrowband, low-power, mission-critical, polar-capable, weather-resilient applications — IoT trackers, GMDSS maritime safety, aviation cockpit data, DoW handsets. It does not extend to consumer smartphone D2D, where Apple+Globalstar and SpaceX+EchoStar will compete with bigger spectrum and bigger balance sheets. Do not confuse Iridium's L-band oligopoly with a defense against Starlink — they overlap less than the news cycle implies.

Remember: The operating leverage cuts both ways. Service revenue growth of 3% drove operating income up 18% in 2025. A 3% decline would do the inverse. Always model service revenue with both a sub growth and an ARPU input, separately for IoT (volume), voice (price), and broadband (mix shift).

Fight about this: Whether NTN Direct — Iridium's 2026 launch of 3GPP-standard direct-to-device service over the existing constellation — is real product or a press release. If real, it transforms Iridium from a niche utility into the cheapest way to play global D2D. If a press release, the valuation gap to GSAT and ASTS is justified and likely permanent. The next four earnings calls (Q1 2026 onward) will produce evidence either way: chipset partner names, MNO contract values, NTN Direct subscriber counts. That evidence — not the slide deck about it — is the thesis-changer.

What would change the thesis, in order of importance:

1. EMSS renegotiation outcome (2026–27). A 10–20% step-up validates the moat; a flat or down renewal would force a re-rating.
2. NTN Direct commercial traction by H2 2026 — chipset SKUs shipped, MNO partner ARPU disclosed.
3. Maritime broadband stabilization. This line has rolled over for two years; if companion-plan migration ends, it adds 100–200 bps to service growth.
4. Capital allocation discipline. Buybacks resumed at ≤$30/share are accretive; debt paydown below 3x leverage opens the door to a larger return-of-capital program.
5. A spectrum or M&A deal — Iridium's 8.725 MHz of L-band has option value if the next round of D2D consolidation includes it, similar to EchoStar's S-band sale to SpaceX in 2025.