Business

Know the Business

Iridium is a fixed-cost satellite tollbooth that finished a $4 billion capital cycle in 2019 and has spent the seven years since converting depreciation into cash. The economics that matter are not telecom-like — they are pipeline-like: a near-zero marginal-cost network, a wholesale channel of ~520 partners, and a 17.5-year runway before the next big capex bill comes due. The thing the market most often gets wrong is treating the share price as a referendum on next quarter's voice-and-data ARPU when the real drivers are spectrum value, the EMSS recompete, and what cash does between now and 2031.

1. How This Business Actually Works

Iridium owns a 66-satellite low-Earth-orbit constellation that is the only mobile-satellite network with full polar-to-polar coverage and inter-satellite links — meaning it does not depend on a dense ground-station network like Globalstar or Orbcomm. The cost of operating that network is essentially fixed. Every additional billable subscriber, hosted payload, or engineering contract drops to the bottom line at near-network-marginal cost, which is why EBITDA margins sit above 50% on $872M of revenue.

The customer-facing model is wholesale, not retail. Iridium does not chase end-users. Roughly 120 service providers, 310 value-added resellers, and 90 value-added manufacturers (Garmin, Thales, Honeywell, Cobham, Intellian, Collins, etc.) embed the Iridium chip or terminal into industry-specific products and resell airtime. Iridium captures the recurring monthly access fee and lets distributors bear the customer-acquisition cost, churn risk, and product-development effort. This is what gives it 71% gross margins on a network with only ~975 employees.

Three things to internalize about the revenue mix:

One large, multi-year U.S. government contract anchors the floor. The Enhanced Mobile Satellite Services contract pays a fixed $110.5M per year for unlimited DoD usage through September 2026, with a six-month government extension option that management already expects to be exercised. A successor contract is in negotiation. EMSS is unusual: it is not priced per minute or per subscriber, so the marginal cost of the ~121,000 government users on the network is close to zero. This is why government service revenue at 12% of total has had outsized influence on cash flow.

IoT is the volume story; broadband is the headache. IoT subscribers (1.998M) are 83% of the total subscriber count but only 35% of commercial service revenue because ARPU is $7.78. That low ARPU is the point — Iridium is targeting use cases where a $20/month cellular plan is uneconomic (heavy-equipment telematics, asset tracking, scientific buoys). Maritime broadband, by contrast, is the only segment in active decline: ARPU fell from $282 to $259 as customers moved Iridium Certus from primary to backup-companion service behind Starlink. Management acknowledges this and is leaning on GMDSS safety regulation to defend the floor.

Engineering revenue is real but lumpy. The 18% engineering-and-support line is largely the SDA (Space Development Agency) subcontract through General Dynamics — $240M to Iridium over five years for ground-segment work on Tranche 1 of the Proliferated Warfighter Space Architecture. It is high-revenue, low-margin pass-through work, not core service economics, and should be valued differently than recurring airtime.

2. The Playing Field

There is no clean peer set, and that itself is the point. Iridium is the only profitable, free-cash-positive, pure-play global mobile-satellite operator left standing.

The peer table makes one thing inescapable: in the satellite-services universe, market value has decoupled from current profitability. AST SpaceMobile trades at a $22B market cap on $71M of revenue and $340M of net losses. Globalstar — about to be acquired by Amazon for D2D spectrum — trades at a $6.5B market cap on $273M of revenue. EchoStar is in restructuring and just sold $26B of S-band spectrum to SpaceX. Iridium, the only operator generating $300M of free cash flow per year, has the smallest market cap on the list except Comtech.

That is not a market mistake — it is a different bet. Investors paying premium multiples for ASTS and GSAT are buying a future where consumer smartphones connect directly to satellites and pay AT&T-like ARPUs to whoever owns the spectrum. Iridium has chosen the opposite path: stay specialized in mission-critical, low-power, narrowband, and trust that "the boring 50%-margin cash business" is worth more than the option on consumer D2D. Whether that is right is the central debate.

What "good" looks like in this industry is not revenue growth — it is return of capital without diluting shareholders while the constellation runs. By that yardstick, Iridium has bought back ~21% of its shares since 2021, pays a growing dividend, and is the only one of these companies meeting that bar. Viasat, EchoStar, and ASTS are all net dilutive or net debt-issuing.

3. Is This Business Cyclical?

Iridium is not cyclical in the conventional macro sense. It has its own internal cycle, and that cycle is the 15-to-20-year constellation replacement super-cycle. Forget the business cycle; this is the only cycle that matters.

The story this chart tells is the company's defining feature. From 2010–2018, Iridium spent roughly $3.3 billion building the second-generation Iridium NEXT constellation — capex ran 80%–120% of revenue every year, and free cash flow was deeply negative for nine consecutive years. The last NEXT satellites launched in early 2019. Capex dropped 70% in twelve months. It has stayed there since.

In Q4 2023 management extended the depreciable life of the constellation from 12.5 to 17.5 years based on the in-orbit health data — meaning the next major rebuild does not begin in earnest until at least 2031. Until then, every dollar of operating cash flow above ~$100M of maintenance capex is genuinely free. That is what is funding the dividend, the buyback (paused October 2025), the Satelles acquisition, and the deleveraging.

The risk hidden in this cycle is capex creep. NTN Direct, the PNT ASIC, and any Aireon-related satellite enhancements all consume incremental capex. 2025 capex of $100M is already 43% above the FY2021 trough. Watch this number — if it drifts toward $150M+ before 2030 without a corresponding revenue uplift, the FCF guide ($1.5–1.8B cumulative through 2030) starts looking aspirational.

4. The Metrics That Actually Matter

Standard telecom metrics — ARPU, churn, net adds — describe Iridium poorly because the business is not one thing. The five metrics below are what actually drive value creation here.

A few things deserve sharper framing. The P/E ratio is misleading here, because depreciation of the satellite asset base swallows reported earnings. EBITDA margin and FCF margin tell a more honest story; FCF/net income has run above 2.0x for five straight years. Investors who anchor on a 36x P/E are misreading the asset.

Net leverage is the discipline metric, not the credit metric. Iridium's term loan has no maintenance covenants and a SOFR + 2.25% rate. The 3.4–3.7x leverage range is a choice, not a constraint — management debt-financed the 2024 Satelles acquisition and ~$400M of buybacks. The pause in buybacks in October 2025 (with $245M still authorized) is a tell that capital allocation has shifted toward optionality, possibly for a spectrum-related transaction.

Subscriber metrics matter only by segment. A net-add of 100,000 IoT subs adds maybe $9M of annualized revenue at $7.78 ARPU. A net-add of 1,000 broadband subs at $259 adds $3M. Treating "billable subscribers grew 3%" as a single number is the most common analyst error in this name.

5. What Is This Business Worth?

The right way to value Iridium is not as one business. The consolidated income statement blends a stable recurring cash machine, a long-cycle infrastructure asset, a 39.5% equity stake in a private aviation-data company, and a strategic spectrum holding whose value just got re-rated by the SpaceX–EchoStar deal. A single multiple on EBITDA misses three of those four.

The economics that actually underwrite the equity are the recurring service-revenue stream and the cash flow from a fully-built constellation that does not need rebuilding for ~6 more years. Apply a normalized 50% EBITDA margin to commercial + government service revenue (~$634M FY25), and the core network throws off ~$317M of segment EBITDA. At 11–13x EV/EBITDA — a reasonable multiple for a global infrastructure asset with 90%+ revenue retention — that produces a core-business EV of roughly $3.5–$4.1B. After subtracting ~$1.66B of net debt, that puts equity value for the core business alone at $1.8–$2.4B, or roughly $17–$23 per share.

The remaining ~$15+ in the current $39 share price is, effectively, what the market is paying for the spectrum optionality, the Aireon stake, and the NTN Direct / PNT growth call. That is the right way to think about underwriting risk here: the floor is the cash-machine multiple, the ceiling depends entirely on whether the optionality items convert. The bear case is straightforward — none of those convert, capex creeps higher into the next constellation cycle, the EMSS recompete prices down, and the stock compresses to its FCF-multiple floor. The bull case is that L-band spectrum gets re-rated, NTN Direct delivers a second commercial channel, and PNT becomes a $200M+ revenue line.

What would cheap or expensive look like? The stock at a 10% FCF yield (~$30) is unambiguously cheap if the constellation is in steady-state. At a 5% FCF yield (~$60+) the market is fully pricing the optionality. The current ~7.5% FCF yield is in the middle — it says the market believes some, but not all, of the optionality.

6. What I'd Tell a Young Analyst

Stop modeling this like a telecom and start modeling it like a midstream infrastructure asset on a 17-year asset cycle. The right framework is: how much cash does this throw off between now and the next major capex event, and what is the residual asset worth at that point? Net income and P/E are not the lens.

Watch four numbers more than the press releases. Capex / revenue (creep above 15% breaks the FCF guide); net debt / EBITDA glidepath (a re-acceleration into buybacks signals nothing big is brewing on spectrum, a continued pause signals something is); EMSS recompete pricing (the 2026/2027 announcement could re-rate or de-rate the multiple by 20%); and Aireon's revenue trajectory (because the redemption-right exercise is the only near-term hard catalyst on the SOTP).

The market probably has the central question wrong. Most coverage is fixated on whether Starlink, AST SpaceMobile, and Amazon Kuiper will eventually erode mobile-satellite-services pricing. They probably will, slowly. But the dominant variable is whether L-band spectrum gets re-rated by spectrum-scarcity dynamics — and the recent SpaceX–EchoStar transaction is a much bigger signal than the D2D consumer narrative. Management has explicitly told investors they will entertain "business alliances that leverage our unique spectrum real estate." That is not a casual remark; it is a flag.

Things that would change the thesis. A successful, scaled SpaceX or AST consumer D2D service that meaningfully cannibalizes mission-critical channels (slow but real). An EMSS recompete priced materially below $110M/year. A capex announcement for next-generation satellites earlier than 2031. Conversely, a strategic spectrum transaction at a multiple anywhere near the SpaceX–EchoStar comp. Any of these matters more than next quarter's IoT net-adds.

The single best test of whether you understand this name. If you can explain — in two sentences — why a company with a 36x reported P/E should be considered for a value portfolio, you understand it. If you can't, you're modeling the income statement instead of the asset.