Iridium Communications Inc. (NASDAQ: IRDM) — Half Cash, Half Rocket Lab, and the Collar Just Broke

Report date: 2026-08-01 · Coverage: Initiation Price at analysis: $47.34 (2026-07-31 close) · Shares outstanding: 105,960,383 (10-Q cover, 2026-07-15) · Market capitalisation: ~$5.02bn Enterprise value (pro forma for Aireon): ~$7.1bn · Sector: Communication Services — Satellite Communications / Mobile Satellite Services Status: Pending acquisition by Rocket Lab Corporation (merger agreement dated 2026-06-28)

Sections 1–15 of this article contain no investment recommendation and no price target. The single exception is the clearly labelled Claude's Take block immediately below.

⚡ Claude’s Take

This block is the author’s own independent opinion. It is general information, not investment advice, and not a recommendation to buy or sell any security. Everything below it (Sections 1–15) is deliberately position-free and carries no price target.

Verdict: AVOID as a new position. HOLD if owned, but understand you are now long Rocket Lab with the calls sold. Not a short. Valuation zone: the security is worth its deal payoff — currently ~$52.98 and moving $0.40 for every $1 of Rocket Lab. Standalone value if the deal breaks: ~$30–43, centred near $38.

Tag: “You sold the upside and kept the downside, for 11.9%.”

Iridium stopped being a satellite company on 28 June 2026, when it agreed to be acquired by Rocket Lab for $27.00 in cash plus a collared number of Rocket Lab shares — a headline $54.00, a 24.1% premium to the $43.52 unaffected close, an ~$8.0bn enterprise value. What matters now is the structure, and the structure has quietly changed against holders. Rocket Lab closed at $64.95 on 31 July, having first slipped below the $67.50 collar floor on 16 July. Below that floor the exchange ratio is pinned at 0.4000, so the stock leg is no longer worth a fixed $27.00 — it is worth $25.98 and falling. Every $1 Rocket Lab loses now costs an Iridium holder $0.40, about 0.85% of the share price. On the way back up, holders get nothing: the stock leg is fixed at $27.00 across the entire $67.50–$112.50 range, so Rocket Lab could rise 73% from here and the payoff would not move. You are long all of Rocket Lab’s downside below $67.50 and short the whole strip of its upside to $112.50 — and Rocket Lab is sitting 3.9% from the hinge, with 62% annualised idiosyncratic volatility. That is the trade you are buying at $47.34 for a gross spread of 11.9% over roughly eleven months.

I would not underwrite it, for three reasons. First, the compensation is thin for the risk taken: on a delta-hedged basis ~13% annualised is fair-to-wide for FCC transfer-of-control and foreign-clearance timing, but almost nobody holds it hedged, and unhedged you are simply long a volatile acquirer with the upside removed. Second, the thing underneath is not worth $47.34 on its own. Iridium earns ~9% ROIC — and only ~6% if you undo the 2023 decision to stretch satellite depreciation from 12.5 to 17.5 years, which is where roughly half of FY2025’s GAAP profit came from. It guides to flat-to-2% service revenue growth and a lower OEBITDA in 2026, its voice, broadband and government subscriber counts all fell in 2025, and its FY2025 10-K contains not one word about the ~$3bn constellation replacement that a 17.5-year life puts in the mid-2030s. It traded at 7.8x EV/EBITDA eight months ago; Rocket Lab is paying ~15.4x. Nothing fundamental improved in between — the guidance got worse. Third, if you want the Rocket Lab exposure, buy Rocket Lab: at $64.95 it has fallen roughly 40% from its June 2026 level, and bought directly it comes with the upside attached.

To be fair to the other side, and it is a real other side: this deal has genuine strategic logic and will probably close. Iridium’s single largest problem — an unfunded multi-billion-dollar constellation rebuild — is literally Rocket Lab’s product line, and building and launching the replacement in-house at cost rather than buying it from Thales Alenia and SpaceX at a vendor’s margin is a saving plausibly worth $1bn+ across a cycle. That is the only argument that makes $8.0bn coherent, and it accrues to Rocket Lab, not to a standalone Iridium holder. There is no financing condition, the board approved unanimously, and the sector precedent is strong: Amazon agreed to buy Globalstar ten weeks earlier. Conviction: medium-high. The single fact that would turn me constructive is Rocket Lab trading durably back above $67.50 with FCC consent and the foreign clearances in hand — at which point the payoff re-fixes at $54.00 and the residual becomes a genuine coupon. The single fact that would turn me bearish is an FCC hearing designation or a foreign-investment referral that pushes closing past the June 2027 outside date, because the standalone business the market would then re-price has declining subscribers in three of four lines and a rebuild bill nobody has costed.

📈 Stock Price Action — Five-Year Event Map

Iridium has round-tripped violently. It closed at $47.34 on 31 July 2026, inside a 52-week range of $15.84–$55.72 and 25.0% below its 52-week high, against a five-year high of $67.26 (April 2023) and a five-year low of $15.84 (20 November 2025) — a 76% peak-to-trough decline followed by a 199% recovery in eight months. The risk-adjusted record explains why that matters: three-year annualised return −0.19%, five-year +4.5%, both through a −75.3% maximum drawdown. Essentially all realised return sits in the last twelve months. Prices below are unadjusted closes, used for event attribution.

Cycle narrative. (1) The 2021–2023 advance was the harvest phase of the Iridium NEXT programme: the constellation was fully deployed by January 2019, capex fell to ~$100m a year against a $3.4bn asset, and free cash flow and buybacks did the rest. (2) July 2023 was the first crack — growth decelerating in a business the market had begun to price as a compounder. (3) The 22.1% fall on 24 July 2025 was the single largest one-day decline in five years and was entirely self-inflicted: the company cut its own full-year service-revenue growth guidance from 5–7% to 3–5% in the earnings release. (4) The September 2025 collapse was sector-wide and structural — SpaceX agreeing to acquire EchoStar spectrum made a global Starlink direct-to-device service credible, and Iridium’s own FY2025 10-K now names exactly that development in its competition section. (5) The 20 November 2025 low of $15.84 valued the whole enterprise at roughly 7.8x EBITDA. (6) The 21.3% jump on 12 February 2026 is the most revealing session in the series: the company simultaneously guided FY2026 service revenue to flat-to-2% growth and OEBITDA down to $480–490m, and the stock rose a fifth — a multiple event, not an earnings event. (7) The March–June advance carried no company disclosure at all on its four biggest up-days (25 Mar, 2 Apr, 16 Apr, 11 Jun); the 14.2% gain on 16 April came two sessions after Globalstar’s Amazon merger 8-K, and the read-across is the obvious interpretation. (8) On 29 June 2026 Rocket Lab agreed to acquire the company; Iridium rose 25.4% on 12.7m shares and Rocket Lab rose 15.9% — an unusual reaction for an acquirer, and evidence the market saw acquiring ~$300m of annual free cash flow as de-risking for a company burning ~$322m. (9) Everything since has been mechanical: Iridium is now a levered claim on Rocket Lab’s share price plus a fixed $27.00, and Rocket Lab’s slide through the collar floor on 16 July converted the stock leg from fixed-value to fixed-ratio.

1. Executive Summary

Iridium Communications operates the only commercial satellite constellation providing genuinely global, weather-resilient, low-latency narrowband voice and data coverage — 66 cross-linked LEO satellites in six polar planes, licensed across up to 8.725 MHz of contiguous L-band spectrum. It generated $871.7m of revenue and $495.3m of Operational EBITDA in FY2025, converting roughly $300m into free cash flow after ~$100m of capital expenditure. Gross margin has sat between 71.5% and 75.5% in every year since 2016. On the surface this is a high-quality infrastructure asset.

Underneath, three facts govern the analysis. First, the returns are mediocre. ROIC is ~9% on the reported numbers and ~6% once the 2023 extension of satellite depreciable life from 12.5 to 17.5 years is reversed — at or below any defensible cost of capital for a 3.3x-levered single-asset operator. Roughly half of FY2025’s $114.4m GAAP net income traces to that accounting change, which conveniently does not touch OEBITDA, the metric on which management is guided and paid. Second, growth has stopped. FY2026 guidance is flat-to-2% service revenue growth with OEBITDA down year over year; voice-and-data subscribers fell 3% in FY2025, broadband subscribers and ARPU both fell, and U.S. government subscribers fell 14%. Only IoT is growing, at an ARPU of $7.78. Third, there is an undisclosed liability. The constellation was deployed 2017–2019; a 17.5-year life implies economic end of life around 2035–2036, requiring a replacement build to start in the early 2030s. The original programme cost ~$3bn and gross book cost is $3.38bn today. The FY2025 10-K contains no mention of a next-generation constellation, end of life, or replacement funding — while the $1.8bn Term Loan matures in September 2030.

On 28 June 2026 the company agreed to be acquired by Rocket Lab Corporation for $27.00 in cash plus a collared number of Rocket Lab shares, a $54.00 headline at signing, a 24.1% premium, and an enterprise value of approximately $8.0bn — ~15.4x pro-forma OEBITDA, against the 7.8x at which the same business traded eight months earlier. The strategic logic is coherent and is the only thing that makes the price coherent: Iridium’s unfunded rebuild is Rocket Lab’s product line, and vertical integration converts a multi-billion-dollar capital liability into an internal transfer at cost.

For a holder today the security is a package, not an equity. At Rocket Lab’s 31 July close of $64.95 — below the $67.50 collar floor, which fixes the exchange ratio at 0.4000 — the payoff is $52.98, of which 51% is cash and 49% floats with Rocket Lab. The gross spread is 11.9% to an expected mid-2027 close. The structure is asymmetric against holders in the current range: full participation in Rocket Lab’s downside below $67.50, and none in its upside between $67.50 and $112.50.

This article takes no position and sets no price target. It concludes that Iridium is a genuinely differentiated but structurally challenged asset whose narrow moat has never translated into returns above its cost of capital; that management allocated capital defensibly but not well, buying back stock with borrowed money at prices far above the subsequent low while paying itself on a metric that ignores the constellation’s consumption; and that the current share price is a function of deal mechanics and Rocket Lab’s share price rather than of Iridium’s fundamentals.

2. Business Overview

2.1 What the company does

Iridium operates a constellation of 66 operational cross-linked low-Earth-orbit satellites (plus in-orbit and ground spares) in six polar planes at ~780 km. Two architectural features distinguish it from every other commercial network. The orbital planes are polar, giving literal 100% surface coverage including both poles — no other commercial MSS or broadband constellation does this. And the satellites are inter-linked by K-band crosslinks, forming a mesh that routes traffic satellite-to-satellite to a gateway, so service does not require a ground station within the coverage footprint. A ship in the Southern Ocean or an aircraft over the pole is served identically to one off Miami.

The current fleet — Iridium NEXT — was built by Thales Alenia Space and launched on eight SpaceX Falcon 9 missions between January 2017 and January 2019, replacing the original 1990s constellation. Gross book cost of the satellite system at 31 December 2025 is $3,379.2m.

2.2 Revenue architecture

FY2025 revenue of $871.7m splits three ways:

Within commercial service, the FY2025 detail is where the business reveals itself:

Service revenue is the recurring core and is genuinely sticky: customers are maritime fleets, airlines, defence users, oil and gas, mining, forestry, emergency services and outdoor recreation, and Iridium sells almost entirely wholesale, through a network of more than 500 service providers, value-added resellers and manufacturers rather than direct to end users. That distribution structure is itself a modest moat contributor — the partner installed base and the certification of devices onto the Iridium network create real friction — but it also means Iridium does not own the customer relationship or the pricing surface.

Engineering and support is almost entirely U.S. government work: maintenance of the government’s dedicated Iridium gateway, plus ground-system and operations-centre work for the Space Development Agency’s Proliferated Warfighter Space Architecture. It grew 26% in FY2025 and is the fastest-growing line, but it is contract-based professional services revenue, not network revenue, and it carries materially lower margins than airtime.

Hosted payload revenue deserves a note. Aireon’s ADS-B receivers and L3Harris’s customer payloads fly on the NEXT satellites. Aireon contracted to pay $200.0m in hosting fees (of which $126.5m received by 31 December 2025) plus ~$23.5m a year in power and data services; L3Harris added $66.6m, fully paid. This revenue is recognised over the satellites’ estimated useful life — which is why the 2023 life extension reduced hosted payload revenue by $2.3m in 2023 even as it reduced depreciation by far more.

2.3 The U.S. government relationship

The Enhanced Mobile Satellite Services (EMSS) contract is a seven-year, $738.5m fixed-price airtime agreement signed with the U.S. Space Force in September 2019, under which an unlimited number of Department of Defense and other federal users get access to specified Iridium airtime through a government-owned dedicated gateway. The fee is fixed at $110.5m a year for the remainder of the term and is explicitly not based on subscribers or usage — which is why government service revenue rose 2% in FY2025 while government subscribers fell 14%. The contract expired in September 2026 with a unilateral six-month government extension right; as of the Q2 2026 release Iridium “continues to expect a renewal of its EMSS Contract with the U.S. Space Force by March 2027.” The U.S. government is the largest single customer at 17% of service revenue and nearly all of engineering and support.

2.4 Aireon

On 13 May 2026 Iridium agreed to acquire Aireon LLC — the operator of the world’s only space-based ADS-B air traffic surveillance system, which flies as a hosted payload on Iridium’s own satellites, is EASA-certified, tracks an average of 190,000 flights per day with 100% global coverage, and serves air navigation service providers covering more than 50% of global airspace. The deal closed 2 July 2026 for approximately $366.7m — half cash at closing, half as a $183.4m non-interest-bearing seller loan maturing in one year — plus assumption of Aireon’s $154.7m credit facility, with $100m drawn on Iridium’s revolver to fund the cash half. NAV CANADA and NATS signed extended data agreements “through 2035 and beyond.” Guided contribution: “at least an additional consolidated $100 million of service revenue and $30 million of OEBITDA on an annualized basis.”

Verdict (Business Overview): A real, differentiated, cash-generative infrastructure business with genuinely recurring service revenue, an unusually stable gross margin, and defensible positions in safety-of-life and government applications. But the composition is deteriorating beneath the aggregate: the high-ARPU lines are shrinking, the growth is concentrated in a $7.78-ARPU IoT product, and the fastest-growing revenue line is lower-margin government engineering services rather than network airtime.

3. Industry Dynamics

3.1 Structure of mobile satellite services

Satellite communications divides into segments with sharply different economics. Fixed satellite services (SES, Eutelsat, Intelsat) use large GEO birds and big stationary terminals for video and trunking — a structurally declining, oversupplied pool. LEO broadband (Starlink, OneWeb) uses K-band and requires comparatively large, powered antennas; it is a genuine growth market but is capital-intensive and increasingly price-led by SpaceX. Mobile satellite services (MSS) — Iridium’s segment — uses lower-frequency L- and S-band spectrum for voice, narrowband data and IoT where the terminal must be small, low-power, weather-resilient and mobile. The named MSS peer set in Iridium’s own FY2025 10-K is Globalstar, ORBCOMM, and portions of Viasat (following its acquisition of Inmarsat).

MSS has historically been an attractive niche for a simple physical reason: L-band spectrum is scarce, internationally coordinated, and cannot be manufactured with capital. Iridium’s licence to up to 8.725 MHz of contiguous L-band plus 200 MHz of K-band for crosslinks and 400 MHz of Ka-band for feeder links is a genuine, government-granted barrier to entry. Nobody can simply outspend it.

3.2 The direct-to-device disruption

That barrier is being attacked from the side rather than head-on, and this is the industry’s defining development. Direct-to-device (D2D) services connect ordinary smartphones to satellites, and the leading implementations do it using terrestrial cellular spectrum, not MSS spectrum — sidestepping the L-band scarcity that protects Iridium entirely.

Iridium’s own FY2025 10-K names the threat with unusual directness: “New entrants are emerging as well such as Starlink’s D2D offerings, which currently utilizes terrestrial cellular frequencies that extend cellular coverage within defined, limited markets, but has plans for a global service in the future utilizing spectrum purchased from EchoStar in 2025.” Word-frequency check across the filing: “Starlink” appears twice; “Globalstar” appears ten times. The company’s disclosure still frames the competitive set around the incumbents it has beaten rather than the entrant that repriced its stock 24% in two sessions in September 2025.

The scale of capital now aimed at this end demand is the Marathon capital-cycle warning in its clearest form. SpaceX operates roughly 650 D2D-capable satellites, delivers service in the U.S. through T-Mobile, has filed for a 15,000-satellite Starlink Mobile constellation and is acquiring ~$19.6bn of EchoStar spectrum. Amazon’s Leo/Kuiper had ~241 of a planned 3,236 satellites in orbit as of April 2026 and in July 2026 filed with the FCC to build a further 5,105-satellite D2D network using Globalstar’s spectrum, deploying from 2028. AST SpaceMobile has non-exclusive partnerships with AT&T, Verizon and Orange. None of this capital is competing for Iridium’s spectrum; all of it is competing for Iridium’s end demand.

3.3 What D2D can and cannot take

An honest read separates the segments. D2D over terrestrial spectrum is well suited to consumer messaging and emergency SOS on handsets — Iridium has essentially no direct exposure there — and, over time, to low-cost consumer-adjacent IoT, where Iridium does compete and where price will fall. It is much less suited, at least this decade, to applications requiring certification, guaranteed availability, sovereign control, polar coverage, or GPS-independent assurance: EASA-certified aviation surveillance, GMDSS maritime distress, assured PNT for contested environments, and dedicated national-security communications. Those are regulated or physics-bounded niches and they are precisely where Iridium has been redirecting effort — the four “strategic initiatives” announced in February 2026 are satellite IoT, assured PNT, national security missions, and aviation safety.

3.4 The consolidation verdict

The most eloquent evidence about this industry’s structure is what has happened to its participants. Inmarsat was absorbed by Viasat in 2023. Globalstar agreed to be acquired by Amazon on 13 April 2026. Iridium agreed to be acquired by Rocket Lab on 28 June 2026. Within roughly three years, every independent Western MSS operator of scale has been acquired or agreed to be acquired, in each case by a buyer that builds and launches its own hardware or wants the spectrum for its own constellation.

The common mechanism is the same in each case: a narrowband MSS operator owns scarce spectrum, a sticky installed base, and a fully-depreciating asset — but cannot fund the next constellation on its own cash flow against competitors who manufacture and launch at cost. Independent MSS is not a viable long-run standalone structure. That is not a cyclical observation; it is a statement about the industry’s cost curve.

Verdict (Industry Dynamics): Structurally bad and deteriorating for an independent operator. Enormous, well-capitalised capacity is being built into adjacent technology aimed at the same end demand, while the incumbent’s protective spectrum barrier is being flanked rather than breached. The exception — and it is genuine — is the set of regulated, safety-of-life and sovereign applications where certification and physics still mandate a dedicated network. Those niches are defensible and profitable, but they are not large enough to carry a $3bn constellation rebuild on their own, which is precisely why the industry is consolidating into vertically integrated owners.

4. Competitive Position

4.1 Naming the moat

In Greenwald’s taxonomy the question is which of three genuine advantages Iridium possesses: supply-side cost advantage, demand-side customer captivity, or economies of scale combined with captivity.

Iridium has a real intangible-asset barrier and a narrow supply-side advantage; it does not have economies of scale plus captivity.

The intangible is the spectrum licence — up to 8.725 MHz of contiguous, globally coordinated L-band, obtainable only by regulatory grant. The supply-side advantage is architectural: the polar orbital planes and K-band crosslinks deliver true global coverage with small, low-power terminals and low latency, and are weather-resilient in a way K-band broadband is not. A competitor wanting to replicate this must obtain equivalent spectrum, build ~66 crosslinked satellites, and launch them — roughly a $3bn, decade-long undertaking with a regulatory prerequisite that money cannot buy. That is a genuine barrier and it is why Iridium has held its niche for twenty years.

Customer captivity is real but shallow and asymmetric. Switching costs exist at the device level — terminals are Iridium-specific, integrated into vessels, aircraft and industrial equipment, and certified through a 500-plus partner ecosystem — so a maritime fleet does not re-terminal casually. But Iridium sells wholesale, which means the partner owns the end customer, and the evidence in the numbers is that captivity is not strong enough to defend price or volume outside IoT.

4.2 The tests the moat fails

The subscriber test. In FY2025, three of four subscriber lines shrank: voice and data −13k (−3%), broadband −0.5k (−3%), U.S. government −20k (−14%). Only IoT grew, +111k (+6%). A network with strong captivity does not lose subscribers in its highest-ARPU lines while the overall market for connectivity expands.

The pricing test. Commercial voice-and-data revenue rose 3% in FY2025 — but the 10-K attributes the increase to “price increases in the third quarter of 2025” on a falling subscriber base. Iridium is using price to offset volume. That is what a business with a defensible niche and no growth does, and it works until it does not. Broadband is the leading indicator: ARPU fell from $282 to $259 (−8%) because customers are increasingly buying Iridium as a lower-priced companion to a primary VSAT or Starlink connection rather than as the primary link. The moat is intact; the role is being demoted.

The returns test — the decisive one. A moat that cannot be tied to a financial outcome that would deteriorate without it is not a moat. Iridium’s ROIC is ~9% on reported figures and ~6% on the pre-2023 depreciation basis. Against a WACC that cannot credibly be below 8–9% for a 3.3x-levered, single-asset, technologically-threatened satellite operator, the company earns at or below its cost of capital. Over the full NEXT cycle — a $3bn+ investment made 2010–2019 and harvested 2019–2026 — Iridium has not generated meaningful cumulative economic profit. Gross margin of 72% tells you the service has pricing power relative to its marginal cost; ROIC of 9% tells you the asset consumed all of it.

4.3 Against the named competitors

Globalstar operates a smaller, non-crosslinked LEO constellation with no polar coverage and depends overwhelmingly on a single customer relationship with Apple; its non-Apple revenue has declined three years running. Iridium is the better network, and it is the better business. ORBCOMM is IoT-only and subscale. Viasat/Inmarsat competes in maritime and aviation broadband where Iridium is now often the companion rather than the primary. Starlink is not a like-for-like competitor today — different spectrum, different terminals, different use case — but it is the entity whose adjacent expansion sets the ceiling on Iridium’s pricing and its terminal value.

The comparison that matters most is not with any of them. It is with the capital being deployed at the problem. FactorsToday’s related-stocks endpoint returns MeiraGTx, Fubo, Schrödinger and JELD-WEN at >0.997 similarity — an absurd peer list on its face, but the correct statistical reading: with model R² of only 11.9–13.5% and idiosyncratic volatility of 62.4% annualised, the factor model finds no genuine peers for Iridium at all. Roughly 88% of the variance in this stock is company-specific. “No real peers” is itself the finding.

Verdict (Competitive Position): A narrow, real, but non-compounding moat. The spectrum-plus-architecture barrier is genuine and durable and will not be breached this decade. But it protects a niche, not a growth franchise, and it has never produced returns above the cost of capital. The advantage is sufficient to defend the safety-of-life, sovereign and polar applications indefinitely; it is not sufficient to defend price in the commercial lines against adjacent capacity, and the broadband ARPU decline shows the erosion has already begun.

5. Growth History and Forward Opportunities

5.1 The record

Seven-year revenue CAGR is 7.6%. The trajectory is unambiguously decelerating, and the deceleration is in the recurring line: service revenue growth 5.2% → 3.1% → guided flat-to-2%. FY2026 is guided to a year-over-year OEBITDA decline ($495.3m → $480–490m), of which $17m is the switch to paying incentive compensation entirely in cash; even excluding that, the guided range of $497–507m is barely above FY2025.

The 2022 spike (+17%) and the 2025 engineering line (+26%) both deserve context. The former reflects post-COVID recovery plus equipment; the latter is Space Development Agency ground-systems work — real and welcome, but lower-margin services revenue that does not carry network economics.

5.2 Quality of the growth

The growth has been low-quality and increasingly so, on three tests.

Mix. The growth is concentrated in IoT, where ARPU is $7.78 per month against $47 for voice and $259 for broadband. Adding 111,000 IoT subscribers generates roughly $10m of annualised revenue; losing 13,000 voice subscribers costs roughly $7m. Iridium is running hard to stand still in dollar terms.

Organic vs. acquired. A meaningful share of recent growth is bought, not built. Satelles (2024, ~$110.7m) created the PNT line. Aireon (closed July 2026, ~$366.7m plus $154.7m assumed debt) adds “at least $100m of service revenue” — which, on the FY2025 base of $634.0m, is a 16% step-up in service revenue purchased for roughly $520m of consideration and assumed debt. That is not organic growth; it is the consolidation of a business Iridium already partly owned and already hosted.

Price vs. volume. As noted, FY2025 voice growth came from price increases on a shrinking base.

5.3 Forward opportunities

Four initiatives were announced in February 2026 and reaffirmed in July 2026.

Satellite IoT. The Iridium 9604 tri-mode module (launched 23 June 2026) combines Iridium Short Burst Data, LTE-M cellular and GNSS in one power-efficient package — a sensible product that lowers device cost and hedges the terrestrial/satellite boundary. Iridium NTN Direct, a 3GPP standards-based D2D and low-cost-IoT service running over Iridium’s own licensed L-band, is the strategically important one: it is the company’s attempt to compete in D2D on standards rather than proprietary protocols. It was promised “later this year” as of July 2026 and has not launched commercially. This is the single most important product in the standalone story and it is unproven.

Assured PNT. A commercially available PNT ASIC extends the addressable market for GPS-independent positioning, navigation and timing — genuinely differentiated, genuinely in demand given growing GPS jamming and spoofing, and a natural fit with Aireon’s jamming/spoofing detection. Revenue is not separately disclosed, which is itself a tell about scale.

National security missions. Building on EMSS and the SDA ground-systems work. Real, but government-budget dependent and services-margin.

Aviation safety. The Aireon acquisition plus the prospect of space-based VHF communications. This is the most credible growth leg: regulated, certified, contracted to 2035 and beyond, with a customer base of national air navigation service providers that cannot readily switch.

Verdict (Growth): Low-quality and decelerating to zero. The company’s own FY2026 guidance — flat-to-2% service revenue and falling OEBITDA — is the most honest statement in this section. The forward opportunities are individually sensible and the aviation-safety leg is genuinely attractive, but the two with the largest addressable markets (NTN Direct and PNT) are unproven and undisclosed respectively, and the one with contracted economics (Aireon) was purchased rather than built. Growth from here requires the market to believe in products that do not yet have revenue lines.

6. Financial Quality

6.1 Margins and their stability

Iridium’s gross margin has been extraordinarily stable — 71.5% to 75.5% in every year from 2016 through 2025 — and OEBITDA margin has run 48.6% to 58.0%. This is what a fixed-cost network with high incremental margins looks like, and it is the genuinely attractive feature of the business.

Operating margin tells a very different story: 1.8% (2019), 6.1% (2020), 7.5% (2021), 10.6% (2022), 10.3% (2023), 24.1% (2024), 27.1% (2025). The line that moved was not revenue or cost of service. It was depreciation.

6.2 The depreciation change — the central quality-of-earnings issue

In the fourth quarter of 2023 Iridium extended the estimated useful life of its satellites from 12.5 years to 17.5 years, based on “the health of the constellation and related engineering data.” KPMG flags the useful-life estimate as a critical audit matter in the FY2025 10-K.

The effect is arithmetic and large:

Consolidated D&A confirms the shift: $303.5m (FY2022) and $320.0m (FY2023) → $203.1m (FY2024) and $210.2m (FY2025).

Undoing it: FY2025 operating income of $236.0m becomes ~$158.8m; pre-tax income of $144.8m becomes ~$67.6m; and net income of $114.4m becomes roughly $55–60m. In other words, approximately half of FY2025’s reported GAAP profit is attributable to the 2023 decision to depreciate the constellation over five more years.

Two observations make this material rather than merely technical. First, the change does not touch OEBITDA, which is struck above D&A — and OEBITDA is the metric on which management guides, on which the credit agreement’s leverage covenant is measured, and on which management is paid. Second, the extension is not obviously wrong: the satellites are healthy and Iridium’s first-generation constellation famously outlived its design life. But an accounting judgement that adds $77m a year to reported profit, is identified by the auditor as the single critical audit matter, and is invisible in the metric everyone quotes, deserves to be stated plainly rather than absorbed.

6.3 Returns on capital

Rebuilt from the filed statements: NOPAT = EBIT $236.0m × (1 − 19.07% effective tax rate) = $190.9m. Invested capital = total debt $1,760.5m + total equity $462.6m − cash $96.5m = $2,126.6m. ROIC = 9.0%. On the pre-2023 depreciation basis, NOPAT falls to ~$128.5m and ROIC to ~6.0%.

Do not quote the ROE. The 22.0% figure is an artefact of a book equity base of $462.6m that has been hollowed by $1.2bn of share repurchases against $880.6m of remaining additional paid-in capital and a $418.6m accumulated deficit. It measures the buyback, not the business.

The honest summary is that a business with 72% gross margins and 51% EBITDA margins earns 9% on invested capital because the asset that produces those margins cost $3.4bn. That is the whole moat question answered in one ratio.

6.4 Cash generation

FY2025 operating cash flow was $400.1m. Capital expenditure was $100.3m (including $4.6m of capitalised interest), per the company’s own proxy. Free cash flow ~$299.8m — a genuine and substantial number, and the reason the equity was never a going-concern question even at $15.84.

A data caution: ROIC.ai reports cf_free_cash_flow of $400.1m for FY2025 and returns cf_cap_expenditures: null; capex is buried inside cf_other_investing_act_detailed (−$100.3m). Any screen using the vendor’s FCF field overstates Iridium’s free cash flow by a third. The filing governs.

6.5 The sustaining-capex problem

Reported capex of ~$100m a year against a $3.38bn satellite asset with a 17.5-year life is roughly half the book depreciation charge of $193m, and book depreciation is itself the floor on the true economic charge, since a replacement constellation will be built at future prices rather than 2015 prices.

This is the “reported capex is meaningless; the charge is fleet replacement” problem in satellite form. The NEXT constellation was deployed January 2017 to January 2019. A 17.5-year life from a ~2018 midpoint puts economic end of life at ~2035–2036, and a replacement programme of this scale requires design, procurement and launch manifesting to begin roughly five years earlier — i.e. the early 2030s. The original programme cost approximately $3bn.

Against that, the FY2025 10-K contains zero occurrences of “next-generation” [constellation], “end of life”, “deorbit” or “future constellation.” The liquidity discussion runs only as far as the Term Loan, states that “capital expenditures in 2026 will be consistent with 2025,” and does not contemplate a rebuild. Meanwhile the $1.8bn Term Loan matures in September 2030 and the revolver in September 2028.

So the standalone financial plan, as disclosed, requires Iridium to refinance $1.8bn in 2030 and then fund a ~$3bn constellation replacement in the 2030s out of roughly $300m a year of free cash flow, while paying a dividend and buying back stock. That is the reason this company was sold, and it is not stated anywhere in the filings.

6.6 Balance sheet

At 30 June 2026: cash $184.2m; short-term secured debt $12.5m; long-term secured debt $1,749.3m; net debt $1,577.7m; total equity $472.5m; net leverage 3.3x trailing OEBITDA. Pro forma for Aireon (closed 2 July), net debt rises to approximately $2,099m — the $183.4m seller note, the $154.7m assumed Aireon facility, the $100m revolver draw, and $183.4m of cash out.

Verdict (Financial Quality): Economics do not improve with scale — they have already peaked. The margin structure is excellent and stable, cash conversion is real, and the balance sheet is manageable at 3.3x with no near-term maturity. But returns on capital are at or below cost of capital, half of reported GAAP profit comes from a depreciation-life extension, reported capex understates the true sustaining charge by roughly half, and the largest capital commitment in the company’s future — the constellation rebuild — is entirely absent from the financial statements and the MD&A.

7. Capital Allocation

7.1 The buyback

Iridium retired 34.8 million shares for approximately $1.2bn between February 2021 and mid-2025, and 6.8m more for ~$185.0m in FY2025. Weighted-average share count fell from ~133.5m (FY2020) to 105.96m (July 2026) — a 21% reduction.

Judged on its own terms this was disciplined and consistent. Judged on outcome it was not good. The repurchases were executed across 2021–2025 at prices spanning roughly $25 to $60, into a stock that traded at $15.84 in November 2025. And they were partly debt-funded: FY2024 shows $419.8m of new Term Loan borrowing against $407.7m of buybacks in the same year, taking net leverage to 3.6x by mid-2025.

Borrowing at a SOFR-based cost to retire equity in a business earning ~9% ROIC — while a multi-billion-dollar constellation replacement sits undisclosed on the horizon — is at best a marginal trade and at worst a transfer of future flexibility into past EPS. The counterfactual is uncomfortable: had the $1.2bn been retained, Iridium would have entered the 2030s with the replacement substantially pre-funded and would not have needed a buyer.

7.2 Dividends

Initiated in 2023 and raised every year since; $0.15 per quarter in 2026, $62.9m paid in FY2025, dividend payout ratio 55% of GAAP earnings. Reasonable in isolation, and comfortably covered by the ~$300m of free cash flow. It is the same capital-return question as the buyback in a smaller denomination.

7.3 M&A

Satelles (2024, ~$110.7m) created the PNT franchise and generated a $19.8m one-time bargain-purchase gain in Q2 2024 — a well-priced, strategically sensible tuck-in whose revenue is still not separately disclosed two years later.

Aireon (agreed 2026-05-13, closed 2026-07-02, ~$366.7m plus $154.7m assumed debt) is the best capital-allocation decision in the record. Iridium bought out the minority of a business it co-founded, that flies on its own satellites, that holds EASA certification, that is contracted with NAV CANADA and NATS through 2035 and beyond, and that occupies a regulated niche no D2D constellation can casually enter. At ~$521m of total consideration and assumed debt for ~$30m of OEBITDA, the headline multiple of ~17x is full — but the revenue is contracted, regulated, and strategically inseparable from the constellation.

The timing deserves a flag. The Aireon agreement was signed 13 May 2026; the Rocket Lab merger agreement was signed 28 June 2026; Aireon closed 2 July 2026 — four days after Iridium agreed to sell itself. The company completed a ~$521m acquisition while under a signed merger agreement and a no-shop. That is not improper — it was signed first and would have required Rocket Lab’s consent under the interim-operating covenant — but it means the buyer explicitly endorsed the transaction, and it tells you the aviation-safety leg is central to Rocket Lab’s thesis.

7.4 Incentive design — read the proxy

The DEF 14A filed 2 April 2026 contains zero occurrences of “TSR”, “Total Shareholder Return”, “ROIC”, “return on invested capital”, or “Free Cash Flow”. “OEBITDA” appears 38 times.

Management is paid on a measure struck above the depreciation that records the constellation being consumed and entirely before the capital expenditure required to replace it. That is precisely the metric design under which a decade of sub-WACC returns registers as success. It is a common failure mode: incentive plans built on EBITDA-style measures and revenue, with no return-on-capital, free-cash-flow or total-shareholder-return metric anywhere in the plan.

And then there is the 2026 change, which is the single most revealing disclosure in this engagement. For 2026 the Compensation Committee eliminated performance-based equity entirely, replacing a 50% PSU / 50% RSU structure with 100% service-based RSUs vesting ratably over five years. The stated reason, verbatim:

“Performance goals that are unreasonably difficult to achieve or insufficiently challenging do not appropriately incentivize management. The Compensation Committee believes that uncertainty in the industry makes it increasingly more difficult to establish multi-year financial performance goals with sufficient certainty to strike the right balance…”

In a proxy filed twelve weeks before agreeing to sell the company, the board stated in writing that it could no longer set two-year financial targets for its own business. Read as a sequence, the last twelve months are coherent: cut guidance (July 2025) → guide to flat-to-2% and lower OEBITDA (February 2026) → remove performance conditions from all equity because the future cannot be forecast (April 2026) → sell the company (June 2026) → stop holding earnings calls and stop giving guidance (July 2026).

7.5 Insider behaviour — a complete census

All 413 Form 4/4A filings from 2021-08-01 to 2026-07-31 were downloaded and parsed from the raw XML, yielding 645 transactions with zero fetch failures.

Open-market purchases (code P) — three in five years:

Open-market sales (code S), excluding one unpriced 500,000-share Baralonco transfer in 2021: 72 transactions, 1,177,838 shares for $61.0m at a weighted-average $51.78, distributed as:

The pattern is unusually clean and reads well for management. 91% of all insider selling occurred in 2022–2023, at weighted-average prices of $47.65 and $60.76 — into the five-year high of $67.26. Selling then stopped almost completely as the stock fell: $1.4m, $1.6m and $0.2m in 2024, 2025 and 2026 respectively. And in the last week of October 2025, within 48 hours of each other, the Chairman and the CEO bought stock in the open market at $17.49 and $17.33 — three weeks before the all-time-low close of $15.84.

Both readings should be stated. The favourable one: this is a genuine conviction cluster at the bottom by the two people who knew the business best, and it was followed by a sale of the company at $54.00 — vindication. The uncomfortable one: the Chairman and CEO bought at ~$17.40 and agreed eight months later to sell the company at $54.00, a 3.1x on their personal purchases. The purchases were made in an open window following the 23 October 2025 Q3 release, and there is no evidence in the record of a sale process at that date; the Rocket Lab agreement is dated eight months later. It is nonetheless a fact that belongs in the file. It is also worth noting that the $54.00 deal price is only 4% above the $51.78 weighted-average price at which insiders were selling in 2022–2023 — the company is being sold for approximately what management thought it was worth three years ago.

Verdict (Capital Allocation): Mixed, tilting negative — competent stewardship of a business whose central capital problem was never addressed. The dividend is covered, Aireon is a genuinely good acquisition, Satelles was well priced, and insider behaviour is honest and well timed. But $1.2bn was spent retiring stock — partly with borrowed money, at an average price far above the subsequent low — by a company that knew it faced an undisclosed multi-billion-dollar constellation rebuild, while paying management on a metric that excludes both the depreciation and the capex that rebuild represents. The buyback bought a 21% lower share count; retaining the cash would have bought independence.

8. Changes and Headwinds — Last Two Years

Q4 2023 — the depreciation-life extension. Satellite useful life extended from 12.5 to 17.5 years, adding ~$77m a year to operating income from FY2024 onward. Strengthens reported earnings; changes no cash flow. Weakens the thesis in the sense that it made a decelerating business look like an improving one for two years.

2024 — Satelles and the debt-funded buyback. ~$110.7m for the PNT business; $419.8m of incremental Term Loan borrowing against $407.7m of buybacks; net leverage to 3.6x.

24 July 2025 — the guidance cut. FY2025 service revenue growth guidance reduced from 5–7% to 3–5% in the Q2 release. The stock fell 22.1% in one session on 10.3m shares — the largest single-day decline in five years. This is the moment the market stopped treating Iridium as a compounder.

8–9 September 2025 — the SpaceX/EchoStar spectrum transaction. Sector-wide repricing of the direct-to-device threat; IRDM fell 14.9% and then 10.3% on consecutive sessions. The FY2025 10-K subsequently names this development explicitly in its competition section. Weakens the thesis materially — it converts D2D from a handset-messaging curiosity into a credibly-funded global service.

20 November 2025 — the low. $15.84, roughly 7.8x EV/EBITDA, a ~76% drawdown from the 2023 high.

28–30 October 2025 — insider buying. Chairman and CEO purchase at $17.49 and $17.33.

4 December 2025 — board expansion. Louis Alterman appointed, taking the board to twelve (nine independent).

12 February 2026 — FY2025 results and FY2026 guidance. Service revenue guided flat to 2%; OEBITDA guided $480–490m versus $495.3m delivered; four strategic initiatives announced. Stock +21.3%. Weakens the fundamental thesis; strengthened the stock — a distinction worth holding on to.

2 April 2026 — the proxy. Performance-based equity eliminated for 2026 in favour of five-year service-based RSUs, on the stated ground that multi-year goals could not be set with sufficient certainty.

13 April 2026 — Amazon agrees to acquire Globalstar. Not an Iridium event, but the decisive sector event. IRDM rose 14.2% on 16 April, two sessions after the Globalstar 8-K.

13 May / 2 July 2026 — Aireon. Agreement and then closing of the ~$366.7m acquisition plus $154.7m of assumed debt. Strengthens the thesis — the most defensible growth leg the company has.

28–29 June 2026 — the Rocket Lab merger agreement. $27.00 cash plus collared stock, $54.00 headline, 24.1% premium, ~$8.0bn enterprise value, expected close mid-2027. IRDM +25.4%; RKLB +15.9%.

16 July 2026 — Rocket Lab breaks the collar floor. RKLB closes at $67.35, below $67.50, pinning the exchange ratio at 0.4000 and converting the stock leg from fixed-value to fixed-ratio. This is the most consequential development for a current holder and it was not announced by anybody.

22 July 2026 — Q2 2026 results and the information blackout. Revenue $225.2m (+4%); operating income $34.0m versus $50.3m; OEBITDA $119.1m versus $121.3m — down year over year; $14.3m of transaction expenses. The company states it “does not intend to hold conference calls to discuss its quarterly financial results or to update or provide financial guidance.” Weakens the thesis for anyone underwriting a break scenario: there is now no forward information.

Verdict (Changes): Net materially negative for the standalone business and decisive for the security. Over twenty-four months, growth decelerated to zero, the competitive threat became credibly funded, the board withdrew its own ability to forecast, and the company sold itself. The two positives — Aireon and the depreciation extension — are respectively a good acquisition and an accounting change. The single most important change for a holder today, the 16 July collar break, is a market event rather than a corporate one.

9. Risk Analysis

The two that matter. Risks 1 and 2 together are the position: a holder is long RKLB’s downside and short its upside, on 49% of the consideration, for eleven months. Risk 4 is the reason the company is being sold and the reason a break scenario is worse than a simple return to the unaffected price — a standalone Iridium in mid-2027, with no guidance, no earnings calls, flat revenue and an unfunded rebuild, is a harder security to value than it was in June.

10. Valuation Discussion (Embedded Expectations)

10.1 Own-history percentiles — and why most of them are void

AZI’s valuation_index at 2026-07-31 reports: price $47.34, TTM EPS $0.9916, BVPS $4.3953, TTM sales per share $8.2237; P/E 47.7 (71.98th percentile), P/B 10.77 (98.95th percentile), P/S 5.76 (65.06th percentile), composite 78.66, n_components 3.

Two of the three components must be discarded.

P/B is void. Book equity of $472.5m has been hollowed by $1.2bn of buybacks against $880.6m of remaining paid-in capital and a $418.6m accumulated deficit. A 98.95th-percentile P/B measures the repurchase programme, not the valuation.

P/E is distorted. Roughly half of trailing GAAP earnings comes from the 2023 depreciation-life extension. On the pre-2023 basis, trailing EPS would be roughly $0.52 and the P/E ~91x — the percentile is not measuring a comparable earnings stream across the ten-year window.

P/S at the 65.06th percentile is the only clean component, and it is the one to quote: Iridium is expensive against its own sales history but not extraordinarily so. The composite of 78.66 averages one valid signal with two corrupted ones and should not be used.

10.2 The re-rating, in one table

In seven months the same business re-rated from 7.8x to 15.4x. Over that window: FY2026 service revenue was guided to flat-to-2%, FY2026 OEBITDA was guided below FY2025, Q2 2026 OEBITDA came in down year over year, and the board removed performance conditions from executive equity because it could not forecast the business. Every piece of fundamental news in the re-rating window was negative. The re-rating is a control-premium and sector-consolidation event, not an earnings event.

10.3 Building the deal enterprise value

Reconciling to Rocket Lab’s stated “approximately $8.0 billion”:

On pro-forma OEBITDA of ~$515m (FY2026 guidance $480–490m plus Aireon’s ~$30m), that is ~15.4x. Current market EV at $47.34 is ~$7,115m, or ~13.8x.

10.4 What must be true to justify $8.0bn

Not an Iridium forecast. On Iridium’s own guidance — flat-to-2% service revenue growth, declining OEBITDA, ~9% ROIC — no discounted cash flow supports 15.4x. A generous standalone DCF using ~$300m of free cash flow growing at 2%, discounted at 9%, capitalises to roughly $4.3bn of equity value before any charge for the 2030s rebuild — and the rebuild is precisely what a standalone DCF cannot ignore. Charge even $2bn of present-value replacement capex and the standalone equity is worth well under $3bn, or under $28 a share.

The price is justified only by a Rocket Lab synergy, and there is exactly one that is large enough. Iridium must replace ~66 satellites in the 2030s. Rocket Lab manufactures satellites and builds launch vehicles. Buying that constellation from Thales Alenia and launching it on Falcon 9 — the 2010s arrangement — means paying a vendor’s margin twice. Building it in-house means paying cost. On a ~$3bn programme, capturing the manufacturing and launch margin plausibly saves $1bn or more across a replacement cycle, and converts Iridium’s largest liability into Rocket Lab’s revenue. Add ~$300m a year of free cash flow that partially funds a company which burned ~$322m in FY2025, and the strategic arithmetic is coherent.

That value accrues to Rocket Lab, not to a standalone Iridium holder — which is why the acquirer’s stock rose 15.9% on announcement, an unusual and telling reaction.

10.5 The payoff — what a holder actually owns

At Rocket Lab’s 2026-07-31 close of $64.95, below the $67.50 floor, the exchange ratio is fixed at 0.4000:

The collar payoff, stated plainly:

Between $67.50 and $112.50 — a 67% range — the payoff is a flat $54.00. Below $67.50 it falls at $0.40 per dollar. Rocket Lab is currently 3.9% below the floor, sitting on the hinge.

10.6 Embedded completion odds

Solving p × $52.98 + (1 − p) × break = $47.34:

The spread cannot be read as a clean deal-odds market, and the spread of implied probabilities across plausible break prices — 40% to 75% — is itself the point. A signed strategic deal with a unanimous board, no financing condition and a $223.62m break fee does not carry a 40% completion probability. What the market is actually pricing is a blend of modest deal risk and substantial Rocket Lab price risk, plus the cost of carry for anyone hedging (shorting 0.40 RKLB for eleven months is not free). On a delta-hedged basis, ~11.9% gross over eleven months is roughly 13% annualised — wide for deal risk alone, and appropriate compensation for FCC transfer-of-control and foreign-clearance timing rather than for a coin flip.

10.7 Standalone value if the deal breaks

Pro-forma OEBITDA ~$515m; pro-forma net debt ~$2,099m; 105.96m shares:

A break would not return the stock all the way to 7.8x — Aireon has closed and the sector has re-rated — nor all the way to the $43.52 unaffected price, which itself embedded M&A anticipation. A defensible break zone is ~$30–43, centred near $38, i.e. a fall of roughly 10% to 37% from $47.34.

Verdict (Valuation): The security trades at a 11.9% discount to a deal payoff that is itself a moving target, and at ~13.8x pro-forma OEBITDA on a standalone basis against 7.8x eight months ago for a business now guiding to zero growth. The deal price of ~15.4x is not supportable by Iridium’s own cash flows and is explicable only by a vertical-integration synergy that belongs to the buyer. The asymmetry embedded in the collar — full downside participation below $67.50, none between $67.50 and $112.50 — is not compensated by an 11.9% spread.

11. Variant Perception

11.1 What consensus believes

The consensus view, visible in the tape and in the sell-side framing of the deal, is roughly: Iridium was a mispriced infrastructure asset that fell too far on an overstated direct-to-device fear; a strategic buyer validated its true worth at $54.00; the spread is a modestly attractive merger-arbitrage return with a strategic acquirer, a unanimous board and no financing condition. The corollary is that the 11.9% spread is mostly deal-risk compensation and will grind in as clearances arrive.

11.2 The strongest bull case

A serious bull case exists and it is not the consensus one.

The deal closes in mid-2027; Rocket Lab recovers above $67.50 by the measurement date — entirely plausible given it traded at $105.05 as recently as 10 June 2026 and at $98.01 on announcement day — and the payoff re-fixes at $54.00, a 14.1% return from $47.34. In that scenario the collar works exactly as designed: it protected holders through a 34% acquirer drawdown and restored full value on recovery. Below the floor, holders are effectively buying Rocket Lab exposure at a discount, with a $27.00 cash cushion underneath, at a moment when Rocket Lab has already de-rated some 40% from its June 2026 level. If Neutron flies successfully in Q4 2026 — the single largest swing factor in Rocket Lab’s equity — the stock leg re-rates and holders capture it up to $67.50 and hold a fixed $54.00 above.

The standalone leg of the bull case: Iridium’s decline is overstated. Aviation safety via Aireon is regulated, certified and contracted to 2035; assured PNT addresses a large and growing GPS-jamming problem; EMSS renews with the Space Force; NTN Direct turns the D2D threat into an Iridium product on Iridium’s own licensed spectrum. ~$300m of free cash flow does not evaporate.

11.3 The strongest bear case

Rocket Lab’s equity is the bear case, and it is a real one. Independent analysis of Rocket Lab in June 2026, when the shares traded near ~$108, flagged that ~$61.4bn of enterprise value — ~90x trailing revenue — already capitalised five years of flawless execution, with 75–80% of that value attributable to Neutron, a rocket that has never flown and that ruptured a fuel tank on a test stand in January 2026. Since then Rocket Lab has fallen 40%, Neutron has slipped to a Q4 2026 target, and a $3bn at-the-market equity programme has been announced. At $64.95 the stock is cheaper but the binary is unchanged: a Neutron failure or another multi-quarter slip takes Rocket Lab materially lower, and an Iridium holder eats $0.40 of every dollar with no offsetting upside participation until $112.50.

The standalone bear case is the constellation. Iridium’s entire free cash flow between now and the mid-2030s is, economically, pre-committed to rebuilding the asset that produces it — and no filing acknowledges this. A holder underwriting a break scenario is underwriting a business with declining subscribers in three of four lines, zero guided growth, a $1.8bn maturity in 2030, no earnings calls, no guidance, and a ~$3bn capital requirement that has never been costed in public.

11.4 The 3–5 assumptions that actually matter

1. Rocket Lab’s share price at the measurement date. This is the largest single driver of the payoff and it is entirely exogenous to Iridium. Falsified by: Neutron’s Q4 2026 flight outcome.
2. FCC and foreign-clearance timing. The deal has no financing condition and a friendly board; the binding constraint is regulatory calendar, not deal logic. Falsified by: an FCC hearing designation order or a foreign-investment referral.
3. Whether the collar is protection or a trap. Consensus reads the collar as downside protection. Below the floor it is the opposite — it is a fixed-ratio equity stake with the upside sold. Falsified by: RKLB recovering durably above $67.50, which restores the $54.00 fixed value.
4. The true sustaining capital charge. Whether ~$100m/yr or ~$193m+/yr is the right number determines whether Iridium’s free cash flow is ~$300m or ~$200m, and therefore whether standalone value is $40 or $25. Falsified by: any disclosure of a next-generation constellation programme and its cost.
5. Whether NTN Direct is a defence or a me-too. Falsified by: a commercial launch with a named mobile-network-operator partner and disclosed revenue.

11.5 Where consensus is most likely offsides

On the structure, not the fundamentals. The market has correctly identified that this is a deal spread. What appears under-appreciated is that the security stopped being a fixed-value deal spread on 16 July 2026 and became a levered, upside-capped claim on Rocket Lab — and that nobody announced this, because it happened through a price threshold rather than a filing.

The tape corroborates. Applying the GSAT volatility test: when Globalstar signed with Amazon, three-month realised volatility collapsed to 11.8% annualised against a five-year 76.8% — the fingerprint of a fixed-consideration deal. Iridium’s idiosyncratic volatility remains 62.4% annualised and three-month realised volatility is 76.5%. This security has not stopped breathing like a satellite stock, because half of it is still an equity — and a highly volatile one.

The second place consensus may be offsides is the reference point. The debate is framed as “$54.00 versus $43.52 unaffected — a 24% premium.” The more informative comparison is $54.00 versus $15.84 eight months earlier, and 15.4x versus 7.8x EV/EBITDA, on a business whose guidance deteriorated throughout the interval. Consensus is anchored on the premium; the relevant question is whether the base was ever $43.52.

Verdict (Variant Perception): Consensus is right that the deal will probably close and right that the strategic logic is real. It is most likely wrong in treating the security as a conventional fixed-value arbitrage. The variant view is narrower and more mechanical: below $67.50, IRDM is a 0.40-delta long position in Rocket Lab with a $27.00 cash floor and the upside sold to $112.50 — and it is being paid 11.9% to take that.

12. Fact vs. Interpretation Table

13. Open Questions

1. When will Rocket Lab file the Form S-4, and what will it disclose? It will contain Iridium’s board-approved projections and the fairness opinion — the two documents that would settle whether $54.00 is generous or cheap against management’s own forecast. Neither exists publicly today. This is the single highest-value future document on this name. (In the Globalstar precedent, the S-4 revealed the seller’s own banker valuing the company at roughly half the takeout price.)
2. What is the actual cost and timing of the next-generation constellation? Nothing is disclosed. Does Rocket Lab intend to build it on Neutron, and on what schedule? This is the entire synergy case.
3. Was there an auction? The 8-K describes a no-shop with a fiduciary out but the background-of-the-merger chronology appears only in the proxy/S-4. Did other parties bid? Given Amazon had just bought Globalstar, the absence of a competing bidder for the better asset would be surprising.
4. Did Rocket Lab consent to the Aireon acquisition, and was Aireon part of the deal thesis? Aireon closed four days after the merger agreement was signed. The interim-operating covenant would have required consent; the disclosure schedules are not public.
5. What are the specified foreign investment and satellite/telecom clearances? The 8-K refers to them generically. Rocket Lab’s New Zealand operations and Iridium’s national-security role make the review perimeter non-trivial.
6. Will EMSS renew, and on what terms? $110.5m a year of fixed-price revenue with an expired original term and a renewal now expected “by March 2027” — after the deal’s expected close. A renewal at lower value would not change the merger consideration but would matter enormously in a break.
7. What is PNT and radar revenue? Satelles was acquired in 2024 and PNT is one of four strategic pillars, but revenue has never been separately disclosed in 2.5 years — the AMBA pattern.
8. Why did the stock rise 21.3% on guidance that was worse than the prior year? The 12 February 2026 session remains the least-explained material move in the series.
9. What is the borrow cost on RKLB? It determines whether the 11.9% spread is genuinely available to a hedged arbitrageur or is substantially consumed by carry.
10. Does Baralonco (9.53%) intend to vote for the transaction? It is not bound by the Support Agreement and has been a holder since 2009.

14. What Must Be True

14.1 The bull case

For a holder at $47.34 to earn an acceptable return, all of the following must hold:

1. The deal closes — FCC transfer-of-control consent, HSR expiry, and the specified foreign investment and telecom clearances all arrive before the 2027-06-28 outside date (or its extensions), and Iridium’s stockholders approve.
2. Rocket Lab’s ten-day VWAP at measurement is at or above ~$64.95, so the stock leg does not deteriorate further — and ideally at or above $67.50, restoring the full $54.00.
3. No material adverse effect arises at either company — including, at Rocket Lab, a Neutron failure severe enough to qualify.

Falsification test (bull): Rocket Lab’s ten-day VWAP prints below ~$55 at the measurement date, or the FCC designates the transfer-of-control application for hearing. Either outcome caps the payoff below ~$49.00 or pushes closing past the outside date. Concretely: watch Neutron’s Q4 2026 maiden flight and the FCC docket. A Neutron loss of vehicle is the most likely single event that would falsify this case.

14.2 The bear case

For the bearish view — that $47.34 overpays for this package — to be right, at least one of the following must hold:

1. Rocket Lab falls materially further — below ~$50 the payoff drops under $47.00 and a holder loses money even on a successful close. Given ~$61bn of enterprise value at initiation was 75–80% attributable to an unflown rocket, and the rocket has since slipped again, this is not a remote scenario.
2. The deal breaks or is materially delayed, returning the security to a standalone valuation of ~$30–43 against a business with no guidance, no earnings calls, declining high-ARPU subscribers, and an uncosted ~$3bn rebuild.
3. The standalone asset proves worth less than the market assumed — NTN Direct fails to launch commercially, EMSS renews at reduced value, or a further satellite useful-life revision reverses the depreciation benefit.

Falsification test (bear): Rocket Lab trades durably above $67.50 into the measurement window with FCC consent and foreign clearances granted. At that point the payoff re-fixes at $54.00, the spread becomes ~14% of genuine, low-risk return, and the bearish structural argument about the collar is moot. Concretely: a successful Neutron orbital flight in Q4 2026 followed by RKLB re-rating through $67.50 falsifies the bear case outright.

14.3 The test that distinguishes them

The two cases are separated by one variable that has nothing to do with Iridium: Rocket Lab’s share price at the measurement date, which is dominated by a single binary — Neutron’s maiden flight. An investor in IRDM at $47.34 is, whether they intend it or not, expressing a view on a rocket that has not yet flown.

15. Source Appendix

(See Appendix B below for the full annotated source list.)

Primary — SEC filings (EDGAR, CIK 0001418819): FY2025 Form 10-K (2026-02-12); Form 10-Q Q2 2026 (2026-07-22) and Q1 2026 (2026-04-23); Form 8-K (2026-06-29, merger agreement); Form 8-K (2026-05-14, Aireon); DEF 14A (2026-04-02); Schedule 13D/A No. 9 (2026-04-02); earnings releases filed as EX-99.1 (2025-07-24, 2025-10-23, 2026-02-12, 2026-04-23, 2026-07-22); complete Form 4/4A corpus (413 filings, 2021-08-01 to 2026-07-31).

Primary — Rocket Lab (CIK 0001819994): Form 8-K and Rule 425 communications (2026-06-29); Rocket Lab press release, “Rocket Lab to Acquire Iridium in Historic Deal” (2026-06-29).

Quantitative data (cross-checked and reconciled to filings): AZI price history and valuation_index; ROIC.ai (income statement, cash flow, profitability ratios, enterprise value); FactorsToday (leaderboard, stock loadings, stock info, specific volatility, related stocks).

Note on availability: Iridium ceased holding earnings conference calls as of Q2 2026, so no post-announcement management commentary exists beyond the written releases.

Sections 1–15 contain no investment recommendation and no price target. The Claude's Take block at the head of this article is the author’s own independent opinion and is general information only, not investment advice. Do your own research.

APPENDIX A — Standard Diligence Questionnaire

IRDM — Standard Diligence Questionnaire Appendix

Iridium Communications Inc. (NASDAQ: IRDM) · Report date 2026-08-01 · Price at analysis $47.34

Supplemental to the main article. Labels: F = Fact, I = Interpretation, A = Assumption.

Context that governs every answer below: On 2026-06-28 Iridium agreed to be acquired by Rocket Lab Corporation for $27.00 in cash plus a collared number of Rocket Lab shares, expected to close mid-2027. Several questions have materially different answers in the pending-deal case and the standalone (deal-break) case; both are given where they diverge.

General

What thoughtful questions have other investors asked about this company?

The serious questions cluster in four places. (1) The direct-to-device question: does Starlink’s D2D service — now credibly funded by the acquisition of EchoStar spectrum — eventually take Iridium’s commercial IoT and voice base, or is Iridium protected by certification, polar coverage and terminal physics? This is the question that took the stock from $32.43 to $15.84 between July and November 2025. (F/I) (2) The replacement question, which is asked far less often and matters more: what does the next constellation cost, when must it be ordered, and how is it funded on ~$300m a year of free cash flow with $1.8bn of Term Loan maturing in September 2030? (F) (3) The earnings-quality question: how much of the 2024–25 operating-margin expansion from ~10% to 27% is the Q4 2023 extension of satellite depreciable life from 12.5 to 17.5 years rather than operating improvement? Our arithmetic says roughly half of FY2025 GAAP net income. (I) (4) Post-announcement, the only live question: is the 11.9% spread adequate compensation for holding an unhedged 0.4000-share claim on Rocket Lab for eleven months with the upside capped at $112.50? (F)

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Neither in a conventional cyclical sense — Iridium is a subscription infrastructure business, not a cyclical. But reported earnings are at a structural high that is partly an accounting artefact: FY2025 operating margin of 27.1% versus 10.3% in FY2023, of which ~$77.2m a year is the depreciation-life extension. On the pre-2023 basis FY2025 operating margin would be ~18%. (F/I)

Driven by the external environment or internal actions? Internal, and in an unflattering way. The margin expansion came from an accounting estimate and from price increases levied on a shrinking subscriber base (the 10-K attributes FY2025 commercial voice growth to “price increases in the third quarter of 2025” while voice subscribers fell from 415k to 402k). The external environment — direct-to-device competition — is deteriorating. (F/I)

How stable are revenues? Genuinely stable and among the best features of the business. Service revenue is 73% of total and recurring, gross margin has held between 71.5% and 75.5% every year since 2016, and $110.5m a year of government service revenue is fixed-price and usage-independent under EMSS. Total revenue has risen every year for at least eight years. (F)

Outlook for products/services? The company’s own FY2026 guidance is flat-to-2% service revenue growth with OEBITDA guided below FY2025 — the most honest available answer. Growth is concentrated entirely in IoT at $7.78 monthly ARPU; voice, broadband and government subscriber counts all declined in FY2025. (F)

How big will this market be — growing, shrinking, domestic or international? Mobile satellite services is a modest, slow-growing global pool being flanked by a vastly larger, better-capitalised direct-to-device build-out (SpaceX filing for 15,000 satellites; Amazon filing for 5,105 D2D satellites on Globalstar spectrum). The addressable niche that is genuinely defensible — certified aviation surveillance, GMDSS maritime distress, assured PNT, sovereign national-security communications — is real, regulated and growing, but is not large enough on its own to fund a ~$3bn constellation replacement. Revenue is international in customer base; the largest single customer is the U.S. government at 17% of service revenue. (F/I)

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Sharply more, and by flanking rather than by frontal attack. Iridium’s L-band spectrum barrier is genuine and unbreachable, but the leading D2D services use terrestrial cellular spectrum, sidestepping it entirely. The evidence of structural distress is that every independent Western MSS operator of scale has now been acquired or agreed to be acquired: Inmarsat by Viasat (2023), Globalstar by Amazon (April 2026), Iridium by Rocket Lab (June 2026). (F/I)

How profitable is the business (ROIC, ROE)? ROIC ~9.0% rebuilt from the filings (NOPAT $190.9m ÷ invested capital $2,126.6m), and ~6.0% on the pre-2023 depreciation basis. ROIC.ai independently reports 8.32% for FY2025 and 7.51% for FY2024. Do not use the 22.0% ROE — book equity of $462.6m has been hollowed by $1.2bn of buybacks, so ROE measures the repurchase programme. Against a WACC that cannot credibly be below 8–9% for a 3.3x-levered single-asset satellite operator, Iridium earns at or below its cost of capital. (F/I; the WACC is an A)

How profitable is the industry — how many competitors, what barriers to entry? Barriers are among the highest in any industry: globally coordinated L-band spectrum obtainable only by regulatory grant, plus a ~$3bn, decade-long constellation build. Direct MSS competitors number perhaps three or four of scale. And yet industry profitability is poor — Globalstar’s non-Apple revenue declined three years running; Iridium earns ~9% ROIC. High barriers protected the niche without producing high returns, because the asset required to occupy the niche costs more than the niche earns. (F/I)

Can the business be easily understood? Yes, unusually so. One constellation, one spectrum position, four revenue lines, disclosed subscriber and ARPU tables by segment. The two things that are not easy to understand are deliberately obscured: the true sustaining capital charge, and the effect of the depreciation-life change on reported profit. (I)

Can it be undermined by foreign low-cost labour? No. This is a capital-and-spectrum business; labour is a minor cost (FY2025 revenue per employee ~$894,000). The relevant substitution threat is capital and technology, not labour. (F/I)

Do brands matter? Modestly. “Iridium” carries genuine weight in maritime, aviation and defence procurement as a byword for guaranteed global coverage, and the certification of devices onto the Iridium network is a real credential. But Iridium sells wholesale through 500-plus partners who own the end customer, so the brand accrues partly to the channel. (F/I)

What is the nature of competition? Historically capability-based (who can deliver global, polar, low-latency, small-terminal service — only Iridium). Increasingly price- and bundle-based at the margin, as Starlink and VSAT become the primary link and Iridium becomes the companion. The clearest evidence is broadband ARPU falling from $282 to $259, which the 10-K attributes to “the increased prevalence of use of lower-priced companion plans.” The moat is intact; the role is being demoted. (F/I)

Customers’ switching costs? Real but shallow and asymmetric. Terminals are Iridium-specific and physically integrated into vessels, aircraft and industrial equipment; re-terminalling a fleet is expensive and slow. But the FY2025 evidence is that captivity is insufficient to defend either volume or price outside IoT: three of four subscriber lines shrank. (F/I)

Financial Condition & Balance Sheet

Assets not fully recognised on the balance sheet? Yes, and it is the most valuable thing the company owns: the L-band spectrum licences (up to 8.725 MHz contiguous, plus 200 MHz K-band and 400 MHz Ka-band) carry no meaningful balance-sheet value but are the entire barrier to entry. The Globalstar precedent is instructive — Amazon paid roughly twice the top of the seller’s own banker’s standalone valuation range specifically for D2D spectrum. (F/I)

Off-balance-sheet liabilities? No material ones in the accounting sense — operating lease commitments total only $45.1m undiscounted. But there is a very large undisclosed economic liability: the constellation replacement. The fleet was deployed 2017–2019 on a 17.5-year life, implying end of life ~2035–36; the original programme cost ~$3bn and gross book cost is $3,379.2m. The FY2025 10-K contains no occurrence of “next-generation” constellation, “end of life”, “deorbit” or “future constellation.” It is not a liability under GAAP and it is not disclosed anywhere. (F for the absence; A for the ~$3bn/2035–36 estimate)

How conservative is the accounting? Aggressive on the single item that matters most. Extending satellite depreciable life from 12.5 to 17.5 years in Q4 2023 added ~$77.2m a year to pre-tax income — roughly half of FY2025 GAAP net income — and KPMG identifies the useful-life estimate as the critical audit matter. The extension is defensible on engineering grounds (Iridium’s first-generation fleet famously outlived its design life) but it is not conservative. Elsewhere the accounting is unremarkable; the non-GAAP measure (OEBITDA) is clearly defined and reconciled, though it excludes share-based compensation of ~$52m a year. (F/I)

How CapEx-hungry is the business? This is the central question and the honest answer is: far more than it currently appears. Reported FY2025 capex was $100.3m — roughly half the $193.1m book depreciation charge on the satellites alone, and book depreciation is itself a floor on the true economic charge because replacement will occur at future prices. The business is in the harvest phase of a ~$3bn investment cycle: capex looks trivial precisely because the asset was paid for a decade ago and has not yet been replaced. Reported capex is not the sustaining charge; constellation replacement is. (F/I)

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy? FY2025 operating cash flow $400.1m less $100.3m capex = ~$299.8m of free cash flow. (Caution: ROIC.ai reports “free cash flow” of $400.1m and returns a null capex field — that figure is operating cash flow and overstates FCF by a third.) Uses in FY2025: $185.0m of buybacks and $62.9m of dividends — i.e. ~83% of free cash flow returned to shareholders. The philosophy is explicit and consistent: return capital, reduce leverage toward 2.0x by the end of the decade. (F)

Significant acquisitions recently? Two. Satelles (2024, ~$110.7m) created the PNT franchise and produced a $19.8m one-time bargain-purchase gain; its revenue has never been separately disclosed in 2.5 years. Aireon (agreed 2026-05-13, closed 2026-07-02, ~$366.7m — half cash, half a $183.4m non-interest-bearing one-year seller loan — plus $154.7m of assumed debt), guided to add “at least” $100m of service revenue and $30m of OEBITDA annually. Aireon is the best capital-allocation decision in the record: a regulated, EASA-certified, contracted-to-2035 aviation-safety monopoly that already flies on Iridium’s own satellites. Note it closed four days after the company agreed to sell itself. (F/I)

Buying back shares? Aggressively. 34.8 million shares retired for ~$1.2bn since February 2021, with 6.8m more for ~$185.0m in FY2025; share count down ~21% from ~133.5m to 105.96m. Judged on outcome this was poor: purchases spanned roughly $25–$60 into a stock that reached $15.84 in November 2025, and FY2024 shows $419.8m of new Term Loan borrowing against $407.7m of buybacks in the same year. Borrowing to retire equity in a ~9%-ROIC business facing an uncosted ~$3bn rebuild is at best marginal. Had the $1.2bn been retained, Iridium would have entered the 2030s with the replacement substantially pre-funded. (F/I)

Issuing large amounts of new shares to insiders? No — this is a genuine positive. Across 413 Form 4 filings over five years, grants totalled 3,088,850 shares against 852,777 withheld for taxes, on a ~106m share base, while the company retired 34.8m shares. Dilution is a rounding error against the buyback. Share-based compensation of ~$51.6m (FY2025) is ~5.9% of revenue — meaningful but not egregious, and the company has now moved annual incentive compensation entirely to cash, costing $17m of OEBITDA in 2026. (F)

Compensation policy of directors/management? The weakest part of the governance record, and the most revealing document in the engagement. The DEF 14A filed 2026-04-02 contains zero occurrences of “TSR”, “Total Shareholder Return”, “ROIC”, “return on invested capital” or “Free Cash Flow”; “OEBITDA” appears 38 times. Management is paid on a measure struck above the depreciation that records the constellation being consumed and entirely before the capex to replace it — the metric design under which a decade of sub-WACC returns registers as success. And for 2026 the Compensation Committee eliminated performance-based equity altogether, moving to 100% service-based RSUs vesting over five years, because “uncertainty in the industry makes it increasingly more difficult to establish multi-year financial performance goals with sufficient certainty.” (F for the quotes; I for the reading)

Motivations of management? The insider record is honest and reads well. A complete census of all 413 Form 4/4A filings (645 transactions, zero fetch failures) shows three open-market purchases in five years: CEO Matt Desch 28,000 shares at $37.01 (2023-11-20); Chairman Robert Niehaus 30,000 at $17.49 (2025-10-28); CEO Desch 20,000 at $17.33 (2025-10-30) — the latter two within 48 hours of each other and three weeks before the all-time-low close of $15.84. Sales totalled 1,177,838 shares for $61.0m at a weighted-average $51.78, with 91% of it in 2022–23 near the five-year high, after which selling essentially stopped. Both readings deserve stating: this is a genuine conviction cluster at the bottom by the two people who knew the business best; it is also true that the Chairman and CEO bought at ~$17.40 and agreed eight months later to sell the company at $54.00. The purchases were made in an open window after the 23 October 2025 Q3 release and no sale process is evidenced at that date. (F)

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No. Iridium Communications Inc. is a Delaware corporation, common stock, par $0.001, listed on the Nasdaq Global Select Market. Standard Form 1099 treatment. (F) Note for the transaction: the merger is “generally intended to qualify as a tax-free reorganization” — but only if the stock/cash mix at closing meets the required threshold. If it does not, the second-step merger will not occur and the whole transaction is taxable. With Rocket Lab below the collar floor the stock share of consideration is falling, which moves in the direction of jeopardising tax-free treatment. (F/I — a real and under-discussed risk)

Dividend policy? Initiated 2023 and raised every year since; $0.15 per share quarterly in 2026 (~1.27% trailing yield at $47.34). FY2025 dividends paid $62.9m, a 55% payout of GAAP earnings and ~21% of free cash flow. Comfortably covered. (F)

How profitable is the business? Gross margin 71.5%; OEBITDA margin ~57% of revenue on the company’s definition; GAAP operating margin 27.1%; net margin 13.1%. Superb margins, mediocre returns — the gap is the $3.4bn constellation. (F)

Is net income diverging from cash from operations? Yes, and structurally rather than suspiciously. FY2025 operating cash flow of $400.1m is 3.5x net income of $114.4m, because $210.2m of D&A and $51.6m of share-based compensation are non-cash. This is the normal signature of a heavy-infrastructure business post-build and is not a red flag. The genuine caution runs the other way: operating cash flow overstates distributable cash because the depreciation being added back is a real economic charge that will become a real cash outflow when the constellation is replaced. (F/I)

Risks & Downside

What factors would cause the stock to decline? In order of current relevance: (1) Rocket Lab’s share price — below the $67.50 collar floor the exchange ratio is fixed at 0.4000, so every $1 of RKLB decline costs $0.40 of consideration, ~0.85% of the IRDM price. RKLB has fallen from $98.01 on announcement day to $64.95. (2) Regulatory delay or failure — FCC transfer-of-control consent plus specified foreign investment and telecom clearances, against a 2027-06-28 outside date. (3) Deal break, returning the security to a standalone ~$30–43. (4) In a standalone world: EMSS non-renewal ($110.5m/yr, 17% of service revenue, renewal now expected “by March 2027”), accelerating D2D substitution, or disclosure of the constellation replacement cost. (F/I)

Risk of a catastrophic loss? Low but non-zero, and unusually concentrated. Iridium is a single-asset company: one constellation, on the balance sheet as one line. A cascading debris event, a design-common-mode failure across the fleet, or a hostile anti-satellite action would be uninsurable in practice and existential. Financial catastrophe is a lesser risk — 3.3x net leverage (~4.0x pro forma for Aireon) with no maturity before September 2028 (revolver) and September 2030 (Term Loan), against ~$300m of annual free cash flow. (I)

Chance of a total loss? Very low. Even in a break scenario the business generates ~$300m of free cash flow against ~$2.1bn of pro-forma net debt, holds irreplaceable spectrum, and has a fixed-price government contract. The realistic bad case is a 10–37% decline to the ~$30–43 standalone zone, not impairment. The genuine long-run risk is not bankruptcy but a slow squeeze: an asset base that must be rebuilt for ~$3bn out of a cash flow stream that is no longer growing. (I)

Recent News & Events

Has the business environment changed recently? Profoundly, and more than once in twelve months. In July 2025 the company cut its own service-revenue growth guidance from 5–7% to 3–5% and the stock fell 22.1% in a session. In September 2025 SpaceX’s agreement to acquire EchoStar spectrum made a global Starlink direct-to-device service credible and the stock fell 14.9% and 10.3% on consecutive days, reaching $15.84 on 20 November 2025. In February 2026 the company guided FY2026 service revenue to flat-to-2% with OEBITDA below FY2025 — and the stock rose 21.3%. In April 2026 Amazon agreed to acquire Globalstar, and Iridium rose 14.2% two sessions later on the read-across. In June 2026 Rocket Lab agreed to acquire Iridium. And in July 2026 the company announced it will no longer hold earnings calls or provide guidance. (F)

Significant acquisitions? Iridium is now both an acquirer and a target. As acquirer: Aireon LLC, closed 2026-07-02, ~$366.7m plus $154.7m of assumed debt. As target: Rocket Lab Corporation, merger agreement dated 2026-06-28 — $27.00 cash plus a collared number of Rocket Lab shares (ratio 0.4000 below $67.50; $27.00 ÷ price between $67.50 and $112.50; 0.2400 above $112.50), a $54.00 headline at signing, a 24.1% premium to the $43.52 unaffected close, and an enterprise value of approximately $8.0bn. Company termination fee $223.62m; outside date 2027-06-28, extendable to 2027-09-28 and 2027-12-28; no financing condition; conditions include the stockholder vote, HSR, FCC consent to transfer of control, specified foreign investment and telecom clearances, and S-4 effectiveness. Directors’ Support Agreements cover only ~1.6%; the largest holder, Baralonco Limited (9.53%), is not bound. Expected close mid-2027. (F)

Change in accounting policies? The material one predates the window but still dominates the reported numbers: in Q4 2023 the estimated useful life of the satellites was extended from 12.5 to 17.5 years, reducing annual depreciation by ~$77.2m on the satellite asset alone and flowing consolidated D&A down from ~$305m (FY2022) to ~$205m (FY2024–25). Separately, in 2026 the company changed practice to pay annual incentive compensation entirely in cash rather than a mix of equity and cash, reducing OEBITDA by ~$17m. (F)

Recent changes — new markets, facilities, management? No change in senior management: Matt Desch has been CEO since 2006 and Robert Niehaus chairs the board. Louis Alterman joined the board on 2025-12-04, expanding it to twelve (nine independent). New markets are the four February 2026 strategic initiatives — satellite IoT (the Iridium 9604 tri-mode module launched 2026-06-23, and Iridium NTN Direct, a 3GPP standards-based D2D service still not commercially launched), assured PNT (a commercially available PNT ASIC), national security missions (SDA ground systems), and aviation safety (Aireon, plus prospective space-based VHF). Under the merger agreement, outstanding Iridium RSUs and PSUs convert into Rocket Lab RSUs on the same vesting schedule, with full vesting on a without-cause termination within twelve months of closing. (F)

APPENDIX B — Source Appendix

IRDM — Source Appendix

Iridium Communications Inc. (NASDAQ: IRDM) · CIK 0001418819 · Report date 2026-08-01 All URLs taken verbatim from scripts/edgar.sh since output (never hand-constructed) and HTTP-verified (200) on 2026-08-01.

A. Primary — Iridium SEC filings (EDGAR, CIK 0001418819)

B. Primary — Iridium earnings releases (Exhibit 99.1 to the earnings 8-Ks)

The earnings 8-K primary document is a one-page cover; the substance is in EX-99.1 and was fetched separately.

C. Primary — Rocket Lab Corporation (acquirer; EDGAR CIK 0001819994)

Material absence, recorded deliberately: As of 2026-08-01 Rocket Lab has not filed the Form S-4 for the shares to be issued. Consequently no board-approved projections, no fairness opinion, and no background-of-the-merger chronology are public. All embedded-expectations analysis in the Valuation section was therefore built by hand from the filed financial statements and the collar terms. The S-4 is the highest-value future document on this name.

D. Industry, competitor and market context (secondary — public)

E. Quantitative data feeds (cross-check; every material figure reconciled to a filing)

Position disclosure: the author holds no position in IRDM, RKLB or any company named in this article, and nothing here should be read as implying one.

Management transcripts: Iridium stated in its Q2 2026 release (2026-07-22) that it “does not intend to hold conference calls to discuss its quarterly financial results.” No earnings call has been held since the merger announcement, and ROIC.ai’s transcript tools returned no usable Iridium calls. Management commentary in this report is therefore drawn from the written earnings releases and the Rule 425 employee-communication transcript (#8), and is treated as hypothesis rather than evidence throughout.