Globalstar

Phase A — Understand the business

Lens 1 · Company Overview

Globalstar is a Mobile Satellite Services (MSS) operator: a constellation of Low Earth Orbit (LEO) satellites + global gateways + ITU-harmonized L/S-band spectrum, providing voice/data beyond terrestrial coverage. Four revenue lines:

1. Wholesale capacity services — the franchise. Globalstar sells satellite network capacity to Apple ("the Customer"), which powers iPhone Emergency SOS / satellite messaging. This is 63% of FY25 revenue and 66% of Q1-26 revenue. No other customer >10%.
2. Commercial IoT — asset-tracking modules/chips (STX-3, ST-150, ST100, RM200M two-way launched Oct 2025) sold to FEMA, U.S. Army/Air Force, NOAA, BP, Shell, U.K. MoD. FY25 $27.3M service rev.
3. SPOT — consumer retail SOS/tracking devices (Gen4, X, Trace). FY25 $37.3M, declining under competitive pressure (Garmin inReach).
4. Duplex (legacy voice, $15.2M, being wound down) + Government & terrestrial spectrum (Parsons partnership; Band 53/n53; XCOM/Virewirx RAN IP).

Contract structure (the key term): the Apple Updated Services Agreements pay Globalstar a fixed service fee + cost reimbursements (opex & capex) + expanded-service fees + bonuses, and crucially fund the entire ~$1.5B Extended MSS Network build via prepayments. The agreements are indefinite-term but terminable by the Customer at any time on notice — the structural Achilles' heel that the Amazon deal has now largely neutralized (see Lens 12/13).

~791,000 MSS subscribers; 477 employees; HQ Covington, LA; auditor Ernst & Young (PCAOB ID 42), unqualified opinion.

Lens 2 · Supply Chain

Named, end-to-end:

• Satellites (upstream): MDA Space (Maxar/MDA, Canada) — two contracts: (a) 2022 agreement for 17 (up to 26) HIBLEO-4 replacement satellites, amended price $329.3M; (b) Feb-2025 agreement for 50+ third-generation C-3 satellites, $775.0M. MDA missed delivery milestones — Globalstar is owed (and negotiating) liquidated damages; first replacement set delivered April 2026, launching May 2026.
• Launch: SpaceX — Launch Services Agreements (Aug-2023, Jun-2025) for the replacement birds + Oct-2024 agreement for the C-3 constellation. (Note the irony: Globalstar depends on SpaceX to launch the very network that competes with Starlink, and that Amazon is buying.)
• Ground network: ~90 new antennas across 35 ground stations in 25 countries under construction (North America, Asia, Europe).
• Devices/silicon: proprietary ASIC + transmitter modules; third-party contract manufacturers ($9.1M open inventory POs); "Big Box" retail distribution (Amazon, Bass Pro, Cabela's, REI, Academy, West Marine).
• Downstream buyer (the chokepoint): Apple — single ~63% customer. Post-merger, Amazon's Acquisition Sub II takes Apple's 20% SPE stake, and Amazon signed a new long-term Apple connectivity agreement simultaneously.
• Controlling owner / financier: Thermo (James Monroe III, Executive Chairman) — guarantor of certain obligations.

Single-source dependency = Apple demand + MDA supply. Both are mid-transition (Apple → Amazon-backed; MDA late).

Lens 3 · Competitive Advantages (moats)

• Spectrum is the moat — and it is regulator-protected. Globalstar holds globally harmonized Big LEO L-band (1610–1618.725 MHz) + S-band (2483.5–2500 MHz) MSS spectrum, plus 11.5 MHz of terrestrial Band 53/n53, and ~12.0 billion MHz-POPs of terrestrial authority across 12 countries. In April 2026 the FCC reaffirmed Globalstar's and Iridium's exclusive Big LEO rights and rejected spectrum-sharing as impractical. That exclusivity is precisely the asset Amazon is paying $11.6B for — "digital real estate" to do direct-to-device without specialized terminals.
• Barriers to entry: licensing lead-time, constellation capex, ITU coordination — genuine and high.
• Bargaining power: historically weak vs. Apple (one customer, terminable contract) and weak vs. capital markets (chronic dilution/refinancing). The Apple deal inverted some of this by pre-funding the network. Amazon's acquisition is the ultimate expression: the moat (spectrum) is worth more to a strategic (Amazon Leo) than the standalone cash flows ever justified.
• IP: XCOM/Virewirx RAN + peer-to-peer connectivity patents (exercised purchase right Jan 2026); pledged as collateral.

Lens 4 · Segments

Single reportable segment (MSS), but revenue disaggregates cleanly:

Geography: US is ~84% of service revenue ($228.7M of $257.3M FY25); Canada/Europe/LatAm the rest; ~13% of total revenue foreign-denominated. The story is one number: Apple wholesale up, everything subscriber-facing flat-to-down. Globalstar is an Apple-capacity company with a legacy device business attached.

---

Phase B — Measure performance

Lens 5 · Earnings Result (FY2025 10-K + Q1-26 10-Q)

FY2025:

• Revenue $273.0M (+9% YoY).
• Operating income $7.4M (2.7% margin) — barely positive at the operating line; FY24 was a $(0.9)M loss.
• D&A&accretion $87.4M; SBC $23.4M (down from $35.5M — a 2023 executive RSU grant rolled off).
• Below the line: interest income/expense net $(40.9)M (up from $(13.6)M — driven by a $40.6M non-cash significant-financing-component charge under ASC 606 on the Apple prepayment); FX gain $15.7M; derivative gain $15.0M.
• Net loss $(8.7)M; net loss to common $(19.3)M after $10.6M preferred dividends; EPS $(0.15).

Q1-2026:

• Revenue $70.1M (+17% YoY); operating income $8.2M (vs $(8.5)M loss PY) — operating leverage turning.
• Adjusted EBITDA $33.5M (vs $30.4M PY). Note: Adjusted EBITDA reconciliation table is not in the markdown extract of the filing on disk; the $33.5M figure is web-sourced from the Q1 press release.
• Net loss $(17.4)M; EPS $(0.16) — the non-cash financing/interest drag keeps GAAP red despite operating profit.

Balance sheet:

• Cash $447.5M (FY25) → $358.4M (Q1-26, drawn down by capex with no new Apple prepayment draw in Q1).
• Debt principal $410.0M (2024 Debt Repayment $221.6M + 2023 Funding $182.1M + 2021 Funding $6.3M), weighted stated rate up to 9%. Repaid via offsets against Apple service fees, not cash — interest paid in cash in FY25 was $0.
• Property & equipment net surged $674M → $1,305M (satellite/ground build).
• Deferred revenue (contract liabilities) $869.0M (FY25) → $891.7M (Q1-26) — overwhelmingly the Apple Infrastructure Prepayment.
• Equity $355.7M; retained deficit $(2.14)B (history of losses).

Market reaction / what's priced: academic by now — the stock is pinned near the $90 deal price ($81 spot). The 2024–25 re-rate (from a $22.35 low) already capitalized the Apple-network + takeout optionality.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research shelf (transcripts=0). Web-sourced: management tone shifted decisively from "execution + dilution survival" (2023–24) to "deliver the Extended MSS Network milestones and close the Amazon deal" (2026). The Q1-26 call narrative is dominated by (a) the May-2026 first replacement launch, (b) the Amazon merger mechanics, and (c) record IoT activations. The recurring phrase is now "the Customer" / "the Mergers"; the thing they stopped emphasizing is standalone equity-value creation — because at $90 it is fixed. Sentiment: management is now effectively running a closing playbook, not a growth narrative.

Lens 7 · Comps

GSAT's own valuation is deal-pinned, not fundamentals-pinned, so the relevant comp work is (a) the implied takeout multiple Amazon paid, and (b) the only clean public MSS peer, Iridium, as a standalone-value sanity check.

• Implied takeout multiple: $11.6B EV / $273M FY25 revenue ≈ ~42× sales; / ~$130M annualized run-rate Adj EBITDA (4× the $33.5M Q1) ≈ ~89× EBITDA. These are spectrum-and-strategic-control multiples, not cash-flow multiples — Amazon is buying Big LEO spectrum + a closing shortcut to D2D, not Globalstar's P&L. Do not anchor on them as "fair value."
• Standalone sanity check (Iridium): at IRDM's 14.55× EV/EBITDA, GSAT's ~$130M run-rate Adj EBITDA → ~$1.9B EV, i.e. roughly $13–15/share ungrounded standalone-ish — a fraction of the $90 deal. The entire premium above ~$15 is the spectrum/strategic bid. This is the number that matters for deal-break downside (see Lens 12).
• Provenance discipline: Iridium multiples are; GSAT "fair-value" standalone is with arithmetic shown; no fabricated GSAT multiple is presented as sourced.

Lens 8 · Stock-Price Catalysts (≥5% moves, last ~5y)

:

• Sep 2022 — Apple unveils iPhone 14 Emergency SOS via satellite (Globalstar the operator): the foundational re-rate.
• Oct/Nov 2024 — Updated Services Agreements + $400M Apple SPE investment + refinancing: large move; de-risked the balance sheet.
• Feb 2025 — 1-for-15 reverse split (10 Feb 2025); $775M C-3 satellite order.
• Throughout 2025 — grind from a $22.35 low to a $84.70 high on Apple-capacity growth + takeout speculation.
• 14 April 2026 — Amazon acquisition at $90 announced: GSAT jumps toward the deal price.
• Pattern: for five years GSAT has traded almost entirely on Apple/strategic-customer news and capital-structure events, not on its subscriber P&L. The market has always valued the option on the spectrum, not the cash flows — which is exactly what the Amazon bid monetized.

---

Phase C — Judge people & books

Lens 9 · Management

• James Monroe III — Executive Chairman & controlling stockholder via Thermo (~57.6–58%). Track record: repeatedly rescued Globalstar through bankruptcy-adjacent distress, dilution, and refinancings over ~15 years; landed the Apple deal; engineered the reverse split and now the Amazon exit. Skin in the game: enormous — Thermo owns ~58% common + $136.7M of the Series A Preferred + warrants (666,668 @ $30 strike). The Apple/Amazon outcome is, in large part, his outcome.
• Capital allocation: historically dilutive and distress-driven (chronic equity/debt issuance, retained deficit $(2.1)B), but the 2024 Apple restructuring + 2026 Amazon sale is a franchise-defining capital-allocation win — he converted a terminable single-customer contract and a perennially under-capitalized balance sheet into an $11.6B strategic exit at ~42× sales.
• Governance / related-party: this is a controlled company. A standing Strategic Review Committee (SRC) of independent directors must approve extraordinary/related-party transactions; the SRC + Board unanimously approved the merger, and Thermo delivered a Written Consent that already constitutes full stockholder approval. Thermo leases Globalstar its HQ ($1.7M/yr) and guarantees obligations — textbook related-party density, but disclosed and SRC-gated.
• Archetype: founder-operator/controlling-financier hybrid. For a closing situation this is a feature — Monroe's 58% removed the shareholder-vote risk entirely.

Lens 10 · Forensic Red Flags

Forensic read of the income statement, balance sheet, and cash flow:

1. Operating cash flow is wildly flattered by the Apple prepayment — do not read $621.7M OCF as business cash generation. FY25 OCF $621.7M includes $430.6M of Apple Infrastructure Prepayment recorded as deferred revenue, which is then spent on capex. Investing outflow was $(550.4)M the same year. The prepayment is a financing-like inflow routed through operating + deferred revenue. Underlying operating cash generation is a fraction of the headline. This is the single most important forensic point.
2. GAAP revenue contains large non-cash / financing constructs. Wholesale capacity revenue includes cost-reimbursements and a $49.2M significant-financing-component estimate; the $40.6M non-cash ASC 606 financing charge in interest expense is the mirror image. Revenue quality is "real but structured" — heavily dependent on judgment around the Apple contract.
3. Off-balance-sheet-ish VIE / NCI presentation. The Globalstar SPE (Globalstar Licensee LLC) is a consolidated VIE; Apple's $400M / 20% interest is recorded as a $451.6M non-current liability ("Other non-current liabilities"), not equity. Legitimate under ASC 810, but it means reported equity ($356M) and leverage optics understate the Apple-linked obligations. SPE assets $824.6M (FY25) → $916.6M (Q1-26).
4. Interest "paid" in cash was $0 in FY25 — debt services via offsets against Apple service fees, and a chunk of interest is non-cash accretion/financing-component. Cash interest understates the true cost of the capital structure.
5. Satellite impairment risk is real and recurring: a $7.0M Q1-25 loss on a second-gen satellite power-control anomaly; a one-year useful-life cut would add $5.1M/yr depreciation. Aging constellation + late MDA replacements = live impairment exposure.
6. Chronic dilution + reverse split — $(2.1)B retained deficit, 1:15 reverse split Feb 2025; warrants to Apple and Thermo are further overhang (mooted by the cashless-exercise-at-close mechanics).
7. SBC $23.4M FY25 (8.6% of revenue) flatters nothing on non-GAAP here because GAAP is already a loss, but it is a real ongoing cost.

Regulatory findings (required sub-section):

• SEC Litigation Releases / AAERs: None. Verified via SEC EDGAR EFTS (LR + AAER) for 2021-06-21→2026-06-21 — total_sec_findings: 0.
• 10-K Item 3 (Legal Proceedings): "In management's opinion, there is no pending litigation, dispute or claim, which could be expected to have a material adverse effect on the Company's financial condition…". The 10-K notes historical exposure to OFAC / BIS / ICE / export-control inquiries inherent to a global satellite operator, but none deemed material.
• Non-SEC enforcement (web search): No material FTC/DOJ/FDA/CFPB enforcement actions or consent decrees against Globalstar surfaced. The relevant regulatory action is the merger review itself — FCC license-transfer review (comment deadline ~6 July 2026; 12–18 mo expected), HSR antitrust, and foreign-investment screening across 120+ jurisdictions. This is deal-process risk, not misconduct.
• Verdict: No material regulatory or legal findings of wrongdoing — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-21. Clean books; an unqualified E&Y opinion; the "risk" is entirely structural (Apple/VIE/prepayment accounting) and process-driven (merger clearances), not fraud.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection

A standalone EPS model is the wrong tool — the value is the deal. I therefore frame the projection as the arbitrage payoff, with a standalone-value floor for the break scenario. No forecast.ts EPS line is logged (per --watchlist rules, and because the security resolves on a deal, not a fiscal-year EPS).

Spot: $81. Deal: $90 cash (≤40% proration; default stock 0.3210 AMZN capped at $90), less up to $110M aggregate milestone clawback ($0.85/share across ~129M shares ). Close: 2027.

• Expected value ≈ 0.60×$90 + 0.20×$89.15 + 0.20×$52 ≈ $82.2 vs $81 spot → only a slightly positive EV at these probabilities, but the gross spread to $90 is ~10.8% over ~12–18 months (~7–10% annualized) if you believe close odds are higher than my conservative 80%.
• Break-downside is the whole game. A clean deal-break would not return GSAT to its $22 low — the Apple network is now more entrenched (Amazon re-signed Apple), the FCC just protected the spectrum, and the $592M reverse termination fee lands on Globalstar if antitrust/FDI kills it. Reasonable break floor ~$45–60, not $22. That asymmetry (small spread up, contained downside, fee cushion) is the trade.
• The standalone fundamental value (~$13–15/sh at Iridium multiples ) is irrelevant unless the deal breaks AND Apple walks AND the spectrum re-rates down — a triple-conditional tail.

Lens 12 · Bull vs Bear (the deal, not the company)

Bull (deal closes ~$90):

• Shareholder approval is already done — Thermo's 58% Written Consent is all that Delaware/the charter require. No proxy fight, no vote risk. This is the single biggest de-risker.
• Apple converted from blocker to partner — Amazon signed a new long-term Apple connectivity agreement and takes Apple's 20% SPE stake; Apple endorsed the deal and waived its ROFO. The historical "single-customer can terminate" risk is largely retired.
• FCC is structurally constructive on the asset — it reaffirmed exclusive Big LEO MSS rights in April 2026 and rejected sharing. The thing being transferred is regulator-blessed.
• $592M Amazon reverse break fee signals Amazon's commitment and cushions the downside.
• Strategic logic is airtight for Amazon: spectrum + 24 operational sats + ground network = a shortcut for Amazon Leo into D2D vs. Starlink, bypassing years of buildout.

Bear (the spread doesn't close, or closes late/low):

1. Regulatory drag is the real risk — not approval, but time. FCC 12–18 mo + HSR + foreign-investment screening across 120+ countries can push close past 2027 or attach conditions. Time decay erodes the annualized return even if it ultimately closes.
2. The Amazon–SpaceX FCC war is a wildcard. Although SpaceX and Globalstar are aligned on Big LEO exclusivity (June 2026 filing), the broader Amazon-vs-SpaceX spectrum/datacenter brawl and an "FCC chair attacks Amazon" dynamic could politicize this specific license transfer. A politicized FCC is the scenario that breaks the deal.
3. Milestone clawback ($110M) if HIBLEO-4 satellite milestones slip — and MDA is already late; the first launch is May 2026, mid-process. Small ($/sh), but a live haircut.

• Pre-mortem (18 months out, thesis broke): the FCC, under SpaceX pressure and antitrust scrutiny of Amazon's spectrum concentration, either blocks the license transfer or extracts divestiture conditions Amazon won't accept; Amazon walks, pays the $592M fee, GSAT re-rates to ~$50; the spread buyer's IRR is negative because of the time + the gap-down.
• Are multiples too high? Irrelevant for the arb — at $90 the multiple is whatever Amazon agreed; the question is binary close/no-close, not re-rating.
• Contrarian view (what the market is refusing to see): the consensus treats this as a near-certain close with a "free" 11% spread. The under-priced risk is regulatory duration, not regulatory denial — this can be a value-destroying slow deal even if it eventually closes, and the FCC has every incentive to take its time given the Amazon/SpaceX feud.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the "buy the spread" case:

• What breaks the model: the deal is a spectrum-concentration transaction handed to regulators in the middle of the most contentious satellite-spectrum fight in the industry (Amazon vs SpaceX at the FCC). That is not a benign HSR filing — it is a politically charged license transfer where a sitting FCC chair has publicly attacked Amazon. Concentration risk: 100% of the thesis is one regulatory outcome.
• Revenue concentration unchanged: 66% of the actual business is still one customer (Apple), now re-papered through Amazon. If the Amazon deal breaks and the Apple economics get renegotiated in the chaos, standalone Globalstar is a low-EBITDA, high-capex, $410M-levered satellite operator worth a small fraction of $90.
• Most dangerous competitor bulls underestimate: SpaceX/Starlink in D2D — the entire reason Amazon is overpaying is that it is behind. If D2D economics compress (T-Mobile/Starlink, AST SpaceMobile), the strategic premium that justified $90 looks like the top tick.
• Worst capital-allocation history: chronic dilution, $(2.1)B retained deficit, a 1:15 reverse split to cosmetically fix the share price — a controlling shareholder (Thermo) whose preferred/warrant/lease interests don't perfectly align with common, now cashing the franchise out.
• Assumptions that must hold for $90: (1) FCC clears the transfer without prohibitive conditions; (2) FDI clears in 120+ jurisdictions; (3) HIBLEO-4 milestones largely met (MDA late); (4) Amazon doesn't get cold feet on its own spectrum strategy. Any one failing impairs the spread.
• If growth/close-odds disappoint 20–30%: at a 60% (not 80%) close probability, EV ≈ 0.60×$90 + 0.40×$52 ≈ $74.8 — below the $81 spot. The spread is only attractive if you're confident the deal closes; at merely "coin-flip-plus" odds you lose money.
• Single permanent-impairment scenario: FCC blocks the license transfer on spectrum-concentration grounds (plausibility: low-but-not-trivial given the Amazon/SpaceX politics) → Amazon walks → GSAT to ~$45–55 → ~35% downside from spot, only partly offset by the break fee. The risk/reward is ~9% up vs ~35% down; you need >~80% close confidence to justify it.

Lens 14 · Management Questions (ordered by information value)

1. What is the expected FCC license-transfer timeline, and what specific conditions (divestiture, spectrum caps, build-out commitments) is Amazon prepared — and not prepared — to accept?
2. Across the 120+ foreign-investment jurisdictions, which are the long-pole / highest-risk approvals, and what is the realistic outside date for close?
3. Given MDA's delivery delays, what is the current status of the HIBLEO-4 satellite milestones that govern the $110M clawback, and how much of that clawback is now likely?
4. If the Amazon deal were to break, what is the standalone capital plan — does the Apple prepayment + service-fee structure fully fund the C-3 build without new equity?
5. What are the exact triggers and amounts of the $419.8M (GSAT→Amazon) and $592.07M (Amazon→GSAT) termination fees, and which scenarios fall outside the reverse-fee protection?
6. Post-merger, what are the economics of the new long-term Apple agreement Amazon signed — pricing, capacity reservation (vs the old 85%), duration?
7. How does Amazon intend to use the Big LEO + Band 53/n53 spectrum — D2D only, or terrestrial/private-network expansion via the XCOM/Virewirx RAN IP?
8. What liquidated damages are recoverable from MDA Space for the delivery delays, and are they netting against the $775M C-3 contract?
9. What is the orbital-health status of the second-generation constellation after the Q1-25 power-control anomaly — any further at-risk satellites / impairment exposure?
10. How should investors think about standalone Adjusted EBITDA run-rate (Q1 $33.5M) converting to cash given the non-cash financing components and offset-based debt service?
11. What is the SpaceX legal/FCC posture toward this specific transaction (as distinct from the broader spectrum feud), and the perceived risk it poses to close?
12. What happens to the Series A Preferred and the Thermo/Apple warrants mechanically at close, and the cash impact of the $1,000/share liquidation distribution?
13. What are the Commercial IoT unit economics and the realistic ceiling on activations now that the two-way RM200M has shipped?
14. Is there any scenario in which Thermo's interests (preferred, warrants, HQ lease, guarantee) diverge from common holders during the closing process?
15. What is the integration/retention plan for the XCOM engineering team and the spectrum-engineering IP that underpins Amazon's stated rationale?

---