EchoStar Corporation (NASDAQ: SATS) — A Spectrum Liquidation in Disguise: $42B of Forced Sales, a SpaceX Lottery Ticket, and a Levered Melting Holdco at a Premium to Its Own Sum-of-the-Parts

An independent equity research note. Report date: 2026-06-12. Company: EchoStar Corporation · Ticker: NASDAQ: SATS · Sector: Communication Services (Satellite / Pay-TV / Wireless) · CIK: 0001415404 · HQ: Englewood, Colorado · FY-end: December · Auditor: KPMG LLP (going-concern qualified) Price reference: ~$114 (yfinance, intraday 2026-06-12) to ~$128.13 (prior close 2026-06-11) · 52-wk range: $16.73 – $147.25 · Market cap: ~$33–37B (~288.9M economic shares) · Net debt (pre-deal): ~$22.6B · Short interest: ~36% of float (40.6M sh, ~8x days-to-cover)

⚡ Claude’s Take

This block is the author’s own subjective opinion. It is general information and analysis, not investment advice. The body of the article that follows takes no position and carries no price target.

Verdict: HOLD / AVOID-paying-up here — a fascinating special situation that has already priced in the happy ending. Not a short (the 36% short interest and 8x days-to-cover make it squeeze-prone on any catalyst), but not a margin-of-safety buy either. The defensible “you-get-paid-to-wait” entry is in the ~$80–105 zone (roughly the conservative, SpaceX-at-cost sum-of-the-parts); above ~$130 you are underwriting an unrealized SpaceX re-rate and Charlie Ergen’s redeployment of ~$18B that he has not yet made. Conviction: medium.

The tag: “Ergen pulled a rabbit out of the hat — and the audience already paid for the trick.” Cornered by an FCC that threatened to revoke his licenses in mid-2025, Charlie Ergen turned a going-concern zombie into ~$42B of spectrum sales (AT&T $22.65B cash; SpaceX ~$20B, up to $11B of it in SpaceX stock) and the stock went from a $16.73 low to ~$147. That is a genuinely brilliant escape. But the framing matters: this is a liquidation dressed as a growth company. What the market is buying at ~$114–128 is not an operating business — every segment is a no-moat, declining or sub-scale asset (Pay-TV is a melting ice cube, Hughes consumer broadband has been destroyed by Starlink, Boost is a perennial 4th-tier MVNO that has lost money for four years). It is buying (a) ~$18B of net cash after taxes and decommissioning, (b) a private, illiquid, not-yet-owned SpaceX stake struck at a ~$400B valuation that secondaries now mark at $800B–$1T, and © a bet that Ergen redeploys the cash brilliantly — the same Ergen who just incinerated ~$16–18B building a fourth wireless network the FCC then forced him to dismantle. My conservative sum-of-the-parts (deals close, SpaceX held at its $11B cost, full $5–7B tax/decommission bill) is ~$104/share; my base case is ~$130–135. The current price straddles fair value. There is no free conglomerate discount here — if anything the market is paying a modest premium to hard asset value for the SpaceX optionality and Ergen’s reputation.

What would flip me bullish: a SpaceX IPO or primary round that prints ≥$800B with EchoStar’s stake intact and a clear path to liquidity, OR confirmation that IRC §1033 involuntary-conversion treatment shelters most of the spectrum gain (cash taxes far below $5–7B) — either would re-rate the conservative SOTP toward ~$160+ and make today’s price cheap. What would flip me bearish: the AT&T deal slipping or repricing toward its $18.6B floor, a coupon skipping past its grace period before the AT&T cash lands, or a SpaceX down-round/IPO below the $212 strike — any of which re-exposes the live going-concern hole and collapses the de-risking narrative. This is event-driven, two-sided, and high-variance; size accordingly.

1. Executive Summary

EchoStar is no longer the company its 10-K segment descriptions suggest. Through 2024 it was Charlie Ergen’s debt-funded bet on becoming America’s fourth facilities-based wireless carrier, bolted onto the declining DISH satellite-TV business and the Hughes satellite-broadband franchise. That bet failed. In mid-2025 the FCC, under Chairman Brendan Carr, opened an investigation into whether EchoStar had met its 5G build-out obligations and signaled it might revoke the company’s spectrum licenses outright. With ~$26B of debt, a going-concern qualification looming, and a $326M coupon it chose to skip in May 2025, EchoStar had no leverage. Ergen’s response was to sell the crown jewels: all of its 3.45 GHz and 600 MHz spectrum to AT&T for $22.65B cash (August 2025), and its AWS-4, H-Block and unpaired AWS-3 spectrum to SpaceX for ~$20B (September–November 2025), of which up to $11B is paid in SpaceX Class A stock struck at $212/share. EchoStar then abandoned and began decommissioning the 5G network it had spent over $10B building — triggering a $17.6B non-cash impairment and a FY2025 GAAP net loss of −$14.5B that wiped stockholders’ equity from $20.2B to $5.8B.

The result is a corporate identity crisis the market is still digesting. What remains (“RemainCo”) is: a melting but cash-generative Pay-TV business (6.998M subs, −10%/yr, ~$2.7B OIBDA — the only real profit pool); a sub-scale, finally-growing-but-still-loss-making Boost wireless reseller (7.5M subs) that now rents its radios from the very MNOs it competes with; a structurally-disrupted Hughes unit that has conceded consumer broadband to Starlink (and now resells it) while clinging to a small enterprise/aviation/government niche; plus, pro forma, ~$24B of cash, ~$13B of remaining (silo’d) debt, a ~$8.5–11B SpaceX equity stake, and ~45 MHz of retained spectrum. Management has rebranded the whole thing an “asset-light growth company” and created an internal capital-allocation vehicle, “EchoStar Capital,” run by CEO Hamid Akhavan, to redeploy the proceeds.

The investment question is therefore not “is this a good business?” (it is not — there is no durable operating moat in any segment) but “is the post-transformation balance sheet plus Ergen’s optionality worth the ~$33–37B equity value the market assigns?” Our conservative sum-of-the-parts — deals close, SpaceX held at its $11B cost basis, full $5–7B of cash tax and decommissioning — lands at ~$30B (~$104/share), roughly the current price. A base case crediting a partial SpaceX re-rate and reasonable spectrum/Boost values lands at ~$130–135. The entire bull premium above ~$110 rests on two unrealized, unproven things: the SpaceX stake re-rating from its $212/$400B strike toward the $800B–$1T marks recent secondaries imply, and Ergen accretively deploying ~$18B of dry powder. The ~36% short interest is a coherent deal-risk-plus-valuation-plus-no-moat thesis; the 8x days-to-cover makes it dangerous to press. This is a high-variance, event-driven holdco, not a quality compounder and not a clean short.

2. Business Overview

EchoStar reports four segments. The first three are operating businesses in secular decline or structural disadvantage; the fourth is a graveyard.

Pay-TV (≈63% of revenue; the cash engine). DISH TV (satellite direct-broadcast) plus SLING TV (over-the-top streaming). 6.998M subscribers at 12/31/2025 (5.022M DISH TV + 1.976M SLING), down from 7.778M a year earlier. FY2025 segment revenue $9.70B (−9.2% YoY), OIBDA $2.69B (−10.0%). This is overwhelmingly recurring subscription revenue from a loyal, older, rural residual base. It is a melting ice cube — but a slow, high-margin one that throws off the cash that keeps the lights on.

Wireless / Boost (≈25% of revenue). Boost Mobile and Gen Mobile — prepaid-heavy retail wireless with a building postpaid base. 7.511M subscribers at 12/31/2025, up +576k YoY (the only segment growing subs). FY2025 revenue $3.80B (+5.6%), but OIBDA −$378M (a widening loss). Crucially, having sold its own spectrum and abandoned its 5G network, Boost completed its transition to a “Hybrid MNO” on 11/15/2025: it retains and operates its own cloud-native network core and billing systems, but rents radio access and spectrum from AT&T (a Network Services Agreement) and T-Mobile (a Master Network Services Agreement) — both carrying minimum volume commitments. Economically, Boost is now a flanker reseller on competitors’ networks, not a fourth national carrier.

Broadband & Satellite Services / Hughes (≈10% of revenue). HughesNet consumer satellite broadband (739k subs, −16% YoY) plus enterprise, aviation and government managed services. FY2025 revenue $1.46B (−7.6%); GAAP operating loss −$1.61B (including a $1.15B impairment), but Adjusted OIBDA still positive at +$327M. The consumer franchise has been gutted by Starlink’s low-earth-orbit (LEO) constellation; EchoStar has effectively conceded, running a fee-based program that refers HughesNet and new customers to Starlink and installs Starlink for them. The durable remnant is enterprise/aviation/government multi-orbit connectivity (~$1.8B aviation backlog: Delta, Turkish Airlines).

Other (the 5G graveyard; ≈2% of revenue). The legacy nationwide 5G network, abandoned in August 2025 and being decommissioned. FY2025 it carried the bulk of the $17.6B impairment ($16.1B). Revenue $295M is residual; this segment is being wound to zero.

Brands: Boost Mobile, DISH, Gen Mobile, Hughes, HughesNet, Sling. Business model: subscription/recurring across all three live segments, with a small minority of equipment sales. The post-2025 reality: EchoStar is best understood as a Charlie Ergen-controlled holding company — a pile of cash and a SpaceX stake wrapped around a Pay-TV cash cow in runoff, a money-losing MVNO, and a disrupted satellite-broadband stub.

Segment scoreboard (FY2024 → FY2025). The operating picture is one large profitable-but-shrinking segment subsidizing two that lose money or barely break even:

*Other FY2025 OIBDA includes the $16,102M non-cash 5G impairment; Hughes FY2025 GAAP OIBDA includes a $1,530M impairment (shown adjusted above). Adjusted (ex-impairment) consolidated OIBDA was $1,494M. Source: 10-K FY2025 Note 16.

The takeaway is stark: Pay-TV generates ~$2.7B of OIBDA; everything else combined is a ~$1.2B drag. The consolidated $1.49B of Adjusted OIBDA does not cover the ~$1.5–2.4B run-rate cash interest. The entire enterprise, stripped of impairment noise, is a Pay-TV annuity in decline plus a collection of cash-consuming options.

3. Industry Dynamics

EchoStar competes in three distinct industries, and the structural verdict on all three ranges from hostile to terminal.

Linear Pay-TV: terminally declining. Cord-cutting and cord-shaving are structural and accelerating, driven by direct-to-consumer streaming and skinny bundles. The number of US linear-video households shrinks every year; there are no new entrants, only runoff and consolidation. The 2024 attempted DISH/DirecTV combination — which collapsed over a DBS bondholder exchange — reflected the industry’s only rational play: merge two declining satellite assets to slow the bleed and gain leverage on programming costs. The recent Paramount–Warner consolidation worsens the distributor’s position: content owners increasingly go direct-to-consumer and compete with their own affiliates, squeezing the margin of distributors like DISH (Ergen flagged this on the Q4-2025 call). Verdict: structurally bad industry — terminal decline. The only value is runoff cash; the correct strategy is to harvest, not reinvest, which is exactly what management is doing.

Wireless: a mature 3-player oligopoly — attractive for the incumbents, hostile for resellers. Verizon, AT&T and T-Mobile are the only nationwide facilities-based MNOs; they enjoy genuine scale economics, spectrum depth, and network-density advantages. Everyone else — cable MVNOs (Spectrum Mobile, Xfinity Mobile), the carriers’ own flanker brands (Metro, Cricket, Visible, Cricket, Mint, Tracfone), and Boost — are guests on those three networks. The industry is structurally attractive for the MNOs and structurally hostile for resellers, who buy wholesale capacity from the same companies they compete against and can have their access “limited, reduced or terminated” (EchoStar’s own risk factor). Having sold its spectrum, EchoStar moved itself from the (failed) facilities-based side to the disadvantaged reseller side. Verdict: good industry for three players; Boost is not one of them.

Satellite broadband: a violent technological transition. Geostationary (GEO) incumbents — Hughes and Viasat — are being destroyed by LEO constellations (Starlink today, Amazon’s Kuiper next). GEO’s ~600ms latency is structurally uncompetitive with LEO’s ~25–40ms for consumer broadband. This is a textbook Marathon “capital-cycle breakdown via technology disruption”: the incumbents’ sunk GEO satellite investments are stranded by a superior architecture. The consumer segment is a bad business in runoff; the enterprise/government “resilient connectivity” niche (multi-orbit failover, vendor-neutral aggregation) is contestable and sub-scale. Verdict: consumer is structurally bad and disrupted; enterprise/gov is a thin, defensible-ish niche too small to matter.

Direct-to-device (satellite-to-phone): emerging, capital-intensive, winner-take-few. SpaceX/Starlink leads; AST SpaceMobile is the other serious contender. EchoStar spent 17 years building a D2D position and then exited as a principal — it sold the spectrum and now participates only indirectly, via its SpaceX equity stake and a Boost reseller agreement for Starlink Direct-to-Cell. It is a spectator, not a player.

Net industry verdict: EchoStar operates across one terminally-declining industry (Pay-TV), one structurally-hostile-to-its-role industry (wireless reselling), and one disrupted industry (satellite broadband). None offers a structurally attractive home for capital. This is the central reason the thesis has migrated entirely to the balance sheet.

4. Competitive Position

Applying the Greenwald framework — barriers to entry as the dominant question, and the three genuine advantage types (supply/cost, demand/captivity, economies-of-scale-plus-captivity) — EchoStar’s operating segments score poorly across the board.

Pay-TV. The moat is customer captivity, and it is genuine — DISH TV’s monthly churn has actually improved every year (1.69% in 2023 → 1.46% in 2024 → 1.31% in 2025), among the lowest in the industry, and ARPU has risen ($104.56 → $108.90 → $110.39) through price increases on a loyal base. But this is captivity in liquidation: the remaining subscribers are an older, rural cohort with no fiber/cable alternative who are aging out and not being replaced (the company is deliberately throttling acquisition spend — gross adds fell to 202k from 282k, and SAC rose to $1,204). The Greenwald share-stability test is passed only in the trivial sense that a shrinking pie shared among shrinking players is “stable.” It is a moat that protects a melting asset. Verdict: no durable advantage — a high-quality runoff.

Boost. Improving churn (4.17% → 3.00% → 2.84%) and finally-positive net adds reflect execution and promotion, not a barrier. Boost owns no network scale economics; it differentiates on its retained cloud-native core and an eventual Starlink Direct-to-Cell hook, but Greenwald is explicit that differentiation without a barrier does not protect profits. After four years and billions invested, the segment still loses money (−$378M OIBDA) and management can only claim it is “very, very close to breakeven.” The “fourth national carrier” ambition died with the spectrum sale; what remains is a perennial 4th-tier flanker MVNO. Verdict: structurally disadvantaged, no moat.

Hughes. Consumer satellite broadband is the clearest moat-destruction story in the portfolio — a GEO incumbent rendered obsolete by LEO, now reduced to reselling its own disruptor’s product. The enterprise/aviation/government managed-services business has some switching costs and relationship/certification barriers, and a differentiated vendor-neutral multi-orbit (LEO+GEO) electronically-steered-antenna offering, but it is sub-scale against a vertically-integrated SpaceX and too small (~$1.5B revenue, barely profitable) to anchor a thesis. Verdict: consumer no moat; enterprise weak/niche moat.

Net competitive verdict: there is no defensible operating franchise here that compounds value. EchoStar is a collection of one melting cash cow, one sub-scale loss-making reseller, and one disrupted unit with a small niche stub. Any durable value in the post-2025 entity resides in the balance sheet (cash, SpaceX stake, retained spectrum) and in Charlie Ergen’s capital-allocation optionality — explicitly not in operating-business quality. An honest assessment says so plainly.

5. Growth History and Forward Opportunities

History. EchoStar’s revenue has declined every year recently: $17.0B (FY2023, reflecting the full-year DISH combination under reverse-merger accounting) → $15.8B (FY2024) → $15.0B (FY2025, −5.2%). This is the arithmetic of three businesses in decline (Pay-TV −9%, Hughes −8%) only partly offset by Boost (+6%). The “growth” chapter of EchoStar’s recent history was the 5G network build — and it destroyed ~$16–18B of value (see).

Forward operating growth is thin. Honestly assessed, the organic opportunities are:

• Boost is the only unit adding subscribers (+576k in 2025) and approaching breakeven. Real, but low-ARPU (~$37), sub-scale, and structurally dependent on competitors’ networks. Marginal new subscribers are profitable; the segment is not.
• Hughes enterprise/aviation/government — a ~$1.8B aviation backlog (Delta, Turkish), multi-orbit ESA hardware, and DoD “resilient connectivity.” Modest, niche, slow.
• Boost + Starlink Direct-to-Cell — optionality to offer satellite-to-phone text/voice to Boost subscribers, dependent on SpaceX’s network and timeline. Not a near-term P&L driver.

The “growth” management now sells is a capital-allocation story, not an operating one. Having relabeled EchoStar an “asset-light growth company,” the genuine forward case is: deploy ~$18B of net cash, monetize the SpaceX stake, and sell the retained spectrum into new “platforms” via EchoStar Capital. That is an Ergen-optionality bet — the quality of which depends entirely on a man whose last big bet (5G) was a catastrophe — not a franchise-growth bet. Verdict: low-quality / largely absent organic growth; the only credible upside is inorganic redeployment, which is unproven and carries real negative precedent.

6. Financial Quality

EchoStar’s FY2025 financials are dominated by a single non-cash event and an unsustainable cash-interest burden. Reading them requires separating the impairment noise from the operating signal — and the operating signal is weak.

The −$14.5B loss is ~$17.6B of impairment, but the underlying business still doesn’t cover its interest. The FY2025 net loss of −$14,506,939K bridges as follows: operating loss −$17,723,146K includes a $17,632,011K non-cash impairment ($16.1B for the abandoned 5G network in “Other,” $1.5B for Hughes); ex-impairment operating result was roughly −$91M. Below the line: net interest expense −$1,521,713K, interest income +$228,733K, other +$122,812K → pre-tax −$18,893,314K; a non-cash deferred-tax benefit of +$4,386,375K (impairments reversing deferred tax liabilities) → net −$14,506,939K. The tax benefit is non-cash and provides no liquidity.

Adjusted OIBDA tells the real story: a ~$2.7B Pay-TV engine dragging a ~$1.2B deadweight. The company’s own ex-impairment Adjusted OIBDA was $1,494,414K in FY2025: Pay-TV +$2,688,094K (the entire engine), Wireless/Boost −$377,519K, Hughes +$327,223K, residual 5G/Other −$1,101,384K. So the only profitable business is Pay-TV, in −10%/yr decline, and it is partly consumed by Boost’s losses and the dying 5G unit’s residual costs.

This does not cover debt service. Free cash flow was −$1.74B. Using the company’s own definition (operating cash flow − capex − capitalized interest): FY2025 = −$99,374K OCF − $965,730K PP&E − $676,311K capitalized interest = −$1,741M burn, versus +$1,253M (FY2024) and +$2,433M (FY2023). Operating cash flow collapsing to negative is the single most important quality-of-earnings fact: even stripping the impairment, the enterprise as constituted in 2025 burned cash, because ~$1.5–2.4B of run-rate cash interest exceeds the ~$1.5B of Adjusted OIBDA. Total cash (incl. restricted) fell from $4.59B to $2.18B over 2025, and to $1.34B by 3/31/2026.

The interest burden just stepped up. Net interest expense jumped from $482M (FY2024) to $1,522M (FY2025) — not because debt grew, but because the company stopped capitalizing spectrum-construction interest after the August 2025 5G termination (capitalized interest was $676M in 2025, $952M in 2024, $1,162M in 2023). Q1-2026 interest expense was $592,660K — a ~$2.37B annualized run-rate — against an operating business generating ~$1.5B of Adjusted OIBDA. The math does not work without the spectrum proceeds.

The 5G sunk cost, quantified. Net PP&E fell from $9,187,132K (12/31/2024) to $2,243,515K (12/31/2025) — a ~$6.9B network write-down — while the “Other”/5G segment was impaired $16.1B. Regulatory authorizations (spectrum, third-party valued) still stand at $34.5B on the balance sheet — but that is precisely the asset being sold for ~$42B in cash and SpaceX stock.

The debt wall and liquidity bridge — the heart of the going-concern question. Total debt was $25.98B at 12/31/2025 ($7.32B current), reduced to ~$24.25B by 3/31/2026 as 2026 maturities were repaid. It is spread across structurally-separate silos, which matters enormously because Pay-TV cash services the DISH DBS notes while the spectrum being sold collateralizes the SATS-level “Seller Notes”:

The maturity wall is front-loaded: 2026 principal $7.28B (+$2.07B interest = ~$9.35B cash need) against just ~$1.34B of cash at 3/31/2026; 2027 $4.22B (the 11.75% DISH notes); 2028 $3.51B. This is why KPMG’s going-concern qualification is real and not boilerplate: the company cannot fund 2026 from cash on hand or operations — it is entirely dependent on the spectrum proceeds arriving in time. The bridge has three planks: (1) the AT&T cash (~mid-2026) retires the near-term wall; (2) SpaceX’s ~$2B of auto-cancellable interim-debt-service loans cover the Seller Notes through ≥Nov-2027; and (3) the Seller Notes are repaid outright by SpaceX at closing. If plank (1) slips materially, the company is forced back to distressed capital markets — the precise scenario the ~36% short interest is positioned for.

Quality-of-earnings flags for the analyst:

• GAAP EPS of −$50.41 and any trailing “P/E” are meaningless (impairment-driven). yfinance/AZI multiples (EV/EBITDA ~41x, profit margin −98%) are artifacts — do not value this on an earnings multiple.
• Stockholders’ equity fell from $20,191,372K to $5,765,784K; accumulated deficit flipped from +$11.6B retained earnings to −$2.9B. Book value per share (~$19) is not a meaningful floor — it neither captures the spectrum’s market value nor the impairment-distorted asset base.
• The OBBBA tax law (enacted 7/4/2025) restored 100% bonus depreciation and interest deductibility — relevant to the cash-tax question on the spectrum gains, not to FY2025 reported results.

Verdict: economics do NOT improve with scale — they are deteriorating. The only profitable segment is in secular decline, the growth segment loses money, and the consolidated entity burns cash after interest. This is not a business whose quality justifies its valuation; the valuation is a balance-sheet and transaction story, full stop.

7. Capital Allocation

Capital allocation is the bridge between business value and shareholder value, and at EchoStar it is the whole story — both the catastrophe that created the crisis and the optionality that defines the bull case. The track record is, candidly, one of the worst major-cap capital-allocation records of the past decade, partially redeemed by a brilliant forced escape.

The 5G catastrophe. After recombining DISH Network into EchoStar (all-stock, 12/31/2023 — an Ergen holdco reshuffle), the combined entity pursued a debt-funded nationwide 5G build to meet FCC deployment deadlines. The verdict is unambiguous: the “Other” segment was impaired $16.1B in 2025, the network was abandoned and decommissioned, and all traffic was migrated to AT&T’s radios by 11/15/2025. This is a textbook Marathon capital-cycle disaster — pouring >$10B into building a fourth network against three entrenched scale players with no captivity, no cost advantage, and no scale economics, financed with high-coupon, PIK-laden, spectrum-secured emergency debt. ROIC on the 5G investment is deeply negative (impaired to ~zero). The Nov-2024 debt exchanges that funded the death-throes created the SATS-level secured “Seller Notes” (10.75% due 2029, 6.75% due 2030, 3.875% convertible due 2030) plus a web of PIK intercompany loans at 11.50%.

The forced escape (genuinely well-executed). Cornered by the FCC’s May–September 2025 license-revocation threat, Ergen extracted ~$42B for spectrum the market had written down: $22.65B cash from AT&T and ~$20B from SpaceX (up to $11B in SpaceX stock). He even amended the SpaceX deal upward (adding unpaired AWS-3 for $2.6B, all in SpaceX stock) — taking more equity in a private company whose valuation he believed would compound. Whether this is vindication or luck is debatable, but the outcome — converting a going-concern zombie into a deleveraged, cash-rich holdco — is a real achievement. It is, however, forced de-risking, not opportunistic value creation: Ergen sold because he had to, not because he found a great price.

Insider behavior gives no bullish confirmation. Across 159 Form 4 filings (June 2021 – June 2026), there is not a single open-market purchase (code P) by Ergen or any insider — not at the 2025 distress lows (when the stock traded near $17), not on the way up. Recent insider activity is exclusively option-exercise sales (codes M/S), estate-planning gifts and entity transfers (the “Ergen Two-Year May 2025 SATS GRAT,” “Telluray Holdings LLC”), and a string of June-2026 Form 144 proposed sales. A SC 13D/A on 2026-05-15 reflects an estate-planning reshuffle, not accumulation. Ergen already controls the company through super-voting Class B shares, so he has no need to buy — but the absence of any conviction buy at the lows, combined with using the rally for estate planning and routine monetization, is a neutral-to-mildly-negative signal that does not corroborate the bull case.

Governance. EchoStar is a NASDAQ “controlled company.” At 2/25/2026 there were 157,527,391 Class A shares (1 vote) and 131,348,468 Class B shares (10 votes); Ergen holds ~51.3% of the equity and ~90.4% of the voting power (~89.5% effective under an Amended Support Agreement). Minority holders have essentially zero governance leverage. The equity thesis is, quite literally, a bet on one man’s redeployment of ~$18B.

Verdict: a value-destructive history (the 5G write-off is among the largest self-inflicted capital losses in recent telecom) redeemed operationally — but not yet financially — by a skillfully-negotiated forced sale. Paying a premium to hard asset value for this management’s next allocation decision requires crediting the escape and discounting the catastrophe that preceded it. The market is doing the former; a skeptic weights the latter.

8. Changes and Headwinds — Last Two Years

The two-year window contains essentially the entire transformation. In rough chronology:

• 12/31/2023: DISH Network recombined into EchoStar (all-stock). Combined debt ~$26B; 5G build underway.
• Through 2024: Subscriber declines across Pay-TV and Hughes continue; Boost still bleeding. Debt exchanges (Nov-2024) create the SATS-level secured spectrum notes and PIK intercompany loans — emergency refinancing at punitive coupons. A proposed DISH/DirecTV merger collapses over a bondholder exchange.
• May 9, 2025: FCC “May 9 Letter” questions 5G build-out compliance and spectrum utilization.
• May 30, 2025: EchoStar elects not to pay ~$326M of interest on the 10.75% Seller Notes — an indenture default — using the 30-day grace period; cured June 27, 2025. (Management skipped interest on a portion of senior notes through early Q3 2025, all cured within grace.)
• August 25–26, 2025: AT&T License Purchase Agreement — sell all 3.45 GHz + 600 MHz spectrum + a 99-year Hawaii lease extension for $22.65B cash (floor $18.6B).
• August 2025: EchoStar abandons and begins decommissioning the 5G network, triggering the $17.6B impairment ($16.5B in Q3, $1.2B in Q4).
• September 2025: SpaceX License Purchase Agreement — sell AWS-4 + H-Block for ~$17B (cash + SpaceX stock); SpaceX provides ~$2B of interim debt service via auto-cancellable loans.
• September 8, 2025: FCC “September 8 Letter” — Chairman Carr directs staff to conclude the investigation favorably (confirm AWS-4 rights, deem build-out satisfied), resolving the revocation threat contingent on the sales.
• November 2025: SpaceX deal amended upward to ~$20B (adds unpaired AWS-3 for $2.6B, all stock; up to $11B total in SpaceX stock @ $212). Boost completes Hybrid-MNO transition (11/15/2025).
• March 2, 2026: FY2025 10-K filed with a KPMG going-concern qualification. Management creates “EchoStar Capital” (Hamid Akhavan, CEO) as the proceeds-redeployment vehicle.
• Q1 2026 (filed 5/11/2026): Cash down to $1.34B; debt reduced ~$1.7B as 2026 maturities are repaid; going-concern qualification persists. Net loss −$147M (a clean operating quarter, +$393M operating income, swamped by interest).

Open headwinds: (1) regulatory/closing risk — no HSR or FCC approval for either deal was satisfied as of the 10-K date; (2) the >18-month gap between the AT&T cash (~mid-2026) and the SpaceX cash (~Nov-2027); (3) cash taxes and decommissioning of ~$5–7B on the spectrum gains; (4) tower-lease force-majeure litigation (EchoStar stopped tower payments arguing the FCC action was a force-majeure event); (5) the xAI/SpaceX merger muddying the value and dilution of the SpaceX stake. Verdict: the changes are existential and net-positive for survival (the company will very likely avoid bankruptcy) but the headwinds keep the outcome binary until the AT&T cash actually lands.

9. Risk Analysis

Risk summary. The dominant near-term risk is liquidity timing (rows 1–3): the company is solvent contingent on the spectrum cash arriving before the 2026 debt wall overwhelms it. Base case is that the AT&T cash (~mid-2026) retires the near-term maturities and removes the going-concern qualification — but it is not yet contractually closed. The dominant medium-term risk is valuation/redeployment (rows 4, 8): the price embeds a SpaceX re-rate and accretive deployment that are unrealized and, given the 5G precedent, unproven. Catastrophic loss (row 10) requires a double deal-break and is low-probability but not zero.

10. Valuation Discussion (Embedded Expectations)

Earnings multiples are useless here; this is a per-share sum-of-the-parts / asset-value problem. GAAP EPS is −$50 (impairment-driven); EV/EBITDA (~41x reported) and P/E are artifacts. The right lens is: build the pro-forma balance sheet after both deals close, value the runoff operating businesses on asset/runoff value, add the cash and the SpaceX stake, subtract the surviving (silo’d) debt and the tax/decommissioning bill, and compare to the ~$33–37B equity value the market assigns. Pro-forma, parent-level (SATS) debt is roughly zero; the residual ~$13B sits in the DISH DBS (Pay-TV) and HSSC (Hughes) silos, serviced in-silo.

Sum-of-the-parts (≈288.9M economic shares; price reference ~$114–128):

What the price implies. At ~$114–128 the market assigns ~$33–37B of equity value against a conservative SOTP of ~$30B (~$104/share) that assumes deals close, the SpaceX stake at its $11B cost, and the full $5–7B tax/decommission bill. In other words, the current price straddles the base case and sits at a modest premium to the conservative case. Everything above ~$110 is the market already paying for two unrealized things: (1) the SpaceX stake re-rating from its $212/$400B strike toward the $800B (Dec-2025 secondary) to ~$1T (post-xAI Forge mark) private valuations — at cost the stake is ~$38/share, re-marked it could be ~$70–110/share; and (2) Ergen redeploying ~$18B accretively inside EchoStar Capital — pure optionality, no realized return yet. No operating moat is being paid for.

The SpaceX stake — the swing factor and the crux of the whole valuation. EchoStar receives up to $11B of SpaceX Class A stock struck at $212/share, which at signing implied a ~$400B SpaceX valuation and a ~2.8% stake. Since then, secondary prints have re-rated SpaceX sharply: ~$421/share (~$800B) in a December 2025 insider sale, and post-xAI-merger marks near ~$604/share (~$1.03T, Forge, April 2026), with 2026 IPO chatter at $1–1.5T. Re-marked at $800B the ~$11B cost becomes ~$22B; at ~$1T, ~$28B — an unrecognized ~$11–17B (~$38–60/share). But the caveats are severe and the bears are right to press them: EchoStar does not yet own the stock (“we don’t have the equity yet… we actually don’t have that equity,” Ergen, Q4-2025 call); it receives it only at the Spectrum Acquisition Closing (~Nov 2027); it cannot transact, pledge or hedge until then; there is no anti-dilution provision; the xAI merger muddies the effective stake (~80/20 xAI/Starlink by Ergen’s rough estimate); and the $800B–$1T marks are thin, hype-adjacent secondaries. The stake is carried at $212 cost (a cost-method investment with no readily-determinable fair value) regardless of secondary marks. It is a genuine, large, but illiquid and unrealized call option on the most valuable private company in the world — and the stock has front-run it.

Embedded-expectations conclusion. For ~$114–128 to prove conservative, you need: both deals to close on schedule, SpaceX to hold ≥~$800B with the stake intact, cash taxes at the low end (§1033 working), and Ergen to redeploy accretively. For it to prove expensive, any one of: a deal slip/break, a SpaceX down-round toward $212, a high tax/decommission bill, or value-neutral redeployment. The price is not a margin of safety; it is a fair-to-slightly-rich quote on a high-variance set of binary outcomes. No price target and no recommendation — but the embedded expectations are demanding, not cheap.

11. Variant Perception

Consensus (post-doubling). The bull narrative now dominates the tape: “Ergen pulled a rabbit out of the hat.” Forced by the FCC, he monetized otherwise-stranded spectrum for ~$42B, turned a going-concern zombie into a deleveraged, cash-rich holdco with a SpaceX stake that has since roughly doubled on paper, retained the “most valuable” AWS-3 spectrum for a future sale, and has ~$18B to redeploy under a proven operator. The popular framing is “the SpaceX trade you can actually buy in size.”

The strongest bull case. (1) Deleveraging from ~$26B toward ~$13B (silo’d, self-funding), with parent-level debt near zero and the going-concern qualification removed once the AT&T cash lands. (2) A SpaceX stake struck at $212/$400B now marked at $800B–$1T — ~$11–17B of unrecognized value with a potential 2026 IPO catalyst. (3) ~$18B of dry powder under a 51%-aligned, demonstrably aggressive capital allocator. (4) More spectrum to sell (paired AWS-3, CBRS, 12 GHz). (5) Embedded optionality: Boost + Starlink Direct-to-Cell, DISH/DirecTV runoff consolidation, and an “EchoStar Connect”-style pure-play SpaceX vehicle floated by sell-side analysts. With ~36% of the float short and 8x days-to-cover, any positive catalyst is a squeeze.

The strongest bear case (why ~36% of the float is short). (1) A still-levered melting holdco until close — KPMG’s going-concern qualification is live; ~$9.35B of 2026 debt-plus-interest sits against ~$1.34B of cash; the company already skipped a coupon; the SpaceX cash is ~18 months out and bridged only by a SpaceX loan. A slip or a capital-markets freeze re-exposes default. (2) Deal-break binary — no approval is satisfied; the price embeds near-certain completion. (3) The SpaceX stake is illiquid and may be over-marked in investors’ heads — not yet owned, no liquidity until ~Nov-2027, no anti-dilution, xAI ambiguity; the stock has front-run an unrealized, unhedgeable mark that could deflate. (4) Tax leakage — ~$42B of low-basis asset sales generate real cash taxes; the $5–7B estimate plus decommissioning eats a quarter-to-third of net proceeds, and §1033 relief is unproven. (5) No moat / runoff base — every operating segment is no-moat and declining or loss-making, and you are paying a premium to hard SOTP for Ergen’s redeployment immediately after he destroyed ~$16–18B on 5G (a Marathon capital-cycle red flag). (6) Governance — Ergen ~90% of the vote, minorities powerless, zero insider buying at the lows.

The 3–5 assumptions that matter, with falsification tests:

1. Both deals close (AT&T ~mid-2026; SpaceX ~Nov-2027). Bull falsified if a counterparty terminates, HSR/FCC blocks, or AT&T adjusts toward the $18.6B floor on a large exclusion. Bear falsified if AT&T regulatory consent is granted and funds are received.
2. SpaceX stake value (cost $11B vs $800B–$1T marks). Bull falsified if a SpaceX IPO/round prices below $212/$400B or xAI dilution shrinks the effective stake. Bear falsified if a 2026 SpaceX IPO/secondary prints ≥$800B with the stake intact.
3. Net cash after tax/decommission (~$18B; tax $5–7B). Bull falsified if the bill exceeds ~$8–9B (§1033 denied/overruns). Bear falsified if §1033 is confirmed and taxes land near the low end.
4. Ergen redeploys accretively. Bull falsified if EchoStar Capital sits in cash or does a value-neutral/destructive deal. Bear falsified if an accretive redeployment or a large buyback below SOTP is announced.
5. Going concern bridged to close. Bull falsified if another coupon skips past grace or a covenant trips before the AT&T cash. Bear falsified if the AT&T proceeds land in 2026, retire the near-term wall, and the KPMG qualification is removed.

Variant-perception synthesis. The honest read is that the bull and bear cases are both coherent and the truth is genuinely uncertain — which is exactly why ~36% of the float is short and the stock has been a multi-bagger. This is not a mispricing to exploit with confidence in either direction; it is a high-variance event-driven situation where the current price fairly reflects a probability-weighted blend of “deleveraged SpaceX-stake holdco” and “levered melting asset that front-ran an unrealized mark.”

12. Fact vs. Interpretation

13. Open Questions

1. §1033 tax treatment — does the FCC’s forcing action qualify the spectrum sales as an involuntary conversion, deferring/reducing the cash-tax bill? If yes, net cash rises materially (bull); if no, the $5–7B estimate is likely conservative.
2. Exact cash-vs-stock split of the ~$20B SpaceX consideration — depends on the Seller-Note payoff amount and EchoStar’s election; determines how much is liquid cash vs the illiquid stake.
3. Hard MHz-pop value of the retained spectrum (paired AWS-3, CBRS, 12 GHz) — undisclosed (Auction 113 quiet period); the lowest-confidence SOTP block.
4. xAI/SpaceX merger — does it dilute EchoStar’s effective stake, and at what blended valuation does the stock deliver at ~Nov-2027?
5. EchoStar Capital’s first move — buyback below SOTP, a new “platform” acquisition, or idle cash? This single decision could re-rate the equity in either direction.
6. Pay-TV terminal decline rate — how fast does the ~$2.7B OIBDA cash engine erode as programmers go direct-to-consumer?
7. Boost’s true unit economics under full SAC — is the “near-breakeven” claim real, and do the T-Mobile/AT&T minimum commitments become a fixed-cost trap if growth stalls?
8. Hughes enterprise/aero/gov size and durability — the only piece with a niche moat; the consumer-vs-enterprise revenue split is not cleanly disclosed.

14. What Must Be True

For the bull case (today’s price proves conservative):

• Both spectrum deals close on schedule, with AT&T paying at or near $22.65B (not the $18.6B floor). Falsification test: a termination notice, an HSR/FCC block, or an AT&T floor election on a material exclusion.
• The SpaceX stake is delivered intact and re-rates toward ≥$800B, ideally with a 2026 IPO providing a liquidity path. Falsification test: a SpaceX round/IPO below the $212/$400B strike, or material xAI dilution of the stake.
• Cash taxes + decommissioning come in at or below ~$6B (§1033 helps). Falsification test: a disclosed tax/decommission bill above ~$8–9B.
• Ergen redeploys ~$18B at a return above cost (buyback below SOTP or an accretive platform). Falsification test: 12+ months of idle cash or a value-neutral/destructive acquisition.

For the bear case (today’s price proves expensive):

• The going-concern hole bites before the AT&T cash lands, OR a deal breaks. Falsification test: AT&T proceeds land in 2026, retire the near-term maturities, and KPMG removes the qualification.
• The SpaceX stake disappoints (down-round, dilution, or deflating private marks) and the unrealized re-rate the stock front-ran reverses. Falsification test: a SpaceX secondary/IPO ≥$800B with EchoStar’s stake confirmed intact.
• Operating decline + tax leakage + value-neutral redeployment leave hard per-share value below the price. Falsification test: a realized SOTP (post-tax cash + marked SpaceX stake + spectrum sale) demonstrably above ~$135/share.

The single most important swing variable is the SpaceX stake’s realized value at ~Nov-2027, followed closely by whether the AT&T cash arrives in 2026 to neutralize the going-concern risk. Watch those two, plus EchoStar Capital’s first capital-allocation decision, above all else.

15. Source Appendix

See the separate detailed source appendix (SATS_source_appendix.md) for the full citation list. Primary sources relied upon:

• EchoStar Corporation FY2025 Form 10-K (filed 2026-03-02, CIK 0001415404): auditor report (KPMG, going concern, F-2/F-3); consolidated financials; Note 1 (going concern, impairments, AT&T/SpaceX transactions, interim debt service); Note 10 (debt schedule, skipped coupon); Note 11 (taxes/OBBBA); Note 15 (maturities); Note 16 (segments, OIBDA/Adjusted OIBDA); MD&A (subscribers, ARPU, churn, FCC chronology, FCF); capital-stock item and cover (share counts).
• EchoStar Corporation Q1-2026 Form 10-Q (filed 2026-05-11): cash $1.34B, current debt $6.24B, going-concern persistence, clean operating quarter.
• Earnings/event transcripts: Q4-2025 (2026-03-02), Q3-2025 (2025-11-06), Special Call (2025-09-15), AT&T Special Call (2025-08-26), Q2-2025 (2025-08-01).
• SEC EDGAR: Form 4 corpus (159 filings, 2021–2026); Form 144s (June 2026); SC 13D/A (2026-05-15); submissions JSON (CIK 0001415404).
• External SpaceX valuation cross-checks (accessed 2026-06-12): July-2025 tender (~$400B/$212); December-2025 secondary (~$800B/$421, Fortune 2025-12-13, SatNews 2025-12-06); post-xAI Forge mark (~$1.03T/$604, April 2026).
• Frameworks: investment-research-frameworks skill (Greenwald Competition Demystified; Marathon Capital Returns). Peer cross-reads: ASTS (D2D/spectrum), CHTR (Pay-TV/MVNO).

Note on data feeds: AZI and yfinance multiples (EV/EBITDA, P/E, margins) are artifacts of the impairment-driven GAAP loss and were not relied upon for valuation; all financial figures are reconciled to EDGAR filings. The AZI three-statement arrays were unreliable for this recently-restated, multi-class filer, consistent with prior experience on post-merger/restated names.

APPENDIX A — Standard Diligence Questionnaire

EchoStar Corporation (NASDAQ: SATS) — as of 2026-06-12

Supplemental to the research memo. Answers grounded in the FY2025 10-K (filed 2026-03-02), Q1-2026 10-Q, transcripts, and SEC filings. Fact / Interpretation / Assumption labels applied where material.

General

What thoughtful questions have other investors asked about this company? The dominant investor questions cluster around the transformation: (1) Will both spectrum deals close, and when does the cash actually arrive? (AT&T ~mid-2026; SpaceX ~Nov-2027). (2) What is the SpaceX stake really worth, and when can EchoStar monetize it? (struck at $212/$400B, marked $800B–$1T, not owned until ~Nov-2027, no anti-dilution). (3) What is the all-in cash tax + decommissioning bill on ~$42B of low-basis sales? (management: $5–7B; §1033 relief unproven). (4) What does Ergen do with ~$18B of dry powder? (5) Is the going-concern risk truly bridged, or is there a 2026 liquidity gap? (6) Why is 36% of the float short if the stock has been a multi-bagger? The variant-perception tension (deleveraged SpaceX-stake holdco vs. levered melting asset that front-ran an unrealized mark) is the crux.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Internal or external drivers? Neither cyclical high nor low — EchoStar’s results are driven by secular forces (cord-cutting, LEO disruption) and idiosyncratic events (the FCC-forced spectrum sale, the 5G abandonment), not the economic cycle. FY2025’s −$14.5B loss is a one-time impairment event (internal/regulatory-driven), not a cyclical trough. The underlying Pay-TV cash engine is in steady secular decline regardless of the macro cycle. (Interpretation.)

How stable are revenues? Declining but predictable: −5.2% in FY2025, with Pay-TV (−9%) and Hughes (−8%) in structural runoff partly offset by Boost (+6%). Revenue is ~95% recurring subscription — stable in quality (recurring) but negative in trajectory. (Fact/Interpretation.)

Outlook for products/services; how big is this market — growing or shrinking? Linear Pay-TV: shrinking (terminal). Satellite consumer broadband: shrinking (disrupted by LEO). Wireless reselling: the market is large and growing, but EchoStar’s sub-scale reseller position captures little economics. Enterprise/gov satellite connectivity: modest growth, niche. Net: the company’s served markets are mostly shrinking; the one growing market (wireless) is one where it is structurally disadvantaged. (Interpretation.)

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More, in every segment that matters. Pay-TV faces programmer direct-to-consumer disintermediation (Paramount-Warner). Wireless: three scale MNOs plus cable MVNOs squeeze flankers like Boost. Satellite broadband: Starlink and (soon) Kuiper. (Interpretation.)

How profitable is the business (ROIC, ROE)? Poor to negative. FY2025 ROE is −1.12 (GAAP, impairment-distorted, not meaningful). Pre-impairment, the business barely breaks even at the operating line (−$91M) and burns cash after interest (−$1.74B FCF). ROIC on the 5G investment was deeply negative (impaired to ~zero). Only Pay-TV earns a positive segment return, and it is shrinking. (Fact.) Bank/utility-style metrics do not apply; the correct lenses are segment OIBDA and per-share asset value.

How profitable is the industry — competitors, barriers to entry? Wireless is highly profitable for the three MNOs (high barriers: spectrum, scale, capital); hostile for resellers. Pay-TV is a declining-profit-pool industry. Satellite broadband profitability is collapsing for GEO incumbents. (Interpretation.)

Can the business be easily understood? The segments individually, yes. The consolidated entity is unusually complex: a multi-class controlled company, four segments, ~5 debt silos with intercompany PIK loans, two staged multi-year spectrum transactions with cash-and-private-stock consideration, and a going-concern overlay. This is a hard-to-model special situation, not a simple business. (Interpretation.)

Undermined by foreign low-cost labor? Not directly — these are domestic network/subscription services. (Fact.)

Do brands matter? Nature of competition? Switching costs? Brands matter modestly (DISH, Boost, HughesNet have recognition but compete largely on price/coverage). Switching costs are low-to-moderate: Pay-TV has habit/equipment friction (real but fading); Boost prepaid has near-zero switching cost; Hughes consumer has none (customers leave for Starlink). (Interpretation.)

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Yes — the SpaceX equity stake (to be received, carried at $212 cost / ~$11B, arguably worth $22–32B at recent marks) and the retained spectrum (carried at historical value; market value undisclosed, Auction 113 quiet period). Both are sources of hidden value. Conversely, the regulatory-authorization line ($34.5B) overstates retained-spectrum value since most is being sold. (Fact/Interpretation.)

Off-balance-sheet liabilities? Minimum volume commitments under the T-Mobile MNSA and AT&T NSA (wholesale wireless capacity); operating/tower leases (with active force-majeure litigation where EchoStar stopped paying); decommissioning obligations (~$5–7B combined with taxes). (Fact.)

How conservative is the accounting? Mixed. The $17.6B impairment was taken aggressively and promptly (conservative). But the company uses a non-standard “OIBDA”/“Adjusted OIBDA” metric, capitalized large amounts of construction interest until Aug-2025 (which flattered prior-period interest expense), and relies on intercompany PIK loans. The cost-method carrying of the SpaceX stake at $212 is conservative relative to marks. (Interpretation.)

How CapEx-hungry is the business? Historically very (the ~$10B+ 5G build); prospectively low. With the 5G network abandoned and satellites already in orbit, FY2025 PP&E capex was ~$966M and falling. RemainCo is relatively capital-light — part of management’s “asset-light” framing. (Fact.)

Capital Allocation & Management

How much FCF, and how is it used? Currently negative (−$1.74B FY2025) — the business consumes cash after interest. Going forward, the ~$24B of spectrum proceeds is the relevant capital, to be deployed via “EchoStar Capital.” Management’s stated philosophy: pay down expensive/maturing debt, manage tax, and weigh investments vs. returning capital to shareholders — “dynamic and interrelated,” explicitly undecided. (Fact.)

Significant acquisitions recently? The reverse: significant divestitures (~$42B of spectrum). The last major acquisition was the 12/31/2023 DISH recombination. (Fact.)

Buying back shares? Minimal — $48.5M of Class A repurchases (1.79M shares) in FY2025, immaterial. A large buyback below SOTP would be a positive future signal but has not occurred. (Fact.)

Issuing shares to insiders? No large issuance; option exercises are routine. Ergen’s activity is estate planning (GRATs, entity transfers), not new issuance. (Fact.)

Compensation / incentive alignment; motivations of management? Ergen controls ~51% of equity and ~90% of votes — his economic alignment with shareholders is total in direction, but minority holders cannot check him, and recent estate-planning transfers suggest succession/wealth-transfer is on his agenda. (Interpretation.) Hamid Akhavan (CEO, EchoStar Capital) runs the redeployment.

Valuation & Market Data

ADR, MLP, or K-1 issuer? No — a standard US C-corp (Delaware), NASDAQ-listed, multi-class common. No K-1. (Fact.)

Dividend policy? No dividend; none expected near-term given the debt wall and going-concern status. (Fact.)

Is net income diverging from cash from operations? Yes, massively — FY2025 net income −$14.5B vs. OCF −$0.1B, the gap being the $17.6B non-cash impairment and a $4.4B non-cash deferred-tax benefit. Both are non-cash; the cash reality is the −$1.74B FCF burn. (Fact.)

Risks & Downside

What would cause the stock to decline? A deal break or repricing toward the $18.6B floor; a going-concern event before the AT&T cash; a SpaceX down-round/IPO below $212 or stake dilution; a high tax/decommission bill; value-neutral redeployment of the cash; an accelerated Pay-TV decline. (Interpretation.)

Risk of catastrophic / total loss? Low but non-zero. A total loss would require both the AT&T and SpaceX legs to fail before liquidity, leaving the 2026 debt wall unbridged and forcing a restructuring that wipes the (controlled) equity. The DISH DBS and HSSC silos add bankruptcy-remoteness complexity. Base case is bankruptcy avoidance once the AT&T cash lands. (Interpretation.)

Recent News & Events

Has the business environment changed recently? Profoundly — see. The FCC’s mid-2025 revocation threat, the AT&T ($22.65B) and SpaceX (~$20B) sales, the 5G abandonment and $17.6B impairment, the Hybrid-MNO transition, the going-concern qualification, and the creation of EchoStar Capital all occurred within roughly nine months. This is one of the most transformed large-cap situations in the market. (Fact.)

Significant acquisitions / accounting-policy changes / new markets? Divestitures (spectrum), not acquisitions; the cessation of interest capitalization (Aug-2025) materially raised reported interest expense; the “Other” segment was carved out to isolate the dead 5G build. (Fact.)

APPENDIX B — Source Appendix

EchoStar Corporation (NASDAQ: SATS) — Research Sources, as of 2026-06-12

All financial figures reconciled to primary SEC filings. Third-party aggregator data (AZI, yfinance) was used only for orientation and was not relied upon for valuation, as the impairment-driven GAAP loss makes standard multiples meaningless.

Primary — SEC filings (EDGAR, CIK 0001415404)

1. EchoStar Corporation FY2025 Form 10-K — filed 2026-03-02 (accession via sats/tmb-20251231x10k.htm). Sections relied upon:
   • Auditor’s report (KPMG LLP, Denver; going-concern qualification), F-2/F-3.
   • Consolidated Balance Sheet (F-5), Statement of Operations (F-6), Cash Flows (F-7/F-8).
   • Note 1 — Organization, going concern, impairments, AT&T/SpaceX transactions, SpaceX interim debt service.
   • Note 10 — Debt (carrying-value table by issuer, the $326M skipped 10.75%-note interest), F-67–F-71.
   • Note 11 — Income taxes (OBBBA, deferred-tax benefit).
   • Note 15 — Debt maturities; Omega license exchange.
   • Note 16 — Segment reporting (revenue, OIBDA, Adjusted OIBDA), F-110–F-114.
   • MD&A — subscriber/ARPU/churn tables, SAC, FCF definition, FCC May-9/Sept-8 chronology, “elected not to make” interest disclosure (~p.73–100).
   • Item 1A risk factors; capital-stock item and cover (share counts by class, Ergen ownership ~p.56).
2. EchoStar Corporation Q1-2026 Form 10-Q — filed 2026-05-11 (sats-20260331x10q.htm): cash $1,343,780K, current debt $6,237,306K, accrued interest $626,229K, net loss −$147,300K, operating income +$392,847K, interest expense −$592,660K; going-concern persistence; subscribers 7.527M Wireless / 6.632M Pay-TV.
3. Prior-year Forms 10-K (FY2021–FY2024) and 10-Qs — for multi-year revenue, subscriber, ARPU, churn, debt, and equity trends (mirrored in output/SATS/sources/).
4. Form 4 corpus — 159 filings, June 2021 – June 2026 (insider-transaction sweep; no code-P open-market purchases identified).
5. Form 144s — June 2026 (proposed insider sales, option-exercise sourced).
6. Schedule 13D/A — filed 2026-05-15 (accession 0001104659-26-062583; Ergen-group estate-planning amendment).
7. EDGAR submissions JSON — data.sec.gov/submissions/CIK0001415404.json (filing index, form-type enumeration).

Primary — Earnings & event transcripts (AZI feed; mirrored to output/SATS/transcripts/)

8. Q4/Year-End 2025 Earnings Call — 2026-03-02 (transcriptid 3667583): SpaceX stake (~2.8% / $400B), xAI ~80/20, no anti-dilution, “don’t have the equity yet,” $5–7B decommission+tax, EchoStar Capital framing, capital allocation “dynamic… interrelated.”
9. Q3 2025 Earnings Call — 2025-11-06 (3582106): AWS-3 amendment ($2.6B all-stock @ $212), AWS-3 paired “most valuable,” §1033 tax theory, proceeds held at EchoStar Capital.
10. Special Call — 2025-09-15 (3549129): RemainCo pro-forma cap structure (~$24B cash, ~$13B debt, ~$8.5B SpaceX equity), Hybrid-MNO mechanics, Hughes enterprise pivot, ~$1.8B aviation backlog, “asset-light growth company.”
11. AT&T Inc. Special Call — 2025-08-26 (3540037): close “around the middle of 2026,” ~36 months to 2.5x leverage; regulators want fallow spectrum deployed.
12. Q2 2025 Earnings Call — 2025-08-01 (3512751); plus the historical earnings-call series (2011–2025) for capital-allocation and 5G-buildout history.

Secondary — external valuation cross-checks (accessed 2026-06-12)

13. SpaceX July-2025 tender / secondary at ~$212/share (~$400B) — the strike for EchoStar’s stake consideration.
14. SpaceX December-2025 secondary at ~$421/share (~$800B): Fortune, 2025-12-13; SatNews, 2025-12-06.
15. Post-xAI-merger SpaceX mark at ~$604/share (~$1.03T): Forge Global, April 2026; SpaceX 2026 IPO chatter at $1–1.5T.

Analytical frameworks & peer cross-reads

16. Analytical frameworks: Greenwald & Kahn, Competition Demystified (barriers to entry, the three advantage types, share-stability/ROIC tests); Marathon Asset Management, Capital Returns (supply-side capital-cycle analysis; the 5G build as a capital-cycle value-destruction case).
17. Peer analysis: AST SpaceMobile (direct-to-device / spectrum-stake valuation framing) and Charter Communications (Pay-TV decline, cable-MVNO economics).

Data reliability notes

• Third-party data aggregators’ multi-period statement arrays were unreliable for this recently-restated, multi-class, post-merger filer; all financial figures in this article are reconciled to EDGAR filings (10-K/10-Q), not to aggregator data.
• Standard valuation multiples (EV/EBITDA ~41x, P/E n/m, profit margin −98%) are artifacts of the impairment-driven GAAP loss and were explicitly not used in valuation.
• Share count: common screening sources report sharesOutstanding of 158.5M (Class A only); total economic shares (~288.9M = 157.5M Class A + 131.3M Class B) were taken from the 10-K cover and Q1-2026 weighted-share count.