MDA Space

Phase A — Understand the business

Lens 1 · Company Overview

MDA Space is Canada's national space champion — a ~55-year-old prime contractor for spacecraft, robotics, and geospatial intelligence, best known to the public as the maker of the Canadarm (Shuttle), Canadarm2/Dextre (ISS), and now Canadarm3 (lunar Gateway). It is not a "robotics company" in the industrial-automation sense the coverage bucket implies; it is a space-systems prime whose three business areas are:

1. Satellite Systems — designs and manufactures satellites, increasingly whole LEO constellations, built on its proprietary MDA AURORA digital software-defined satellite platform (digital beamforming). This is now ~68% of revenue and the entire growth engine. FY2025 revenue CA$1,109.5M, +85.5% YoY.
2. Robotics & Space Operations — space robotics (Canadarm3, the commercial MDA SKYMAKER robotics line derived from it), ISS operations support, lunar rovers/landers, satellite servicing, in-space assembly. FY2025 revenue CA$309.3M, +10.5% YoY.
3. Geointelligence — Earth-observation satellites and the data/analytics business, anchored by the CHORUS SAR (synthetic aperture radar) constellation (successor to RADARSAT-2) launching late 2026, plus the new MDA MIDNIGHT space-domain-awareness/defense platform. FY2025 revenue CA$214.4M, +6.1% YoY.

How it makes money: long-dated, milestone-based prime contracts — mostly cost-plus/fixed-price government and commercial constellation builds (design → manufacture → assemble → integrate → test). Revenue is recognized over time against percentage-of-completion, which is why a swelling backlog "converts" into revenue over several years. This is a project/backlog business, not a recurring-subscription business — the exception being the nascent CHORUS geospatial-data revenue, which is subscription-like and being pre-sold (41 early customer commitments before launch).

Key customers: Telesat (Lightspeed constellation), Globalstar/Amazon (D2D constellation), Canadian Space Agency (Canadarm3, CHORUS/RADARSAT replenishment), NASA (Gateway), Airbus (OneWeb antennas), and — pending — the US DoD via the Blue Canyon acquisition. Customer concentration is high and lumpy: two commercial constellation programs (Telesat, Globalstar) drive the Satellite Systems surge.

Contract-structure caveat (payment terms): these are not take-or-pay annuities. As the 2025 EchoStar cancellation proved, a commercial customer can walk — MDA's protection is milestone billing plus contractual termination-cost recovery, not guaranteed revenue. Concentration risk is real (see Lens 13).

Governance: MDA emerged from Maxar in 2020 via a Northern Private Capital-led Canadian buyout (CA$1B), IPO'd on the TSX April 2021, and completed a US IPO/NYSE dual-listing March 2026.

Lens 2 · Supply Chain

(Web-only — supply-chain.md in the KB is a generic robotics-topic file, not company-specific; names below are from primary reporting.)

Upstream inputs → MDA → end customer:

• Semiconductors / ASICs (the critical chokepoint). AURORA's digital beamforming depends on custom silicon. On the Q1 2026 call MDA flagged it "received production-ready chips for Telesat Lightspeed" — i.e. chip readiness gated the constellation ramp. Single-source advanced-node ASIC dependency is the tightest link in the chain; a foundry or design slip cascades into constellation delivery slips (and, per Globalstar, liquidated damages).
• Space-grade components — radiation-hardened electronics, RF front-ends, structures, propulsion, solar arrays — sourced from a specialized global supplier base (the same pool Airbus/Thales/Lockheed draw on). MDA's OneWeb antenna win with Airbus shows it also sits inside competitors' supply chains as a component/subsystem supplier, not only as a prime.
• Launch — MDA builds satellites, it does not launch them. The Globalstar 17-sat replenishment is a Rocket Lab-launched program; MDA's delivery is coupled to a third-party launch cadence.
• MDA (the integrator) — three main facilities (Montreal/Ste-Anne-de-Bellevue satellite manufacturing; Brampton, ON robotics; plus Geointelligence). The Blue Canyon acquisition (Colorado) adds a US smallsat bus/component line and — critically — a US-soil manufacturing footprint for ITAR/defense work MDA cannot do from Canada.

End customers / demand side: Telesat, Globalstar→Amazon, CSA, NASA, US DoD (pending BCT), Apple (indirect, via Globalstar D2D for iPhone SOS), commercial EO buyers (maritime, defense, mining, energy, forestry) for CHORUS.

Chokepoints & single-source dependencies:

• Advanced ASIC supply for AURORA — the single hardest bottleneck; ramp-rate-limiting.
• Skilled space-systems labour — MDA has hired thousands; the sector is talent-constrained.
• Customer solvency/strategy — the "supplier" MDA most depends on is a paying customer's continued commitment (EchoStar walked; Amazon's Globalstar takeover backstops the other).

Lens 3 · Competitive Advantages (moats)

MDA's moat is real but narrow and contested — it sits in the awkward middle between the primes and the disruptors.

Durable advantages:

1. Flight heritage + sovereign trust. 55 years, Canadarm as national iconography, and status as the Canadian space prime. For Canadian/allied-government work (Canadarm3, CHORUS/RADARSAT, defense) this is a near-captive position — a genuine regulatory/geopolitical moat. Governments buy sovereignty, not just satellites.
2. AURORA digital-beamforming platform. MDA claims an "industry first" software-defined satellite with on-board digital beamforming, productized across Telesat, Globalstar and (previously) EchoStar. If AURORA is genuinely a reusable, cost-advantaged bus, it's a scale/IP moat: win one constellation, amortize the platform across the next. The bull case is AURORA-as-platform.
3. Robotics IP (Canadarm lineage → SKYMAKER). Decades of flight-proven space-robotics know-how that is hard to replicate; the emerging commercial market (space stations, lunar, satellite servicing, in-space assembly) is a genuine option MDA is uniquely positioned to sell into.
4. Switching costs mid-program. Once a constellation build is underway on AURORA, re-competing to another prime is enormously costly — this protects the installed backlog (though not new awards).

Where the moat is weak (bargaining power):

• MDA needs the big customers more than they need MDA. Telesat and Globalstar each had alternatives (Airbus, Thales, Loral heritage, the smallsat crowd). EchoStar demonstrated a customer can cancel a CA$1.8B program with a strategy pivot and MDA is left with termination fees, not revenue. That is low bargaining power over marquee commercial customers.
• The disruptor squeeze. Astranis (10x cheaper/faster GEO), and vertically-integrated constellation owners who build in-house (SpaceX/Starlink, Amazon Leo/Kuiper) are eroding the merchant-satellite-builder TAM. If hyperscalers build their own buses, the addressable commercial-prime market shrinks structurally.

Net: MDA has a strong sovereign/heritage moat and a plausible platform moat, but limited pricing power against its largest commercial buyers. The moat protects the base and the government franchise; it does not guarantee the growth.

Lens 4 · Segments

(All `` — MDA press releases; no segments.csv on shelf.)

FY2025 revenue by business area:

Q1 2026 revenue by business area:

Read of the mix:

• Satellite Systems is the flywheel and the risk. +85.5% in 2025 was almost entirely Telesat Lightspeed + Globalstar work-volume ramp. The Q1'26 deceleration to +41% (vs +85% full-year) is the first sign the constellation-build S-curve is passing its steepest point — consistent with FY2026 guidance implying company-wide growth collapsing to ~10% (see Lens 5/11).
• Robotics & Geointelligence are the ballast — slower but higher-quality/stickier (government, and future CHORUS data subscriptions). The bull's "second act" is Geointelligence inflecting when CHORUS launches (late 2026) and SKYMAKER commercial robotics scaling — but neither is material yet.
• Geography: MDA does not cleanly break out revenue by geography in the summary releases (n/a at the segment×geography level). Directionally: heavily Canada + allied-government, growing US (Blue Canyon is an explicit US-market entry) and European ambition (IPO proceeds earmarked for US/European M&A). This is a North-America/allied revenue base, not global-commercial.

---

Phase B — Measure performance

Lens 5 · Earnings Result (latest print — Q1 2026, reported 2026-05-07)

(All `` unless noted.)

Top line: Revenue CA$464.1M, +32.2% YoY (Q1'25: CA$351.0M). Driven by Satellite Systems +41% (Telesat Lightspeed + Globalstar volume).

Profitability:

• Gross profit CA$115.2M; gross margin 24.8% (up from 22.7% Q1'25) — margin expansion, a genuine positive as programs mature.
• Adjusted EBITDA CA$90.6M, +32.1%; margin 19.5% (in the guided 18–20% band).
• Adjusted net income CA$50.7M, +32.0%; adjusted diluted EPS CA$0.38 (+26.9%).
• GAAP net income CA$29.6M, −10.0% YoY (Q1'25: CA$32.9M); GAAP diluted EPS CA$0.22, −11.5%.

The GAAP-vs-adjusted divergence is the single most important flag in the print. Adjusted metrics grew +32% while GAAP net income fell 10%. The wedge (CA$50.7M adj vs CA$29.6M GAAP = CA$21.1M of add-backs in one quarter, ~42% of adjusted NI) is the recurring forensic question (Lens 10) — it includes SBC, IPO/transaction costs, FX, and financing. A CA$0.38 adjusted vs CA$0.22 GAAP EPS gap means the "37–40x P/E" the market pays is on a number that is 73% larger than GAAP earnings. On trailing GAAP EPS (~CA$0.80 run-rate) the P/E is closer to ~70x.

Backlog & bookings — the yellow flag:

• Backlog CA$3.7B at 2026-03-31, down from CA$4.8B a year earlier. Management frames this as "strong conversion," which is partly true — but it also reflects the CA$1.8B EchoStar cancellation (Sep 2025) and, more tellingly, weak new bookings.
• Q1'26 order intake was only CA$143.9M against CA$464.1M of revenue → book-to-bill ≈ 0.31x. A prime burning backlog 3x faster than it replaces it. One weak quarter isn't a thesis-breaker (bookings are lumpy; the CA$688M CHORUS/RADARSAT and other awards land in specific quarters), but backlog has now fallen four consecutive quarters (CA$4.8B → CA$4.0B YE'25 → CA$3.7B) and that trend is the bear's exhibit A.

Balance sheet (post-US-IPO, strong):

• Cash CA$544.0M; net cash CA$299.3M (−0.9x net debt/EBITDA); total liquidity CA$1.2B; long-term debt CA$244.7M. The March 2026 US IPO added US$300M gross / net proceeds CA$441.5M incl. over-allotment (11,180,136 shares at US$30.50). This war chest funded the CA$874M Blue Canyon deal.
• Free cash flow guided "neutral to negative" for 2026 — capex CA$225–275M (constellation capacity + CHORUS) exceeds operating cash generation. This is a building company, not a harvesting one; don't expect FCF yield support for the multiple near-term.

Guidance / tone: FY2026 reaffirmed — revenue CA$1.7–1.9B, adj EBITDA CA$320–370M (18–20%), capex CA$225–275M, FCF neutral-to-negative. Midpoint revenue ~CA$1.8B on CA$1.633B actual = ~10% growth — a sharp deceleration from +51%. Management tone: confident on execution, explicit that quarter-to-quarter will be lumpy.

Market reaction: stock has risen through 2026 (to ~CA$58.80 by late June) despite decelerating growth — the market is paying up for the Amazon-Globalstar de-risking, the Blue Canyon US-defense pivot, and CHORUS optionality, not for near-term growth.

Lens 6 · Earnings Calls (sentiment trend)

(Web-only; Seeking Alpha transcript 403'd, synthesized from Via Satellite / SpaceQ / Investing.com call coverage.)

Recurring management themes (Greenley, CEO; Lavoie, CFO):

• "Disciplined execution" / "operational momentum" — the dominant frame every quarter; MDA sells reliability of delivery as the differentiator vs. cheaper/riskier disruptors.
• "$40 billion opportunity pipeline" — the forward-demand narrative that replaces the shrinking backlog in investor attention. Introduced/emphasized at FY2025; ~CA$10B of it is government down-selects or follow-ons. Blue Canyon adds ~US$3.5B more.
• "AURORA / digital" and "production-ready chips" — proof the platform is manufacturable at scale.
• "Defense" / "space domain awareness" (MIDNIGHT) — a deliberate pivot toward government/defense demand (structurally stickier, higher-margin, less cancellation-prone than commercial constellations). Blue Canyon is this pivot made concrete.

What management stopped saying / tone shift:

• EchoStar disappeared from the growth narrative after Sep 2025 — replaced by "our CA$4.6B backlog excluding EchoStar." The pivot from "commercial D2D land-grab" to "diversified defense + government + commercial" is a direct response to being burned.
• On Globalstar (the Q1'26 investor worry after the Amazon deal), Greenley was emphatic: "no change" to MDA's work — a confident, defensive tone answering a specific fear.

Sentiment trajectory (last ~4 prints): consistently upbeat on execution and margin, increasingly emphatic on de-risking (defense diversification, pipeline breadth) precisely because the growth rate is rolling over and two commercial customers (EchoStar cancelled, Globalstar acquired) shook confidence. Tone = confident operator managing a narrative transition from hypergrowth to durable-compounder.

Lens 7 · Comps

(Provenance-critical. Multiples are `` with source/date or n/a. Space is a hard comp set — pure-play listed satellite primes barely exist; peers are conglomerate segments or sub-scale disruptors.)

Read: On EV/EBITDA (~15–17x fwd) MDA looks reasonable-to-cheap for a 10%+ grower with 20% margins and a fortress balance sheet — this is the bull's "looks cheap for how fast it's growing" case. On P/E (~37–40x normalized, ~70x GAAP) it looks expensive for a business decelerating to ~10% growth. The comp verdict hinges entirely on which number you anchor on — and on whether you believe the adjusted EBITDA add-backs (Lens 10). Against sub-scale/loss-making disruptors (Terran, Rocket Lab, Astranis) MDA is uniquely profitable; against conglomerate primes there is no clean multiple. There is no listed pure-play twin — MDA is close to sui generis, which is itself why it commands a scarcity premium.

Lens 8 · Stock-Price Catalysts (moves >5%, ~5-yr pattern)

(All ``.)

Pattern: MDA trades on (1) constellation-contract wins/losses and (2) the fate of its handful of anchor customers. It is not primarily a macro or a quarterly-EPS name — it is a binary-contract-event and customer-concentration name. The single biggest historical shock (EchoStar, −20%) was a customer decision MDA "was caught off guard by", not an operational miss. That is the risk profile in one line: the tape reacts to whether the whales keep swimming.

---

Phase C — Judge people & books

Lens 9 · Management

(All ``.)

• Mike Greenley (CEO, since Apr 2020). ~6.25-yr tenure; ran MDA through the Maxar carve-out, the CA$1B Canadian buyout, the 2021 TSX IPO, and the 2026 US IPO. Named Globe & Mail "Innovator of the Year". Track record is genuinely strong on the metrics that matter here: doubled sales, won Telesat/Globalstar/Canadarm3 (~CA$1B)/CHORUS-RADARSAT (CA$688M), launched AURORA/CHORUS/SKYMAKER/MIDNIGHT product lines, and repatriated a foreign-owned asset into a national champion. This is a builder-operator who has actually delivered the growth, not a promoter.
• Skin in the game: Greenley directly owns ~0.17% (~CA$13.5M). Comp is CA$6.45M, 86.9% bonus / 13.1% salary — heavily variable/incentive-weighted, which is directionally good (pay-for-performance) but the equity ownership is modest for a founder-scale narrative — he is a professional manager, not a founder. Guillaume Lavoie is CFO.
• Ownership overhang: Northern Private Capital (John Risley / Andrew Lapham) led the buyout, with Jim Balsillie and Senvest as co-investors. The PE sponsor's ongoing stake and eventual exit is a supply overhang — worth tracking; a large secondary sell-down would pressure the stock independent of fundamentals.
• Capital-allocation history: thus far reinvestment-led (constellation capacity, CHORUS), now turning acquisitive — the US$620M all-cash Blue Canyon deal is the first major M&A test. It is strategically coherent (buys US-soil defense manufacturing MDA can't build fast organically; accretive to 2027 EBITDA/EPS per management ) but it also spends the IPO war chest and imports integration + goodwill risk. The next 18 months are the verdict on whether this management can buy as well as it can build.
• Red flags (management): none egregious. The heavy adjusted-vs-GAAP add-back culture (Lens 10) and the modest insider equity are the watch-items, not the related-party/promotional patterns that sink theses.

Archetype: capable professional-manager operator with a strong delivery record and a diversification strategy explicitly designed to fix the concentration flaw the market punished (EchoStar). The bet on management is a bet that the defense/US pivot lands.

Lens 10 · Forensic Red Flags

(Web-only — no filings on shelf; flags are analytical, sourced to reported metrics. Where a 10-K/6-K figure would normally ground this, it is n/a — not on shelf.)

1. Adjusted-vs-GAAP wedge (the headline forensic issue). FY2025 adjusted net income CA$189.9M vs GAAP CA$108.5M — a CA$81.4M / 43% gap. Q1'26: adjusted CA$50.7M vs GAAP CA$29.6M (42% gap), and GAAP NI fell 10% while adjusted rose 32%. The add-backs (SBC, IPO/transaction costs, FX, financing/interest) are large and recurring. Every valuation multiple quoted in the market (P/E, EV/EBITDA) is on the flattering adjusted number. Requires reading the actual 6-K reconciliation (not on shelf) to judge how much is truly one-time vs. structurally recurring dilution/financing. This is the number-one thing a forensic reader must pull from the filings before underwriting the multiple.
2. Backlog quality & burn. Backlog CA$4.8B → CA$4.0B → CA$3.7B over four quarters, with book-to-bill ~0.31x in Q1'26. A prime's backlog is its forward revenue — sustained burn without replacement is the leading indicator of a growth cliff. The CA$1.8B EchoStar cancellation shows backlog can also evaporate non-operationally. Watch bookings each quarter more than revenue.
3. Revenue recognition on POC constellation contracts. Percentage-of-completion accounting on multi-year fixed-price constellation builds is inherently estimate-laden (cost-to-complete assumptions drive recognized revenue and margin). A cost overrun on Telesat/Globalstar (or the chip-supply slip that already caused a Globalstar delay) could force margin true-downs. n/a on the specific contract-loss provisions without the filings.
4. Working capital / cash-flow-vs-earnings divergence. FCF guided neutral-to-negative in 2026 while adjusted EBITDA is CA$320–370M — the gap is capex + working-capital build on constellation programs. Not a red flag per se (it's a building phase), but it means earnings are not converting to cash near-term — receivables/contract-asset and inventory build on POC contracts is the place to watch for the classic "earnings outrunning cash" pattern.
5. Liquidated-damages / penalty exposure. The Globalstar 17-sat replenishment was delayed to mid-2026 and Globalstar was reportedly set to impose liquidated damages. Delivery slips on fixed-price programs carry direct penalty cost — a margin risk sitting inside the "strong execution" narrative.
6. Goodwill/intangibles (forward). Blue Canyon at US$620M all-cash will add material goodwill + acquired intangibles; integration and any future impairment is a 2027+ watch-item.

Regulatory findings (required sub-section):

• SEC (EDGAR): MDA had no CIK at the time the research-layer regulatory scan ran (2026-07-06), so no SEC Litigation Releases or AAERs are indexed. (Note: MDA acquired a CIK via its March 2026 US listing and now files 6-Ks; a future scan should re-run against CIK 0001857047 — but no EDGAR enforcement history exists for a company that only just began filing.)
• Non-SEC enforcement (web): No material FTC/DOJ/FDA/CFPB/competition enforcement, consent decrees, fines, or penalties surfaced for MDA Space in web search as of 2026-07-06. As a government space contractor MDA operates under ITAR/controlled-goods regimes; no public sanction found.
• Litigation/contract disputes: The EchoStar cancellation was a contractual termination (MDA was compensated for termination costs/fees per contract) not litigation; the Globalstar liquidated-damages matter is a commercial contract remedy, not regulatory action. Item 3 (Legal Proceedings) disclosure n/a — 10-K/AIF not on shelf.
• Verdict: No material regulatory or legal enforcement findings — verified via SEC EDGAR EFTS (no CIK/0 findings), web search, as of 2026-07-06. Contract-level disputes (EchoStar termination, Globalstar LDs) are commercial, disclosed, and — for EchoStar — compensated.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (three fiscal years: FY2026–FY2028)

(Bottom-up from FY2025 actuals + FY2026 guidance. Output ``; inputs labeled. Reporting currency CAD.)

Anchors:

• FY2025 actual: revenue CA$1,633.2M, adj EBITDA CA$323.9M (19.8%), adjusted diluted EPS CA$1.46, GAAP EPS CA$0.84.
• FY2026 guidance: revenue CA$1.7–1.9B, adj EBITDA CA$320–370M, capex CA$225–275M.
• Consensus sanity checks: FY2026 EPS pulled down to ~CA$0.86 (GAAP) / adj ~CA$1.45; net income +15% expected next year. Blue Canyon accretive to 2027 adj EBITDA/EPS.
• Share count rising (~11.2M new shares from US IPO on a ~137M base ≈ +8% dilution).

Base-case logic: FY2026 is the air-pocket year — the constellation-build S-curve flattens (guidance ~10% rev, flat-ish adjusted EPS as dilution/interest offset EBITDA growth). FY2027–28 re-accelerate if three things convert: (a) Blue Canyon closes and delivers US-defense revenue + accretion, (b) CHORUS launches (late 2026) and the 41 pre-committed EO-data customers turn into recurring revenue, (c) the CA$40B+ pipeline converts new bookings back above 1.0x book-to-bill. Bear path: FY2026 ~CA$1.7B (low end), bookings stay <0.5x book-to-bill, backlog falls below CA$3B, CHORUS slips, EPS flat-to-down → the multiple compresses toward 20x on a decelerating name (~CA$29 ). Bull path: a new multi-billion constellation award (the pipeline delivers) re-ignites Satellite Systems, BCT + CHORUS beat, adj EPS ~CA$1.70 in FY2027 at a maintained ~35x → CA$60+.

Valuation frame: at ~CA$58.80 the market pays ~40x FY2026E adjusted EPS / ~70x GAAP for ~10% near-term growth. That only works if you underwrite the FY2027–28 re-acceleration and trust the adjusted number. The price already embeds the bull's second act.

Lens 12 · Bull vs Bear

Bull case. MDA is the only profitable, at-scale, listed pure-play space prime — a national champion with a fortress balance sheet (net cash, CA$1.2B liquidity), 20% EBITDA margins, and a genuinely differentiated digital-satellite platform (AURORA) proven across two live constellation programs. The growth "cliff" is a temporary S-curve pause, not a peak: a CA$40B+ pipeline (CA$10B government down-selected/follow-on) plus Blue Canyon's ~US$3.5B US-defense pipeline gives multiple shots at the next multi-billion award. Three secular tailwinds compound — (1) the LEO-constellation buildout (Telesat, Amazon-Globalstar, and more D2D constellations coming), (2) the defense/space-domain-awareness spend cycle (MIDNIGHT + Blue Canyon + allied-sovereignty demand), and (3) the commercial-robotics + Earth-observation-data second act (SKYMAKER for lunar/servicing/in-space assembly; CHORUS recurring EO data with 41 pre-launch customers). The Amazon-Globalstar takeover removed the biggest customer-solvency risk and Amazon committed to continue MDA's 50+ sat build. On EV/EBITDA (~15–17x) it is not expensive for the durability. Management has delivered — doubled sales, won the marquee programs, repatriated the asset, executed two IPOs. Earnings surprise potential: a single new mega-constellation award re-rates the whole Satellite Systems line.

Bear case (2–3 permanent-impairment risks + expectations).

1. The merchant-satellite-prime TAM is structurally shrinking. Constellation owners (SpaceX/Starlink, Amazon Leo/Kuiper) increasingly build in-house; disruptors (Astranis, 10x cheaper GEO) undercut the merchant model. If the hyperscalers vertically integrate, MDA's largest addressable market — building constellations for others — erodes independent of execution. Amazon owning Globalstar is double-edged: it backstops today's contract but plants a future in-house builder.
2. Customer concentration is an existential-scale risk, proven live. EchoStar cancelled CA$1.8B (option to CA$3.5B) with a strategy pivot and MDA "was caught off guard." Two-to-three programs drive the growth. Any one anchor pivoting, getting acquired-and-redirected, or slipping (Globalstar already slipped + faces liquidated damages) knocks 20%+ off the stock in a day — the tape has shown this.
3. The growth deceleration is real and the multiple hasn't discounted it. +51% → guided ~10%. At ~40x adjusted / ~70x GAAP EPS on a decelerating base, the expectations bar is set for the FY2027–28 re-acceleration to happen — if bookings stay sub-1.0x and backlog keeps bleeding, the re-rate reverses toward 20x.

Pre-mortem (18 months out, thesis broke): Backlog fell below CA$3B because no new mega-constellation award landed and CHORUS/EO-data monetization disappointed; Blue Canyon integration hit snags and US-defense revenue ramped slower than the ~US$3.5B pipeline implied; a Telesat or Globalstar program slipped again, triggering margin true-downs and liquidated damages; the adjusted-EBITDA add-backs drew scrutiny and the market re-based valuation onto GAAP. The stock is back in the low-CA$30s — exactly where EchoStar left it in Sep 2025.

Are multiples too high? On GAAP, yes (~70x for ~10% growth is indefensible without the second act). On adjusted EV/EBITDA, defensible only if you trust the add-backs and the pipeline conversion. The honest answer: the valuation is priced for success and offers little margin of safety at ~CA$58.

Contrarian view (what the market refuses to see): The market is treating MDA as a growth stock (paying 40x) while the company is quietly becoming a defense/government sovereignty compounder (Blue Canyon, MIDNIGHT, Canadarm3, CHORUS-for-defense). If that transition completes, the right comp is not high-multiple space-growth but mid-teens-multiple defense-prime — which means either the earnings must grow into the multiple or the multiple de-rates to the sector. Bulls and bears are both arguing about the wrong franchise: the interesting question is whether MDA can trade its cancellable-commercial-constellation revenue for stickier allied-defense revenue fast enough to justify the price. The EchoStar scar is what's driving the strategy, and the market hasn't fully repriced MDA as the lower-growth-but-more-durable business it is deliberately becoming.

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

The structural break in how MDA makes money: MDA is a merchant builder in a market whose biggest buyers are integrating vertically. Every constellation MDA wins today is a customer that, at scale, has an incentive to bring building in-house (SpaceX already does; Amazon-Kuiper does; Amazon now owns a Globalstar it could redirect to Amazon Leo's own manufacturing over time). The merchant-satellite-prime is the arms dealer in a world where the superpowers are building their own arsenals.

Revenue concentration — and what happens if it shifts: ~68% of revenue is Satellite Systems, itself dominated by two commercial constellation programs (Telesat Lightspeed, Globalstar). EchoStar already demonstrated the failure mode: CA$1.8B gone in a single strategy pivot MDA didn't see coming, −20% in a day. Now model Telesat (which has its own funding/competitive pressures) slipping, or Amazon rationalizing the Globalstar D2D roadmap toward its own Leo constellation post-2027-close. Two customers = the whole growth story, and both have shown they can move.

Why the moat is weaker than bulls think: AURORA's "industry-first digital beamforming" is a lead, not a fortress — Airbus, Thales, Lockheed, and well-funded disruptors are all building software-defined buses; MDA's own OneWeb-antenna win for Airbus shows the tech is componentizable and the primes cooperate/compete fluidly. The durable moat is the Canadian-sovereign-government franchise (Canadarm3, CHORUS-RADARSAT, defense) — which is real but low-growth and budget-dependent. The high-growth commercial franchise is the contestable part.

Most dangerous competitor bulls underestimate: not the legacy primes — it's the vertically-integrated constellation owner (Amazon/SpaceX) who removes the need for a merchant prime entirely, and secondarily Astranis-class disruptors who reset the cost curve MDA's AURORA is competing on. Amazon is simultaneously MDA's customer (Globalstar) and its most dangerous long-term substitute (Leo/Kuiper in-house).

Worst capital-allocation risk: the US$620M all-cash Blue Canyon deal spends the entire US-IPO war chest on a Raytheon carve-out at a moment when MDA's own organic bookings are running sub-0.5x book-to-bill. If BCT's US-defense pipeline (~US$3.5B) converts slower than sold, MDA has swapped a fortress balance sheet for goodwill and integration risk to paper over a domestic-growth air-pocket. Buying growth is what companies do when they can't book it organically.

Assumptions that must hold for today's price (~CA$58, ~40x adj): (1) FY2027–28 re-accelerates to mid-teens revenue growth; (2) at least one new multi-billion constellation award lands from the pipeline; (3) Blue Canyon closes, integrates, and delivers accretion on schedule; (4) CHORUS launches on time (late 2026) and EO-data monetizes; (5) no further anchor-customer defection or major program slip; (6) the market keeps valuing on adjusted, not GAAP, earnings. That is six things that all have to go right for a business decelerating from 51% to 10%.

If growth disappoints by 20–30%: FY2027 revenue ~CA$1.7B instead of ~CA$2.05B, adjusted EPS flat ~CA$1.45 → at a de-rated 22–25x that's ~CA$32–36, a 40%+ drawdown from ~CA$58, with the EchoStar low (CA$35.25) as the obvious support test.

Single scenario that permanently impairs the business: the LEO-constellation buildout the whole growth story rests on consolidates into a handful of vertically-integrated hyperscaler-owned systems (Starlink, Amazon Leo, one or two others) that build in-house — leaving MDA with a Canadian-government franchise (~CA$0.5–0.7B, low-growth) plus a shrinking merchant-commercial book. Plausibility: medium over 5+ years — it is the direction of travel of the industry, and Amazon buying its own D2D operator (Globalstar) is a concrete step down that path. Not imminent, but not tail-risk either.

Lens 14 · Management Questions (15, ordered by information value)

1. Book-to-bill was ~0.31x in Q1'26 and backlog has fallen four straight quarters to CA$3.7B — what specific new awards from the CA$40B pipeline do you expect to convert to signed backlog in the next 12 months, and what book-to-bill do you need to sustain the FY2027–28 growth you're implicitly guiding the market toward?
2. Walk me through the CA$81M FY2025 gap between adjusted (CA$189.9M) and GAAP (CA$108.5M) net income line by line — how much of the add-backs (SBC, financing, FX, transaction costs) is genuinely non-recurring versus structural, and why should we value the company on adjusted rather than GAAP?
3. Amazon now owns Globalstar and is building Amazon Leo in-house — over a 3–5 year horizon, does Amazon remain a customer for MDA-built satellites, or does it become a competitor that internalizes constellation manufacturing? What have they contractually committed to beyond the current 50+ satellite build?
4. EchoStar (CA$1.8B, option to CA$3.5B) was cancelled with a strategy pivot you said you were "caught off guard" by. What have you changed in customer selection, contract structure (non-cancellable minimums? larger termination fees?), and concentration limits to prevent a repeat?
5. On Blue Canyon: what is the specific US-defense revenue and margin ramp you underwrote at US$620M, what book-to-bill does BCT bring, and what is your integration risk plan given it's a Raytheon carve-out onto a Canadian parent under ITAR?
6. What percentage of the CA$40B pipeline is (a) firm government down-select, (b) commercial-competitive, (c) early/speculative — and what's your realistic historical win-rate by category?
7. The merchant-satellite-prime model faces vertical integration by constellation owners (SpaceX, Amazon). What is your structural answer — why will there be a large, growing market for someone else to build constellations in 2030?
8. AURORA is your platform moat — how much of the bus is genuinely reusable/amortizable across programs versus re-engineered per customer, and how defensible is the digital-beamforming lead against Airbus/Thales/Lockheed software-defined buses?
9. Globalstar's 17-satellite replenishment slipped to mid-2026 with liquidated damages reportedly imposed — what caused the slip (chip supply?), what did it cost, and what's your confidence in the delivery schedule for the larger 50-satellite build?
10. FCF is guided neutral-to-negative in 2026 on CA$225–275M capex — when does MDA turn structurally FCF-positive, and how do you fund the next constellation-capacity build if a large new award lands?
11. Your advanced-ASIC supply gated the Lightspeed ramp ("production-ready chips") — how single-sourced is that silicon, and what's your second-source or inventory strategy against a foundry slip?
12. CHORUS launches late 2026 with 41 pre-committed EO-data customers — what recurring, high-margin data revenue does that represent at scale, and how does the Geointelligence segment's growth/margin profile change once it's a data business, not a build business?
13. What is Northern Private Capital's remaining ownership, and what is the expected timeline/mechanism for any sponsor sell-down that would create a supply overhang?
14. SKYMAKER commercial robotics targets space stations, lunar, and in-space servicing — what is the realistic revenue timeline, and how much of it depends on markets (commercial space stations, lunar economy) that don't yet exist at scale?
15. If the LEO-constellation market consolidates into a few vertically-integrated hyperscaler systems, what does MDA look like — and is the strategic endgame to be the allied-sovereign-government-and-defense prime, with commercial as opportunistic upside?

---