Abstract. On June 18, 2026, the Federal Energy Regulatory Commission (FERC) acted on Docket RM26-4-000 to establish, for the first time, a standardized federal framework for interconnecting very large electricity consumers — data centers above all — to the interstate transmission system. The proceeding originated in an October 2025 directive from the Secretary of Energy under Section 403 of the Department of Energy Organization Act, and was paralleled by a December 2025 order (193 FERC ¶ 61,217) compelling the PJM Interconnection to create new tariff services for co-located load. Together these actions compress and standardize interconnection study procedures, formalize the co-location of load with on-site generation, and propose to assign the full cost of network upgrades to the connecting load. This paper situates the reform within the institutional architecture of U.S. transmission and tests a single proposition: that interconnection procedure is not the binding constraint on the rate at which firm, interconnected electrical load can be added, but rather a comparatively minor term in a system governed by physical manufacturing, transmission construction, generation adequacy, and capital-allocation constraints. Using PJM — the regional grid hosting the densest concentration of data-center load in North America — and drawing on interconnection-queue census data, three years of capacity-auction results, the North American Electric Reliability Corporation's reliability assessment, equipment lead-time surveys, and the FERC primary docket, we decompose the speed-to-power problem and quantify each constraint. We find that the procedural stage the reform targets occupies the shortest segment of a five-to-seven-year critical path; that the reform's most consequential effect is therefore distributional (through cost allocation) rather than temporal; and that the binding constraints — gas-turbine and transformer lead times measured in years, a single domestic source of grain-oriented electrical steel, and a capacity market that has now cleared below its reliability requirement for the first time in its history — lie almost entirely outside the reform's reach. The reform is necessary; it is not, by itself, sufficient.

Keywords: electricity markets; transmission interconnection; data centers; capacity markets; cost allocation; resource adequacy; PJM; FERC. JEL classification: L94, Q41, Q48, L51, K23.

1. Introduction

1.1 An electricity-demand shock without modern precedent

For roughly two decades, U.S. electricity consumption was essentially flat; utilities planned around demand growth of well under one percent per year. That regime has ended, and the proximate cause is computation. Lawrence Berkeley National Laboratory's congressionally-mandated assessment documents U.S. data-center electricity consumption rising from 58 terawatt-hours (TWh) in 2014 to 176 TWh in 2023 — 4.4% of national consumption — and projects 325 to 580 TWh by 2028, or 6.7% to 12% of all U.S. electricity (Shehabi et al., LBNL, 2024). The Electric Power Research Institute's 2026 scenarios place the 2030 figure higher still, at 9% to 17% (EPRI, 2026). On a global basis, the International Energy Agency estimates data-center consumption growing from about 415 TWh in 2024 to roughly 945 TWh by 2030 — a doubling — and notes that by 2035 U.S. data centers alone will consume more electricity than the production of aluminum, steel, cement, and chemicals combined (IEA, Energy and AI, 2025).

The defining feature of this load is not only its magnitude but its velocity. An artificial-intelligence training campus is conceived, financed, and brought toward commissioning on a horizon of eighteen to thirty months, while the transmission system that must serve it is planned, permitted, and built over five to ten years. The mismatch between the clock speed of compute and the clock speed of the grid is the organizing tension of contemporary U.S. electricity policy, and it is the problem to which Docket RM26-4-000 is addressed.

1.2 The policy response

On October 23, 2025, Secretary of Energy Chris Wright invoked Section 403 of the DOE Organization Act — a seldom-used authority permitting the Secretary to direct FERC to consider a rulemaking — to instruct the Commission to ensure "the timely and orderly interconnection of large loads to the transmission system," citing AI data centers as an "urgent issue." FERC transmitted the resulting Advance Notice of Proposed Rulemaking (ANOPR) on October 27, 2025 (Accession 20251027-3056), opening Docket RM26-4-000. More than 3,500 pages of comments followed. On April 16, 2026, FERC issued a procedural "Order Regarding Intent to Act," committing to a final action by the end of June 2026 in a manner "quick, efficient, and legally durable," and on June 18, 2026, the Commission took up the matter as Agenda Item E-1.

In parallel, on December 18, 2025, FERC issued a unanimous order — PJM Interconnection, L.L.C., 193 FERC ¶ 61,217 (2025) — finding PJM's tariff "unjust and unreasonable" for lacking clear terms for co-located load, and directing the creation of three new transmission-service constructs. The political framing of both actions is straightforward: regulatory friction is the obstacle to powering the AI build-out, and removing it will unlock supply.

1.3 Thesis and contributions

This paper subjects that framing to scrutiny. Our central proposition is that the procedural reforms, although sound and overdue, operate on the least binding constraint in the speed-to-power system. We make five contributions. First, we provide an integrated account of the reform and its institutional antecedents (Sections 2–3). Second, we develop a constraint-decomposition framework that locates interconnection procedure within a broader critical path (Section 4). Third, we assemble the empirical record — queue economics, demand projections, the PJM case study, and the equipment supply chain — and quantify each constraint (Sections 5–6, 9). Fourth, we analyze the distributional economics of participant funding, which we argue is the reform's most consequential feature (Section 8). Fifth, we develop scenarios and a research agenda (Sections 11–12). The analysis is positive rather than prescriptive, and deliberately avoids security-specific or investment recommendations.

1.4 Roadmap

Section 2 establishes the institutional background. Section 3 details the reform and its primary record. Section 4 develops the conceptual framework. Section 5 presents the queue evidence. Section 6 is the PJM case study. Section 7 addresses co-location and the Susquehanna precedent. Section 8 analyzes cost allocation. Section 9 examines generation adequacy and the supply chain. Section 10 synthesizes; Section 11 offers scenarios and sensitivity; Sections 12 and 13 address limitations and conclude. An appendix provides a glossary and methodology note.

2. Institutional Background: The Architecture of Interconnection

2.1 Generators, loads, and the pro forma agreement

Across roughly two-thirds of the United States, the bulk power system is administered by independent system operators and regional transmission organizations (ISOs/RTOs) operating under FERC-jurisdictional tariffs. Historically, interconnection reform has concentrated on generators: a power plant seeking to inject energy executes an interconnection service agreement (ISA) whose terms derive from a federal pro forma standard established in Order No. 2003 (2003). Agreements that deviate from that pro forma — "non-conforming" agreements — bear a high justificatory burden, requiring demonstration of reliability concerns, novel legal issues, or unique operational factors. This doctrinal point, seemingly technical, becomes decisive in the co-location analysis of Section 7.

2.2 Order No. 2023 and the generator queue

By the early 2020s, generator interconnection queues had become severely congested, dominated by speculative renewable and storage projects. FERC's Order No. 2023 (2023) responded by replacing the serial, "first-come, first-served" study process with a "first-ready, first-served" cluster-study approach, imposing commercial-readiness deposits and study deadlines to deter non-viable applications. Crucially for the present analysis, Order No. 2023 reformed the generator queue. It did not establish a comparable national framework for large loads, which had long been treated under the retail jurisdiction of states and the bespoke practices of individual utilities.

2.3 The procedural gap and the jurisdictional question

It is precisely this gap that RM26-4-000 targets. A hyperscale data center is, electrically, a large load rather than a generator; yet a 500 MW load can stress a transmission network as severely as a 500 MW plant. In the absence of a standardized federal large-load interconnection process, treatment varied by jurisdiction, producing uncertainty, litigation risk, and delay. Filling the gap, however, raises a contested legal question that runs through the entire proceeding: whether FERC possesses jurisdiction over load interconnection at all. The Department of Energy argues that the Federal Power Act, Section 201(b) — governing wholesale transmission access and practices "directly affecting" wholesale rates — supplies the authority; states and many utilities counter that load interconnection is a retail matter reserved to them (American Bar Association, 2026). As we discuss in Section 3, FERC's parallel PJM order suggests a strategy designed to sidestep, rather than resolve, this collision.

3. The Reform in Detail

3.1 Provenance and the six principles

RM26-4-000 originated not in an industry petition but in the Secretary of Energy's October 2025 directive. The ANOPR set out six reform principles (Mayer Brown, 2025; CSIS, 2025):

1. Standardized study deposits, readiness requirements, and withdrawal penalties, modeled on the generator-interconnection discipline of Order No. 2023, together with an "option to build" network upgrades;

2. A joint study process for hybrid facilities that co-locate load with generation;

3. System Support Resource / Reliability-Must-Run (SSR/RMR) studies when a large load co-locates with an existing generator, to capture the reliability consequences of diverting that generator's output;

4. An expedited interconnection study — potentially 60 days — for curtailable or dispatchable loads, an explicit reward for flexibility;

5. One hundred percent participant funding, under which the large load bears all assigned network-upgrade costs; and

6. Transition planning for projects already in study.

3.2 The 20 MW threshold and the federalism problem

The ANOPR provisionally defines a "large load" as one exceeding 20 MW, while expressly soliciting comment on whether that threshold is appropriate or should be eliminated. The figure is contested precisely because it is low: hyperscale facilities routinely exceed several hundred megawatts, so a 20 MW floor could "sweep in mid-sized manufacturers and facilities traditionally served under state tariffs, effectively federalizing retail matters" (CSIS, 2025). The comment record fractured along this line (Table 6): the Edison Electric Institute urged a floor of at least 100 MW; PJM's transmission owners argued for 200 MW; the R Street Institute opposed a size threshold entirely, proposing instead a test based on wholesale-market participation; and the Data Center Coalition supported the 20 MW figure and the expedited timelines.

3.3 Participant funding and cost causation

The reform's economic core is the 100% participant-funding model: the connecting large load pays the full cost of the network upgrades its interconnection triggers, on a cost-causation rationale. We analyze the welfare properties of this choice in Section 8; here we note only that it is a significant departure from the prevailing treatment of transmission as shared infrastructure recovered through regional rates — a structure under which, by one leading estimate, more than 90% of U.S. transmission investment is presently socialized rather than assigned to specific beneficiaries (Pfeifenberger / Brattle Group, 2024).

3.4 The PJM co-location order (193 FERC ¶ 61,217)

The December 2025 order is the reform's operational vanguard within PJM. Finding the PJM tariff "unjust and unreasonable" for lacking "clear rates, terms, and conditions of service" for co-located loads — a deficiency that produced "uncertainty" and "disparate treatment" — FERC directed PJM to create three new transmission-service constructs (Table 7) and to reform behind-the-meter generation (BTMG) rules it likewise found no longer just and reasonable, imposing a three-year transition, a materiality threshold, and grandfathering of existing contracts. Co-located loads must pay for Regulation and Black Start service on a gross demand basis, addressing the cost-shifting concern directly. Compliance filings were ordered for January 17 and February 16, 2026.

Chairman Laura Swett framed the order's animating purpose in consumer terms — "I don't want any American family to feel that anxiety because of anything that we do" — while Commissioner Judy Chang observed that "the broader challenge of reliably, efficiently, and fairly interconnecting large loads … remains before the Commission, the States, and industry." Commissioner Rosner's concurrence is analytically telling: it described the order as achieving "the ANOPR's goals without jurisdictional confrontation." This suggests that FERC's preferred strategy is to accomplish large-load reform through its unambiguous authority over wholesale transmission-service tariffs, rather than by asserting the legally novel and contested jurisdiction over retail load interconnection discussed in Section 2.3.

3.5 The June 18, 2026 action

At its June 18, 2026 open meeting, FERC acted on RM26-4-000 (Agenda Item E-1), advancing the large-load framework consistent with the ANOPR record — standardized study procedures, co-location pathways, an expedited-study incentive for curtailable load, and the participant-funding principle. Consistent with the Rosner concurrence, the instrument is best read as grounded in FERC's wholesale-service authority, preserving legal durability at the cost of comprehensive coverage of loads that never participate in wholesale markets. (The precise instrument — a notice of proposed rulemaking advancing toward a final rule, versus a direct final rule — and its operative provisions should be confirmed against the issued order; pre-meeting materials did not definitively resolve the instrument type.)

4. Conceptual Framework: Decomposing Speed-to-Power

4.1 The interconnection timeline as a critical path

We model the interval from a developer's decision to build to energization at scale as a critical path of partially overlapping stages. For a greenfield large load — or the generation that must accompany it — this interval typically spans five to seven years. Table 2 decomposes it and identifies which stages the reform reaches.

4.2 A taxonomy of constraints

It is useful to classify the constraints on speed-to-power into five categories. Procedural constraints are study queues and agreement execution. Physical constraints are the manufacturing and construction of transformers, turbines, and transmission lines. Supply-chain constraints are the upstream inputs — grain-oriented electrical steel, high-voltage bushings, skilled labor. Capital constraints are the financing and cost-allocation terms that determine whether a project proceeds. Reliability constraints are the engineering limits planners impose to keep the system stable. RM26-4-000 operates almost entirely on the procedural constraint, with a secondary effect on the capital constraint through cost allocation. It does not touch the physical or supply-chain constraints, and it bears on reliability only insofar as the expedited-study incentive presumes a flexibility that, as Section 7 shows, large loads have been reluctant to provide.

4.3 The marginal value of relaxing the procedural constraint

The policy-relevant quantity is the marginal reduction in total elapsed time from relaxing the procedural constraint while the others bind. If the procedural stage is the shortest element of the critical path — and Table 2 indicates it is — then compressing it yields a correspondingly small reduction in total time. The intuition is captured in the practitioner observation that the regulatory delay amounts to "a three-to-six-month edit on a five-to-seven-year wait": grid connections take years not because approvals take years, but because the manufacturing and construction of physical assets does. Figure 4 visualizes this asymmetry.

This is not an argument against the reform. Removing a discretionary, variable, and litigation-prone source of delay has genuine value: it improves cost discovery, reduces uncertainty, and materially accelerates the minority of projects for which the procedural stage is on the critical path — most importantly co-located projects that avoid network upgrades entirely (Section 7). The point is narrower and more precise: a reform confined to the procedural stage cannot, by construction, relax the physical and supply-chain constraints that dominate the remaining years of the path.

5. Queue Economics: The Evidence on "Shovel-Ready" Capacity

5.1 The scale of the queues

A recurring claim in public discourse — visible in the response to the very Bloomberg report that frames this paper — is that vast capacity sits "ready to connect." The empirical anchor is Lawrence Berkeley National Laboratory's Queued Up: 2025 Edition, which reports approximately 2,290 GW of generation and storage actively seeking interconnection at the end of 2024 — roughly 1,400 GW of generation plus 890 GW of storage across some 10,300 projects, nearly twice the entire installed U.S. generating fleet (Rand et al., LBNL, 2025). The peer-reviewed treatment in Joule reports a comparable figure of about 2,600 GW against an installed base of roughly 1,280 GW (Gorman et al., 2024). On its face, this supports the abundance narrative.

5.2 Attrition and the "ready to connect" fallacy

The interpretation collapses under the completion data. Of the capacity that requested interconnection between 2000 and 2019, only 13% had reached commercial operation by the end of 2024; approximately 77% had been withdrawn; and 10% remained active (LBNL, 2025). The Joule analysis, spanning more than 38,000 project-level observations, reports a withdrawal rate of roughly 78–80% (Gorman et al., 2024). Far from clearing, the funnel is lengthening: the median time from request to commercial operation for projects that energized in 2024 was 55 months (4.6 years), up from 22 months for the 2008 cohort and 36 months for the 2015 cohort — a near-tripling. And the attrition is not merely early-stage speculation: in 2024 a record 700 GW withdrew — exceeding the roughly 500 GW of new requests — with one-third of withdrawals occurring at the late-stage facility-study or interconnection-agreement phase, after the system had already incurred planning and study costs (LBNL, 2025).

A queue, properly understood, is not an inventory but a funnel with heavy attrition; a project with secured equipment, financing, transmission, and offtake is the exception. The headline gigawatt figure overstates deliverable near-term supply by nearly an order of magnitude once historical completion rates are applied.

5.3 The load–generation distinction

A further subtlety is decisive. The queue census describes generation. Data centers are load, and the large-load interconnection pathway is both procedurally distinct from and far less mature than the generator queue that Order No. 2023 reformed. Extrapolating generator-queue abundance to load-interconnection readiness is therefore a category error. In PJM specifically, the load side is already congested in its own right: the Independent Market Monitor reports more than 3,300 projects awaiting interconnection — the largest backlog of any RTO (Monitoring Analytics, 2025), and PJM's recent cycle drew applications representing roughly 220 GW, of which only about 21 GW had reached engineering/procurement and 8.2 GW were under construction as of January 2026.

6. The PJM Interconnection: A Case Study

6.1 Why PJM

PJM coordinates the bulk power system across thirteen states and the District of Columbia and administers the largest competitive wholesale electricity market in the world. It contains "Data Center Alley" in Northern Virginia, the single densest cluster of data-center load on the continent. PJM therefore offers the sharpest available test of how a grid absorbs concentrated, rapid large-load growth — and, as the data below show, the first system in which the strain has become acute enough to break a previously reliable market mechanism.

6.2 Demand: a forecast dominated by data centers

PJM's load forecast has been transformed in three years. The aggregate of utility ten-year demand forecasts for the region rose roughly six-fold, from 24 GW of projected summer growth in 2022 to 166 GW in 2025 (LBNL Large Load Literature Review, 2025). PJM's own 2025 Long-Term Load Forecast projects peak load growth of 32 GW between 2024 and 2030, of which data centers account for approximately 94%; over 2025–2040, data centers represent 422 TWh of a projected 605 TWh of annual load growth — about 70% (PJM, 2025). NERC's reliability assessment, applying a more conservative lens, still projects PJM summer peak rising by 56 GW and winter peak by 62 GW by the mid-2030s (NERC, 2026). The composition is the salient fact: large loads account for essentially all net peak growth, as conventional demand is flat.

It is worth noting that even these figures embed downward corrections. PJM tightened its data-center counting methodology to avoid double-counting speculative or duplicative requests — an admission that "announced" load materially overstates probable load, a theme we return to in Section 11.

6.3 The capacity market and the price signal

PJM secures resource adequacy through a forward capacity market, the Reliability Pricing Model (RPM), whose Base Residual Auction (BRA) procures committed capacity three years ahead. As surging demand projections met a tightening, retiring supply stack, the clearing price escalated across three consecutive auctions, as shown in Figure 1.

Three features of the 2027/28 auction (held December 17, 2025) deserve emphasis. First, the entire RTO — every locational deliverability area, and even the Fixed Resource Requirement areas — cleared at the price cap of $333.44/MW-day. Second, the auction procured 134,479 MW through the BRA plus 11,299 MW of FRR resources, for 145,777 MW combined, and still fell 6,623 MW short of the one-event-in-ten-years reliability requirement — the first such shortfall in PJM's history, leaving a reserve margin of 14.8% against a 20% target. Third, approximately 5,100 MW of the 5,250 MW year-over-year increase in the demand forecast — about 97% — was attributable to data-center load (PJM, 2025).

6.4 The price cap and the structural trap

The 2027/28 result reflects a price-collar settlement negotiated after Pennsylvania challenged the prior auction at FERC; the floor was set at $179.55/MW-day. A counterfactual simulation indicates the auction would otherwise have cleared at roughly $529.80/MW-day RTO-wide, and $542.83 in the Dominion zone — implying the cap suppressed the price by about 59% (Avalon Energy Services, 2026). This produces a structural trap. The cap was designed as a consumer protection for a market with surplus capacity; it now suppresses the very investment signal needed to close a 6,623 MW shortfall. The Independent Market Monitor has accordingly declared the capacity auctions for the 2025/26, 2026/27, and 2027/28 delivery years "not competitive," "in significant part as a result of forecast demand for data centers" (Monitoring Analytics, 2026). The market is simultaneously pricing scarcity at its ceiling and prevented from pricing it high enough to resolve it.

6.5 Consumer incidence and the wider market

The escalation reaches consumers through both capacity and energy channels. The IMM's 2025 State of the Market report records the full-year real-time locational marginal price rising 50.4%, from $33.74 to $50.73/MWh, and the total cost of wholesale power rising 48.9%, from $55.52 to $82.67/MWh, with the capacity-cost component alone up 262.3% year-over-year (Monitoring Analytics, 2026). Real-time load rose 3.7% in 2025, and PJM set new winter and summer peak records in the same year. These are not abstractions: as Section 8 details, the increases have already produced double-digit monthly bill increases across PJM states and have forced administrative intervention in a market mechanism.

6.6 Reliability and the supply side

NERC's 2026 Long-Term Reliability Assessment classifies PJM at "elevated risk," with the anticipated reserve margin falling below the reference margin level beginning in 2029; the 2026 summer anticipated reserve margin deteriorated from 35.7% in the prior assessment to 29.7% in a single year (NERC, 2026). NERC also revised its ten-year North American summer peak-demand growth forecast upward by 69%, from +132 GW to +224 GW. Notably, the greatest risk of unserved energy in PJM falls in winter, under below-normal temperatures with natural-gas infrastructure constraints — a reminder that resource adequacy is a seasonal and fuel-security problem, not merely a nameplate-capacity problem. An adversarial review argues NERC overstates the risk by excluding likely-to-connect queue resources, and that counting them would resolve most identified shortfalls (Grid Strategies, 2026) — a useful caution, but one that itself depends on the queue-completion assumptions Section 5 shows to be historically generous.

7. Co-location: The Escape Valve and Its Precedent

7.1 The Susquehanna proceeding

The cleanest way to avoid a congested transmission network is not to use it — the logic of co-location, in which a large load is sited at a generator and served directly. Its defining test was the arrangement between Talen Energy and Amazon at the 2.5 GW Susquehanna nuclear plant in Pennsylvania. In November 2024, FERC, by a 2–1 vote, rejected an amended interconnection service agreement that would have increased behind-the-meter supply to a co-located Amazon Web Services facility from 300 MW to 480 MW. The majority held that PJM had not met the Order No. 2003 burden of justifying a non-conforming agreement; the chairman dissented, characterizing the configuration as "first-of-a-kind" and warranting approval (Utility Dive, 2024). Talen sought review in the Third Circuit.

7.2 Net versus gross: the analytical crux

Co-location's central technical dispute concerns the basis on which a hybrid load-plus-generation facility is studied. Developers argue for a net basis — measuring grid impact by net injections and withdrawals — on the ground that on-site generation reduces system burden. Utilities and planners insist on a gross basis, arguing the grid must withstand contingencies when on-site generation fails, and that net accounting shifts reliability risk and cost onto other ratepayers (CSIS, 2025). The December 2025 order's requirement that co-located loads pay for Regulation and Black Start on a gross basis is a partial resolution in favor of the conservative view. The stakes are not trivial: an analysis of PJM territory finds that traditional firm interconnection of new large load imposes roughly $764 million per gigawatt in system supply costs, whereas a flexible "bring-your-own-capacity" approach — accepting curtailment of only 40–70 hours per year while maintaining grid availability above 99% — can reach full operation in about two years, three to five years faster than firm service (Camus Energy, via Institute for Progress, 2026).

7.3 The restructured template

Market behavior is instructive. In June 2025, Talen and Amazon restructured their arrangement into a long-dated power-purchase agreement — reported at roughly $18 billion over 17 years for up to 1,920 MW — delivered front-of-the-meter, through the grid rather than behind it. The restructuring preserved the economic substance while conforming to FERC's reliability and cost-allocation concerns, and the December 2025 order generalized the lesson by building standardized service options. RM26-4-000 carries that template to the national level. The episode illustrates both the appeal of co-location and the regulatory friction it must still navigate — friction that the reform is, genuinely, designed to reduce.

8. Cost Allocation: The Distributional Heart of the Reform

8.1 Cost causation versus beneficiary pays

Beneath the engineering lies a distributional question with a substantial economics literature. Two principles compete. Cost causation assigns upgrade costs to the party whose connection triggers them — the basis for participant funding, and the position defended on both legal and efficiency grounds by, among others, the R Street Institute (2024) and formalized in recent scholarship (Ribeiro et al., 2025). Beneficiary pays assigns costs in proportion to benefits received, recognizing that transmission, once built, lowers congestion and improves reliability for all users; it is the principle embodied in FERC's Order No. 1920 (2024), which requires that costs be allocated to beneficiaries "at least roughly commensurate with estimated benefits," subject to a benefit-to-cost threshold no greater than 1.25 to 1.

8.2 The efficiency case and its limits

Participant funding has a clean efficiency rationale: it confronts the connecting load with the social cost of its location decision, sharpens siting incentives toward unconstrained areas, and protects existing ratepayers from subsidizing private expansion. Yet even its proponents acknowledge the logic is incomplete. Transmission exhibits positive externalities and economies of scale; charging the first mover the full, lumpy cost of an upgrade that subsequently benefits many can deter socially valuable investment (CSIS, 2025). And the equity incidence is uneven: capital-intensive industrial loads — an electric-arc-furnace steel mill, for instance — "often cannot" absorb full network-upgrade costs in the way a hyperscaler can (CSIS, 2025), a concern the industrial commenters pressed directly (IECA/ELCON, 2025). FERC's Order No. 1920, with its beneficiary-pays orientation, sits in tension with a strict participant-pays model and frames the unresolved question.

8.3 Incidence: the cross-subsidy is already running

The most important empirical point is that the distributional stakes are not hypothetical. The cross-subsidy is already operating at multi-billion-dollar scale. PJM's Independent Market Monitor attributes 63% of the 2025/26 capacity-price increase — about $9.3 billion — to data-center load, and finds that across the last three base auctions data-center forecast load above existing levels accounted for $21.3 billion, or 45%, of the $47.2 billion in cleared capacity costs (Monitoring Analytics, 2025–2026). Separately, the Union of Concerned Scientists documents $4.4 billion in 2024-approved local transmission projects serving data centers — more than 150 projects across seven states, with nearly $2.2 billion in Virginia alone — socialized across all ratepayers (UCS, 2025). Because more than 90% of U.S. transmission investment is currently categorized as "reliability-driven" and rolled into broad zonal rates (Brattle Group, 2024), the existing structure provides almost no mechanism to trace these costs back to the load that triggered them.

8.4 The household bill

The incidence reaches households directly. Synapse Energy Economics, in studies commissioned by state consumer advocates, documents capacity-driven monthly bill increases of approximately $21 in Washington, D.C., $18 in western Maryland, and $16 in Ohio, beginning in June 2025 (Synapse, 2025), and projects that PJM residential bills could rise nearly 60% by 2036–2040 absent interconnection reform (Synapse / Evergreen Action, 2025). The NRDC estimates cumulative data-center-driven capacity costs of $100–163 billion through 2033 — about $70 per month for the average PJM household — with Dominion-zone ratepayers facing capacity fees 65% above the PJM average (NRDC, 2025). A White House "ratepayer protection pledge" signed by hyperscalers in March 2026 carries no enforcement mechanism; absent a binding FERC rule, it is, as one analysis observes, legally unenforceable (Latitude Media, 2026). Cost allocation is therefore not a peripheral feature of the reform but its most consequential lever.

8.5 Compromise designs

A middle path is emerging in the comment record and at state commissions: large loads fund upgrades upfront but receive partial refunds or credits if their facilities later deliver system-wide benefits. Such designs attempt to preserve the incentive properties of participant funding while internalizing the positive externalities that pure cost causation ignores — a pragmatic reconciliation of the two principles that bracket the debate.

9. Generation Adequacy and the Equipment Supply Chain

If the binding constraint is physical, the supply chain is where the analysis must conclude. The evidence here is the strongest in the paper, and it is the reason the procedural reform cannot, on its own, move aggregate speed-to-power.

9.1 Gas turbines

The fast-ramping gas capacity utilities would normally deploy to serve new load is effectively unavailable on the relevant horizon. GE Vernova's gas-power backlog grew from 83 GW to 100 GW in the first quarter of 2026 alone, with management guiding to at least 110 GW by year-end and stating that reservation slots will be "sold out through 2030" by the end of 2026 (GE Vernova Q1 2026 Form 8-K). New orders placed today are not deliverable until late 2028 at the earliest; Mitsubishi Power quotes 2028–2030; and Siemens Energy carries a record backlog of roughly €131 billion (about $148 billion), with three manufacturers serving more than 75% of projects under construction (RMI, 2025). S&P Global reports U.S. turbine wait times of as much as seven years, and procurement costs rising sharply — Duke Indiana's Cayuga project cleared at $2,340/kW, 36% above the prior year, a roughly $900 million overrun on a 1,476 MW facility (S&P Global, 2025).

9.2 Transformers and grain-oriented electrical steel

Transformers are the second binding constraint. Large power transformer lead times average roughly 120–160 weeks (about 2.5–3 years), with specialized units reaching 210 weeks; the U.S. Department of Energy's 2024 Report to Congress cites commonly quoted lead times of 36 months and maxima of 60 (DOE, 2024). Generator step-up transformers average about 144 weeks. The IEA reports that procurement now takes up to four years — double the 2021 baseline — with transformer prices up roughly 75% since 2019 and DC-cable lead times beyond five years (IEA, 2025). Demand for high-voltage transformers has risen 116% since 2019, and for generator step-up units 274%. The upstream chokepoint is acute: transformer cores require grain-oriented electrical steel, of which the United States has a single domestic producer (Cleveland-Cliffs), so that more than 80% of new large power transformers are imported (DOE, 2024; NIAC, 2024). Within the transformer itself, high-voltage bushings and on-load tap changers — specialized, application-certified components served by few qualified suppliers — are the most consistent build-schedule bottlenecks (IEA, 2025).

9.3 The systemic consequence and the price feedback

These constraints aggregate into a build-out that lags its own demand signal. Industry analyses estimate that more than half of planned U.S. data centers may be delayed or cancelled for lack of electrical equipment — of roughly 12 GW expected to come online in the U.S. in 2026, only about one-third was under active construction, the remainder awaiting equipment with three-to-five-year delivery (trade synthesis, 2025–2026). Wood Mackenzie projects the U.S. data-center electrical-equipment market expanding from about $20 billion to $65 billion by 2030, a tripling that manufacturing capacity cannot match in time. The feedback into prices is direct: the Brattle Group's 2025 Cost of New Entry study for PJM found that the scarcity of turbines, transformers, and switchgear drove the cost of a new gas-fired plant 43–46% higher than in the prior study (Brattle, 2025) — which is precisely why the capacity market of Section 6 cleared where it did. The binding constraint, in short, is not capital: collective AI-infrastructure capital expenditure for 2026 exceeds $650 billion. It is energized megawatts.

9.4 New firm supply: nuclear, SMRs, and on-site gas

The fastest path to firm power has become bringing one's own. This takes three forms. Nuclear restarts and uprates — the Three Mile Island Unit 1 / Crane restart contracted to Microsoft; Palisades; Duane Arnold — add carbon-free firm capacity but on multi-year timelines. Small modular reactors remain years from first commercial megawatts. On-site gas, frequently aeroderivative units used as "bridge" generation, is presently the quickest deployable firm option and the principal beneficiary of the co-location reforms — though it, too, is gated by the turbine constraint of Section 9.1, and the IEA cautions that reliable on-site gas service requires overbuilding generation by 30–70% relative to demand, with an estimated 15–27 GW of on-site gas potentially serving U.S. data centers by 2030 (IEA, 2025).

10. Synthesis

10.1 For policy

The reform correctly identifies, and fills, a genuine institutional gap: the absence of a standardized large-load interconnection framework. But policymakers should hold realistic expectations about its effect on aggregate speed-to-power. Standardizing procedure will not, by itself, materially accelerate the rate at which firm load can be added while equipment and transmission constraints bind. The higher-leverage levers lie upstream — in transformer, turbine, and electrical-steel manufacturing capacity; in workforce; and in transmission permitting and construction — none of which is within the reform's reach. The reform is best understood as removing a real but secondary friction, and as setting the terms (through cost allocation) on which the binding constraints will be financed.

10.2 For reliability

By easing co-location and behind-the-meter pathways, the reform shifts part of the resource-adequacy question off the centrally-planned grid and onto private arrangements. This can accelerate individual projects, but it raises the net-versus-gross question of Section 7 at system scale, and it presumes a load flexibility that hyperscalers have been reluctant to provide in practice. The conservative, gross-basis treatment in the December 2025 order mitigates, but does not eliminate, the risk that planning models understate firm-capacity needs — a risk made concrete by the first RTO-wide capacity shortfall in PJM's history.

10.3 Economic incidence

The reform's most consequential effect is distributional. Participant funding would reallocate the cost of network upgrades from the general ratepayer base toward the connecting load — a shift with efficiency merits and real equity costs, and one whose final calibration (including any refund-for-shared-benefit mechanism) will determine the incidence of a sum already running into the tens of billions of dollars. Given that the cross-subsidy is already documented at $21.3 billion across three auctions and $4.4 billion in socialized transmission, the cost-allocation question — not the study-timeline question — is where the reform's stakes are highest.

11. Scenarios and Sensitivity

Base case. RM26-4 standardizes large-load interconnection and modestly compresses study timelines; co-location accelerates a meaningful minority of projects; equipment and transmission lead times keep aggregate speed-to-power gated through the late 2020s. PJM capacity prices remain at or near the cap; behind-the-meter generation proliferates; the cross-subsidy debate intensifies as bills rise.

Upside. Aggressive participant funding plus standardized co-location catalyzes faster private build-out of generation alongside load, easing the network-upgrade burden; manufacturing capacity for turbines and transformers expands on the strength of multi-year backlogs; the supply curve bends by approximately 2028–2029, and capacity prices ease as committed supply arrives.

Downside — and it is well-supported. A demand air-pocket. Goldman Sachs estimates that only about 60% of announced data-center capacity will materialize on schedule in the near term, falling toward 50% over two years, and frames the resulting stranded-asset risk explicitly (Goldman Sachs, 2025); the IEA independently flags that around 20% of planned projects are at risk of delay (IEA, 2025); and PJM's own methodology tightening concedes that announced load overstates probable load. If AI capital expenditure moderates, or if efficiency gains compress per-query energy intensity faster than base cases assume, the committed generation, equipment orders, and participant-funded upgrades would face a softer demand signal, concentrating stranded-cost risk. Rhodium's high-growth scenario, conversely, places data centers at 14% of U.S. power by 2030 and raises system costs 13–15% by 2035 (Rhodium, 2026) — the band of outcomes is wide in both directions.

What would change the conclusion. The thesis — that procedure is not the binding constraint — would weaken if (i) equipment lead times compressed sharply through manufacturing expansion; (ii) co-location and behind-the-meter generation scaled fast enough to bypass the transmission constraint for a large share of load; or (iii) the procedural stage proved, on better project-level data, to occupy a larger share of the critical path than Table 2 assumes. Each is a testable empirical question, and each is the right place to look for evidence against the argument advanced here.

12. Limitations and Future Research

The stage durations of Table 2 are representative ranges, not point estimates; they vary by region, voltage class, and equipment type, and a more rigorous treatment would estimate them econometrically from project-level data. The capacity-price counterfactuals depend on the price-collar settlement and on PJM's evolving load-counting methodology. The precise legal instrument issued on June 18, 2026 governs the reform's ultimate effect, and compliance litigation — together with the unresolved Section 201(b) jurisdictional question — will shape implementation for years. Promising avenues for further research include: project-level survival analysis of large-load interconnection requests; estimation of the cross-elasticity between capacity prices and behind-the-meter adoption; welfare modeling of alternative cost-allocation designs, including refund-for-benefit mechanisms; and empirical measurement of realized large-load flexibility against the assumptions embedded in expedited-study eligibility.

13. Conclusion

FERC's large-load interconnection reform is sound policy directed at a real institutional gap, and it arrives at a moment of genuine urgency: PJM's capacity market has cleared at its cap three years running and, for the first time in its history, has cleared below the reliability requirement. But the reform should be understood for what it is — the relaxation of the shortest constraint in the speed-to-power system. The binding constraints are physical and financial: gas-turbine slots sold out to 2030, transformer lead times of two-to-four years resting on a single domestic source of electrical steel, a capacity market whose price cap suppresses the very signal needed to close a 6,623 MW shortfall, and tens of billions of dollars in cross-subsidy whose allocation the reform must still decide. The procedural fast-track is a necessary permission slip. It is not, by itself, the power.

References

• American Bar Association, Infrastructure & Regulated Industries. (2026). The Jurisdictional Collision over Large Loads and Data Center Interconnection. Spring 2026.

• Avalon Energy Services. (2026). PJM 2027/2028 Base Residual Auction Clears at Price Cap.

• Baker Botts LLP. (2025). FERC Issues Order Providing Guidance for "Co-locating" Power Plants with Data Centers within PJM.

• Brattle Group / Pfeifenberger, J. (2024). Transmission Cost Allocation for Order 1920 Compliance. NARUC–NASEO–DOE Webinar, Dec. 6, 2024.

• Brattle Group. (2025). 2025 Cost of New Entry (CONE) Report for PJM. April 2025.

• Center for Strategic and International Studies (Yang, A., Cai, R., Majkut, J., Zacarias, M.). (2025). What's at Stake in FERC's Large Load Proposal? Dec. 9, 2025.

• Federal Energy Regulatory Commission. Interconnection of Large Loads to the Interstate Transmission System, Docket No. RM26-4-000 (ANOPR transmitted Oct. 27, 2025; Order Regarding Intent to Act, Apr. 16, 2026; action June 18, 2026). Order No. 2003 (2003); Order No. 2023 (2023); Order No. 1920 / 1920-A (2024).

• Federal Energy Regulatory Commission. PJM Interconnection, L.L.C., 193 FERC ¶ 61,217 (Dec. 18, 2025); Fact Sheet and Commissioner concurrences (Chang, Rosner).

• Goldman Sachs Research / GS SUSTAIN. (2025). Data Center Power Demand: The 6 Ps Driving Growth and Constraints; and US Data Center Power Demand Projected to Double by 2027.

• Gorman, W., Mulvaney Kemp, J., Rand, J., Seel, J., Wiser, R., et al. (2024). Grid connection barriers to renewable energy deployment in the United States. Joule 9(2). DOI: 10.1016/j.joule.2024.11.008.

• Grid Strategies LLC. (2026). Review of NERC's 2025 Long-Term Reliability Assessment.

• Institute for Progress. (2026). Fast and Secure Grid Interconnection for American AI Leadership (citing Camus Energy PJM study).

• International Energy Agency. (2025). Energy and AI (World Energy Outlook Special Report); Building the Future Transmission Grid. (2026). Electricity 2026.

• Lawrence Berkeley National Laboratory (Rand, J., et al.). (2025). Queued Up: 2025 Edition. OSTI:3008763.

• Lawrence Berkeley National Laboratory (Shehabi, A., et al.). (2024). 2024 United States Data Center Energy Usage Report.

• Lawrence Berkeley National Laboratory. (2025). Large Load Literature Review, December 2025 Update; and Electricity Rate Designs for Large Loads (Jan. 2025).

• Mayer Brown LLP. (2025). FERC Large-Load Interconnection Preliminary Rulemaking: Key Takeaways. McGuireWoods LLP. (2025). FERC Establishes Proceeding to Consider DOE Directive.

• Monitoring Analytics, LLC (PJM Independent Market Monitor). (2025–2026). State of the Market Report for PJM — 2025 (Mar. 12, 2026); Q3 2025 State of the Market; IMM Comment, Docket RM26-4-000 (Nov. 25, 2025); IMM Complaint re Data Center Loads.

• Morgan Stanley Research. (2025–2026). Powering AI: Markets Race to Invest in AI Energy Solutions.

• North American Electric Reliability Corporation. (2026). 2025 Long-Term Reliability Assessment.

• NRDC (Rutigliano, T.). (2025). PJM Board to Decide on Higher Bills for 67 Million Americans Due to Data Centers.

• NIAC / CISA. (2024). Addressing the Critical Shortage of Power Transformers. (Draft, June 2024.)

• PJM Interconnection. (2024–2025). Base Residual Auction Reports (2025/26, 2026/27, 2027/28); 2025 Long-Term Load Forecast; Regional Transmission Expansion Plan.

• R Street Institute (Chandler, K.). (2024). Cost-Causation Is No Cause for Concern; and Initial Comments, Docket RM26-4-000 (2025).

• Rhodium Group (Carter, T., King, B., Larsen, J., van Brummen, A., Kolus, H.). (2026). The Impacts of Rising Electricity Demand from Data Centers on US Energy and Emissions. Apr. 7, 2026.

• Ribeiro, L., Street, A., Arroyo, J. M., Moreno, R. (2025). A Causation-Based Framework for Pricing and Cost Allocation of Energy, Reserves, and Transmission. arXiv:2505.24159.

• RMI (Cohen, J., Fitch, T., Shwisberg, L.). (2025). Gas Turbine Supply Constraints Threaten Grid Reliability. June 18, 2025.

• S&P Global Commodity Insights. (2025). US Gas-Fired Turbine Wait Times as Much as Seven Years; Costs Up Sharply.

• Synapse Energy Economics. (2025). For DC Office of People's Counsel, Drivers of PJM's Capacity Market Price Surge; for MD Office of People's Counsel (2026); for Evergreen Action, Tackling the PJM Electricity Cost Crisis.

• Union of Concerned Scientists (Jacobs, M.). (2025). Connection Costs Loophole Costs Customers Over $4 Billion.

• U.S. Department of Energy. (2024). Large Power Transformer Resilience Report to Congress.

• Utility Dive; American Nuclear Society. (2024–2025). Talen–Amazon Susquehanna interconnection coverage. GE Vernova Inc., Form 8-K / Q1 2026 (SEC EDGAR, CIK 0001996810).

Appendix A. Glossary

ANOPR — Advance Notice of Proposed Rulemaking. BRA — Base Residual Auction (PJM's forward capacity auction). BTMG — behind-the-meter generation. CONE — Cost of New Entry. Co-location — siting a large load at a generator to be served directly. FRR — Fixed Resource Requirement (an alternative to RPM participation). GOES — grain-oriented electrical steel. ISA — Interconnection Service Agreement. LDA — Locational Deliverability Area. LMP — Locational Marginal Price. RML/ARM — Reference Margin Level / Anticipated Reserve Margin. RPM — Reliability Pricing Model (PJM's capacity market). RTO/ISO — Regional Transmission Organization / Independent System Operator. SSR/RMR — System Support Resource / Reliability-Must-Run. UCAP — Unforced Capacity.

Appendix B. Methodology note

Quantitative claims are drawn from primary sources (FERC dockets and orders, PJM auction and load reports, the PJM Independent Market Monitor, NERC, LBNL, DOE, and equipment-manufacturer disclosures), peer-reviewed literature (Joule; arXiv), and reputable institutional research (Brattle, IEA, EPRI, Goldman Sachs, Morgan Stanley, Rhodium, RMI, Synapse). Where figures vary across sources (for example, data-center demand projections), ranges are reported rather than point estimates. Critical-path stage durations (Table 2) are representative syntheses intended to convey relative magnitude, and should be read as ordinal rather than cardinal.

This paper is for informational and educational purposes only and does not constitute investment, legal, or engineering advice.

That is the trade in one paragraph. The market is still pricing a failed Bitcoin miner. The insiders are pricing the power. We are siding with the insiders, and we have started buying.

This is Part 1. It lays out what the company is, how it got here, what the asset is worth, and the one event that closes the gap. Parts to follow will track the catalysts in real time.

How we got here

You cannot understand the opportunity without the wreckage that created it.

The company used to be Mawson Infrastructure, ticker MIGI, and it was one of the worst-run names in the public Bitcoin-mining sector. While direct peers compounded more than 600 percent on the shift to AI infrastructure, Mawson destroyed roughly 97 percent of its value. Management churned. Creditors filed an involuntary bankruptcy petition. The stock was staring at a Nasdaq delisting.

Then an investor group called Endeavor started buying. Through late 2025 it accumulated stock and issued shareholder letters arguing the obvious: the infrastructure was worth far more than the equity, and the problem was leadership, not assets. By April 2026 the campaign was over. Endeavor had won board control, the prior CEO was gone, five new directors were seated, the company was renamed Big Digital Energy, and the company set itself on a path back into Nasdaq compliance. The board has since retired the takeover poison pill, now that the new owners are the ones in control. The new team pointed the company at the highest-value use of its core asset: low-cost, energized power for AI and high-performance compute.

The stock has already moved off the lows, from a roughly 16 million dollar market cap to roughly 40 million. We think that is the first leg of a much larger re-rate.

What you are actually buying

Strip away the ticker change and the narrative, and BGDE is a power-first infrastructure operator with a real, verifiable footprint. From the company's own SEC filings, not a pitch deck:

The flagship is Midland, one of the largest sites in PJM operated by a public miner, with roughly 38,810 machine positions and carbon-free power supplied under a fixed-price agreement with Energy Harbor, owner of the adjacent Beaver Valley nuclear station. Bellefonte is a smaller complementary site, and Corning is an additional development parcel that carries optionality the market ignores. A second 250-megawatt substation already sits on the site, currently offline, next to an option on 200 adjacent acres. That is development upside, not operating capacity, and none of it is counted in the 129-megawatt figure.

This is not a shell. The infrastructure is real enough that the company runs production data-center management tooling and live energy metering across its sites. You are buying a genuine operator, not a story.

The power economics

Here is the part the market is not modeling. BGDE buys power under a fixed-price contract at a blended cost in the low 40s of dollars per megawatt-hour, roughly 4 cents per kilowatt-hour, while the 2025 PJM all-in wholesale price ran closer to 79 dollars per megawatt-hour. That spread is the entire business model: control cheap, contracted power in a market where power is repricing violently upward.

It monetizes that position two ways today. First, hosting and colocation, renting capacity and racks to customers. Second, and more interesting, energy management: getting paid by the grid to curtail and provide demand response when PJM is short. That second line is small relative to what it becomes if the company converts capacity to AI, but it is real, it is growing, and it is the proof that scarce power is a cash-generating asset, not just a cost center.

The business already generates real revenue

The bear reflex on any ex-miner is "no revenue." That is wrong here. In 2025 the company produced roughly 38 to 40 million dollars of revenue across two real lines: digital colocation at about 26 million, and energy management at 11.8 million, up 56 percent year over year.

That energy-management line is the important one. It is how a power-rich operator turns grid scarcity into cash. As power gets scarcer, that line gets more valuable, and it is already the second-largest revenue source in the company. And the platform has shown it can monetize the assets themselves: it previously sold its Sandersville, Georgia site to CleanSpark for up to 42.5 million dollars, more than the company's entire market cap going into this takeover.

The macro: power is the bottleneck, and BGDE sits on it

The AI build-out is not constrained by chips or capital. It is constrained by energized megawatts with grid interconnection already in place, the thing that takes years to permit and build. Nowhere is that tighter than PJM, and the price proves it.

PJM capacity cleared at 28.92 dollars per megawatt-day for 2024 to 2025. It then cleared 269.92, then 329.17, and the most recent auction hit 333.44, the regulatory cap, the first time the entire grid fell short of its reliability requirement. Data centers drove the majority of that move, and PJM is warning of a shortfall of up to 60 gigawatts over the decade. BGDE controls a scarce, grid-connected position in the exact market where everyone now needs power and nobody can get it fast.

The tell: the people who know the most are buying the most

Cheap is not a catalyst. Aligned owners buying their own stock is. And the insider record here is as one-directional as we have ever underwritten.

From late November 2025 through June 2026, the controlling group bought stock almost continuously. They did not buy the bottom and stop. They paid 4.07, then 4.84, then 5.37, then 6.01, then 8.38, and on one print 12.00. They paid up into their own thesis. Then, in June, the entire newly seated board followed with personal cash in the high 6 to mid 7 range. Total open-market insider buying runs past 6 million dollars. Real insider selling is zero, into a float of under 6 million shares. That is accumulation by the most-informed buyers in the building.

The prize: AI conversion

This is where the asymmetry lives. The same megawatt earns very different money depending on what runs on it. Bitcoin and legacy hosting generate roughly 20 to 45 million dollars of revenue per 100 megawatts at mining-grade margins. The publicly disclosed AI and HPC colocation contracts that former miners have signed imply 140 to 225 million dollars per 100 megawatts at 40 to 65 percent site-level margins.

This is not theoretical. Former miners have signed enormous, long-dated AI contracts that re-rated their entire equities. These are the precedents the market will use to price BGDE the day it signs its first deal.

Be clear-eyed: BGDE has not signed an AI offtake yet. That is the un-hit catalyst, and it is exactly what the new team was installed to deliver. But that is also why you are getting 129 megawatts at a price that assumes it never happens.

What is $BGDE worth?

We value this on the asset, because that is what a re-rate prices. Today BGDE trades around 0.3 to 0.5 million dollars per megawatt. The miners that have signed AI contracts trade at 11 to 30 million per contracted megawatt. That gap is the opportunity, and the risk. On the company's own platform math the existing assets carry a stated net asset value near 75 million dollars, so the stock changes hands at roughly half of that book value before any AI conversion is priced in.

The variable is dollars per megawatt, and the swing factor is whether the company converts capacity into contracted AI compute.

• Base case, roughly 18 dollars. The miner-plus-demand-response business stabilizes, the market gives partial credit for 129 megawatts of PJM power, and the stock re-rates toward 1.5 to 2 million per megawatt. Roughly 2.5 times from here, and that is the case where the AI thesis never even lands.

• Bull case, roughly 55 dollars. The team signs its first real AI or HPC offtake, even 30 to 50 megawatts, and the contracted capacity reprices toward 5 to 7 million per megawatt, still a steep discount to peers for a first contract. Roughly 7 to 8 times, after the shares issued to fund the conversion.

• Moonshot, 150 dollars and up. Multiple sites convert to executed contracts and the platform re-rates toward the peer band of 11 to 30 million per megawatt. Twenty times and up. This is the home run, and it requires the team to execute against the clock, but it is the deck's own platform math, not a fantasy.

Against a price around 7 dollars, the base case is a winner and the bull case is a multi-bagger. The targets are aggressive on purpose. They are also anchored to the asset and to peer multiples, not to hope.

The team

The thesis rests on whether this group can do the one thing the old regime never did: convert power into contracted compute. The control group, operating as Endeavor and SixThirty.AI, are the chairman, chief executive, and operating chief, and they collectively own about 29 percent of the company. Their backgrounds run across energy, AI and HPC infrastructure, real estate, and operations, and one built and sold a cybersecurity company. The most important fact about them is not a resume line. It is that they have bought more than 6 million dollars of stock with their own money and put it under multi-year performance milestones. Their outcome is the public shareholder's outcome.

What are the catalysts?

This is an event-driven setup. The things that re-rate it, in order of importance:

1. The first executed AI or HPC offtake with a named compute partner. This is the single re-rating event. It converts a few-hundred-thousand-per-megawatt valuation into a contracted multiple overnight.

2. A new long-dated power contract replacing the December 2026 agreement at a still-competitive rate, defusing the cost-reset clock.

3. Refinancing or terming-out the near-term loans, removing the going-concern overhang.

4. A clean, independently priced combination of the controlled development pipeline into the public company, turning the platform story into public-company megawatts.

5. Litigation resolution at de minimis cost, clearing the last legacy overhangs.

What are the risks?

We are bullish, not blind. At any real position size you hold these in view.

• The cheap-power agreement reprices at the end of 2026. Midland's fixed-price contract runs through December 2026; beyond it, power costs move toward the PJM market. A new long-dated contract is the catalyst we are watching, and its absence is the clock on the cost advantage.

• Funding and dilution. The balance sheet carries a going-concern note and roughly 26.5 million dollars of near-term loans, and the funding answer is mostly debt rather than fresh stock. The executive chairman's own entity has extended the company a secured revolving credit line of up to 40 million dollars, while the public at-the-market equity program is capped at just 9.6 million, which holds near-term dilution down. Two items inside that stack are worth watching: a defaulted legacy promissory note with about 11.1 million dollars outstanding that now sits in active litigation, and a Nasdaq condition that requires at least 5 million dollars of stockholders' equity in every quarter through the middle of 2027. The value per share still has to outrun the share count, and so far the insiders are buying faster than the company is issuing.

• No AI contract is signed yet. The thesis lives or dies on the first real offtake. A prior 2024 attempt did not convert. This is a bet that this team gets it done.

• Related-party structure. Part of the growth pipeline and the flagship colocation deal involve entities affiliated with the same principals who control the company, reviewed under a stated independent-committee framework with the conflicted principals recused. The alignment here is structural, not cosmetic. The affiliate's upside on the colocation deal is paid in warrants struck at 20 dollars, nearly three times the current price, so those insiders only monetize that piece by driving the stock to 20 and above. It still deserves ongoing scrutiny, and the offset is that the same principals hold common stock alongside outside shareholders.

None of these is disqualifying. All of them are why a 129-megawatt PJM asset is still available at a fraction of replacement cost, and why we treat this as a high-conviction, defined-risk position rather than a sure thing.

Is $BGDE a buy? The bottom line

Big Digital Energy is a real 129-megawatt power operator in the most power-starved grid in America, with a demand-response revenue line growing 56 percent, nuclear-adjacent power, and an activist team that put more than 6 million dollars of its own money in from 4 to 8 dollars with zero selling. It trades at a few hundred thousand dollars per megawatt while contracted peers trade in the tens of millions. The one thing it has not done, sign an AI offtake, is the one thing that closes that gap, and it is the catalyst we are positioned for.

We are buyers. This is the kind of asymmetry the portfolio exists to find.

Disclosure

Shawarma Capital holds a position in BGDE and intends to add to it. This is research and opinion, not investment advice. Microcap stocks are volatile and can go to zero. Do your own work and size accordingly.

Written 17 June 2026. As I write this, Trio-Tech trades near $14.82, up roughly 18 percent on the day, against a 52-week range of $2.44 to $21.38. That range is the first thing to understand. This is a roughly $150 million company on about 10.1 million shares with no sell-side coverage, and it moves violently on small flows in both directions. Read the position-sizing and the bear section before you act on anything here. The stock has already run hard. The point of this letter is not the candle today. It is the gap between what Trio-Tech earns and what the market that already loves this exact business is willing to pay for it.

Let's go.

The layer below the layer everyone already bought

The AI trade has already paid up for everything one rung beneath the models. The compute. The memory. The networking. The foundries. That framing is consensus now, and consensus is priced.

There is a rung below that one, and it is still not priced, because it is unglamorous and almost nobody covers it.

When a chip is built, it does not ship straight to a customer. The expensive ones, the AI accelerators, the automotive controllers, the memory that has to live for years inside a data center, get stress-screened first. The most important screen is burn-in. The chip is mounted on a specialized circuit board, the burn-in board, then run hot, at elevated temperature and voltage, under electrical load, for hours or days. The purpose is to force the weak units to fail in the lab instead of in the field. A part that survives burn-in is far less likely to die early inside a customer's product.

For a cheap consumer part you can argue about whether that screen pays for itself. For an AI GPU that sells for tens of thousands of dollars, an automotive chip where a field failure is a safety event, or memory that runs continuously for years in a data center, the screen is not optional. The economics of the part demand it.

Trio-Tech International sells the burn-in boards. It sells the reliability-test equipment around them. And it runs the testing itself as a service, including final test for advanced chips, from its own floors across Asia. It does not design chips and it does not compete with the chip makers. It sells to them and to the houses that assemble and test for them. When AI and automotive chip volume rises, demand for burn-in and reliability capacity rises with it, and that demand lands on a very small number of specialists.

The whole thesis is one sentence. Every expensive AI and automotive chip has to survive the oven first, and Trio-Tech sells the oven, the boards that go in it, and the oven time, for a fraction of the multiple the market already pays the one burn-in name it has discovered.

Why the screen is becoming mandatory right now

The reason this is a today story is that the AI parts are breaking the old way of testing.

The cleanest read on the category does not come from Trio-Tech. It comes from its larger, pure-play public cousin, Aehr Test Systems, which has spent the past year telling the market two things. First, that rising power densities in next-generation GPUs and accelerators are overwhelming legacy test tools and forcing customers onto new high-power burn-in platforms. Second, the number that frames the entire opportunity, that only about 5 percent of ASICs and roughly 50 percent of AI accelerators currently go through production burn-in at all.

Sit with that. Half of AI accelerators and the overwhelming majority of custom AI silicon ship today without the production burn-in screen. As these parts get more expensive, hotter, and more mission-critical, and as the hyperscalers and the automakers demand field reliability they can underwrite, that attach rate only goes one way. The burn-in attach rate on the most valuable chips on earth is low and climbing. The screen is also spreading into new device classes, silicon carbide power devices and silicon photonics among them. That is the tide. Aehr is the high-multiple way the market already plays it. Trio-Tech is the way the market has not noticed yet.

The inflection is in the filings, and it is accelerating

This is not a story about what might happen. It is reported.

In fiscal Q3 2026, the quarter ended 31 March 2026, revenue was $16.5 million, up 124 percent from $7.4 million a year earlier. The Semiconductor Back-End segment, the part that matters, did $13.1 million of that, up 141 percent. Industrial Electronics added $3.4 million, up 76 percent. Across the first nine months of fiscal 2026 revenue reached $47.7 million, up 85 percent, with the back-end segment at $36.9 million, up 104 percent.

The shape of that growth matters as much as the size. The year-over-year rate went 58 percent, then 82 percent, then 124 percent across the three quarters of fiscal 2026. This is a curve, not a single print.

And the nine-month figure already sits above any full year Trio-Tech has reported since fiscal 2022. This is the part the market is still missing.

Read what that says. Revenue sat on a roughly $42 to $44 million plateau from fiscal 2022 through 2024, then slid to $36.5 million in fiscal 2025. The market, reasonably, priced a tired, cyclical, slowly declining industrial. Then nine months of fiscal 2026 came in at $47.7 million, already past the old plateau, with the rate of growth still rising. The stock is being valued on the company that was declining. The opportunity is the distance to the company the filings now describe.

The segment doing the work is unambiguous.

The customers, by name of theme, and one the company already disclosed

The growth has three sources, and one of them is bigger than the founder letter let on.

The first is final test for a leading AI chip maker. In its fiscal Q1 2026 release, Trio-Tech disclosed its entry into providing final testing services for next-generation high-performance AI devices for what it called a leading AI chip manufacturer, and management guided explicitly to additional revenue from that AI customer for final test services. This is not an inference. The company put it on the record. That is a recurring services relationship with a top-tier AI chip vendor for production test of next-generation parts. It is the highest-quality leg of the story, and it is exactly the kind of relationship that compounds as the customer's volume ramps.

The second is the AI GPU burn-in board order book, which has kept filling after the quarter closed.

The June 4 order matters more than its size. It landed after the quarter ended and after the earnings print, which is the cleanest possible answer to the bear worry that this is one front-loaded burst. The demand is still arriving.

The third is automotive, and it is structured as a service, not a one-time sale. On March 4 Trio-Tech secured an initial roughly $2.5 million production burn-in order from a leading automotive integrated device manufacturer, described as an ongoing program that ramps in phases through calendar 2026, supported by targeted investment in capacity, workforce, and process controls to meet automotive reliability standards. Automotive burn-in is recurring by nature. Every production lot gets screened. That is the steadier counterweight to the lumpier board sales.

The hidden leg: Penang, and the memory customer the market has never tied to this name

Here is the part that turns a good setup into an asymmetric one, and it did not come from a press release. It came from the company's own footprint.

On April 28 Trio-Tech's Malaysian subsidiary signed a lease for about 104,000 square feet of industrial space in Perai, Penang, at a base rent near $115,000 a month, with the term running from June 1, 2026, into 2028 and a one-year extension option. That is a large block of new testing capacity in the heart of the Penang semiconductor cluster, funded by a roughly $10 million raise the company closed days earlier. You do not lease that much floor and fund it fresh unless you intend to fill it.

The more telling signal is in the hiring. Trio-Tech is staffing roughly thirty positions in Batu Kawan, Penang, a park where the company has not historically run a facility, and at least one advert states the role is based at MMP. In the Penang cluster, MMP is the routine internal shorthand for Micron Memory Penang, Micron's memory assembly-and-test operation anchored at Batu Kawan, where Micron has committed billions of dollars and built a center of excellence for memory and storage.

A burn-in and reliability specialist staffing engineers to sit physically inside a memory plant is the signature of a supplier being qualified into that plant's supply chain.

Be precise about confidence here. Micron has not named Trio-Tech. Trio-Tech has not named Micron. This is an inference drawn from public hiring data, the new Perai lease, and the timing of the raise, three signals pointing the same direction at the same time. Treat it as a strong, well-supported lead, not a fact. But the logic is clean, and the direction is memory, the third giant end-market of the AI cycle alongside compute and automotive, and the one Trio-Tech has never been associated with. Penang is also where Coherent has built a datacom and silicon-photonics research presence and where the broader AI back-end is concentrating. Trio-Tech is expanding straight into the densest reliability-test cluster on earth. If the memory read is right, a fourth leg is being added to a story the market is already underpricing. If it is wrong, the AI final-test, AI board, and automotive legs still stand on their own disclosed numbers.

The number that defines the trade: the comp pays 75 times sales

To value Trio-Tech you do not need a discounted cash flow you can argue with. You need the comp.

Aehr Test Systems is the public, pure-play, AI-burn-in name the market has already found. Here is what the market pays for it, next to what it pays for Trio-Tech.

This is the whole letter in one table. The market has decided burn-in levered to AI is worth paying for. It pays Aehr roughly 75 times sales for the privilege, while Aehr loses money. It pays Trio-Tech roughly two times sales, while Trio-Tech makes money and grows faster. The two companies sit on adjacent rungs of the same ladder. Aehr sells the capital systems. Trio-Tech sells the consumable boards that populate those systems, the surrounding reliability equipment, and the test service, including final test for a leading AI chip maker. On burn-in boards specifically, there are only a handful of specialist suppliers in the world. Trio-Tech is one of them.

I am not arguing Trio-Tech should trade at 75 times sales. I am arguing that the distance between 2 and 75, for a profitable, faster-growing company in the same niche with no coverage to close the gap, is the largest mispricing I have put in front of you in this book.

The one thing the bulls are not modeling: the margin

Now the honest center of the analysis, the part that separates a careful read from a press-release read.

The revenue doubled. The gross margin did not come with it. In fiscal 2025 Trio-Tech earned a roughly 25 percent gross margin. Across the first nine months of fiscal 2026 the gross margin ran closer to 16 percent, and the March quarter itself produced a small operating loss of $81 thousand and a net loss to common of $38 thousand, despite record revenue.

That is the swing factor, and it cuts both ways. The bear read is that the explosive revenue is coming through lower-margin burn-in board hardware and new-capacity startup cost, that the surge therefore does not drop to the bottom line, and that a microcap selling capital boards at 15 percent margins is not the high-quality compounder the headline growth implies. The bull read is that this is exactly what an inflection looks like before it converts. The lower-margin board orders are the door opener. The higher-margin work, the final-test services for the AI chip maker, the multi-phase automotive program, and the apparent Micron memory engagement, are recurring service revenue that scales into the new Penang floor and carries a structurally better margin. If the mix shifts back toward services as that capacity fills, the gross margin recovers toward and through the old 25 percent line on a much larger revenue base, and the operating on a near-fixed Asian cost base is violent. The entire profit case rests on that recovery. Watch the gross margin line on every print from here. It is the single most important number in the model, and it is the one the headline-reading bulls are ignoring.

The balance sheet does not have the usual microcap problem

A company inflecting this fast usually carries a stretched balance sheet. This one does not. At the March quarter Trio-Tech held $13.0 million of cash, and on April 24 it priced a registered direct offering of 1,052,632 shares at $9.50, raising about $10 million gross, which closed April 27. That puts pro-forma cash near $23 million against a company with almost no debt, a current ratio near 3.0, and total debt to equity around 0.11. It sits in a net-cash position.

The dilution was real, a bit over ten percent, and the stock sold off on the news, which is the normal reaction to a raise. The constructive read is that management raised from strength rather than distress, after the stock had already run, and now has the cash to build the Penang capacity that the AI, automotive, and apparent memory demand all require. There is also a Real Estate segment, a legacy of the company's long history in Malaysia and China that earns rental income and holds tangible property value. It is not the thesis. It is a small cushion under a stock the market still prices on its sleepy past.

Alignment, control, and the float

Trio-Tech has been led since 1990 by S.W. Yong, chairman since 2023 and inside the company since the 1970s. He owns about 16.9 percent of the shares, roughly 1.47 million, and in January he exercised options to buy 80,111 shares at $2.64 each, deep conviction at a price a fraction of today's. This is an owner-operator who has carried the company across decades and is buying, not selling.

Institutions hold only about 16 percent. That is the other half of the setup. A profitable, fast-growing semiconductor reliability play levered to AI, automotive, and possibly memory, with almost no institutional ownership and no sell-side coverage, is a stock where the marginal buyer has not arrived yet. The float is tiny, roughly ten million shares, which is why the stock can run from $2.44 to $21.38 and give back a third of it inside weeks. That illiquidity is the risk and the fuel in the same sentence.

What a re-rate is worth

I will not hand you a single false-precision number. I will hand you the framework and four scenarios, each built the same way: a fiscal 2027 revenue level, a gross-margin assumption, and a price-to-sales multiple that stays far below the 75 times the market already pays the comp. Run the math yourself and change the inputs. The shape is what matters.

Read the moonshot line carefully, because it is the point. An 11 times sales multiple, the one that implies a roughly $150 stock, is still only about one-seventh of what the market pays Aehr today. To reach the upside in this letter, Trio-Tech does not have to be loved more than the comp. It has to be loved a fraction as much, while being profitable and growing faster. That is what makes the asymmetry real rather than rhetorical. The base case alone, an $80 million revenue year at a 4 times multiple the market would consider cheap for any AI-levered name, roughly doubles the stock. The bull and moonshot cases are what happen when coverage initiates, the Micron memory leg gets disclosed, and the gross margin recovers on a much larger base at the same time.

What would make this wrong

You get the bear case in full, because the bull case is only worth anything if it survives it.

The first risk is liquidity. This is a genuinely tiny, illiquid stock that has already run roughly six times off its low and trades a third below its high. It moves hard in both directions on small flows. Position sizing matters more here than in anything else in this book. Do not chase a green candle in a name that can erase a week in an afternoon.

The second risk is the margin, the one the whole bull case rests on. If the revenue mix stays weighted to lower-margin board hardware and the new Penang capacity carries fixed cost faster than it carries revenue, the gross margin does not recover, the company stays at breakeven on a bigger base, and the re-rate never comes because there are no earnings to re-rate. This is the live debate, and the next two prints settle it.

The third risk is cyclicality. Trio-Tech has had revenue surges before that faded back to the plateau. The bull case requires that the AI cycle is structurally larger and more durable than a normal back-end upcycle. The disclosed final-test relationship with a leading AI chip maker, the multi-phase automotive program, and the Penang buildout are the evidence that it is. They are not yet proof.

The fourth risk is the inference. The Micron memory read is drawn from public hiring signals, not confirmed by either party. If it does not convert into disclosed business, the most exciting leg of the story is softer than it looks today. The thesis does not depend on it. The moonshot does.

The fifth is concentration and control. An owner-operator with a 16.9 percent stake gives you alignment and key-person and minority-shareholder risk in the same breath, in a name with thin governance scrutiny because no one covers it.

What to watch

The fiscal year ends June 30, so the quarter happening now is the close of fiscal 2026, and the full-year report around September is the first hard test of whether the order surge converted into sustained revenue and, more important, whether the gross margin started to recover.

The single highest-value catalyst is the one nobody can schedule: the first analyst who picks up coverage. In a name with none, the day the gap gets a voice is the day it starts to close.

Is $TRT a buy? The bottom line

$TRT is a 67-year-old, no-coverage microcap sitting in the one layer of the semiconductor supply chain that every expensive AI and automotive chip has to pass through. Its revenue is accelerating, it runs final test for a leading AI chip maker, its order book keeps filling with AI GPU and automotive work after the quarter closed, it is building new capacity in the heart of the Penang cluster, and the hiring trail points at a memory customer it has never been tied to. The balance sheet is clean, the insider owns one share in six and is buying, the institutions have not arrived, and the closest comp trades at 75 times sales while losing money. Trio-Tech trades at two times sales while making it.

The honest catch is the gross margin, and the next prints decide whether the inflection converts to profit. Size for the illiquidity. But understand the asymmetry: to reach the upside in this letter, this company does not have to be loved more than the burn-in name the market already found. It has to be loved a fraction as much.

The oven does not care how good the chip looks on paper. It only cares whether it survives. Right now the entire AI build-out is feeding more and more expensive silicon into that oven, and almost nobody has noticed who owns it.

The full valuation model and the catalyst-by-catalyst case go out in Part 2. Subscribe so the next one lands in your inbox before the gap closes.

Disclosure. Long $TRT. This is research synthesis for educational purposes, not investment advice. You should not buy or sell securities based on anything written here. I am not a registered investment advisor and I do not owe you a fiduciary duty. The price scenarios above are illustrative model outputs built on stated assumptions, not forecasts or guarantees, and the inputs can be wrong. The Micron memory connection is an inference drawn from public hiring and corporate records, not confirmed by the companies involved, and may be wrong. Figures are drawn from public filings and announcements and may contain errors. Position sizing in a microcap this illiquid is your responsibility. Do your own due diligence.

More in this series

Every AI Chip Has to Survive the Oven First

Read more

Read more

Read more

If you are new to this series, read the next two paragraphs slowly, because everything after them depends on understanding what this company actually is.

Merlin, which trades under the ticker $MRLN on the Nasdaq, builds software that flies airplanes. Not a better autopilot, which simply holds a heading a human sets. Merlin is building what it calls a digital pilot: software meant to taxi, take off, talk to air traffic control over the radio, handle an engine failure or a weather diversion, and land, with fewer humans in the cockpit and, eventually, none. The company calls the product the Merlin Pilot. Its lead customer is U.S. Special Operations Command, the part of the military that runs special-operations missions, and the aircraft they want it on first is the C-130J, a large four-engine cargo plane that is the backbone of Air Force special-operations aviation.

Merlin became a public company on March 16, 2026 through a de-SPAC. A SPAC is a shell company that lists on an exchange with nothing but cash, then goes shopping for a real business to merge with; when the merger closes, the private company becomes public without running a traditional IPO. So the quarter that ended March 31, 2026 was Merlin's very first as a public company, and there is almost no trading history to lean on. Keep that in mind every time you hear someone talk about Merlin's stock chart.

Everything below is built from primary records anyone can pull: federal contract databases, the company's SEC filings, internet certificate logs, granted patents, and Senate lobbying disclosures. Where a number could not be verified against one of those sources, it is not in this post.

The fact that moved, and the fact that did not

On June 4, 2026, Merlin announced it had cleared its Critical Design Review for the C-130J program. The phrase sounds like jargon, so here is the plain version. In any serious aircraft program, the government customer makes the contractor pass through a series of formal checkpoints. Early ones ask whether the requirements are clear. Middle ones ask whether the design is sound. A Critical Design Review, or CDR, is the checkpoint where the customer agrees the detailed design is mature enough to stop drawing on paper and start building and testing on a real aircraft. In one sentence: CDR converts a program from "are we designing the right thing" into "build it and prove it flies."

Merlin cleared the earlier Preliminary Design Review on March 5, 2026 and the Critical Design Review on June 4. Two design gates in ninety days. No C-130 has yet flown with Merlin's autonomy installed. The CDR is the permission slip that lets that finally happen.

Here is how the market reacted. The stock spiked roughly 32 percent in after-hours trading on the news, to about $9.49. Then it faded. Over the following days it drifted back down toward its 52-week low of $5.78 and, as of the most recent close, trades around $6.95, with a total market value near $660 million, against a 52-week high of $17.00. There is no clean year-long chart to compare against, because for most of the past year the ticker was the empty SPAC shell sitting near its cash value, not this operating business. The prior installment of this series was written with the stock near $7.60. The milestone landed, the stock is lower, and that contradiction is the entire opportunity this post is about.

Now the fact that did not move. Merlin cleared the central engineering milestone of its flagship program, and the public record shows no new federal contract and no fresh task order since. The most recent SEC filing is dated June 2. The most recent 8-K, which is the form companies file to disclose a material event, came on May 14. There is no successor SOCOM order on the federal spending databases as of mid-June. The catalyst is real. The follow-through is pending. That is precisely why the price has not re-rated: the market is waiting for the money, not the milestone.

What the company actually has, and what it does not

The single most important number in the Merlin story is its contract ceiling with Special Operations Command, and it is the number most people get wrong. In June 2024, SOCOM awarded Merlin what is called a sole-source IDIQ.

Two pieces of jargon, both worth understanding. Sole-source means the government chose Merlin without a competition, because it decided Merlin was the only suitable supplier. IDIQ stands for indefinite-delivery, indefinite-quantity. It is not a check. It is a master agreement that sets a maximum the government is allowed to spend and the terms it will spend on, and then the actual money is released later in pieces called task orders, as the program hits milestones. The ceiling is the size of the tank, not the fuel in it.

Merlin's ceiling is $105 million, and the vehicle runs through April 6, 2029. Against that ceiling, the government has so far released two task orders. The main one paid for the Preliminary and Critical Design Reviews and was funded at roughly $15.96 million. A small kickoff order adds about $7,500. So of a $105 million ceiling, only about $16 million has actually been spent. That leaves roughly $89 million of headroom that only Merlin can be paid against, with no competitor allowed in, through 2029.

This is the most common error you will see in coverage of Merlin: the "$105 million contract" gets treated as if it were revenue or backlog. It is neither. It is an option the government holds to buy up to $105 million of work from Merlin, of which it has exercised about 15 percent.

Now read one more fact alongside that ceiling. The task order that funded the design work ended its period of performance on May 15, 2026. As of mid-June, no follow-on order has appeared on the public databases. There is a funding gap open right now. And here is the mechanism that makes the CDR matter: future task orders under this vehicle are conditional on the current one succeeding. The CDR clearing on June 4 is the technical event that makes Merlin eligible for the next slice of funded work. The condition has been met. The order has not yet come. That is why the next SOCOM task order is, by a wide margin, the single highest-signal event to watch in this entire name.

The second federal item worth understanding is SHIELD, and it has to be handled carefully because the headline number is enormous and easy to abuse. In December 2025, Merlin won a seat on the Missile Defense Agency's SHIELD vehicle, which is the contracting structure behind Golden Dome, the homeland missile-defense initiative. SHIELD carries a ceiling of $151 billion. That number is not Merlin's. It is a shared ceiling spread across 2,440 different awardees, a pool that includes giants like SAIC, Blue Origin, and L3Harris. Merlin's actual obligated amount on SHIELD is $500. The right way to think about SHIELD is as a hunting license: it gives a fixed-wing autonomy startup the right to bid on homeland missile-defense task orders through 2035 alongside the defense primes. The signal is strategic, that Merlin is now inside that tent at all, not financial.

Two smaller awards round out the federal ledger. An Air Force AFWERX agreement worth $749,499, signed in October 2024, whose official statement of work reads "enhancing reduced-crew flight operations with NLP-driven real-time dynamic replanning." NLP means natural-language processing, the branch of software that understands human speech and text. And a historical $969,000 FAA research contract from 2022 and 2023 for work on flying uncrewed aircraft inside national airspace.

The certification machine, hiding in plain sight

The most analytically interesting part of Merlin is not anything the company announced. It is something it never meant for you to see.

To certify software that flies an aircraft, a company has to build and run a very specific set of testing and proof tools, and it has to keep dashboards that track the results. Between March 18 and May 5, 2026, Merlin stood up exactly that toolchain as internal web dashboards, each on its own web address. Every web address needs a security certificate, and by the design of the internet, every certificate issued is logged, permanently and publicly, in what are called certificate-transparency logs. Nobody at Merlin chose to publish this. It leaked automatically. And the names of those dashboards spell out the precise toolchain used to certify flight-critical software.

Here is why this matters in plain language. Aviation software is graded by how catastrophic its failure would be. The strictest grade, the one reserved for software whose failure could cause a crash, is called DO-178C DAL-A. To certify at that level you do not just test the software, you have to mathematically prove it cannot enter certain failure states. The dashboard named code-prover is the tool that does exactly that kind of proof. The others track which lines of code have been tested and that every piece of code traces back to a written requirement. This is the full apparatus of the highest aviation software-safety standard, and you can watch Merlin assemble it, week by week, in public, in the weeks bracketed by its two design reviews.

There is a second, independent way to confirm this is real. Merlin's own engineering job postings name the identical tools, plus the system-safety standard ARP4754A and the hardware standard DO-254. So two completely unrelated public surfaces, the certificate logs and the hiring listings, describe the same certification machine. It is very hard to fake an entire toolchain into existence across two channels that have nothing to do with each other.

This also dovetails with the regulator's own clock. In October 2025, Merlin reached what is called Stage of Involvement 2 of 4 with the Civil Aviation Authority of New Zealand, which means roughly half of its flight-control software has now been formally evaluated by an aviation regulator, with the U.S. FAA validating in parallel. Stage 1 cleared back in May 2023. There are really two certification clocks running here, and they matter differently. The military clock leads, because a military flight release does not require civilian certification, which means defense revenue can arrive first. The civilian clock, the New Zealand and FAA path, runs behind it and points toward 2027 and later.

The lobbying ramp

There is a third public surface most autonomy stories never bother to check. Under the Senate Lobbying Disclosure Act, every dollar a company spends lobbying the federal government is on the public record. Merlin's spend has climbed roughly fourfold into the C-130J program, and it is split between a defense-appropriations firm and an aerospace-certification firm.

The figures are in thousands of dollars; the all-time total is about $2.42 million. The defense work runs through a firm called Invariant, and the named lobbyist on the account is Paul Arcangeli, who spent twelve years as the Democratic staff director of the House Armed Services Committee, the most senior staff job on the committee that writes defense policy, until 2022. The disclosed issues name the fiscal-year 2026 defense appropriations bills by number. The reading is straightforward. A company that hires the former staff director of the defense committee, while separately paying an aerospace firm to work the certification regulator, is a company trying to do two things at once: get its programs funded and get its product cleared to fly. That is exactly the two-clock structure described above. None of this proves success. It is evidence of intent, and of where the money is going to manufacture catalysts rather than wait for them.

The validators standing around the technology

You do not have to take Merlin's word for the technology, because a series of outside companies have attached their own names and balance sheets to it.

The strongest is GE Aerospace, one of the largest aerospace companies in the world, which in September 2025 agreed to co-develop a certifiable autonomy core with Merlin. GE's flight-management systems already sit on more than 14,000 aircraft, which is a ready-made distribution channel if the joint product works. Honeywell, another aerospace giant, signed an agreement to bring the Merlin Pilot onto its Anthem avionics suite. Northrop Grumman named Merlin one of six autonomy partners on its Beacon flight-test program, alongside well-funded peers like Shield AI and Applied Intuition. And the physical system is built on named, already-certified suppliers: a Green Hills safety operating system cleared to the top aviation tier, a CMC Electronics flight computer, and a Shadin data unit already certified for both the C-130J and the KC-135 tanker. None of these companies sign agreements with a vendor whose product does not exist.

Merlin's own patents match the architecture all of this implies. Its granted U.S. patents cover reading air-traffic-control radio audio with a question-and-answer model, a flight-processing system that checks a command for validity before executing it, identifying a safe landing site from sensor data, and monitoring the human pilot's attention. That last patent is the reduced-crew thesis written directly into intellectual property. The first, a question-and-answer model interpreting ATC audio, is the language-model layer of the autonomy stack, filed as IP rather than merely claimed in a slide deck.

The honest tension on the numbers

Now the part a promoter would bury. When Merlin went public, its own forecast called for $32 million of revenue in 2026, almost entirely from the SOCOM contract, and that $32 million figure is what justified the $800 million valuation the company carried into the merger. That works out to roughly 25 times projected revenue, a rich multiple that only makes sense if the revenue actually shows up.

Set that forecast against what the first public quarter delivered.

Revenue of $1.0 million in the first quarter against a full-year forecast of $32 million is the central tension in this name, and it deserves to be stated plainly rather than spun. A few of these lines need translation for newer readers. Remaining performance obligations is the contracted work a company has signed but not yet delivered or recognized as revenue, the visible near-term backlog; here it is only $1.124 million, which is thin. The GAAP net loss of $90.4 million looks alarming but is heavily distorted: roughly $63 million of it is a non-cash accounting remeasurement of merger-related instruments, a mark on paper, not cash leaving the building. The cleaner number is the adjusted EBITDA loss of $23.3 million, which strips out those non-cash and one-time items and approximates the real cash burn, about $24 million in the quarter.

On the balance sheet, the company held $122.8 million in cash at quarter-end, raised another $80 million in a financing on May 1, and carries no debt. The quarterly filing states that management believes existing cash plus that financing funds operations for at least twelve months, and the accounts were prepared on a going-concern basis, meaning the auditors did not flag a survival risk. One caveat the company disclosed itself: a material weakness in its internal financial controls, which is the reason it posted a senior SEC-reporting role on June 15. That is the kind of housekeeping gap that small, newly public companies routinely carry, but it is a gap and it belongs on the record.

The bridge from $1 million of quarterly revenue to a $32 million year runs entirely through the SOCOM ceiling converting into task orders. That is the whole investment question in one sentence, and it is why the next order matters more than anything management will say in London.

How an independent analyst frames the same path

The equity-research outlet Market Apostle published its own Merlin milestone map this month, and it frames the road ahead the same way this series does, which is worth noting because it is an independent read of the same facts. Its framework runs on two parallel tracks: a systems-engineering track of formal design reviews, and an airworthiness track that clears the modified aircraft to fly. Its line on the contract structure is worth quoting directly: "The headline contract figure is the maximum budget the DoD is allowed to spend on this program. The actual revenue Merlin will get will be based on the task orders issued, while future task orders are conditional on current order success." That is the federal ledger above, arrived at independently, and it treats a cleared design review as necessary but not sufficient for the next order. That is the correct framing.

Scenarios

In the bear case, the funding gap that opened on May 15 simply stays open, revenue limps far below the $32 million forecast, and the roughly $24 million quarterly burn forces a dilutive capital raise that hurts existing holders. In the base case, the CDR converts into a successor task order that draws down a meaningful slice of the $89 million of headroom, and revenue inflects but still lands below the forecast for the year, because task orders take time to obligate and to recognize as revenue. In the bull case, the next order is large and fast, the first autonomous C-130 flights happen on schedule, and a second program converts on top, at which point the market stops pricing Merlin as a pre-revenue de-SPAC and starts pricing it as a funded defense-autonomy platform.

Catalysts to watch, in order of how much they matter

The successor SOCOM task order is first, and by a wide margin. It is the conversion event the entire thesis rests on, the funding gap is open right now, and the CDR just removed the technical blocker. Watch the federal spending databases for it. Second is the ROTH London Conference, running today through June 18, where management holds one-on-one investor meetings and any forward commentary becomes quotable. Third is the first actual C-130 flight with Merlin autonomy installed, the physical proof the CDR was meant to enable; the preliminary airworthiness approach for that demonstration was already accepted back at the March design review, which lowers one of the risks. Fourth is any SHIELD task order that turns the $500 seat into real missile-defense work. Fifth is the next certification stage with New Zealand's authority and the FAA, which advances the civilian clock toward 2027.

What are the risks?

The revenue-to-forecast gap is the dominant risk and is quantified above. Beyond it, a roughly $24 million quarterly burn against a bit over $200 million of cash gives a runway measured in quarters, not years, which turns the timing of the next task order into a financing question as much as a commercial one. The disclosed material weakness in internal controls is real and worth tracking until it is remediated. The SHIELD ceiling invites overstatement and must never be read as Merlin's money. And the competitive field is crowded: Merlin sits alongside Shield AI, Applied Intuition, and others on Northrop's Beacon testbed, which is both a validation of its relevance and a reminder that it is not the only company chasing this.

Is $MRLN a buy? The bottom line

$MRLN cleared the engineering gate that matters on June 4, and the stock sits near its 52-week low while a SOCOM funding gap stays open and London investor meetings happen today. The setup is a company with $89 million of sole-source headroom it has not yet drawn, a certification machine assembling in plain view, a roster of blue-chip validators from GE Aerospace to the U.S. military, and a $1 million quarter sitting under a $32 million forecast. The next task order resolves the tension in one direction or the other. Until it lands, owning this is a wager that a cleared design review converts into funded integration, priced as though it will not. That gap, between a real catalyst and a price that has ignored it, is the trade.

Not investment advice. Do your own work.

More in this series

MRLN: The Complete Thesis

$MRLN: I built Merlin World

MRLN Part 10: The Deep Update

Merlin Intelligence - Live OSINT Tool

I Tracked Every Merlin Aircraft for 180 Days. The Testbeds Just Peaked

Alyeska Is Underwater. None of the Thesis-Breakers Have Hit

MRLN's $105M Ceiling Is Not the Revenue, It Is the Authorization

Alyeska Wrote Both Checks and the Market Missed It

The UAE Deal Changes Everything Merlin Was Supposed To Be

MRLN: Part 4 - The Board, the Preferred Supernova, and the Catalyst Nobody

$MRLN Part 3: The DAWG, the $54.6B, and Why Merlin Is Positioned to Win the

MRLN +21.2% - The C-130J critical design review cleared with USSOCOM on June 4 and moved the program from design into aircraft integration on the hundred-million-plus IDIQ ceiling. Management restated the plan to fly a crewed fixed-wing aircraft autonomously in commercial revenue service from New Zealand in 2027, gated on finishing certification by year end. I shipped Part 10, Part 11, and a live OSINT tool that tracks every Merlin tail by FAA owner record. The tape gave back the design-review pop and trades near $7.31, up about a fifth from the October entry. None of the seven kill criteria have been met.

IQEPF +159.5% - IQE signed a multi-year indium phosphide epiwafer supply agreement with Tower Semiconductor for AI data-center optical connectivity, with first-year minimum purchase commitments, and settled the prior IP dispute in the same stroke. An £81M raise earlier cleared the bank debt and the company guides to more than 20 percent revenue growth this year. The stock is up double digits intraday on the Tower print. Position grew through price action, no fresh capital added.

LPKFF +127.3% - The activist standoff went to a vote at the June 4 annual meeting and management won. LPK.DE popped to €23.60 on the result and settled near €21.50. The glass-substrate AI packaging monopoly thesis is intact and Part 4 is live. Plan is still to size back in on dips once BMM re-rates.

BMM +5.1% - Blue Moon Metals consolidated after Part 3. Springer in Nevada is still the only permitted US-domiciled ammonium paratungstate hub, and Nussir in Norway carries the EU Critical Raw Materials Act strategic designation after closing its C$156.3M bought deal. Building toward the 35 percent target weight, the largest position in the book. The probability-weighted expected value against today's print keeps the asymmetry wide.

AMPG +68.8% - New starter taken in early June at $4.50, about five percent of the book. AmpliTech is a twenty-four-year-old RF house that turned itself into a 5G open-RAN radio OEM with a quantum tail, printing 48 percent gross margins with no debt. The first conversion landed on the record this month: FCC and ISED cleared the radio and the BEAD program money started to move. It already trades near $7.59, through the seven-dollar target I underwrote it against. The asymmetry on a five-dollar microcap is wider than anything else I own.

RBRK +63.4% - Rubrik beat every metric in the Q1 fiscal 2027 print on June 4 and raised full-year guidance across the board. Subscription ARR grew 32 percent to $1.57B, revenue grew 39 percent, and the subscription contribution margin widened to 13.2 percent from 8 percent a year ago. Thesis intact, no exit, the light haircut to fund the BMM build stays.

PGY +44.4% - Pagaya posted its first clean GAAP profit in Q1, $25M of net income against a loss a year ago, on $318M of revenue and $2.6B of network volume, and raised the 2026 guide. A CFO transition takes effect mid-June. Quiet execution, no drama, the tape finally repriced it.

CODA -13.1% - Coda Octopus gave back the year's gain into the microcap pullback and now sits below my basis. Underwater imaging for defense and offshore energy, still profitable and debt-free. Thesis intact, plan is to size back in on dips post-BMM re-rate.

See you mid-July.

Long all positions. Not investment advice.

Now you can.

Today I am releasing MERLIN WORLD, a life-size virtual hangar that puts the entire Merlin Labs campaign in one room at true scale. You sign in, you spawn on the floor, and the fleet is standing in front of you. The C-130J wing runs forty meters over your head. You walk up to it. You read its dossier off the pedestal. Every number on that pedestal is sourced.

This is the same OSINT that powers the Merlin Intelligence terminal. I turned it into a place.

ACCESS IT AT https://shawarmacapital.net

I. The fleet, at true scale

Walk the main aisle and the tails sit exactly where the data puts them. Real aircraft, real dimensions, life-size.

N208B, Big Red. The U.S. certification workhorse. It flew the first fully automated U.S. takeoff on April 11, 2026, and it is the most-flown tail in the campaign.

ZK-MLO, Kerikeri. New Zealand's lead certification tail, flying near-daily. New Zealand is the guided site for Merlin's first autonomous revenue service, targeted for 2027.

N831DL, the jet. An Aero L-39C Albatros, registered May 21. The long-horizon bet, pushing Merlin Pilot toward tactical speeds. The ADS-B tripwire on it is armed.

C-130J, Condor. The flagship. A USSOCOM ceiling north of $100M, Critical Design Review completed June 4, 2026, now entering aircraft integration.

You do not read about these airframes. You stand under them.

II. You can fly them

Every aircraft in the hall is flyable. Climb in and you get a real flight model over real terrain. Take the Caravan off the testbed strip. Push the L-39 toward the speeds the program is chasing. Fly the C-130J. The simulator is the fleet you just walked, in the air.

III. It is live and it is multiplayer

You are not alone in there. Up to 48 subscribers can be on the floor at the same time, in real time, each one a callsign moving through the hangar. You claim your callsign once and it is yours. A leaderboard tracks the people who log the most time on the floor. The live chat runs alongside it, so the room talks while it walks.

The research stopped being a document. It became a room with people in it.

IV. The intelligence is in the walls

Merlin World sits on top of Merlin Intelligence, the OSINT desk behind the coverage.

The orbital globe at the head of the fleet carries every Merlin operating location and route tracked by ADS-B. Two continents flying one certification stack.

The master control console is a walk-up nerve center. A live 2D fleet map, a rolling 24-hour flight log, and the full intel browser, without leaving the floor.

Upstairs, the galleries hold every chart from the research desk. Price action. The runway simulation. The sortie ramp. The cap table. The catalyst clock. The whole model on the walls, where you stand in front of it.

V. Access

Until today, all of this was for paying subscribers only.

Starting now, a free trial gets you in. Start a trial, sign in at shawarmacapital.net with the same email as your Substack subscription, and a six-digit code lets you through the hangar door. You get the whole thing. The hangar, the flight deck, the live floor, the dashboard, the chat.

Here is the honest part. When the trial ends, if you have not converted to a paid or founding subscription, access closes. The door locks behind you. The system enforces it cleanly. I would rather say that up front than surprise you.

Founding members get the same access plus the standard founding benefits.

VI. Bottom line

The entire Merlin Labs campaign is sitting in plain sight, scattered across ADS-B archives, SEC EDGAR, GlobeNewswire, FAA registries, ATS boards. Nobody was assembling it. I assembled it, and then I built you a place to walk through it.

Take the trial. Walk the fleet. Fly the C-130J. Read every filing. Then decide.

$MRLN

I am long the names disclosed at the top of this post. This is research synthesis, not investment advice. You should not buy or sell securities based on anything I write. I am not a registered investment advisor. I do not owe you a fiduciary duty. My conclusions could be wrong in ways I have not anticipated. Financial projections are model outputs based on publicly available data. They are not guarantees. Microcaps are illiquid and can lose 50% or more in a single session. Do your own due diligence. Verify every number against primary sources.

More in this series

MRLN: The Complete Thesis

The Gate Cleared and the Tape Forgot

MRLN Part 10: The Deep Update

Merlin Intelligence - Live OSINT Tool

I Tracked Every Merlin Aircraft for 180 Days. The Testbeds Just Peaked

Alyeska Is Underwater. None of the Thesis-Breakers Have Hit

MRLN's $105M Ceiling Is Not the Revenue, It Is the Authorization

Alyeska Wrote Both Checks and the Market Missed It

The UAE Deal Changes Everything Merlin Was Supposed To Be

MRLN: Part 4 - The Board, the Preferred Supernova, and the Catalyst Nobody

$MRLN Part 3: The DAWG, the $54.6B, and Why Merlin Is Positioned to Win the

Most of what I write about is a sub-billion-dollar hardware company sitting on a physical chokepoint the market has not priced. Ouster does not fit that mold on size. It is a $2.5 billion company, up more than threefold in a year, with seven analysts already at strong buy. The market is wide awake to the lidar story. So before I spend six thousand words on it, let me earn the place it takes in this letter.

The chokepoint here is not a factory, a mine, or a patent. It is a statute. The United States government took the largest lidar manufacturer in the world, named it a Chinese military company, and barred its hardware from American defense work. That company is Hesai. It runs more than four hundred million dollars of annualized revenue, several times Ouster's size, and it builds the cheapest high-performance lidar on the planet. On price and volume it should win. On nationality it just lost the part of the market that matters most. Ouster is the company standing on the other side of that line: American-domiciled, debt-free, growing, and already cleared through the exact procurement framework the ban funnels demand into.

That is the whole thesis in one paragraph. The rest of this post is the proof, the numbers, the bear case, and the parts the market is still misreading. The timing is the reason I am writing now. The stock ran from roughly $26.80 in mid-May to a fifty-two-week high near $47 on June 4, then dropped almost sixteen percent in a single session on June 5. There was no bad news under that drop. A sell-side firm initiated coverage with a buy rating and a $75 target, and the early money sold the news. The thesis did not change. The price did.

Let's go.

I. Why A $2.5 Billion Company Is In A Microcap Letter

I am not going to pretend Ouster is undiscovered. The edge is not obscurity. The edge is that the market is pricing Ouster as one more lidar company in a brutal, commoditizing market, and is underweighting two structural facts that redrew the map in the last six months: the United States removed Ouster's most dangerous competitor from the highest-value slice of demand, and the Western lidar field consolidated from a crowd into a survivor list of about four. Neither of those is a quarter. Both are still working their way into the model.

There is a second reason. Ouster is no longer only a hardware company. Its software-attached bookings more than doubled in 2025, and it just acquired a profitable computer-vision business. The market still values it on a sensor-unit multiple. The company is quietly rebuilding itself into a perception-and-software business with recurring revenue, and that re-rating has not happened yet.

So this is not "nobody knows about this." It is "everybody knows the lidar story, and almost nobody has repriced for the ban, the shakeout, and the software turn happening at the same time." That is the gap.

II. The Print, And The Trajectory Underneath It

Start with the numbers, because they are the foundation and because one of them looks worse than it is.

Two things jump out, and both need explaining honestly, because the casual reader gets both backwards.

First, Q4 2025 looks like a breakout and Q1 2026 looks like a relapse. It is the opposite. Q4 2025 carried approximately $21.2 million of one-time IP-licensing royalties tied to long-term license contracts. That royalty is the entire reason Q4 printed $62.2 million of revenue, a sixty percent gross margin, and a $4.0 million profit, the company's first GAAP-profitable quarter. Strip the royalty and Q4 product revenue was about $41 million. Q1 2026 then printed $48.6 million with no royalty, which means the underlying product business grew sequentially off that $41 million base, by roughly eighteen percent, even as the headline fell twenty-two percent. Management has guided total 2026 royalty revenue to less than $5 million, most of it in the back half, so do not model the royalty as recurring. The forty-three percent GAAP and forty-six percent non-GAAP margins in Q1 are the clean run-rate, and they sit at or above the company's own thirty-five-to-forty percent framework. The ugly sequential chart is a footnote, not a trend. Anyone who reads only the headline reads it backwards.

Second, the business is not profitable on its own yet. Q1 2026 was a $17.5 million GAAP net loss and a $6.9 million adjusted EBITDA loss, both narrower than a year earlier. The direction is toward breakeven. The arrival is not here, and I am not going to round a narrowing loss up to a profit. The lone GAAP-profitable quarter was royalty-driven, and I will say that every time someone calls Ouster profitable.

Full year 2025 was $169.4 million of revenue, up fifty-two percent against $111.1 million in 2024, at a forty-nine percent blended gross margin, with the net loss cut to $60.4 million from $97.0 million the year before. The company shipped more than 25,000 sensors in 2025, up forty-eight percent, and set a quarterly shipment record in every quarter of the year. For the first quarter of 2026 it shipped more than 12,600 lidar and camera sensors, of which lidar was roughly sixty-five percent. The growth is real, it is broad, and it is accelerating in units even when the royalty distorts the dollars.

III. The Balance Sheet Is The Cleanest In The Sector

Cash, cash equivalents, restricted cash and short-term investments totaled about $175 million at March 31, 2026. The company carries no debt. That combination, real cash and zero , is rarer in lidar than it sounds, and it is the single most important defensive fact in the whole story.

One number people misread: cash fell from about $211 million at year-end 2025 to $175 million in a single quarter. That looks like a burn scare. It is not. Operations used only $7.3 million in the quarter. The bulk of the decline is the roughly $35 million of cash Ouster paid to acquire StereoLabs in February. The company spent its cash buying an EBITDA-positive business, not funding losses. Adjust for the acquisition and the operating burn is single-digit millions against a $175 million pile.

A $175 million cash position, no debt, and a single-digit-millions quarterly operating burn is multiple years of runway with no financing cliff. Hold that fact, because the next two sections are about a sector where almost nobody else can say it.

IV. The Chokepoint Is A Statute, And It Is Aimed At The Biggest Player

Here is the fact the model has not absorbed.

Hesai Technology, headquartered in Shanghai, is the largest lidar manufacturer in the world by volume. It supplies commercial autonomous-vehicle programs globally and, by the United States government's own finding, Chinese autonomous-warfare systems. Its revenue is several times Ouster's: Hesai reported roughly $72 million in the first quarter of 2025, $99 million in the second, and $112 million in the third, an annual run-rate comfortably above four hundred million dollars and growing. On cost, scale, and unit volume, Hesai is the gorilla. In an open market it wins the price war.

The market is not open anymore, not for the part that pays best. The Department of Defense has identified Hesai as a Chinese military company under Section 1260H of the FY2021 National Defense Authorization Act, a designation first published in October 2024 and reaffirmed since. Separately, the NDAA's lidar provisions bar lidar developed by Hesai from United States defense-related procurement, with the restriction biting in 2026. Congress named the company. This is not a tariff that raises Hesai's price. It is a wall that removes Hesai from the bid.

That wall matters most precisely where the money is least price-sensitive. Defense and government buyers do not shop on dollars per sensor. They buy on clearance, supply-chain integrity, cybersecurity, and program longevity. They are exactly the buyers who cannot touch Hesai now, and exactly the buyers who must find a vetted Western alternative.

Now the other side in detail. Ouster's OS1 digital lidar is compliant with the National Defense Authorization Act and was the first high-resolution 3D lidar sensor approved under the Defense Innovation Unit's Blue UAS framework, the Pentagon's vetted list for unmanned systems that meet NDAA, supply-chain, and cybersecurity standards. Ouster sensors have been deployed by the Army, the Navy, and NASA. And the long-running patent dispute between Ouster and Hesai was dismissed at the Federal Circuit, removing the one legal overhang that could have clouded Ouster's standing.

Put it together. The government barred the dominant Chinese supplier from defense work and pre-cleared Ouster for exactly that work. When Hesai is banned, the demand does not disappear. It reallocates to whoever is cleared, and at the high-resolution end of the vetted list, that is a short list with Ouster at the top.

V. The Shakeout Is Over, And Ouster Is The Survivor

The Western lidar field spent five years as a graveyard of SPAC-funded promises. The cull has now happened, and seeing the field as it actually stands is most of the thesis.

Luminar was the most-hyped name of the cycle, with Volvo and Mercedes signed and billions raised. It filed for bankruptcy in December 2025 and its core lidar assets were sold to MicroVision for thirty-three million dollars, roughly the price of a mid-size office building. MicroVision, the buyer, runs on a going-concern warning. Aeva holds cash but burns through it at a pace its 2028 production timeline does not cover. Innoviz survives on perpetual dilution.

Against that field, Ouster is the one that does not have to think about running out of money. The shakeout did the competitive work Ouster would otherwise have had to fund with its own losses. It is cheaper to win a market when half the field goes bankrupt around you, and that is the position Ouster is in: more cash than the rest of the listed Western field combined needs to survive, and a customer base that is consolidating onto fewer, safer suppliers precisely because the others keep failing.

VI. The Total Addressable Market Is Four Markets, And Three Of Them Buy Now

Lidar is often pitched as a bet on robotaxis arriving. That framing is what bankrupted the field, because robotaxis kept slipping. Ouster's actual revenue comes from markets that buy today, with the automotive dream as a free option on top.

Take them in order of nearness. The overall lidar market roughly quadruples this decade, from about $3.3 billion in 2025 to about $12.8 billion in 2030, a thirty-one percent compound rate. That is the tide. Counter-UAS is the fastest-funded defense category in the West right now, growing from about $6.6 billion in 2025 to about $20.3 billion in 2030, with defense roughly fifty-five percent of it and laser and optical detection the fastest-growing detection layer. Drone interception needs to close the last hundred meters precisely, and lidar is the sensor that does it. The May 2026 ARGUS Interception agreement, integrating Ouster lidar into the A1-Falke interceptor, is the first visible dollar of that, and the NDAA ban points the rest of the category's spending toward NDAA-compliant suppliers.

Smart infrastructure and industrial are the revenue engine today, and they buy on annual budgets, not on a self-driving timeline. Robotaxi is the long-dated call option. The robotaxi market is forecast to go from roughly two billion dollars in 2024 toward a hundred and eighty-nine billion by 2034, and Motional, the Hyundai-Aptiv venture beginning ride-hailing in 2026, uses Ouster as its exclusive long-range lidar supplier. I will not underwrite automotive revenue, because the sector just watched automotive timelines bankrupt its most-funded player. If automotive comes, it is upside on top of the base case, not inside it.

VII. The Software Turn Nobody Is Pricing

This is the section that separates the 2026 Ouster from the 2022 Ouster, and it is the one the unit-multiple bears miss entirely.

Ouster sells more than sensors now. It sells Gemini, a perception-software platform, and BlueCity, a traffic-intelligence platform, both of which attach recurring software revenue to the hardware. In 2025, software-attached bookings more than doubled, rose more than a hundred and twenty percent year over year, and represented over fifteen percent of sensors shipped. On the Q4 call, management cited significant Gemini renewals including a seven-figure annual license with a leading global technology company.

Read what that means. A lidar company whose customers renew seven-figure annual software licenses is no longer a hardware company with a one-time sale. It is building an installed base that pays every year. BlueCity is deployed at nearly seven hundred sites and the combined Gemini-plus-BlueCity contracted footprint is past twelve hundred sites worldwide. Every one of those is a recurring-revenue anchor and a switching cost. The market multiples a sensor sale once. It multiples a software renewal forever. Ouster is moving from the first to the second, and the tape has not noticed.

VIII. StereoLabs: From Sensor Vendor To Perception Company

In February 2026 Ouster acquired StereoLabs, and the deal matters more than its size.

The terms: roughly $35 million in cash plus 1.8 million Ouster shares, of which about 0.7 million release over four years. The acquisition closed February 4, 2026. StereoLabs builds vision-based perception systems, stereo cameras and the software that turns them into 3D understanding, for robotics and industrial use. It generated approximately $16 million of unaudited revenue in 2025 and, critically, it is EBITDA-positive. Ouster bought growth that helps its path to profitability rather than growth that drags on it.

The strategic logic is the shift from selling a sensor to selling perception. Lidar gives precise range. A stereo camera gives color and texture. Fused, they give a robot or a drone or a vehicle a richer model of the world than either alone. Ouster's new Rev8 line already introduced native color lidar; StereoLabs adds the camera and software layer. The company is repositioning from a lidar vendor into the perception stack for physical AI, the term the industry now uses for everything from humanoid robots to drones to warehouse automation. That repositioning, paired with the software-attach data above, is what supports a multiple above a pure hardware vendor's. The FieldAI collaboration, integrating Rev8 color lidar into general-purpose mobile robots, is the first product of the combined stack.

IX. The Product Stack

The technology case rests on Ouster's digital lidar architecture, built on a custom silicon system-on-chip rather than the discrete-component assemblies most competitors use. Digital lidar means the sensor is, in effect, a chip with optics, which lets Ouster ride the cost-down curve of semiconductors rather than the slow curve of mechanical assembly. That is the structural reason its gross margins sit in the forties while it scales, and why it can eventually meet the price points the volume markets demand without the losses that killed the analog-lidar players.

In May 2026 Ouster launched the Rev8 family, powered by its next-generation L4 and L4 Max silicon, with upgraded OS0, OS1, and OSDome sensors and a flagship 256-channel OS1 Max that the company calls the world's first native color lidar, with up to double the range and resolution of the prior generation, built to automotive functional-safety standards. The next chip, internally called Chronos, has already been fabricated by Ouster's foundry partner and is in in-house testing. A company with its next-generation silicon back from the fab and on the bench is executing its roadmap, not promising it.

X. Ownership And Insiders

Ouster is broadly held: roughly forty-seven to forty-nine percent institutional, with insiders in the high-teens to high-twenties percent depending on the source, and a float near ninety-six percent of shares. The company diluted shareholders by roughly twenty percent over the past year, which is the honest cost of funding a growth ramp without debt, and a real line in the bear case. Insider activity has been mixed-to-buying recently, with a director purchasing shares in the open market in May 2026. This is not a tightly-held founder stock with a squeeze setup. It is a liquid, institutionally-owned name where the return comes from the fundamentals re-rating, not from a float event.

XI. What Could Kill This

A free reveal that skips the bear case is a pitch, not research. Here is what I am watching, in order of weight.

The gross-margin optics are dangerous to the careless. Q4 2025 GAAP margin was sixty percent on a one-time royalty; Q1 2026 was forty-three. The forties are the truth, and forty-six percent non-GAAP is above the company's own framework, but the chart scares people who do not read the footnote, and if the forties do not hold, the semiconductor-cost-curve argument weakens.

It is not profitable on its own yet. The lone GAAP-profitable quarter was royalty-driven. The product business still loses money, narrowing. A growth stock that stops narrowing its loss gets repriced fast.

There is real customer concentration. In Q1 2026 one customer was roughly thirty percent of revenue, and one was twenty-five percent of accounts receivable, down from forty-two percent a year earlier. Improving, not gone. One re-order delay can soften a quarter.

The valuation already prices a lot. It is volatile, with a beta above three and a sixteen-percent single-day move on no fundamental news. It dilutes. And the NDAA tailwind carries two-way geopolitical risk: Chinese retaliation against United States sensor makers, or a softening of enforcement, cuts against the most important leg of the thesis. Finally, the automotive option could stay an option for years; nothing in my base case needs it, but a market that re-rates on a robotaxi headline can de-rate on a robotaxi delay.

XII. The Bull DCF: $90 Base, $120 Bull, $60 In Twelve Months

The Street is at an average target near $47 and a high of $75. I think both are too low, and the reason is structural: the published models ramp Ouster off its pre-ban, pre-software-attach trajectory. They do not fully price the reallocation of a four-hundred-million-dollar competitor's defense demand, the counter-UAS budget inflection, or a software line compounding at triple digits. So I built the bull case out properly, driver by driver, the same way I built the Merlin model that landed within pennies of TD Cowen's independent DCF.

Here are the assumptions. They are aggressive, and I am labeling them aggressive. This is the bull note, not the base-rate.

*DCF inputs: the demand ramp and operating *

Two things make that revenue line defensible rather than fantasy. The lidar market alone quadruples to ~$12.8 billion by 2030; counter-UAS adds a ~$20 billion pool where Ouster is NDAA-cleared and Hesai is banned; physical-AI robotics and the software attach layer on top. A category winner taking a credible share of those compounding markets reaches a billion-plus in revenue without needing a single robotaxi. And the exit multiple that anchors the terminal value, five to nine times forward sales, is below the twelve-point-seven times the market pays today. The bull case does not require multiple expansion. It requires the multiple to compress while revenue compounds nine-fold. That is the whole trick, and it is the opposite of a moonshot input set.

The twelve-month anchor is more conservative and still well above the tape: applying a ten-times multiple to 2027 estimated sales of roughly $380 million, adding net cash, and discounting one year, lands near $60, a third above the current price and below even where the stock traded a week ago. So the framework is a twelve-month path to roughly $60 and a multi-year intrinsic value of $90 in the base case and $120 in the bull, against a $40 tape and a $47 Street average. The downside, if the growth framework breaks and Ouster re-rates to a no-premium grower, is roughly $20 to $25, which sizes the asymmetry: I am risking about forty percent to make one hundred to two hundred.

The full scenario tree, the segment-by-segment revenue build, the software-revenue model, and the position sizing get their own follow-up. Part 1 puts the name, the structural case, and the bull valuation on the table, in full.

XIII. Bottom Line

Ouster is not a hidden microcap. It is a $2.5 billion company in a story the market knows. The edge is that the market prices the lidar story and underprices three things happening at once: the statute, the shakeout, and the software turn.

The largest lidar maker in the world is Chinese, runs four hundred million dollars of revenue, and just got named a Chinese military company and barred from United States defense work. The most-funded Western competitor went bankrupt and sold its assets for thirty-three million dollars. The survivors are running out of money. Ouster is debt-free, grew revenue forty-nine percent, holds $175 million in cash, doubled its software-attached bookings, is cleared through the Blue UAS framework, is deployed by the Army, Navy, and NASA, and just bought an EBITDA-positive vision company to move from selling sensors to selling perception.

It is not profitable on its own yet, it is not cheap on today's numbers, and it can drop a fifth in a session. But the structural map moved in its favor while the field thinned, the software line is compounding, and the bull DCF says a $40 stock is worth $90 in the base case and $120 in the bull, with a twelve-month path to $60. The tape handed out a sixteen-percent discount on a sell-the-news the day after a fresh fifty-two-week high. I am opening a starter this week.

Shawarma Capital

Part 1 said LPKF was a four-tool monopolist on glass-substrate AI packaging hiding inside a fifty-year-old German laser shop. Part 2 said the market had eighteen months and named the phrase to listen for on the April 30 call. Part 3 said the phrase was spoken, the stock re-rated past every published target in three weeks, and there was an activist already on the supervisory board with the chief financial officer buying stock with his own money.

Part 4 is the one where the disagreement Part 3 described stopped being a disagreement and became a vote. The June 4 annual general meeting was not a formality. It was a referendum on the single question that divides this shareholder base: raise capital now and spend into the glass-substrate surge, or hold the cost line and let LIDE scale on the company's own cash. The shareholders who wanted the first wrote it down as formal countermotions. The chief executive recommended rejecting them. The meeting backed the chief executive. The stock went up twelve percent on the day.

That outcome is good news and bad news in the same sentence. And underneath the governance drama sits the thing this series has been building toward: a real, driver-based discounted cash flow that says a twenty-one-euro stock is worth seventy in the base case and ninety-five in the bull. This is the long one.

Let's go.

Read more

Part 2 ended on a sentence I want to be held to. I wrote that the entire bull case still rested on two non-binding letters of intent and a management team with a documented history of guiding up and delivering down, and that the next testable checkpoint was conversion. Letters of intent into purchase orders. Certification into deployment. A guide weighted to the back half into actual back-half revenue. A federal program into actual flowing dollars.

Please DM me if you’d like to get a 12-page sell-side DCF.

Three weeks later, the first pieces of that conversion are on the record. Not the big letter turning into a signed multi-year contract; that has not happened. What happened is more concrete and more checkable: the product the letters are written against cleared the regulatory gate that lets a North American carrier deploy it, and the federal broadband program that the entire Buy-America moat depends on moved from "approved plans" to "signed award agreements with fifty-two states."

This is the deep Part 3. It marks the position to today, lays out the two structural tailwinds in full, the BEAD program and the quantum line, and is honest about the dilution and the management history that keep this a ten-percent position and not a twenty-percent one.

Let's go.

Read more

Read more

Nothing here is investment advice. For educational purposes only. I may hold positions in names I discuss. Do your own research. No liability assumed.

Subscribe now

Executive Summary

Merlin Labs (NASDAQ: MRLN) is the public-equity entry point on crewed-aircraft autonomy at a moment when the defense-autonomy category is re-rating against private comps (Anduril at $14B+, Shield AI at $2.7B, Saronic at $9.25B). The investable structure has six load-bearing surfaces.

One. Capital structure mechanics: $674M market cap, $491M EV ex-preferred, $183M cash, 84.3M shares outstanding, a freely tradeable float of approximately 3.6-5.0M shares after the 90.3 percent redemption at the de-SPAC, and IBKR borrow utilization at 99.93 percent in late April 2026. A tight-float defense-tech name with a real catalyst pipeline.

Two. The September 16, 2026 lockup expiry + Conversion Date Reset. 44.3M insider and PIPE-anchor shares open the same day the Series A Preferred conversion-date reset prices off a 20-trading-day VWAP, floored at $5.00. (The conversion price already reset from $12 to $6.67 on the May 1 PIPE full-ratchet.) The reset can only hold at $6.67 or move lower toward $5.00, so it is neutral-to-dilutive for common, not favorable. Lockup absorption is the dominant pre-September risk surface. Six P0 catalysts sit in the 111-day window between today and September 16.

Three. The board reads as a customer-relationship map. A former Secretary of the Navy (Kenneth Braithwaite, also Chairman of Fincantieri Marinette Marine, Senior Advisor to Saronic + Kongsberg + Chertoff Group). The first Chief Accounting Officer at Amazon (Brannon). The former Air Force Comptroller (Montelongo). The Kingstown Capital PIPE anchor as Lead Independent Director (Blitzer).

Four. The federal contract trail with named award numbers. USSOCOM C-130J IDIQ (H9240824F0010, $15.96M obligated, $105M total program ceiling per CFO Carrithers commentary on Q1 earnings). MDA SHIELD foothold (HQ085926FE898, $500 floor obligation, Merlin's first publicly visible Missile Defense Agency contract). USAF KC-135 19-month integration contract (awarded late 2024, demo target September 30, 2026). GE Aerospace Autonomy Core co-development (announced September 23, 2025 at AFA) targeting the KC-135 Center Console Refresh as a separate procurement track from the 19-month USAF vehicle.

Five. The customer-axis P&L architecture. A three-lane GM architecture (Tactical / Mobility / Maritime), with Mobility the only seat currently filled. The Maritime seat transitioned from a six-month internal-state hold in Merlin's Lever applicant tracking system to publicly recruiting in the same nine-day window that five Washington DC Revenue Operations roles posted. An SVP Engineering with carrier-deck unmanned combat aircraft pedigree (N-UCAS / X-47B / MQ-25 lineage). A SOCOM-grade Chief Revenue Officer.

Six. The comp framework. At the base-case FY2027 revenue ($90M), MRLN trades at 5.5x EV/Revenue versus 13x for Anduril and 5.2x for AeroVironment. A category re-rate toward 15x FY27 implies $13/share base, $27 bull, $58 moonshot.

The Part 10 incremental update over Part 9 also surfaces: a Northrop Beacon program enrollment with a named-flight slot overdue, a UAE distribution agreement with Remah International Group, a 188-day flight dataset showing the certification campaign in its busiest stretch, and a regulatory-cure conversion gap on the converted Caravan (ZK-MLQ) that re-emerged on May 27, 2026 with a 28-minute cert circuit at Kerikeri.

This document is the full update across every visible surface. The structural data carries the thesis. The CEO's May 27 parenthetical reply on X (treated in Section I) is one observation; the reader is invited to weight it independently.

The six P0 events in the 111-day window from June 3 through September 16 are the most catalyst-dense quarter MRLN has carried since the SPAC closed on March 17, 2026. The asymmetry sits in the spread between what is publicly known to be working and what is publicly known to be coming. The first half of this document inventories what is working. The second half models what comes next.

A note on the probability estimates that appear throughout this document (0.78 for the KC-135 CCR RFP, 0.62 for Northland initiation, 0.70 for GM Maritime fill, 0.42 for Riley Loftus board seat, and others). These are Bayesian point estimates that I derive from a combination of three inputs: (a) base-rate analysis for the relevant event category (sell-side coverage initiation cadence, federal contract RFP timing, board-seat appointments at SPAC-graduate stage), (b) signal density in the public corpus around that specific event (LinkedIn engagement, contracting-office activity, named-individual public engagement), and (c) my own pattern-matching against historical analogues. Treat them as order-of-magnitude calibration rather than precise odds. They are not derived from a formal model or from any non-public source. A reader who disagrees with the underlying base rate or the signal-density weighting will reach a different number; the directional conclusion typically holds across reasonable parameter variation.

Section I, An Observation on a CEO Reply (Make of It What You Will)

One small observation before the structural sections take over. On May 27, 2026 at 1:03 PM Pacific, Matt George replied to a critic of the CNN "AI Pilot Era" segment with: "Most of what we're (publicly) working on is augmenting crew, meaning a pilot still in the flight deck (just one instead of two). See our C-130J work, etc."

The word "publicly" inside the parentheses is doing some work. By saying "most" of what is "publicly" being worked on is crew-augmentation, the CEO leaves the negation open: some non-public work is something other than crew-augmentation. The set of plausible meanings is broad. (a) Classified programs that are still crew-augmentation but cannot be announced. (b) Commercial integrations under NDA. (c) Customer-specific work not yet ready for press. (d) Early-stage R&D not on a roadmap slide. (e) Fully autonomous, uncrewed military aircraft work, which is the category Merlin's own product page marks as "Future" state.

Any of (a) through (e) is consistent with the reply. Reasonable readers can stop at (a)-(d) and dismiss (e) as overreach; I will not insist on (e). What I will note is that the product page already enumerates the "Future" category, the public reduced-crew work is what is "publicly" disclosed, and the parenthetical leaves the negation open. Take the observation for what it is. The thesis in this document does not require (e) to hold; the structural surfaces (capital structure, board, contracts, GM org, comp framework) carry the weight independently.

The Part 10 update over Part 9 surfaces the following structural changes in May 2026: a three-GM customer-axis P and L architecture with the Maritime seat transitioning from a six-month internal-state hold (visible only via direct Lever API scrape) to publicly recruiting on the careers landing page; a five-role DC Revenue Operations cluster posted in nine days; a board reading as a customer-relationship map with a former Secretary of the Navy; a Northrop Beacon cohort enrollment via N437VN; a Missile Defense Agency SHIELD contract foothold on USAspending; a documented technical-bench engagement with Boeing Defense via two named contacts; a 188-day flight dataset showing the spring 2026 certification campaign in its busiest stretch; and the May 27 re-emergence of the ZK-MLQ converted Caravan after a three-month silence.

The rest of the document inventories each surface.

Section II, The Three-GM Customer-Axis P and L Architecture

Boeing Defense, Space and Security has a GM for tanker and transport, a GM for rotorcraft, a GM for autonomous systems, a GM for space. Lockheed Aeronautics and Northrop Aeronautics Systems have the same customer-axis P&L architecture. Each GM owns revenue, capture pipeline, and customer relationship for one command axis.

Merlin Labs has been quietly building the same architecture. As of last week, three of the lanes are confirmed and a fourth has been staffed up in support.

Erin Staine-Pyne. Title: GM, Mobility. Filled December 2025. Background: USAF Academy Class of 1998, 3,500-plus flight hours in the C-17 and C-130, commander of the 62nd Airlift Wing at Joint Base Lewis-McChord (a 2,400-person C-17 wing she led through the onset of the COVID pandemic), and most recently Senior Military Advisor to the Under Secretary of the Air Force. That is the number two civilian leadership office in the Air Force, and it owns budget, acquisition oversight, and senior personnel actions for the entire service. Staine-Pyne walked out of the Pentagon weeks before joining Merlin with relationships across every PEO that touches mobility autonomy. Her customer axis is USAF Air Mobility Command and Air Force Materiel Command. Tanker autonomy on the KC-135 is her line. Reduced-crew C-130J is her line. C-17 follow-on autonomy is her line. Her May 19, 2026 LinkedIn post on the KC-135 program drew engagement from four active duty USAF officers in mobility-adjacent roles (Lankford J33, Herzberg 16th Airlift, Clancy HAF/A33D, plus one more) and a public LIKE from retired General Jacqueline Van Ovost, the former commander of US Transportation Command and former AMC commander. The engagement signature is consistent with an RFP in late-stage development. The alternative reading is that AMC-level requirements officers engage with industry posts routinely and the timing is coincidence. Both readings are live.

The GM Maritime seat. The listing first entered Merlin's Lever applicant tracking system on December 4, 2025 (two days after the Mobility seat closed) and was scraped from the Lever API at that timestamp. It only became publicly browseable on the Merlin careers landing page in the past several days. That transition from internal-state to public-state is its own intelligence signal. The seat was held in internal-state for nearly six months while the company waited on a single targeted candidate; the public posting now suggests the company has broadened the candidate pool and is actively soliciting. The job spec is unambiguous about what the seat is. Required: fifty million dollars plus of P and L authority. Required: one hundred million dollars plus of contract awards generated. Required: active Secret clearance. Required: Washington DC location. Mandate: maritime ISR, with named customer commands of US Coast Guard, US Navy, Department of Homeland Security, and global Foreign Military Sales. The job description explicitly names the King Air platform. This is not a sales hire. The dollar thresholds, the clearance requirement, the DC geography, and the customer set point at one thing: a tier-one prime counterpart hire dressed as a maritime line of business. Reading the spec, the natural profile is a recently retired Navy or Coast Guard captain or rear admiral whose last assignment was Pentagon staff on OPNAV or USCG HQ-9 and whose operational community is P-8 Poseidon, P-3 Orion, MH-60R, or HC-144 maritime patrol.

Pablo Gonzalez. Title: Senior Vice President of Engineering. Joined Merlin on November 19, 2025. Came from Sierra Space, where he served as Vice President and Program Manager of the Dream Chaser Cargo System. Before Sierra Space he was at Northrop Grumman for almost twenty years as Program Director on the F/A-18, the F-5, and the Navy Unmanned Combat Air System (N-UCAS). The N-UCAS program is the carrier-launched unmanned ISR and strike effort that became the X-47B and then evolved into the MQ-25 lineage. Pablo's resume already includes shipping naval unmanned combat aircraft into the fleet. His job at Merlin is the engineering scale-up that lets the GM organization run parallel customer execution without dropping integration quality. He was hired two weeks before the GM Maritime seat was posted. That sequencing is deliberate. Engineering scale comes first, capture cadence comes second. His April 23, 2026 LinkedIn post linking back to the original X-47B autonomous aerial refueling work was a small flag, posted with the simple caption "To those of you who walked this journey with me…. Thank you." That post tells the engineering community he brought the carrier-deck autonomy playbook with him.

Mark Brunner. Title: Chief Revenue Officer. Joined April 13, 2026 announced. Background: 22 years in the special operations community, including SOFWERX channel positioning. Krishnan Anand's April 14 LinkedIn announcement made the hire public, drawing 14 reactions. CROs do not get hired to manage seven and a half million dollar revenue businesses. They get hired when an organization expects to close multiple large contracts and needs dedicated commercial infrastructure to capture them. Brunner is a SOCOM operator in a sales-leadership chair. The capture lane he opens is the lane the GM Maritime hire will close on.

Michael Baker. Title: Chief Marketing Officer. Seated March 31, 2026. The Anthem partnership announcement, the Condor product launch on May 14, the CNN Pete Muntean pilot segment that aired May 20, the Merlin Pilot for Commercial Cargo announcement, and the May 21 CNN repost all sit in Baker's communications surface area. He is the corporate megaphone.

Krishnan Anand. Title: Chief Program Officer. The longest-tenured executive in the public-facing operator surface. His February 28, 2026 public post recruiting from the Supernal layoff round (the Hyundai air mobility shutdown that landed approximately the same time as the Merlin de-SPAC) is the playbook of an org that absorbed senior aerospace engineering displaced by a competitor failure. Krishnan publicly recruited "systems engineering, flight controls, autonomy roles, finance and others." David Lasater, the CPO who arrived from Supernal himself, named Liz Pelberg-Schariter, Cory Ronan, Drew Picconi, and Eliot Jackson as the colleagues helping build the team. That is five named senior hires from a single competitor failure within sixty days of the de-SPAC. The acquisition cost was zero. The cycle time was sixty days.

David Lasater. Title: Chief Product Officer. Came from Supernal in Q1 2026. His March 15, 2026 LinkedIn post explicitly recruited from the Supernal layoff round, naming colleagues. His March 17 post (the SPAC listing day) thanked employees, board, and investors and used the phrase "joined this journey at a key moment." The CPO seat is the product roadmap seat. Lasater owning it locks the Condor product family ownership, the Merlin Pilot for Commercial Cargo ownership, and the Anthem integration ownership into a single executive accountability line.

Liz Pelberg-Schariter. Title: VP Talent. The cert engineering hiring burst is one of the highest-conviction operator signals, when a company starts hiring Designated Engineering Representatives (the FAA-authorized signatories who sign off on Supplemental Type Certificates), the certification program is past the document stage and into the signature stage. Liz's hiring footprint is the leading edge of the SOI-3 (Stage of Involvement 3) and STC progression.

A defense services company with three customer-axis GMs, a senior engineering executive of that pedigree, a SOCOM-grade CRO, a post-spin CMO, a long-tenure CPO absorbing Supernal talent, and a VP Talent running a cert-engineering hiring burst is not a startup. It is a defense services provider that has decided what shape it wants to be when it grows up. The shape is identifiably a tier-one prime in scaled formation.

Section III, The Capture Machine Going Up Around Them

Between May 18 and May 22, 2026, Merlin posted five senior roles at the Lever portal. Every single one is Washington DC, every single one is in the Revenue Operations team.

Five senior beltway roles in nine days. Nobody opens this many capture positions in a tight window unless the contract pipeline is real and the company expects to close meaningful business inside twelve months. The GM Maritime seat has been open since December. These five new seats are the supporting infrastructure that lets whoever fills the GM Maritime chair actually convert relationships into signed paper. Proposal Manager writes the responses. National Security Policy positions the company on the right side of every classified roadmap conversation. Business Operations runs the P and L mechanics that the GMs will be accountable for. Strategy maps the program landscape. Government Relations works the appropriations and authorizations cycles.

Stacked together with the three GM seats and the engineering SVP, what the public hiring graph reveals is a defense capture organization standing up in real time. The Lever postings are a leading indicator most public-equity investors do not read. They predate any of the company-side disclosures that the market will eventually price.

The broader Merlin hiring graph since the March 17, 2026 de-SPAC carries the same shape. Cert engineering hiring (Liz Pelberg-Schariter's FAA/DER recruitment footprint). Systems engineering, flight controls, autonomy roles, finance roles (Krishnan Anand's Supernal-displacement recruitment burst). Naval engineering pedigree (Pablo Gonzalez bringing the N-UCAS bench). Public-company communications (Michael Baker). SOCOM revenue capture (Mark Brunner). Pentagon-staff customer-axis leadership (Erin Staine-Pyne). The function-cluster spread is wider than a typical Series C autonomy startup. The seniority is the seniority of a public defense services company in scaling phase, not a venture-backed software firm.

The hiring cadence tells a tighter story than a generic hire-burst framing. Three moments in sequence. December 2025: Merlin closes the GM Mobility seat (Erin Staine-Pyne walks out of the Pentagon Under Secretary's office into the customer-axis chair owning AMC and AFMC). The GM Maritime requisition is opened in Lever the same week, but held in internal-state on the company's applicant tracking system. March 5, 2026: Merlin announces C-130J PDR passage on the USSOCOM IDIQ. The Tactical Autonomy customer-axis is now de-risked at the engineering gate. May 18-22, 2026: Merlin transitions GM Maritime from internal-state to public-state on the careers landing page and simultaneously posts five senior Washington DC Revenue Operations roles in nine days. The combination is the operating signal.

It would be unusual for a company to transition a senior P&L seat from a six-month internal-state hold to public-recruit while concurrently posting five DC capture roles in a generic environment. The combination is consistent with the company positioning ahead of a deal or contract event in the immediate three-to-six-month window. The five DC roles are the infrastructure that whoever fills GM Maritime would need to convert a relationship into signed paper, on this read. The three-to-six-month forward window puts the contract event in the late August to late November 2026 range, which brackets the September 16 lockup expiry, the August 14 Q2 earnings print, and the author-estimated Boeing MOU disclosure window (Q3 2026). On this read, the September 16 lockup acts as the de facto recruiting deadline and the August 14 Q2 earnings print as a natural contract-signing window.

Section IV, The Board, Read As A Customer-Relationship Map

Matt George's April 16, 2026 introduction of the Merlin board led with the phrase "A Secretary of the Navy." That phrasing was not random. Boards send signals. Five board members were named that day.

Kenneth Braithwaite is the 77th Secretary of the Navy. Served May 2020 through January 2021. Retired Rear Admiral. US Ambassador to Norway before serving as SecNav. What makes Braithwaite's appointment load-bearing is what he does at the other end of the day. He is Chairman of Fincantieri Marinette Marine, the US shipyard arm of Fincantieri that builds the US Navy Constellation-class frigate. He is Senior Advisor to Saronic Technologies, the unmanned surface vessel startup that just closed a three hundred ninety two million dollar production contract for Corsair USVs and is expanding its shipbuilding capacity by one and three-quarters of a billion dollars (the comp print Wall Street is watching for the autonomy-on-defense-platforms category). He is Senior Advisor to Kongsberg Defence, to Silicon Valley Bank, and to The Chertoff Group. Every single one of those adjacencies points at maritime defense, with a strong unmanned surface vessel cluster. Braithwaite's outside affiliations (Fincantieri Marinette Marine, Saronic Technologies, Kongsberg Defence, The Chertoff Group) all cluster in maritime defense and unmanned surface vessels. The board appointment is consistent with Merlin pursuing US Navy, US Coast Guard, and global maritime defense FMS business on a horizon that justifies the relationship cost. He did not join to chair the audit committee.

Kelyn Brannon brings Amazon-scale financial discipline. As Amazon's first Chief Accounting Officer (1999), she built the financial controls and accounting infrastructure for what became the largest cloud, retail, and logistics business in the world. The Merlin board uses Brannon as the audit-committee floor. The relevance for a pre-revenue defense company is that she is the executive who tells the CFO when a contract recognition policy is too aggressive and when a cost allocation should be re-categorized. That is the function defense investors care about when the company starts recognizing C-130J integration revenue under percentage-of-completion accounting (which it will, starting in late 2026).

Michael Montelongo. Former Air Force CFO (the 19th Assistant Secretary of the Air Force for Financial Management and Comptroller). Served 2001-2005. Currently sits on multiple defense and capital markets boards. The Air Force CFO seat in the Pentagon is the senior civilian who controls the entire USAF appropriations spend allocation, every Major Program Object Memorandum (POM) submission, and every congressional justification for every PE element in the Air Force budget. Montelongo's name on the board is the bridge to USAF AMC and AFMC at the most senior civilian budget-defending level. Combine Montelongo (USAF budget bridge) with Staine-Pyne (USAF Under Secretary's office) and the customer relationship to AMC and AFMC is the most over-built customer-axis arrangement in the public defense autonomy startup space.

Michael Blitzer is Lead Independent Director. He is co-founder of Kingstown Capital, a hedge fund focused on event-driven and special situations investing. Blitzer anchored approximately 80 percent of the $200 million PIPE at the de-SPAC. The Lead Independent Director seat is the most material non-executive board role in a public company; it is the seat that chairs executive sessions, that controls the Compensation Committee, and that is the natural counterparty to activists if any ever appear. Blitzer's PIPE anchor stake plus his lead-independent-director role means the largest non-management common shareholder controls the most powerful non-executive board chair. That is sponsor continuity at its most explicit. The lockup expiry on September 16 includes Blitzer's PIPE shares. Lead Independent Directors rarely sell through a lockup expiry because the optics weaken the activist-defense value of the seat. Modeling the September 16 lockup as a supply event with Blitzer's PIPE tranche held back is the base-case approach.

Section V, The DC Capture Capital Cluster

The board sits inside a denser DC capital cluster than the public summary suggests. Beyond the five named directors, the broader Merlin-adjacent capital cluster includes:

The Latham & Watkins legal bench is the second cluster. Nick Dhesi (the firm's senior SPAC partner) and Stephen Ranere (PE/M&A) led the Merlin-touch deal work through the de-SPAC. Haim Zaltzman (Vice Chair, Emerging Companies Group) is one degree away, with a deal book in the same window including AEVEX (defense autonomy adjacency), X-energy (nuclear power adjacency), and Cerebras (AI infrastructure adjacency). The firm's partner-celebration cadence around the Merlin close on March 16-17, 2026 was consistent with a high-prestige flagship transaction.

Roth Capital is the sell-side anchor. Sujeeva De Silva initiated coverage on April 14, 2026 with a Buy rating and an initial $15 price target. De Silva re-rated to $25 on April 15 after reading the S-1/A FY2025 financials (a 67 percent intraday revision tells you which number was the pre-information model and which was the post-information model). Roth's pre-IPO research relationship with the C-130J SOCOM customer-axis story is the only published sell-side model on file. Of course, it does not make the stock investable overnight. But it does change the discovery setup. For a small-cap company, published coverage, a third-party model, and a broker-distributed narrative can meaningfully improve visibility with institutional investors. That gate opened on April 14.

Northland Capital Markets is the highest-conviction next-sell-side initiator. Mike Latimore covers the defense autonomy SPAC graduate cohort (initiated RCAT with a Buy, initiated DPRO with an Outperform and a $20 target in January 2026, initiated SPAI on the same template). The probability-weighted prediction set has Northland Outperform initiation on MRLN at approximately 0.62 probability in a 60-120 day window. A Northland initiation on the Latimore template would carry a volume-spike event day-of with no other news; that is the tell that traders are watching.

Beyond the named directors and the sell-side, a flag-officer endorsement stack has been quietly assembling.

The endorsement stack is what hedge funds look at to determine whether a startup has crossed the credibility threshold for institutional procurement decisions. Twelve named endorsers spanning DARPA autonomy lead, retired AMC commander, FAA cert authority chair, active duty USAF mobility staff, Boeing technical bench, and crossover VC named partners is dense for a pre-revenue defense company nine weeks into its public life.

Section VI, The Customer Command Landscape

The customer set MRLN is positioned to service spans seven major commands across the US defense and homeland security architecture, plus the foreign military sales pipeline. Each command has its own budget line, its own program executive office, and its own decision-making timeline. The org chart above maps the customer-axis GMs to specific commands; this section traces each command back to its budget line.

USSOCOM. The first command. The lead contract vehicle is the C-130J autonomy IDIQ with award H9240824F0010 visible on USAspending at $15,955,778.32 obligated for PDR plus CDR. PDR completed March 5, 2026. CDR sits ahead. The total program IDIQ ceiling is $105 million per CFO Ryan Carrithers commentary on Q1 earnings. The program executive office is PEO Rotary Wing (which houses the C-130J inside SOCOM's airframe portfolio for special operations missions). Mark Brunner's SOFWERX channel positioning maps directly to this customer set. The capture pathway is the SOF Week Tampa conference cycle plus the SOFIC industry day cadence.

USAF AMC. The second command. Erin Staine-Pyne owns this customer axis end to end. The KC-135 tanker autonomy work runs under a 19-month integration contract awarded late 2024 to USAF Air Mobility Command and Air Force Materiel Command (under the AFLCMC Mobility Directorate). The 376-tail KC-135 fleet is the addressable population. The targeted demo window is late 2026 (the September 30 P0 catalyst). The CCR (Crew Reduction) RFP on SAM.gov is the highest-probability single contract event in the 30-90 day window, prediction set carries this at 0.78 probability based on the AFLCMC industry day March 25, 2026 plus the Staine-Pyne May 19 post engagement signature.

USAF AFMC. Owned by Erin Staine-Pyne. The C-130J reduced-crew work sits inside the USSOCOM IDIQ vehicle but the airframe sustainment, the integration engineering, and the certification path runs through AFMC. The 19-month tanker contract was a joint AMC-AFMC award. The C-17 follow-on autonomy potential rides on the same axis.

USAF ACC. This axis maps to the Tactical Autonomy lane, whose GM seat is currently vacant. The future-state customer relationship here is the Collaborative Combat Aircraft (CCA) program: Northrop Grumman is the prime on the Beacon program, Lockheed Martin and General Atomics hold separate CCA Increment 1 awards. The Northrop Beacon cohort enrollment via N437VN gives Merlin a CCA-adjacent named-flight slot. The named-flight has not landed. The watchlist tripwire is a Northrop press release naming Merlin on a Talon IQ flight, or a named slot at AFA Air, Space and Cyber, or at Farnborough.

MDA. The Missile Defense Agency contract visible on USAspending. Award number HQ085926FE898, agency MDA, description "SHIELD INITIAL ORDER", obligated $500. SHIELD is the Scalable Homeland Innovative Enterprise Layered Defense program inside the Golden Dome layered missile defense initiative. The $500 obligation is the floor on what is typically a much larger contract-vehicle ceiling. The award number format HQ085926 maps to the MDA contracting office, fiscal year 2026. This is Merlin's first publicly visible MDA contract. It has not been disclosed in any Merlin press release or 10-Q. The thesis implication is non-trivial: a fourth customer command (after USSOCOM, USAF AMC, USAF AFMC) just opened, and it is a command nobody had this company positioned on. Missile defense is not the same axis as crewed military aviation. The contract surface itself is the signal.

US Navy. GM Maritime seat (vacant). The latent customer relationship surfaces via the Braithwaite board seat: the former SecNav's other affiliations are Fincantieri (frigate yard), Saronic (USVs), Kongsberg (naval missiles + autonomy), and Chertoff Group (national security advisory). The natural P-8 Poseidon follow-on autonomy contract, the MH-60R helicopter autonomy contract, and the MQ-4C Triton autonomy contract are all OPNAV N98 (Air Warfare) decision lines. The GM Maritime hire is the seat that opens this customer relationship publicly.

US Coast Guard. GM Maritime seat (vacant). The job specification explicitly names USCG. The natural contract is the HC-144 Ocean Sentry maritime patrol airframe (35 aircraft fleet) and the follow-on King Air maritime patrol fleet. The decision line is USCG HQ-9 (Aviation Forces).

US DHS. GM Maritime seat (vacant). The job specification names DHS. The DHS Science and Technology directorate and Customs and Border Protection's aviation division are the two anchor decision lines. The CBP fleet includes 87 King Airs (per public DHS budget justifications), which makes the King Air autonomy lane structurally over-aligned with the CBP fleet population.

Foreign Military Sales. The Philippine Coast Guard signed for three King Air maritime patrol aircraft in February 2026 via Foreign Military Financing. Merlin as autonomy sub-tier on those airframes would be a Maritime LOB proof point. The UAE Remah International Group exclusivity agreement signed April 23, 2026 names a thirty-year UAE Armed Forces partner; the exclusivity language places Merlin's autonomy distribution into the UAE defense procurement chain through Tawazun Economic Council's offset program. The full international pipeline (Five Eyes plus Gulf plus Indo-Pacific) is treated in Section XIV.

Section VII, The Federal Contract Pipeline, Vehicle By Vehicle

A direct pull from USAspending.gov today surfaced three Merlin Labs DoD contracts and one notable program tie. The KC-135 AMC vehicle is real but not yet visible in the open obligations layer (a normal lag for AMC-line contracting where the contract sits inside a larger ID/IQ ceiling that obligates on task orders).

USSOCOM H9240824F0010. The award is the C-130J autonomy IDIQ. The $15,955,778.32 obligation covers PDR (Preliminary Design Review) plus CDR (Critical Design Review). PDR was completed March 5, 2026. CDR sits in the immediate future. The total program IDIQ ceiling is widely characterized at $105 million in trade press citing CFO Ryan Carrithers commentary on the Q1 2026 earnings call; readers should pull the call transcript directly to verify the exact phrasing. That ceiling, as commonly described, is the development authorization, not the production contract. IDIQ ceilings amend at the program executive office level as the program matures. The next observable task-order steps are the CDR award (covering critical design completion), the SIT (System Integration Test) award (covering ground integration), and the flight test award (covering first autonomy flight). Each task-order step is a separate SAM.gov notice. The watch channel is SAM.gov daily search on "Merlin Labs" or on the H9240824 contract vehicle prefix.

USSOCOM H9240824F0007. $7,535 obligated, marked as kickoff meeting. Administrative housekeeping on the same contract vehicle. Not a separate contract.

MDA HQ085926FE898. SHIELD INITIAL ORDER. $500 obligated. SHIELD is the Scalable Homeland Innovative Enterprise Layered Defense program inside the broader Golden Dome layered missile defense initiative. The $500 floor obligation is the canonical pattern when a federal contracting officer opens a contract vehicle with a placeholder initial order to lock in the awardee identity while the actual scope and ceiling get negotiated. The award number format HQ085926 maps to the MDA contracting office, fiscal year 2026. The program ceiling will become visible on the next obligation step. The thesis-relevant fact is the contract exists. Missile defense is a fourth customer command surface this company has now opened. The watch channel is the MDA HQ085926 contract-vehicle history on USAspending.

USAF KC-135 AMC. The 19-month integration contract awarded late 2024. The award was confirmed in trade press (Flying Magazine) and in USAF public statements. The visibility lag in USAspending is normal for AMC-line work where the contract sits inside a larger IDIQ ceiling. The next obligation step is the formal demo award (the September 30, 2026 P0 catalyst). The next contract step beyond that is the CCR (Crew Reduction) formal RFP on SAM.gov, which the prediction model carries at 0.78 probability in a 30-90 day window based on the AFLCMC industry day cadence and the May 19 Staine-Pyne engagement signature.

A-GRA Layer 3 CRADA. Cooperative R&D agreement under the A-GRA (Air-Ground Robotic Architecture) program. October 2025 scope. The competitive picture against Reliable Robotics is unsettled. The CRADA is not a procurement contract, it is a cost-sharing R&D vehicle that lets the Air Force evaluate both vendors' autonomy stacks against the same mission requirements. The follow-on procurement decision is the binary outcome. The watch channel is any AFRL or AFLCMC RFI or sources-sought notice for CCR aircraft platforms.

Beyond the visible contract layer, the SOFIC industry day cadence (April 14, 2026 announcement of Brunner as CRO), the SOF Week Tampa motion-driver pattern (Brunner LinkedIn engagement), the AFLCMC Mobility Directorate industry day (March 25, 2026), and the AFA Air Space and Cyber Conference cadence (September 14-16, 2026) are the recurring industry events where MRLN's contract pipeline density gets surfaced.

The forward contract pipeline math, modeled at the IDIQ-ceiling-utilization layer:

The Total program revenue column in each row is larger than the sum of the three named contract lines (C-130J + KC-135 + MDA) because each row also includes a residual line for FMS sub-tier work, commercial Condor advisory revenue, and existing FY26 NRE billings not tied to the new task orders. The bear case assumes the C-130J IDIQ steps draw at $20M for FY26, KC-135 demo slips into FY27, MDA SHIELD stays at floor, and the residual line tracks $5M. The base case assumes mid-pace task-order draws on C-130J, a successful demo on KC-135 (no CCR task order yet), a $2M MDA step, and a $5M residual. The bull case assumes accelerated C-130J task orders, demo plus first CCR sub-task on KC-135, a $5M MDA step, and a $10M residual. The moonshot case assumes a Boeing MOU lands, a maritime customer letter lands, the cumulative pipeline draws faster than current cadence suggests, and the residual line absorbs $25M of commercial Condor + FMS sub-tier work. None of these scenarios price the commercial Condor opportunity at its full long-tail.

Section VIII, Capital Structure And The Lockup Mechanics

The capital structure is the load-bearing risk surface in the next 111 days. Every share that unlocks on September 16 is a share that becomes available for open-market sale.

Float math. Total outstanding is 84,262,886 shares. Of that, 44.3M is locked through September 16, 2026 (Matt George 14.9M + Bleichroeder 8.8M + FR Capital 12.3M + SnowPoint 8.3M). The remaining ~40M consists of (a) ~22.4M PIPE common (also locked through September 16), (b) the SPAC IPO public shares that did not redeem, and (c) restricted founder/employee shares. The IPO float started at approximately 45M public shares before the merger; 90.3 percent redeemed at the closing, leaving roughly 4.4M public shares unlocked. The freely tradeable common float as of today is approximately 3.6-5.0M shares depending on how much of the unrestricted employee equity is treated as exercisable. That is the float that drove the short-squeeze dynamics in April 2026 when IBKR borrow utilization hit 99.93 percent. A float this tight on a name with a real catalyst pipeline sets up asymmetric upside response to any announcement.

Short interest. Per the most recent available data, IBKR borrow utilization sat at 99.93 percent in late April 2026. Short interest as a percentage of float is mechanically a multiplier on any catalyst-driven move. The short setup amplifies upside; it does not change the long-term thesis.

The lockup. Per the lockup agreement disclosed in the merger proxy and the Schedule 13D filings from Matt George (March 16, 2026) and Bleichroeder Sponsor 1 LLC (March 24, 2026), insider and sponsor shares are restricted through September 16, 2026. The 180-day count from the March 16 closing falls on September 12; the September 16 expiry date matches the language cited in the 13D filings themselves and aligns with the Conversion Date Reset on the Series A Preferred. The operating date for the lockup expiry is September 16, 2026 per the filings.

The conversion price of the Series A Preferred already reset from $12 to $6.67 on May 1, 2026, when the $80M PIPE priced below the original $12 conversion and triggered the full-ratchet anti-dilution in the Certificate of Designation (confirmed in Quiet Capital’s Schedule 13G). Separately, the Section 7(b) conversion-date reset on September 16, 2026 reprices off the 20-trading-day VWAP and is floored at $5.00. That mechanism can only hold the price at $6.67 or push it lower toward the $5.00 floor; it cannot move in common’s favor, and a lower conversion price means more shares issued on conversion. The $260M liquidation preference and the 12% cumulative dividend (~$31.3M per year, accruing whether paid or not) remain until the preferred actually converts.

The reset does not favor common. Best case it leaves the conversion price at $6.67; the worse cases ratchet it toward the $5.00 floor, each step adding dilution. So the risk to common is twofold: that incremental dilution, plus the supply pressure between today and September 15 if the catalyst stack has not delivered enough re-rate to absorb the lockup open supply.

PIPE warrants. The $253.9M warrant strike is at $11.50 per share (the standard SPAC warrant strike). Warrants are exercisable for one common share each, becoming in-the-money above $11.50. If the stock trades above $18 for 20 of any 30 consecutive trading days, the company can call the warrants for redemption at $0.01 (forcing exercise). This is the standard SPAC warrant "force-the-call" mechanism. At current price ($8.00 today), warrants are out-of-the-money. The warrant overhang adds approximately 22M shares of dilution risk if exercised cash-for-shares post-CDR.

Insider Form 4 history. No insider open-market buys have appeared on Form 4 since the de-SPAC. No insider open-market sells either (all insiders are locked through September 16). The first Form 4 after September 16 will be a load-bearing data point, either an insider buy (a 0.36-probability prediction at this point) or an insider sell (the default base rate for post-lockup behavior). The historical SPAC base rate is approximately 60 percent insider selling within the first 30 days post-lockup, although that base rate is skewed by SPACs that did not deliver on their pipeline.

The lockup supply scenarios:

The Lead Independent Director almost never sells through a lockup. Bleichroeder Sponsor 1 LLC is a SPAC sponsor whose normal exit pattern is gradual disposition over 6-18 months post-lockup (sponsors are not motivated by tax management to sell into the open; they have very low cost basis and gradual disposition is normal). FR Capital Holdings and SnowPoint Ventures are venture vehicles whose limited-partner reporting cycles motivate them to distribute or sell shares once liquid; their behavior post-lockup is the most uncertain variable in the supply scenario set.

The single biggest determinant of how the lockup absorbs is whether the catalyst stack delivers between today and September 16. Six P0 catalysts in the window. Each delivery reduces the lockup-absorption risk. Each miss increases it.

Section IX, Q1 2026 Earnings, Line By Line

Q1 2026 earnings landed May 14, 2026. The headline: revenue $1.0 million, adjusted EBITDA loss $23.3 million, GAAP net loss $90.4 million (96 percent non-cash mark-to-market on convertible promissory notes from the de-SPAC). Cash position post-PIPE approximately $183 million with no debt. CFO Ryan Carrithers said runway extends beyond the near-term certification and commercial milestones. CEO Matt George revealed Condor (the first product family) paired with a non-binding MOU with World Star Aviation for commercial cargo. The tape sold 10.5 percent on the resale registration becoming effective.

The non-cash MTM is the dominant signal in the loss statement. The convertible promissory notes from the de-SPAC are marked to market each quarter based on the underlying stock price. When the stock rises, the MTM loss expands (because the embedded conversion feature becomes more valuable to the noteholder); when the stock falls, the MTM loss contracts. The $86.7M Q1 MTM was driven by the stock's price action in the quarter (it traded from $14.29 on April 15, the day Roth raised the target to $25, through the May 14 Q1 print at $6.97). The MTM dynamics will reverse and unwind when the convertible notes convert (the conversion is forced when certain SOFR-linked triggers fire, or at maturity).

Cash position. The $183M post-PIPE cash position is the runway anchor. At Q1 cash burn of approximately $23M, that is approximately 24 months of conservative runway (no revenue growth, no margin improvement). With FY2026 revenue tracking toward management's $32M guidance (an 8x lift from FY2025's $7.55M), the burn drops materially and runway extends through 2028 without a raise.

Revenue composition. The Q1 2026 $1.0M was predominantly USSOCOM C-130J IDIQ NRE work, with a small commercial advisory contribution. The transition from NRE-dominated revenue to per-aircraft software-licensing revenue is the margin inflection point that defense-tech investors watch for. That inflection has not happened yet. The Q2 2026 10-Q (August 14, 2026 P1 catalyst) is the next observable data point. The watch question is whether Q2 revenue carries any percentage-of-completion accounting recognition on integration work in addition to NRE milestone billings.

The $32M FY2026 revenue guide. Management has not provided explicit per-quarter revenue guidance but the implied path from $1.0M in Q1 to $32M for FY2026 requires materially higher quarterly recognition in Q2/Q3/Q4. Two paths get there:

Path 1: Lumpy NRE milestone billings. Each of the next three CDR-adjacent task-order awards on the USSOCOM C-130J IDIQ triggers an NRE billing event. If three steps land in the FY26 period at $5-10M each, that delivers $15-30M of NRE revenue.

Path 2: Percentage-of-completion recognition. Once the C-130J integration work crosses CDR and enters SIT (System Integration Test), the program shifts from milestone-billing accounting to percentage-of-completion accounting. POC recognition can begin recognizing revenue over the development period rather than at discrete milestone gates. This is the bigger accounting shift and the one that gets defense investors to re-rate the multiple.

Both paths are visible in management commentary. Neither has been documented publicly with timing precision. The Q2 10-Q is the first hard data point.

CFO Ryan Carrithers said on the Q1 call: "Our runway extends beyond the near-term certification and commercial milestones." That is the precise phrasing CFOs use when they want the market to know the company is not raising capital before September. Translation: no PIPE follow-on, no convertible add-on, no equity offering between now and the end of the certification cadence in late 2026. The earliest plausible capital raise window is post-CDR (late 2026) or post-lockup (October 2026 if the price action permits).

Condor. The Condor product family was the most material Q1 reveal. Condor is the multi-crew aircraft product family that wraps Merlin Pilot for Commercial Cargo. The non-binding MOU with World Star Aviation for commercial cargo is the first publicly disclosed commercial partner for Condor. The product targeting is 737 and 767 freighter conversions, an addressable fleet that is materially larger than the C-130J/KC-135 mil-aviation universe. Condor's commercial unit economics, when published, will shift the long-term TAM model from a pure-defense story to a defense-plus-commercial-cargo story.

Section X, Competitor Positioning Matrix

Five competitors and one capability adjacency dominate the autonomous aviation procurement landscape MRLN is competing in. Each occupies a slightly different position; none is a one-for-one substitute.

Reliable Robotics. The most direct technology competitor. Their Cessna 208 autonomy program has demonstrated remote-piloted flight. They sit on the A-GRA Layer 3 CRADA opposite Merlin and are the head-to-head bid for the same Air Force evaluation cycle. Their differentiation is a more cargo-civilian-focused certification posture. Merlin's differentiation is the defense-customer-axis P&L architecture and the broader airframe coverage (Cessna 208 plus King Air plus KC-135 plus C-130J versus Reliable's Cessna-only public footprint). The procurement decision likely splits on customer axis: Reliable Robotics on the FAA-cert-cargo lane, Merlin on the defense-mission lane. The October 2025 CRADA scope is the binary determination.

Shield AI. The named-flight CCA competitor on Talon IQ. Shield AI's Hivemind software flew on N437VN on March 19, 2026 as the most recent rotation slot. Shield AI's customer-axis is the air combat command CCA program; their product is the multi-platform autonomy stack designed for fighter-class missions. Merlin's CCA-adjacency comes through the Beacon partner enrollment (the named-flight has not landed). Shield AI is the fighter autonomy company. Merlin is the airlift and tanker autonomy company. They do not compete on the same airframes. Both are in the Northrop Beacon cohort and both can be customers of Northrop on different airframe lines.

Applied Intuition. Did the software swap-test on Talon IQ in March 2026. Applied Intuition's core competency is autonomy simulation infrastructure. Their CCA play is the test-environment provider for autonomy verification rather than the autonomy stack itself. The competitive overlap with Merlin is minimal; the partnership read is more likely. Applied Intuition's simulation platform could be the cert-environment Merlin uses to do its STC test-card execution.

Anduril. The full-stack defense autonomy company with Lattice and the Fury CCA contender for the USAF ACE (Air Combat Evolution) program. Anduril's customer-axis is the full DoD enterprise, with the largest revenue mass currently in counter-UAS systems. No direct overlap on the C-130J/KC-135/Caravan airframes. Anduril is the comparable in the public-equity-comp question hedge funds ask ("what is Merlin worth in defense autonomy multiples?"). Anduril's most recent reported valuation is $14B+ private; if it goes public via S-1 in the next 12-24 months, it sets the multiple bar for the category.

Aurora Flight Sciences. Boeing's wholly-owned autonomy subsidiary. The internal CCA work and the X-66A Sustainable Flight Demonstrator program both run through Aurora. Aurora is locked into Boeing's commercial and defense roadmap; they cannot easily be a third-party supplier to the rest of the OEM matrix. That structural lock is exactly why Boeing Defense has reached out to Merlin's technical bench (PD Weber + David Zeitouni). Aurora's existence does not preclude Boeing from sourcing a third-party autonomy supplier for non-Aurora-controlled programs.

The capability adjacency to watch: Saronic Technologies (unmanned surface vessels). Saronic is not an air autonomy company. Saronic is the maritime-autonomous comp print. Saronic's Series D at $9.25B private valuation, and the Corsair USV $392M production contract, sets the public-market expectation for what an autonomous-systems-on-defense-platforms company is worth at scale. The Braithwaite-as-Saronic-Senior-Advisor data point links MRLN's board directly to the Saronic comp. The Saronic IPO timing (rumored 2026-2027) is the comp print Wall Street is waiting for; an MRLN-to-Saronic multiple comparison would put MRLN at 5x its current EV based on the Saronic private-market comp.

Section XI, OEM Relationship Matrix

The OEM relationship surface is the second-largest revenue lever after the federal customer surface. Six major OEMs sit in the matrix; each has a different lane to Merlin.

Honeywell. The October 17, 2024 MOU commits Honeywell and Merlin to integrate Merlin Pilot with the Anthem flight deck. The Anthem-Merlin program has its canonical public disclosure venue ahead on June 3, 2026 (HONA Investor Day in Phoenix). Unit economics, addressable Anthem fleet, and the full HONA scenarios are treated in dedicated detail in Section XIII.

Boeing Defense. The technical-bench engagement is documented via two named contacts: PD Weber on the MQ-25 Advanced Development team, and David Zeitouni who carries the title Senior Technical Fellow at Boeing. The Senior Technical Fellow rank inside Boeing's engineering hierarchy is the highest individual-contributor technical grade in the company; there are fewer than 200 Senior Technical Fellows across all of Boeing's commercial and defense business units. A Senior Technical Fellow does not engage with a startup unless there is sustained technical diligence behind the engagement. The latency from Boeing technical-bench engagement to public MOU is typically six to twelve months. The window opens July through September 2026. The watchlist tripwire is any Boeing Defense PR mentioning an autonomy partner, or a joint HONA-Boeing-Merlin co-mention (June 3), or a Farnborough Air Show MOU (July 20-24), or a named program-axis disclosure at AFA Air, Space and Cyber (September 14-16). The Aitech Group joining Boeing's MQ-25 Industry Team announcement (separate from Merlin) is what flagged the second technical lane. Boeing's technical-bench engagement with Merlin is one of several visible third-party autonomy diligence tracks for non-Aurora-controlled program lines.

Lockheed Aeronautics. No public signal. Lockheed's preferred internal autonomy effort runs through the Skunk Works. Lockheed has internal autonomy capacity through the Skunk Works, so a Lockheed-Merlin engagement would be on a non-fighter platform (P-3 follow-on autonomy, KC-130J variant, C-5 variant) where Lockheed's internal capacity does not extend.

Northrop Grumman. The Beacon partner enrollment. Northrop's Beacon program (announced July 30, 2025) is a partner cohort program for autonomy integration on Northrop's CCA-class platforms. Merlin is one of six named Beacon partners. The N437VN airframe (the Scaled Composites Model 437 Vanguard) is the partner testbed; named-flight rotations happen at industry events. Shield AI flew March 19, 2026. Applied Intuition plus Accelint swap-tested in March 2026. Merlin's named-flight slot is overdue. The tripwire is a Northrop press release naming Merlin on a Talon IQ flight, or a named-flight rotation slot at AFA Air, Space and Cyber (September 14-16) or Farnborough (July 20-24).

GE Aerospace. The Autonomy Core co-development initiative was announced at the AFA conference on September 23, 2025. The GE-side counterparties are Matt Burns (GM Avionics Systems) and Jeremy Barbour (GM Connected Aircraft); the Merlin-side counterparty is Matt George (CEO). The agreement pairs GE Aerospace's Flight Management System (FMS) and Modular Open System Architectures with Merlin's autonomy stack, avionics packages, and Datalink. The first-target program is the USAF KC-135 Center Console Refresh (CCR). The expansion path named in trade press includes C-130J, Boeing 737, P-8 Poseidon, and KC-46 (all FMS-installed platforms). GE Aerospace's FMS installed base is approximately 14,000 aircraft.

Critical distinction: the GE Aerospace Autonomy Core / KC-135 CCR program is a separate track from the existing USAF KC-135 Merlin Pilot integration contract awarded under the 19-month AMC/AFMC vehicle in late 2024. The two are adjacent and structurally built on the same airframe, but they are different contract vehicles with different scopes. The 19-month USAF integration contract is the direct USAF-to-Merlin work that the September 30, 2026 demo target window covers. The Autonomy Core / CCR program is the broader cockpit modernization competition that GE Aerospace and Merlin are jointly bidding against; the CCR formal RFP on SAM.gov is the next observable milestone in that program (prediction model carries this at approximately 0.78 probability in a 30-90 day window). Both programs run on the KC-135 airframe; both build on the same Merlin Pilot autonomy stack; but they are separate procurement tracks.

The agreement is co-development, not contractually exclusive in primary source language, but it places Merlin on the GE FMS roadmap for the canonical military airlift and tanker platform set. The strategic implication is that the KC-135 program now has a three-way alignment between USAF (the customer), GE Aerospace (the FMS provider), and Merlin (the autonomy stack) for the CCR competition, in addition to the existing direct USAF-Merlin 19-month integration vehicle. The HONA Investor Day on June 3 may reference the GE-Merlin Autonomy Core in the "additional integration partner" framing, since the partnership is structurally non-overlapping with the Honeywell Anthem MOU (GE FMS sits in commercial-tanker airframes; Honeywell Anthem sits in business-jet airframes). MRLN's OEM matrix is therefore not Boeing-Honeywell-Northrop-Merlin-vs-the-world; it is Boeing+Honeywell+Northrop+GE-Aerospace-and-Merlin-on-FMS, with two separate KC-135 tracks running in parallel.

Pratt & Whitney. No public signal on the autonomy axis.

The OEM matrix update reads as: Honeywell is the documented active partner with the canonical disclosure venue ahead in 6 days; GE Aerospace is the documented active partner via the Autonomy Core co-development announced September 23, 2025, targeting the KC-135 CCR competition; Boeing is the active diligence partner with a 90-180 day MOU window inferred from the technical-bench engagement; Northrop is the cohort partner with a named-flight slot overdue; Lockheed and Pratt have no public engagement signal.

The next two events that move this matrix are the HONA Investor Day on June 3 and a Boeing Defense press release within the July through September window. Either materially re-rates the OEM relationship surface.

Section XII, Northrop Beacon Cohort, Partner By Partner

The Northrop Grumman Beacon program is the named cohort of autonomy partners that Northrop has enrolled for the Talon IQ testbed and the broader CCA-adjacent collaboration. The N437VN Vanguard (Scaled Composites Model 437, hex N437VN) is the primary partner testbed; the Talon IQ is the demonstration airframe family. Named-flight rotations are how Northrop publicizes which autonomy partner's stack is flown.

The six-partner cohort structure means each partner gets a named-flight rotation slot every 8-12 weeks. Shield AI's March 19 flight covered combat air patrol plus target engagement at Mojave. Applied Intuition plus Accelint's swap test demonstrated mid-flight autonomy skill-switching. Merlin's rotation slot has not yet landed. The expected window is summer through fall 2026 based on the cohort's rotation cadence.

The N437VN airframe itself is the most quiet of MRLN's tracked partner testbeds. Six total sorties across the entire 188-day window. Last seen April 15, 2026. Forty-three days dark as of today. The dormancy is consistent with the program being between named-partner flight slots. The dormancy is logged as an open question.

The Beacon program is also the primary public surface for Northrop's CCA-adjacent autonomy partner roster. The CCA program (Collaborative Combat Aircraft) is the USAF's loyal-wingman-class autonomous combat aircraft initiative. CCA Increment 1 was awarded to General Atomics (YFQ-42) and Anduril (YFQ-44). CCA Increment 2 is the next contract round. Northrop is widely expected to bid CCA Increment 2 with a Northrop-led prime structure leveraging the Beacon cohort partners for the autonomy stack subsystem. If Northrop bids and wins, the Beacon partners become structurally embedded in the CCA Increment 2 supply chain.

The Merlin-on-Talon-IQ named-flight is the immediate watchlist event. The deeper Northrop-Merlin relationship to watch is the CCA Increment 2 source selection (anticipated late 2026 or early 2027).

Section XIII, Honeywell And The Anthem Through-Line

The October 17, 2024 MOU between Honeywell and Merlin commits both parties to integrate Merlin Pilot with Honeywell's Anthem flight deck. Honeywell announced the intent to spin off its Aerospace segment in February 2024; as of this writing the separation has not legally closed and HON remains the parent ticker. Industry materials use "HONA" as shorthand for the segment and its forthcoming standalone identity, the convention I use throughout this document.

Anthem is Honeywell's next-generation flight deck system. The aircraft using Anthem in production or commitment include:

The total Anthem-equipped business jet fleet exceeds 2,200 aircraft when you sum Bombardier Global, PC-12 NXG, and the eVTOL development programs. The autonomy unit economics on this fleet (assuming a $100-250K per-aircraft software license over a 15-year deck lifespan) represent a $220M to $550M lifetime revenue opportunity from the Anthem integration alone, without counting new platform additions.

The HONA Investor Day on June 3, 2026 in Phoenix is the inaugural investor event for the post-spin entity. T+6 days from this document. HONA's investor relations will use the Investor Day to introduce the post-spin business segments, the long-term revenue and margin model, and the partnership roadmap. The Anthem segment is the part of HONA's product portfolio most relevant to MRLN.

Two specific scenarios to watch for at HONA Investor Day:

The combined probability of any Merlin-favorable mention (named-flight, named-customer, or additional-OEM footnote) is approximately 0.60. The probability of an unambiguous re-rate event is approximately 0.42. The HONA Investor Day is the canonical disclosure venue. Any positive disclosure at HONA materially affects Q3 MRLN positioning.

Beyond HONA, the Anthem-Merlin integration roadmap has natural fold-in points at the Bombardier Global product release cycle (rolling), the PC-12 NXG OEM update cycle, and the eVTOL certification milestones for Vertical Aerospace and Embraer Eve. Each of those gives Anthem-equipped autonomy a public disclosure surface.

Section XIV, Foreign Customer Pipeline

The international customer pipeline is the second-largest soft-pipeline lever after the US DoD customer set. Five distinct foreign tracks are open or in active diligence.

New Zealand. The certification testbed campaign at Kerikeri. Four MRLN-owned or operated Caravans operate from Kerikeri under NZ regulatory authority: ZK-MLN, ZK-MLO, ZK-MLQ, and ZK-JMP (the last operated under Part 135 lease from Tandem Skydiving). The NZ CAA is the regulatory partner for the parallel certification path; NZ MoD is the latent customer should the certification path lead to operational fielding for NZDF airlift or maritime patrol. The NZ CAA's autonomous aviation regulatory framework is more permissive than the FAA's, which is what enables the certification-cadence advantage of the Kerikeri campaign. The 188-day flight backfill confirmed the NZ certification cohort produced 94 sorties in March 2026 (the biggest single-cohort-month in the dataset). Grant Crenfeldt (CEO NZ) is the on-site operator for the Kerikeri campaign; his March 18, 2026 public statement celebrated the de-SPAC.

United Arab Emirates. The Remah International Group exclusivity agreement signed April 23, 2026. Remah International Group is a three-decade UAE Armed Forces partner with distribution rights to Northrop Grumman and SAAB platforms across the Gulf region. Damian Killeen of Remah was the public spokesperson on the signing. The exclusivity language places Merlin's autonomy distribution into the UAE defense procurement chain through Tawazun Economic Council's offset program. The UAE defense procurement environment carries a $23B FY2025 defense budget; EDGE Group's autonomous portfolio is the in-country prime competitor. The Merlin-Remah pairing is positioned to compete for UAE autonomy procurement on imported airframes (C-17, IL-76MD variant, A330 MRTT). The UAE FMS pipeline is one of the largest unannounced revenue surfaces in the autonomy stack. Watch the Tawazun Economic Council 2025-2030 acquisition plan for autonomous platform commitments.

Philippines. The Philippine Coast Guard signed for three King Air maritime patrol aircraft in February 2026 via Foreign Military Financing. The aircraft delivery cadence and the autonomy sub-tier are the variables. If Merlin is selected as the autonomy sub-tier on those airframes, the contract value would be approximately $5-15M per airframe in autonomy license + integration cost, or $15-45M total over the program. The disclosure pathway is either through the DSCA FMS notification or through a Merlin IR press release. Watch the DSCA monthly FMS announcements for "King Air" plus "Philippines" plus "autonomy" or "crew reduction."

Australia. The RAAF (Royal Australian Air Force) operates a substantial KC-30A tanker fleet, a C-17 mobility fleet, and a P-8 Poseidon maritime patrol fleet. Any RAAF autonomy adjacency would come through a Five Eyes cooperation pathway. No formal engagement is publicly disclosed. The probability of a 12-month engagement disclosure is approximately 0.34.

Japan. The JASDF (Japan Air Self-Defense Force) operates a C-2 transport fleet, a KC-46A tanker fleet, and a P-1 maritime patrol fleet. Japan's defense autonomy posture under the Self-Defense Force Act constraints is more limited than the Five Eyes; the indirect pathway is through Mitsubishi Heavy Industries (the prime on C-2). No formal engagement is publicly disclosed. The probability of a 12-month engagement disclosure is approximately 0.20.

The international pipeline is the strongest single source of long-tail TAM expansion that hedge funds typically under-model. The UAE/Remah exclusivity alone could justify a 10-20% multiple expansion on the long-term revenue projection if it converts to signed contracts.

Section XV, Insider And PIPE Transaction Log

The insider transaction record post-de-SPAC is short by design (the lockup constrains both buying and selling). The relevant historical record is the Schedule 13D and Schedule 13D/A filings around the de-SPAC closing.

No Form 4 open-market buys by insiders since the de-SPAC. No Form 4 open-market sells either (all locked). The first Form 4 after September 16, 2026 will be a load-bearing data point.

The historical SPAC post-lockup behavior pattern:

• Sponsor (Bleichroeder) usually disposes gradually over 6-18 months post-lockup; tax management is not a motivator given the $0.004 cost basis.

• Insiders (George) typically do not sell aggressively post-lockup if conviction is high; his public engagement footprint suggests conviction is high.

• VC vehicles (FR Capital, SnowPoint) typically distribute to LPs post-lockup if mandate timing requires liquidity; this is the most variable category.

• PIPE anchors (Kingstown / Blitzer) typically do not sell into a lockup; the Lead Independent Director seat constrains visible selling for governance optics.

The Form 13F-HR filings from institutional buyers will become the parallel insider read. Watch for Alyeska 13G crossing 5% (a 0.34 probability prediction in the watch set). Watch for Rockefeller GFO (Najar) disclosing an MRLN position in a 13F (0.30 probability). Watch for ParaFi PM (Yedid-Botton) crossover position appearing in 13F or PIPE participation (0.22 probability).

The August 14 Q2 2026 10-Q will disclose share count updates and any post-lockup share class changes. The October-November Q3 10-Q will be the first quarterly print with post-lockup ownership rebalancing visible.

Section XVI, The Full Hiring Graph Since De-SPAC

The Merlin hiring graph since the March 17, 2026 de-SPAC has been the highest-frequency public-data signal of forward operating cadence. Function clusters and seniority distribution are the underlying intel.

The cumulative public-company hiring cadence implies an organization scaling from approximately 200 employees pre-SPAC to a 350-450 employee target by year-end 2026. The fully-loaded annual personnel cost (assuming $200K average fully-loaded compensation for the senior cohort and $150K for the broader engineering cohort) is approximately $60-75M annual personnel expense at the 400-employee target.

The hiring burst is one of the highest-conviction operator signals tracked. The October 2025 RFI for the AFLCMC Mobility Directorate, the March 25, 2026 AFLCMC industry day, the May 19 Staine-Pyne KC-135 post drawing four active duty USAF officers, and the May 18-22 DC capture role burst all sit in a coherent timeline that implies a forward contract pipeline real enough to justify the personnel-cost ramp.

The two specific hiring tells to watch over the next 90 days:

(a) GM Maritime fill, the largest single hiring event in the pipeline. Whoever fills this seat will be a recently-retired Navy or Coast Guard captain/rear-admiral. The 8-K Item 5.02 disclosure is the formal channel; the LinkedIn announcement is the leading-edge signal. Probability of fill in next 90 days: approximately 0.70.

(b) Founding-tier engineering hires for the maritime autonomy program. The hiring burst that follows a GM Maritime fill will be a 5-10 person engineering ramp on the new line of business. Watch the Merlin Lever portal in the 30 days after the GM Maritime announcement for the maritime engineering cluster.

Section XVII, The Per-Airframe Full Ledger

The Merlin Intelligence dashboard tracks 54 airframes across seven mission cohorts. The 188-day flight dataset (November 21, 2025 through May 28, 2026) is the dataset of record.

ZK-JMP. The Tandem Skydiving Caravan under Merlin Part 135 lease. The busiest single tail in the dataset at 254 sorties and 125.1 hours. The March 2026 month produced 85 sorties (the largest single-tail-month in the dataset). The April pause (1 sortie) is consistent with a major maintenance period or instrumentation refit. The May rebound to 31 sorties confirms the testbed is back to operational tempo. ZK-JMP's role is the primary NZ-cert testbed under Part 135 operating authority, which is the most flexible regulatory context for autonomy test cards (Part 135 commercial-pilot operations have broader instrumentation and procedure latitude than Part 91 private operations).

N7198Y. The RU-21H Ute under SIGINT/COMINT mil-variant lineage operating from KSLI Los Alamitos Army Airfield (California Army National Guard). 156 sorties, 334.2 hours. The hours-per-sortie ratio is 2.14 (highest in the priority tail set), consistent with long-loiter ISR mission profiles. The ramp from 8 sorties in January 2026 to 44 sorties in April 2026 marked the steepest accelerating phase. The current 30-sortie last-30-days is a slight cooling but still heavy. The basing from KSLI (Army Guard airfield in California, 1,855 nm from Dynamic Aviation HQ in Bridgewater VA) is consistent with an active Army/CalGuard/domestic ISR contract; the geographic outlier from Dynamic's normal Bridgewater operations is what made this tail re-tier from generic dynamic_ruleout to king_air_isr_watch. The thesis-relevant flag: any flight pattern shift toward Mojave (KMHV) or PAX River (KNHK) would signal maritime LOB demo activation. The 9 nm from Long Beach proximity is the secondary unusual data point.

N208B. The "Big Red" Cessna Caravan at Quonset. The primary US certification testbed. 84 sorties, 68.0 airborne hours, 40 active days. The April 2026 month was its biggest at 25 sorties. The single-sortie March month is anomalous against the rest of the fleet's March peak. The Quonset State Airport (KOQU) basing is the East Coast cert hub.

ZK-MLP. The Cessna TU206F Turbo Stationair at Kerikeri. 78 sorties, 61.9 hours, 43 active days across the 188-day window. The pattern is not load-bearing for the certification-program thesis. ZK-MLP is most likely a Kerikeri-based service / charter aircraft or a separate program rather than a Merlin autonomy testbed. It is tracked here only because the airframe sits inside the Kerikeri ramp footprint; do not weight it.

N43A. The C-12 Huron at KSHD (Shenandoah Valley Regional Airport, near Dynamic Aviation HQ). 62 sorties, 42.5 hours, 32 active days. Activated in February 2026, peaked in April at 28 sorties. The cooling to 18 sorties in May is the current trend. C-12 Huron is the King Air 200 mil-variant; the Dynamic Avlease ramp suggests a contracted ISR cycle.

ZK-MLO. The Super Cargomaster at Kerikeri. 58 sorties, 57.3 hours, 26 active days. Cold for four months, then activated in March 2026 and accelerating every month since. The auto-signal flagged it as ACCELERATING (+13, 33 vs 20). ZK-MLO is in the steepest accelerating phase of any Merlin-controlled airframe currently.

N713CB. The second US certification Caravan at the Caravan Charters facility. 34 sorties, 21.8 hours, 22 active days. Activated February 2026, steady at 10 sorties per month for three months, dropped to 4 in May. The flat-then-fall pattern reads like a discrete test campaign that completed.

ZK-MLQ (with VH-WZY merged). 31 sorties, 22.1 hours, 13 active days. The historical pattern of 21 sorties January + 9 sorties February (matching an independent intel source's "21 sorties Jan + 9 sorties Feb" description) confirms identity merge with the prior VH-WZY registration. The tail was a Cessna 208B re-registered from Australian VH-WZY (hex 7C72FC) to NZ ZK-MLQ (hex C83851) on February 11, 2026. The 3-month silence from February through May 26 was the open conversion gap; ZK-MLQ re-emerged on May 27, 2026 with a 28-minute pattern circuit at Kerikeri (KKE → KKE, 03:51-04:19 UTC). This is treated in dedicated detail in Section XIX.

N437VN. The Scaled Composites Model 437 Vanguard at Mojave. 6 sorties, 10 hours, 5 active days. Last seen April 15, 2026. 43 days dark. Treated in Section XII (Beacon cohort).

ZK-MLN. The original NZ certification Caravan. 4 sorties, 3.5 hours, 2 active days. The lowest-flying merlin_owned tail in the dataset. Was the testbed Part 9 leaned on; in the extended dataset, ZK-JMP under lease has done the lion's share of NZ cert sortie generation while ZK-MLN sits quiet. This is a re-prioritization the company has not announced.

N506DB. The Burton Long-EZ at Mojave. R&D testbed, intermittent transponder behavior, ground-bound for the entire 188-day window per public ADS-B.

Beyond the priority tails, the 10-tail KC-135 host pool at the 171st ARW (treated in Section XXI), the 33-tail Dynamic Aviation King Air pool (rule-out cohort), and the secondary partner platforms round out the full 54-airframe FLEET roster.

Section XVIII, The Spring 2026 Ramp, Read Across Cohorts

Part 9 introduced the finding that Merlin's certification fleet flew the busiest stretch of the entire window in March and April 2026. With the dataset extended through May 28 and the new tails folded in, the spring ramp does not look like a one-off pulse. It looks like a structural pace change.

Read across the rows.

The NZ certification cohort peaked at ninety-four sorties in March 2026, the single biggest cohort-month in the entire dataset. The April number drops to twenty-three (most of that drop is the ZK-JMP maintenance pause, which went from eighty-five sorties in March to one in April). May rebounds to fifty-eight as ZK-JMP picks back up to thirty-one and ZK-MLO accelerates to twenty-seven.

The US certification cohort hit its biggest month in April at thirty-five sorties. April was the month N208B "Big Red" did twenty-five sorties by itself. May settled to twenty-one across both tails. Steady, not a one-month spike.

The mobility tanker cohort (the ten KC-135 host-pool tails at the 171st ARW) ran sixty-four and eighty sorties in February and March, then dropped to nineteen and seventeen in April and May. The 171st ARW's tempo is driven by both the Merlin autonomy demo work and the squadron's own readiness flying, so the drop is consistent with end-of-quarter taskings clearing.

The new ISR watch cohort (the two re-tiered King Air mil-variants, N43A and N7198Y) ramped from zero in November/December to forty-four in February, fifty-three in March, seventy-two in April, and thirty-nine in May. That is a cohort that did not exist before the re-tier, and it has been quietly running heavy ISR sorties through the spring out of California.

The cca_testbed cohort (N437VN, the Vanguard at Mojave) ran just six sorties across the entire window. The last trace was April 15. The Beacon program testbed has been quiet for forty-two days as of today. Treated as an open question.

The dynamic_ruleout cohort (the thirty-three Dynamic Aviation King Airs out of Bridgewater) is included only as a noise filter. It scaled from zero in November/December to two hundred sixty-six sorties in May. None of that is Merlin signal. The cohort is Dynamic's commercial ISR contract business with DEA, CBP, NSA, and CalFire. The dashboard surfaces these tails on the rule-out tier specifically so a casual reader does not mistake their tempo for Merlin's. The exact pattern that made the re-tier of N43A and N7198Y out of this cohort was the hex-database mil-variant flag combined with the SIGINT/COMINT lineage of the airframe types.

The spring ramp Part 9 described is not over. It has redistributed across more airframes. The King Air ISR watch cohort is a new source of tempo. ZK-JMP came back from maintenance hot. ZK-MLO is in steep acceleration. ZK-MLQ re-emerged May 27 with a 28-minute cert circuit. The certification campaign is doing exactly what a certification campaign in its execution phase does: spread the load across more airframes as each subsystem clears its test card.

Section XIX, The Conversion Gap And Its Resolution

The airframe at the center of this section is a Cessna 208B identity-merged between Australian VH-WZY (hex 7C72FC) and NZ ZK-MLQ (hex C83851), re-registered on February 11, 2026. The historical pattern read 21 sorties in January 2026, 9 sorties in February 2026, then a complete public-trace silence from February 4 through May 26.

The pattern was unambiguous in regulatory terms. A certification testbed in the SOI-2 to SOI-3 transition (FAA Stage of Involvement 2 to Stage of Involvement 3) goes silent when:

1. Test article modification, major hardware changes (new instrumentation, new sensor, new flight control package) require ground-test cycles before resumption of flight test.

2. Test card pause, a finding during one test card requires resolution before the next card can be flown.

3. Regulatory hold, the certification authority requires documentation submission and review before the next flight authority is granted.

4. Test data lock, the certification campaign has reached a discrete milestone (a Means of Compliance demonstration, a Stage transition) and the regulatory authority is processing the test report.

All four options were consistent with a roughly 3-month pause after 30 sorties of flight test (21 in January + 9 in February). The 30-sortie boundary is roughly the duration of a single discrete certification test phase. The 3-month duration is roughly the regulatory review period for a Means of Compliance submission.

The gap closed on May 27, 2026. ZK-MLQ flew a 28-minute pattern circuit at Kerikeri (NZKK / KKE) on hex C83851, departing 03:51 UTC, landing 04:19 UTC. That is the textbook profile of a post-modification or post-pause initial-resumption flight: a short local circuit at the cert facility to validate the airframe is safe-to-fly before moving back into the longer test cards. A 28-minute circuit is not data collection, it is system check.

The May 27 trace has one methodological anomaly worth a callout for any analyst reproducing this work. The adsb.lol metadata for hex C83851 has its registration and aircraft-type fields empty, the public hex-to-registration databases have not yet ingested the February 11 NZ re-registration. Any registration-keyed lookup ("show me ZK-MLQ flights") returns zero on adsb.lol, FlightAware, FR24, and ADSBExchange. Only a hex-keyed lookup ("show me hex C83851 flights") surfaces the data. The dashboard's flight-history pipeline is hex-keyed, which is why the May 27 flight populates correctly after refresh.

The thesis-relevant read after the resumption: ZK-MLQ is not a failed testbed. It is a paused testbed that just exited its discrete regulatory cycle. The watch-list state flips from "silent for three months" to "re-emerged May 27, 2026, watch for cadence rebuild." The next observable signal is either a string of follow-on circuits at Kerikeri (cert campaign resuming) or a single shake-out and then another silence (incident or hardware finding). The first two weeks post-resumption are the diagnostic window.

The independent confirmation pattern is the most useful methodological element. An external intel source described "21 sorties Jan + 9 sorties Feb + 3 months silent" without seeing the dashboard data; the dashboard confirms exactly the same numbers. The same source flagged the May 27 resumption flight independently, including the precise UTC timestamps and the airport pair (KKE → KKE). That is a clean third-party confirmation of both the data-extraction methodology and the identity-merge logic.

Section XX, The King Air ISR Watch Cohort

The two re-tiered King Air mil-variants (N43A and N7198Y) are the most thesis-relevant data surface that was not in the public dashboard before the May 2026 update. The re-tier itself is the story.

N7198Y. The RU-21H Ute. The RU-21H is the militarized SIGINT/COMINT variant of the King Air, manufactured for the US Army for SIGINT collection missions in the 1970s-80s. The mil-variant database flag plus the airframe lineage plus the operating pattern combined justified the re-tier from generic dynamic_ruleout (Dynamic Aviation's broad commercial King Air pool) to king_air_isr_watch (the watch-tier for King Air mil-variants on active mission profiles).

N7198Y's operating geography. KSLI is Los Alamitos Army Airfield, an active California Army National Guard installation. The aircraft is operated under Dynamic Aviation's commercial contracting structure but the basing at Army Guard infrastructure 1,855 nm from Dynamic's Bridgewater HQ implies an active Army/CalGuard/domestic ISR contract. The proximity to Long Beach (9 nm), to Mojave (72 nm from the Merlin Mojave test site), and to PAX River and other naval air stations means the operating envelope sits squarely in maritime-coastal-ISR territory.

N7198Y's flight pattern signature. 156 sorties, 334.2 airborne hours, 88 active days across the 188-day window. The hours-per-sortie ratio is 2.14, which is the highest in the priority tail set and consistent with long-loiter ISR mission profiles rather than short pattern flying. The Q4-to-Q2 ramp (0/0/8/40/38/44/26) confirms this is an active, growing operational tail rather than a stale historical record.

N7198Y is the most plausible Maritime LOB demo candidate in the visible fleet. The watchlist tripwire is any flight pattern shift toward Mojave (KMHV) or PAX River (KNHK) corridors, which would signal maritime LOB demo activation. Currently the tail flies out of KSLI on a coastal California envelope.

N43A. The C-12 Huron. The C-12 Huron is the King Air 200 mil-variant, designated for VIP/utility transport with limited ISR capability. N43A operates from KSHD (Shenandoah Valley Regional Airport), which is the airport adjacent to Dynamic Aviation HQ in Bridgewater VA. The contracted ISR cycle ramping from 0/0/0/4/15/28/17 across the period is consistent with a Dynamic Avlease contract execution.

N43A is the lower-conviction Maritime LOB demo candidate (the C-12 Huron airframe is less ISR-capable than the RU-21H Ute), but the activity ramp through spring 2026 was the second-strongest data point that supported re-tiering it from dynamic_ruleout to king_air_isr_watch.

The structural thesis implication of the King Air ISR watch cohort: when the GM Maritime seat fills (estimated July 2026), the first observable contract activity will likely be on a King Air airframe. The Dynamic Aviation Avlease platform is the most likely commercial-operating-authority partner for that work. The transition from N7198Y (current pattern) to a Merlin-operated or Merlin-leased King Air on a named maritime ISR mission is the leading-edge observable signal.

Section XXI, The KC-135 Host Pool At The 171st ARW

The KC-135 host pool at the 171st Air Refueling Wing (Pittsburgh International Airport) is the actual fleet location for the publicly disclosed Merlin autonomy testbed work. The dashboard cohort labels them mobility_tanker. Public DVIDS imagery has identified at least two specific tails (59-1460 confirmed photo-identified; 58-0084 photo-linked).

The eighteen-to-thirty-two sortie range across nine of the ten 171st ARW tails is the structural feature that prevents ADS-B alone from pinning the autonomy testbed. With ten tails operating in a narrow range, the autonomy demonstration sorties get camouflaged inside the squadron's normal readiness flying. Public DVIDS imagery is the supplementary data layer that identifies which specific tails are the autonomy testbeds.

The MacDill candidate at 63-8020 shows a notably lower nine-sortie count, which is consistent with that tail being on a different mission cycle than the Pittsburgh pool. MacDill is the home of US Central Command and US Special Operations Command; a KC-135 testbed at MacDill would tie the autonomy work directly to the SOCOM customer-axis owner (Tactical Autonomy’s lane) and to the CENTCOM operational geography. Worth additional investigation.

The next major observable signal on the KC-135 program is the September 30, 2026 P0 catalyst (USAF KC-135 autonomy demo target window). The deeper signal is the CCR (Crew Reduction) formal RFP on SAM.gov, which the prediction model carries at 0.78 probability in a 30-90 day window based on the AFLCMC industry day cadence and the May 19 Staine-Pyne post engagement signature (Lankford J33, Herzberg 16th Airlift, Clancy HAF/A33D + Van Ovost public LIKE).

Section XXII, Patent Portfolio And Defensible Technology IP

Merlin's defensible technology IP is concentrated in three patent families:

(a) Crew augmentation and reduced-crew autonomy methods, the Merlin Pilot core stack. (b) ATC communication automation, the voice-recognition and pilot-substitute interface methods that enable the autonomous communication with air traffic control demonstrated in the CNN Pete Muntean segment. (c) Flight-deck integration architecture, the Anthem-adjacent integration methods and Honeywell Memorandum of Understanding implementation methods.

The patent portfolio size is not publicly broken out in the 10-K (Merlin lists "intellectual property" as a strategic asset but does not enumerate specific patents). The S-1/A and the 10-K both reference the Merlin Pilot trademark and the company's "proprietary autonomy stack" without listing specific US Patent and Trademark Office filings. The visible USPTO assignee record under "Merlin Labs Inc" is sparse and concentrated in the three families above; a precise count is not actionable for the thesis.

The freedom-to-operate posture vis-a-vis competitors:

(a) Versus Reliable Robotics, Reliable holds patent families in remote-piloted Cessna 208 operations and ground-control-station methods. Merlin's stack is sufficiently differentiated (full onboard autonomy rather than remote-pilot ground control) that the FTO posture is clean. No public patent litigation between the two parties.

(b) Versus Shield AI, Shield AI's Hivemind patents are concentrated in fighter-class autonomy and multi-platform coordination. Merlin's Caravan/King Air/KC-135/C-130J stack is in a different airframe class and a different mission profile. No FTO conflict on the public record.

(c) Versus Aurora Flight Sciences (Boeing), Aurora holds extensive autonomy patent families from the X-66A and ARES programs. The patent overlap is broader, but the airframe class differentiation (Aurora's work is concentrated in the X-plane and tilt-rotor categories) limits the FTO conflict on Merlin's specific airframe lanes.

The patent portfolio is not a primary thesis pillar. The primary defensible competitive position is the customer-axis P and L organizational architecture, the certification campaign progress, and the specific airframe-program access (USSOCOM C-130J, USAF KC-135). The patent portfolio is a secondary moat that prevents commodity competition but does not constitute the bull case.

Section XXIII, Scenario Valuations

The valuation framework, anchored to the reference price at time of writing (May 28, 2026 close ≈ $8.00, market cap ≈ $674M). 84.3M shares outstanding, $183M cash, $260M preferred (with the September 16 CDR mechanics treated separately), $253.9M PIPE warrant overhang at $11.50 strike. Enterprise value approximately $491M ex-preferred and ex-warrants; approximately $751M including the preferred overhang at fair value. The implied price targets below are derived from EV/Revenue multiples on FY2027 estimates and back into a per-share number off the current share count; readers should re-anchor to the live price when reading this document.

The bear case priced fundamentals (revenue + multiple) plus a 20% lockup supply discount. Cash position protects the downside floor; with $183M cash and a $25M-$40M revenue run rate in the bear case, the company would still have 18-24 months of runway at $2.00/share, which is the structural support level for a recapitalization or strategic acquisition.

The base case prices the catalyst stack delivering 3-4 of the 6 P0 events between now and September 16, normal lockup absorption (-10% to -15% post-lockup), and FY2027 revenue tracking at $90M (mid-range of the modeled outcome). The 15x EV/Revenue multiple is consistent with where Anduril and AeroVironment trade on near-term revenue.

The bull case adds a Boeing MOU disclosure within the September 30 window, a maritime customer letter from one of (Philippines, UAE, USCG, USN), and a multiple expansion to 18x EV/Revenue reflecting the entry of a defense autonomy public-market category leader at scale. The $27.00 price target is also consistent with where the prior Roth Capital price target ($25) sat after the April 15 S-1/A read.

The moonshot case adds multiple OEM MOU landings (Boeing + a second OEM partner), conversion of UAE/Remah and Philippines FMS into signed contracts, and a multiple expansion to 22x EV/Revenue reflecting a Saronic-comp valuation. The $58.00 price target is approximately 7x the current share price; this is the asymmetric upside case that justifies the position size.

The probability-weighted expected value across these scenarios (with bear at 0.20, base at 0.45, bull at 0.25, moonshot at 0.10) is approximately $17.30 per share, or roughly 2.2x the reference price at time of writing. Re-anchor the upside multiple to the live price when reading.

The single largest variable that shifts the probability weighting is the HONA Investor Day disclosure on June 3. A positive disclosure shifts probability mass from bear to bull (approximately 0.10 shift). A neutral disclosure leaves the distribution unchanged. A negative absence (no Merlin mention) shifts probability mass from bull to base (approximately 0.05 shift).

The second largest variable is the lockup absorption pattern between September 16 and October 16. A clean absorption (no >10% drawdown) shifts probability mass from bear to base (approximately 0.05 shift). A messy absorption (>25% drawdown) shifts probability mass from base to bear (approximately 0.10 shift).

Section XXIV, Risk Register

The honest risk surface, organized by category.

The seven kill criteria. These are the conditions that, if observed, would prompt thesis exit:

1. C-130J CDR fails outright (not slippage but failure)

2. KC-135 19-month contract is publicly cancelled or de-scoped

3. Boeing technical bench engagement publicly contradicted as not material

4. Major FAA regulatory action that suspends cert program

5. Founder + senior management mass departure

6. Cash burn accelerates beyond modeled runway with no offsetting revenue growth

7. Material adverse PIPE / preferred holder action (such as conversion at unfavorable terms triggering equity restructuring)

None of the seven kill criteria have been met as of May 28, 2026.

Section XXV, Watchlist And Tripwires

Each of the open questions resolves binary, in one direction or the other, on a specific tripwire. The dashboard surfaces five explicit open questions; this document expands to the full set.

Each tripwire has a defined observation window. Tripwires 1, 3, 4, 5, 10, 11 are continuous-observation (any moment). Tripwires 2, 8, 9, 12, 17, 18 are event-window observation (specific calendar windows). Tripwires 6, 7, 13, 14, 15, 16, 19, 20 are discrete-event observation (single PR or filing).

The composite watchlist is the dashboard surface that the Merlin Intelligence terminal renders to paid subscribers and the public-facing /sign-in marketing page.

The Read

A pre-revenue defense company is, by definition, a story you read from the org chart and the airframes because the income statement does not yet say much.

The org chart now reads as three customer-axis P and L lanes (the currently vacant Tactical Autonomy seat, Staine-Pyne on Mobility, and the vacant Maritime seat), backed by a senior engineering executive with carrier-deck autonomy pedigree (Gonzalez), a SOCOM-grade revenue chief (Brunner), a public-company-grade communications chief (Baker), a long-tenure program officer (Anand), a Supernal-displacement absorbing product chief (Lasater), and a cert-engineering hiring lead (Pelberg-Schariter). Five DC Revenue Operations seats posted in nine days. A board with a former SecNav at flag rank skewed maritime (Braithwaite), an Amazon-scale audit-committee floor (Brannon), an Air Force CFO budget bridge (Montelongo), and a PIPE-anchor Lead Independent Director (Blitzer).

The airframes now read as three merlin_owned NZ-based Caravans (ZK-MLN, MLO, MLQ), one operated-under-lease Caravan that is the busiest testbed in the dataset (ZK-JMP), two merlin_owned US Caravans (N208B "Big Red", N713CB), one partner Northrop testbed (N437VN Vanguard), ten KC-135 host-pool tankers, and two King Air mil-variants on the maritime ISR watch list (N7198Y and N43A). The certification campaign is running on both sides of the Pacific. The named-flight moments are stacking inside a single quarter.

The contract surface now reads as a $105M USSOCOM C-130J IDIQ (PDR done, CDR pending), a 19-month USAF KC-135 autonomy integration contract (demo window September 30, 2026), a $500-floor MDA SHIELD contract (first MDA contract, scope expansion ahead), a documented Boeing technical-bench engagement (MOU window July-September 2026), a Honeywell Anthem MOU (HONA Investor Day June 3, 2026), a Northrop Beacon partner enrollment (named-flight slot overdue), a Remah International Group UAE exclusivity (April 23, 2026 signed), and a Philippine FMS King Air sub-tier opportunity (August 31 window).

The capital structure now reads as 84.3M shares outstanding, 3.6-5.0M float, 99.93% IBKR borrow utilization at the April peak, $183M cash, $260M preferred (conversion already reset to $6.67 on the May 1 PIPE; the September 16 reset is floored at $5.00), $253.9M PIPE warrant overhang at $11.50 strike, and a lockup expiry on September 16 that releases approximately 44M shares from four major holders simultaneously. Six P0 catalysts in the 111-day window between now and that lockup expiry.

The question is no longer whether Merlin is positioning to be a tier-one defense autonomy supplier. The org chart, the board, the GM cascade, the engineering executive's resume, the five DC capture roles, the Northrop Beacon cohort placement, the King Air ISR watch tails, the MDA SHIELD contract foothold, the Honeywell Anthem program, the Boeing technical bench engagement, the UAE Remah exclusivity, and the conversion gap on ZK-MLQ all converge on the same answer.

The question is which of the visible lanes opens first, and whether HONA on June 3, the GM Maritime fill in July, Farnborough on July 20, AFA on September 14, or the lockup absorption on September 16 is the first tile that re-rates the stock.

I have built the position to be long whichever lane lands first. The next 111 days are the diagnostic.

$MRLN. Not financial advice.

Appendix A, The Bear Case, Read Honestly

The most credible bear write-up on MRLN landed May 6, 2026: Cupofcoffeecapital's "Merlin: A Contrarian Look at the Bull Case." It pulled 10,900 views. The piece earned its engagement. The bear's core technique is definitional, take a phrase from the bull case, read it in its loosest possible form, show the math does not hold at that loose reading. That is a legitimate analytical move. It is also not the only reading. Here are the five anchor phrases the bear contested, read precisely.

"100+ aircraft under contract." The bear reads this as 100+ aircraft with signed, funded, per-tail production contracts. At $3M per tail, that is $300M in contracted revenue. The bear is correct that no such document exists in the public record. The precise reading is different. "Under contract" in the context of a USSOCOM IDIQ vehicle means aircraft covered by the scope of an active indefinite-delivery, indefinite-quantity agreement. The IDIQ establishes the legal vehicle. Task orders draw against it. The 100+ figure refers to the addressable tail count within the contracted program scope, not 100 funded integrations. That distinction is not rhetorical. It is how IDIQ vehicles work. The ceiling is not the revenue. The ceiling is the authorization.

"$105M contract." The bear's arithmetic is correct in isolation. $105M divided by $3M per integration equals 35 tails maximum if the entire ceiling is consumed on integration work alone. That math is right. What it misses: the $105M is the IDIQ ceiling for the USSOCOM sole-source vehicle. IDIQ ceilings are set at program initiation and amended at the program executive office level as the program matures. The original ceiling is the floor of program scope, not the cap on total program value. Production ceilings amend. The CDR is the binary gate that determines whether the program scales to production. The $105M number is the development authorization. It is not the production contract. The bear is pricing the development ceiling as if it were the production ceiling. Those are different documents.

"300 C-130J planes." The US Air Force operates approximately 300 C-130J aircraft. The bear is correct that Merlin has not signed a 300-tail production contract. Nobody claimed it had. The 300-tail figure is the addressable fleet within the program's platform scope. The USSOCOM IDIQ covers the C-130J. The fleet size establishes the ceiling on what the program can eventually address. Conflating "addressable fleet" with "contracted production" is the definitional move the bear makes here. The bull case does not require 300 funded integrations. It requires CDR passage and a production task order. Those are different thresholds.

"$1.6B ARR." This is the most aggressive bull-case forward extrapolation. The bear is correct that there is no published model justifying this number in the next 24 months. The precise reading: the $1.6B ARR figure is the back-of-envelope long-tail TAM if every Anthem-equipped business jet, every Caravan in the cert-eligible cohort, every KC-135 in the AMC fleet, every C-130J in the SOCOM/AMC inventory, every Condor-eligible 737/767 freighter, and every FMS pipeline contract converted at midpoint software-license unit economics. That is the long-tail TAM ceiling. It is not the near-term revenue model. The bear is right to flag that this number requires multiple years and multiple programs to converge.

"De-SPAC base rate." The bear cites the historical base rate of de-SPAC underperformance (approximately 75% of de-SPAC transactions trade below offering price within 24 months). The bear is correct that the base rate is unfavorable. The precise reading: the base rate captures the population of de-SPAC transactions that lacked product, revenue, contract, or differentiated technology at the merger date. The Merlin merger included a $105M USSOCOM IDIQ already in execution, a Honeywell MOU already signed, a Northrop Beacon enrollment already announced, a $200M PIPE anchored 80% by Kingstown Capital with conviction lock-in, and a 514% revenue growth rate. The base rate applies less cleanly when the de-SPAC carries pre-merger contracted defense revenue and a multi-OEM partnership topology. The right comparison is not all de-SPACs. The right comparison is the defense-tech de-SPAC subset (RKLB, JOBY, BKSY, REKR, SAIL, AERT, RDW), which trades on a different distribution.

The bear's structural argument: even granting the precise reading on all five phrases, the price-to-pipeline ratio could compress if the catalyst stack does not deliver. That is a thesis-stressing point worth holding. The 111-day window between now and lockup is where that compression risk sits. The catalyst delivery cadence in the window is the answer to the bear's structural argument.

The bull's structural argument: a defense services company with three customer-axis GMs, a board with a former SecNav, an N-UCAS-pedigree SVP Engineering, a SOCOM-grade CRO, five DC Revenue Operations seats posted in nine days, a confirmed Boeing technical-bench engagement, a Honeywell Anthem MOU, a Northrop Beacon enrollment, an MDA SHIELD contract, a UAE Remah exclusivity, a 188-day flight dataset confirming the certification campaign is in execution, and $183M cash on hand is not the same animal as the median de-SPAC underperformer. The base rate compresses when you specify the cohort.

The bear case earned its engagement because the bear was careful, the bear was right on the loose readings, and the bear's structural framing (price-to-pipeline compression risk) is a real risk. The bull case responds by specifying the cohort and the catalyst-delivery cadence. Both sides are operating in good faith. The verdict depends on whether the catalyst delivery cadence in the 111-day window matches the bull's specification.

Appendix B, The FAA Certification Path, In Technical Detail

The FAA Type Certification pathway for autonomy software is the regulatory anchor underneath the entire commercial monetization story. The path is well-defined but not well-understood by the broader equity market.

The Stage of Involvement (SOI) framework is the regulatory milestone sequence. SOI-1 is the agreement on the certification basis (the regulatory authority for the project). SOI-2 is the agreement on the means of compliance (the test methods that show the autonomy software meets the certification basis). SOI-3 is the regulatory review of the test articles, test plans, and integration architecture. SOI-4 is the final compliance demonstration and the certification basis closure.

A typical Cessna 208B autonomy Supplemental Type Certificate (STC) program goes through approximately 18-30 months between SOI-1 and STC issuance. The Reliable Robotics program is publicly tracking toward STC issuance via a similar pathway. The Merlin program is targeting parallel issuance via both the FAA pathway and the NZ CAA pathway.

The NZ CAA pathway runs on a parallel but more permissive regulatory cadence. NZ CAA has historically issued autonomous aviation regulatory exemptions and special airworthiness certificates faster than the FAA, which is what enables the NZ certification campaign at Kerikeri to operate ahead of the FAA cert cadence. The ZK-MLN/MLO/MLQ/JMP campaign generates the flight test data that gets re-submitted as evidence within the FAA pathway.

The September 16 to December 15 window contains two cert-relevant catalysts: the FAA Type Cert progress disclosure window (December 15 P1) and any 8-K disclosure of SOI-3 or SOI-4 progress. The SOI-3 progression to SOI-4 is the single biggest cert-pathway derisking event in the calendar.

The conversion gap on ZK-MLQ (Section XX) is most likely a SOI-2 to SOI-3 transition data lock. The 21+9+silence pattern is the regulatory submission and review cycle. The next observable event is the resumption of flight test (re-emergence on hex 7C72FC or C83851).

The Designated Engineering Representative (DER) hiring footprint that Liz Pelberg-Schariter's engagement profile reveals is the cert-pathway leading edge. A DER is the FAA-authorized signatory who can sign off on specific compliance items within a Type Certification or STC project. DERs are scarce; the autonomy-cert-qualified DER cohort is small (fewer than 50 individuals nationally). When a company starts publicly recruiting DERs, the cert program is past the document stage and into the signature stage.

The Special Federal Aviation Regulation (SFAR) for autonomous aviation is the broader regulatory framework that the FAA has been developing since 2023. Wes Ryan (Northrop Grumman Fellow for Airworthiness of Autonomy and AI, Chair of ASTM AC377) sits at the center of the SFAR development conversation. His public engagement footprint with Merlin-adjacent material aligns the cert-pathway policy framework with Merlin's stack development trajectory.

Appendix C, The CNN Pete Muntean Segment, Read Carefully

The May 20, 2026 CNN segment by aviation correspondent Pete Muntean aired in CNN's "AI Pilot Era" series. Muntean is a multi-engine ATP-rated pilot in addition to his aviation correspondent role; his analytical framing of autonomy stories is more technical than general-news correspondents.

The segment showed Muntean experiencing Merlin Pilot taking off, landing, and communicating with air traffic control on its own. Human pilot optional. The visible airframe in the segment was the N208B "Big Red" Caravan at Quonset.

The segment's specific technical demonstration is the load-bearing part. ATC communication automation requires (a) speech-to-text on ATC radio transmissions in real-time, (b) intent extraction from the natural-language ATC instructions, (c) flight-control system response that matches the ATC instruction within FAA-acceptable tolerances, and (d) voice-synthesis pilot-response back to ATC that meets ATC controller expectations for radio brevity and content. None of those four elements is trivial. Doing all four in series in a real-world flight with live ATC is the technical demonstration that defines a commercial-grade autonomy stack versus a controlled-environment test article.

Michael Baker's May 21 LinkedIn repost of the segment (24 reactions) carried the framing "Human pilot optional." That phrasing is the most aggressive public framing Merlin has used to date. It is also accurate. The segment showed Muntean's hands off the yoke for the takeoff, the cruise, the ATC communications, and the landing.

The corollary read for the parenthetical thesis: if the visible "Big Red" demo at Quonset can fly autonomous takeoff-cruise-landing with full ATC communication on a Caravan, the architecture is general enough to operate on any of the larger airframes (King Air, KC-135, C-130J) with airframe-specific tuning. The Caravan demo is not the product. The Caravan demo is the proof that the architecture is the product.

The CNN segment also raised the regulatory question that the FAA pathway answers. Muntean's segment did not minimize that the FAA cert pathway is still ahead; it framed the regulatory question as the deciding factor between cert-track and demo-track operations. That framing tracks with where the Merlin program actually sits in the SOI-3 to SOI-4 transition.

The CNN segment is the first major mainstream-news autonomy demonstration that featured a CO-piloting frame ("Human pilot optional" but optional implies still permitted) rather than a fully-autonomous frame ("Pilot eliminated"). That co-piloting framing is the exact frame that George used in his parenthetical reply ("augmenting crew, meaning a pilot still in the flight deck (just one instead of two)"). The CNN segment is the public exhibit for the public surface; the parenthetical is the public exhibit for the private surface.

Appendix D, The Harpoon Series C Cluster

Harpoon Ventures led Merlin's Series C and continues to surface as a coordinated public-engagement presence around every MRLN operator milestone. Five named Harpoon principals, Riley Loftus, Larsen Jensen, Mat Vogels, Matias Zorrilla, and Jeff Torrance, have visibly engaged with both the May 14 Condor product launch and the May 19 Erin Staine-Pyne KC-135 post. That is the fingerprint of a Series C lead positioning for the next capital event rather than waiting passively for one.

The prediction model carries Riley Loftus formal MRLN board seat probability at approximately 0.42 within 90 days. The 8-K Item 5.02 disclosure channel is the formal observable. A Loftus board seat would be the third VC-backed director seat (alongside FR Capital and SnowPoint's eventual representation), which would shift the board's voting alignment in favor of growth-stage operational discipline.

The strategic implication: when a Series C lead operates a coordinated presence through five named individuals on every operator milestone, the implied behavior is preparation for either a follow-on round or a strategic exit event. The next observable event from the Harpoon cluster is one of: (a) Public statement by Larsen Jensen (the highest-seniority GP) on MRLN-related investment thesis; (b) Loftus board seat 8-K disclosure; (c) Participation in a follow-on capital raise (the prediction model carries Harpoon-led inside-priced follow-on at approximately 0.45 probability by December 31, 2026).

Appendix E, The Latham And Watkins Deal-Flow Analysis

Latham And Watkins is the law firm of record for the de-SPAC. The firm's authored-content engagement signature is one of the most useful intel signals available because partners post deal announcements as the firm's marketing channel.

The firm's partner-celebration cadence around the Merlin close on March 16-17, 2026 was consistent with a high-prestige flagship transaction.

Nick Dhesi is the senior SPAC partner Latham deployed onto MRLN. Dhesi's career deal-list includes the most-watched SPAC graduate transactions in defense and aerospace. The firm-level signal of putting Dhesi on the MRLN de-SPAC is the firm's view that the transaction is a flagship marketing event.

Haim Zaltzman is one degree of separation from MRLN. Zaltzman's deal portfolio in the same window includes AEVEX (defense ISR services), X-energy (nuclear power), and Cerebras (AI infrastructure). Zaltzman is the Latham partner who covers defense-tech / national-security capital raises. His one-degree adjacency to MRLN through Dhesi means the firm has a natural handoff if MRLN moves toward a strategic capital event (a follow-on equity offering or a strategic-acquisition transaction).

The prediction model carries "Haim Zaltzman next deal filing in Merlin orbit" at approximately 0.50 probability on a rolling window. The observable channel is SEC EDGAR for new 8-K filings naming MRLN or for Form D filings showing private capital raises by MRLN-adjacent entities.

The Latham engagement signature is what hedge funds use to triangulate capital-event timing. A defense-tech name with the firm's senior SPAC partner on the de-SPAC, and the firm's senior defense-capital-markets partner one degree away, is structured for additional capital events. The likely next capital event (if it happens before September 16) would be a strategic add-on PIPE; if it happens after September 16, the most likely structure is a follow-on equity offering at a premium to the lockup absorption clearing price.

Appendix F, Sell-Side Sequencing And The Northland Prediction

The sell-side coverage sequence on MRLN matters because published coverage, a third-party model, and a broker-distributed narrative meaningfully improve a small-cap's visibility with institutional investors. The first published initiation does not make a stock investable overnight, but it changes the discovery setup.

Roth Capital. The Sujeeva De Silva coverage is the only published sell-side model. The April 14 initiation at Buy / $15 followed by the April 15 revision to $25 (after reading the S-1/A FY2025 financials) is the canonical pre-information to post-information pattern. The 67% intraday revision on a freshly-public name tells you exactly which number was the pre-information model and which was the post-information model.

The next observable Roth event is either a quarterly update note (which Roth typically issues a few days post-earnings) or a target revision (which Roth has done aggressively in this initiation cycle). The 10-Q on August 14 is the next natural Roth update trigger.

Northland Capital. The highest-conviction next-initiator. Mike Latimore covers the defense autonomy SPAC graduate cohort: he initiated RCAT (Buy), DPRO (Outperform / $20 target in January 2026), and SPAI on the same template. The Latimore template for defense-tech SPAC graduates is approximately a Buy or Outperform rating with a price target in the $15-25 range based on a 2026 EV/Revenue multiple of 12-18x.

The prediction model has Northland initiation at approximately 0.62 probability in a 60-120 day window. The Latimore template suggests an MRLN initiation would land between $18 and $24 price target. The volume-spike day-of an initiation publication is the tradable event; the longer-duration impact is the additional institutional eligibility.

The watch channels are: (a) Northland Capital research portal (latimoreresearch.com or the Northland institutional research site) for new MRLN PDF; (b) Bloomberg analyst-recommendation aggregator for Northland coverage update; (c) corporate IR notifications from Merlin Labs.

Cantor, Cowen, Stifel, Jefferies. These four are the next-most-likely cohort to add coverage in a 90-180 day window. Each has a defense-tech sector analyst team. The probability of any one of them initiating in 90 days is approximately 0.20; the probability of at least one of them initiating in 180 days is approximately 0.55.

The sell-side build is the longer-term institutional-positioning catalyst. By Q4 2026, MRLN should have 3-5 active sell-side coverage notes if the catalyst stack delivers. That coverage build alone is approximately a 10-15% multiple expansion driver as institutional eligibility broadens.

Appendix G, Comp Table And Multiple Justification

The Defense autonomy public-equity comp set is small but instructive. The comp table here is the framework hedge funds use to justify MRLN's price.

The MRLN current EV/Revenue multiple is 61x on TTM ($491M EV / $8M TTM revenue). That multiple is high but not anomalous for a pre-revenue defense services company in cert phase. The right comparison is FY2027 forward, not TTM trailing. At the base-case $90M FY2027 revenue, the current EV implies 5.5x FY27E, which is at the lower bound of the defense services comp set. At the bull-case $150M FY2027 revenue, the current EV implies 3.3x FY27E, which is below the mature defense services comp set.

The implied undervaluation on FY27E is the structural argument: even at base case, MRLN trades at less than half the AVAV multiple and a fraction of the Anduril private-market comp. At bull case, MRLN trades at one-third of the Kratos multiple.

The bridge from the current 5.5x base-case forward multiple to a 15x defense-autonomy growth multiple is approximately 2.7x re-rate, or $22 per share at FY27 base-case revenue. The bridge to a 22x moonshot Saronic-style multiple is approximately 4x re-rate, or $32 per share at FY27 base-case revenue.

The Saronic comp is the most asymmetric reference point. Saronic raised at a $9.25B private valuation; the Corsair USV contract at $392M and the shipbuilding capacity expansion at $1.75B are the comp print. If the public-market re-rate process places MRLN within 25-30% of the Saronic private-market multiple, the implied share price target moves substantially above $30.

The hedge fund framing question: at what point does the defense autonomy public-equity multiple converge with the defense autonomy private-equity multiple? The answer is the question. The convergence event is what re-rates the entire category. The MRLN-specific trigger is whichever of the catalyst stack (Boeing MOU, maritime customer letter, HONA disclosure, KC-135 demo, GM Maritime fill) is the most asymmetric.

Appendix H, The Erin Staine-Pyne May 19 Post, Forensically

The May 19, 2026 LinkedIn post by Erin Staine-Pyne (Merlin's GM Mobility, the December 2025 hire from Pentagon staff) carried more institutional intelligence per character than any other public Merlin communication this year. The post itself was an update on the KC-135 tanker autonomy program. The engagement signature is the story.

Four active duty USAF officers in mobility-adjacent roles plus a retired four-star general engaging on a single LinkedIn post is the fingerprint of a working program. Joint Staff J33 is the operations directorate of the JCS; J33 engagement on a tanker autonomy post implies the program has reached the level where joint-operations planners are tracking the requirements.

HAF/A33D is the headquarters Air Force office for operations plans and requirements. A33D engagement on the post implies the requirements team at HAF level is tracking the program. The 16th Airlift Squadron operates C-17s; their engagement implies operating-unit-level interest in the autonomy capability that maps to their airframe.

Van Ovost's public engagement on the post is the single most senior endorsement signal currently visible. Van Ovost is a retired four-star general who commanded AMC and USTRANSCOM, retired in 2024, and remains active in defense industry advisory roles. Her public engagement with Merlin's GM Mobility on a tanker autonomy post is the equivalent of a senior endorsement letter without the formal letter.

The engagement signature indicates the KC-135 autonomy program is at the requirements-formalization stage and that AMC/AFMC at HAF level is tracking it. The next observable contract event is the CCR (Crew Reduction) formal RFP on SAM.gov, which the prediction model carries at approximately 0.78 probability in a 30-90 day window. The Staine-Pyne post engagement signature is the leading-edge confirmation of that probability.

The post itself did not name the contract vehicle, did not announce a milestone, and did not include any quantified program metric. The post's content was a generic update on the autonomy program. The intelligence is entirely in the engagement signature, not the post content. That is the pattern in mobile-program disclosure: the post is what passes Air Force Public Affairs review; the engagement is the unfiltered institutional response.

Appendix I, Mark Brunner, The SOFWERX Channel, And The SOCOM Capture Cadence

Mark Brunner joined Merlin as Chief Revenue Officer on April 13, 2026. His background is 22 years in the special operations community, including SOFWERX channel positioning. The SOFWERX program is the technology rapid-prototyping center for US Special Operations Command, located in Tampa near MacDill Air Force Base. SOFWERX is the formal industry-engagement vehicle for SOCOM technology procurement; companies that win SOFWERX engagements are positioned to convert to SOCOM contracts.

Brunner's appointment timing is informative. April 13 is six weeks before the May 18-22 DC capture-role hiring burst and seven weeks before the May 19 Staine-Pyne KC-135 post. The CRO seat being filled before the DC roles posted is consistent with the operating sequence: the CRO defines the capture strategy, then the supporting roles are recruited to execute against that strategy.

Krishnan Anand's April 14 LinkedIn announcement of Brunner's hire pulled 14 reactions. That announcement is the public marker for the SOCOM capture readiness alignment. The implied operating sequence: Brunner stands up the SOCOM-axis capture team in Q2 2026, the C-130J CDR fires in Q3 2026, the first production task-order awards land in Q3-Q4 2026, and the SOCOM revenue scales materially in FY2027.

The SOF Week Tampa conference is the canonical SOCOM industry engagement event (typically held in May annually). Brunner's LinkedIn engagement footprint during May 2026 (multiple SOF Week-adjacent posts) is the leading-edge indicator that the SOCOM capture cadence is operating at conference rhythm.

The prediction model has "Brunner drives SOCOM disclosure flow" at approximately 0.66 probability on a rolling window. The observable channel is Brunner LinkedIn for direct evidence; the indirect channel is DoD daily contracts page for any SOCOM task-order awards.

The SOCOM customer-axis is the most-mature contract surface MRLN has. The $105M IDIQ is in execution. The PDR is complete. The CDR is pending. The task-order pipeline is the natural revenue runway through FY2027. Brunner's role is to convert that pipeline into accelerated execution, not to open new SOCOM relationships.

Appendix J, Anthem Unit Economics, Modeled

The Honeywell Anthem flight deck system is the integration point for Merlin Pilot on commercial business jets. The unit economics of an Anthem-Merlin integration are the basis for the Anthem-pathway revenue model.

A typical commercial flight deck software integration carries the following economic structure:

(a) Non-recurring engineering (NRE), the development and integration cost paid by the OEM to the integrator. Typical range $5-15M per platform integration.

(b) Type Certificate amendment cost, the regulatory cost of adding the autonomy software as an approved part of the platform's Type Certificate. Typical range $2-5M per platform.

(c) Per-aircraft software license, the recurring revenue per delivered aircraft. Typical range $100K-300K per aircraft over a 15-20 year deck lifespan.

(d) Subscription / maintenance, the ongoing software update and support contract. Typical range $10K-30K per year per aircraft.

For the Bombardier Global series (approximately 200 operational aircraft):

• NRE: $10M (integration)

• TC amendment: $3M

• Software license: $200K average per aircraft x 200 aircraft = $40M

• Subscription: $20K average per aircraft per year x 200 aircraft x 15 years = $60M

• Lifetime program revenue: ~$113M (NRE + TC + license + 15-year subscription)

For the Pilatus PC-12 NXG (approximately 2,000+ operational aircraft, with continuing deliveries):

• NRE: $8M (integration)

• TC amendment: $3M

• Software license: $150K average per aircraft x 1,000 aircraft (50% adoption assumption) = $150M

• Subscription: $15K average per aircraft per year x 1,000 aircraft x 15 years = $225M

• Lifetime program revenue: ~$386M (with 50% adoption assumption)

For the eVTOL platforms (Lilium suspended; Vertical and Embraer Eve pre-certification):

• NRE: $5-10M per platform integration

• Per-aircraft economics are uncertain pending platform unit economics

The Anthem-pathway lifetime revenue ceiling, summed across Bombardier Global, PC-12 NXG, and one eVTOL platform integration, is approximately $500M cumulative over 15-20 years. That is the Anthem long-tail TAM. It is not the near-term revenue. The near-term revenue from Anthem is the NRE and TC amendment phases (~$25-30M over Q3 2026 through Q4 2027).

The Anthem economic structure is also informative for the Boeing-Merlin opportunity if it materializes. Boeing Defense has approximately 1,000 KC-46A and approximately 100 C-32A/C-40 aircraft in the airlift-tanker portfolio. A Boeing-Merlin integration on this fleet would carry a multi-hundred-million-dollar lifetime program economics structure.

The unit-economics modeling is what hedge funds use to scale the long-tail TAM number. The $1.6B forward ARR figure the bear contested becomes more tractable when broken down: it is the sum across Anthem-Bombardier + Anthem-Pilatus + Anthem-eVTOL + USSOCOM-C130J + USAF-KC135 + USAF-C17 + Condor-737 + Condor-767 + UAE-FMS + Philippines-FMS + USCG-HC144 lifetime program economics. No single program produces $1.6B. The aggregate of 8-12 programs at midpoint adoption produces $1.6B+ lifetime program economics.

Appendix K, Glossary And Source Notes

Glossary

ADS-B, Automatic Dependent Surveillance-Broadcast, the public aviation transponder data stream used for live aircraft tracking.

AFLCMC, Air Force Life Cycle Management Center, the USAF acquisition organization that owns multiple Program Executive Offices (PEOs) including PEO Mobility for the KC-135 program.

AMC, Air Mobility Command, the USAF major command responsible for airlift, air refueling, and aeromedical evacuation.

A-GRA, Air-Ground Robotic Architecture, an AFRL program for autonomous air-ground integration.

CCA, Collaborative Combat Aircraft, the USAF program for autonomous loyal-wingman-class aircraft.

CCR, Crew Reduction, the term-of-art for autonomous flight deck operations enabling reduced-pilot operations.

CDR, Critical Design Review, the second major design gate in a defense acquisition program (after PDR).

CRADA, Cooperative Research and Development Agreement, a cost-sharing federal R&D vehicle.

DER, Designated Engineering Representative, the FAA-authorized signatory for specific aircraft certification compliance items.

DSCA, Defense Security Cooperation Agency, the DoD organization that manages Foreign Military Sales.

DVIDS, Defense Visual Information Distribution Service, the public DoD imagery and video archive.

FMS, Foreign Military Sales, the formal DoD program for selling defense articles to foreign governments.

HONA, Honeywell Aerospace, the segment Honeywell announced in February 2024 it intends to spin off as a standalone entity; spin-off has not yet legally closed, HON remains the parent ticker.

IDIQ, Indefinite-Delivery, Indefinite-Quantity contract vehicle.

ITAR, International Traffic in Arms Regulations, the US export control framework for defense articles.

MDA, Missile Defense Agency.

N-UCAS, Navy Unmanned Combat Air System, the carrier-launched unmanned ISR/strike program that became the X-47B and evolved into the MQ-25 lineage.

PDR, Preliminary Design Review, the first major design gate in a defense acquisition program.

PEO, Program Executive Office.

PIPE, Private Investment in Public Equity, the structured equity investment by anchor investors at SPAC closing.

POC, Percentage-of-Completion accounting.

SAM.gov, System for Award Management, the federal contract opportunity and award database.

SFAR, Special Federal Aviation Regulation.

SOFIC, Special Operations Forces Industry Conference.

SOFWERX, The SOCOM technology rapid-prototyping center in Tampa.

SOI, Stage of Involvement, the FAA certification milestone framework.

STC, Supplemental Type Certificate.

USAID, United States Agency for International Development.

USCG, US Coast Guard.

USSOCOM, US Special Operations Command.

Source notes

Live MRLN dashboard intelligence: shawarmacapital.net/sign-in. Substack research: shawarmacapital.substack.com. USAspending direct pulls: usaspending.gov. SAM.gov contract opportunity tracking: sam.gov. FAA Type Certificate registry: faa.gov/aircraft/air_cert/type_certificates. Schedule 13D filings: sec.gov EDGAR. Q1 2026 earnings call: investors.merlinlabs.com. Independent network analysis across the MRLN orbit (operator-side corpus). ADS-B historical archives: adsb.lol globe_history releases, 188 days backfilled to May 28, 2026.

This document is research synthesis built entirely from public data. Cross-references are available in the structured corpus and the dashboard's Open Questions panel.

Appendix L, Trump Administration OSC Drone-Funding Read-Through

On May 27-28, 2026 the Wall Street Journal reported that the Trump administration is in talks to provide funding to select US drone companies through a mix of debt and equity stakes channeled via the Office of Strategic Capital (OSC), the Biden-era DoD lending unit being repurposed. The named recipients are Unusual Machines (Donald Trump Jr. adviser; FPV components), Neros (Sequoia-backed small autonomous quadcopters for Ukraine-style use), and Performance Drone Works (Army squad-level recon). Not an executive order. Not Defense Production Act Title III. Not new appropriations. Existing OSC authority being deployed against named small-UAS vendors. The action sits under the broader FY27 "drone dominance" budget priority (approximately $55B autonomy line inside the $1.5T defense request; $30B DPA ask separate).

The MRLN read.

This specific OSC capital program is narrow to small-UAS / Group 1-2 attritable drones. The named recipients are FPV component vendors, small quadcopter manufacturers, and squad-level recon contractors. Wrong size class, wrong mission set, wrong customer office for MRLN. None of MRLN's lanes (C-130J USSOCOM IDIQ, KC-135 USAF integration + Autonomy Core CCR, MDA SHIELD, Northrop Beacon, Honeywell Anthem, GE Aerospace Autonomy Core) touch this funding pool. Also: MRLN has $183M cash on hand and does not need OSC equity. Absence from the OSC list is not a negative tell, it is "doesn't need it."

The cross-read that is bullish for MRLN runs through the FY27 budget umbrella, not through this specific capital program:

1. C-130J IDIQ ceiling expansion willingness. Program offices working under a "drone dominance" mandate have political cover to amend IDIQ ceilings upward at the CDR milestone. The $105M USSOCOM C-130J ceiling moves higher faster under this regime.

2. KC-135 CCR RFP acceleration. The AFLCMC industry day cadence plus the September 30 demo target window both benefit from a White House priority signaling autonomy speed. The Autonomy Core / CCR competition becomes more politically attractive to fund.

3. MDA SHIELD scope expansion. The Golden Dome layered missile defense initiative is the canonical recipient of an autonomy / drone-dominance budget tailwind. MDA's HQ085926FE898 contract is positioned to expand its scope under this regime.

4. CCA-adjacent crewed-autonomy. The Northrop Beacon named-flight slot becomes more politically attractive to disclose under a "drone dominance" framing.

5. Multiple expansion through category proximity. Even if the funding doesn't flow to MRLN, the category-wide re-rate (Anduril, Shield AI, Saronic) under the political tailwind lifts MRLN's peer-comp multiple by reflection.

The narrative risk: if the market reads "drone dominance" too narrowly (FPV / attritable Group 1-2 only) and miscategorizes MRLN as "not a drone company," multiple expansion could lag the broader autonomy comp set. The counter-framing for the deep update audience: MRLN is the crewed-aircraft autonomy category, which is structurally larger and politically protected by both the FY27 autonomy line and the existing USSOCOM / AMC / AFMC budget lines.

The thematic context for the MRLN thesis. The OSC capital program itself is not a direct MRLN catalyst. The political regime it sits inside, the FY27 "drone dominance" budget umbrella, is the broader tailwind that supports every MRLN customer command's ability to expand IDIQ ceilings and accelerate program timelines. Whether the equity market over-categorizes MRLN as "not a drone company" and undervalues it relative to the small-UAS cohort, or correctly categorizes MRLN as the crewed-aircraft-autonomy adjacency that benefits from the same regime, is one variable in how the catalyst stack absorbs.

Watch channels for additional thematic signal: WSJ for further OSC capital deployment announcements; SAM.gov for any expansion of MDA SHIELD, USSOCOM C-130J, or AMC KC-135 task-order ceilings; the FY27 defense budget detailed PE-element justifications when they publish; any executive order specifically naming crewed-autonomous airlift or tanker programs, which would convert from umbrella tailwind to direct MRLN catalyst.

The 111-day window between today and the September 16, 2026 lockup is the operational ground truth for this position. Six P0 catalysts, one mechanical risk event, and a parenthetical word from the CEO that says the public surface is not the actual surface. The rest is execution and observation.

Long $MRLN. Research synthesis, not investment advice.

Nothing here is investment advice. For educational purposes only. I may hold positions in names I discuss. Do your own research. No liability assumed.

More in this series

MRLN: The Complete Thesis

The Gate Cleared and the Tape Forgot

$MRLN: I built Merlin World

Merlin Intelligence - Live OSINT Tool

I Tracked Every Merlin Aircraft for 180 Days. The Testbeds Just Peaked

Alyeska Is Underwater. None of the Thesis-Breakers Have Hit

MRLN's $105M Ceiling Is Not the Revenue, It Is the Authorization

Alyeska Wrote Both Checks and the Market Missed It

The UAE Deal Changes Everything Merlin Was Supposed To Be

MRLN: Part 4 - The Board, the Preferred Supernova, and the Catalyst Nobody

$MRLN Part 3: The DAWG, the $54.6B, and Why Merlin Is Positioned to Win the

Almost every microcap that gets pitched as a 5G play is a parts house that drafts off Ericsson, Nokia, or Samsung. They sell components. They do not sell radios. When a hyperscaler or a Tier 1 carrier writes the check, the parts house gets a single line on a bill of materials and a margin the buyer pre-negotiated.

AmpliTech Group is not that company anymore. It used to be. It is not now.

Q1 2026 closed with $5.35 million of revenue against $3.60 million a year earlier, a 48.6 percent year-over-year print. Gross profit doubled. Gross margin moved from 33.0 percent to 48.0 percent in a single year, a fifteen-hundred-basis-point step-up that the tape has not yet priced. Cash plus marketable securities sits at $18.4 million. Working capital is $25.4 million against $10.2 million on December 31. Stockholders' equity is $48.4 million. Long-term debt is zero. The company funded its 2026 build through a $16 million combined rights offering and registered direct, both closed before this quarter ended. The dilution has already happened.

AmpliTech sells more than parts now. It sells a 64T/64R massive MIMO ORAN 5G radio unit, the only fully integrated open-RAN small-cell hardware that a US-domiciled vendor manufactures end-to-end on US soil. It sells cryogenic low-noise amplifiers, the kind every qubit on every superconducting quantum computer in the IBM, Rigetti, and IonQ pipelines requires. It sells GaN power amplifiers and ultra-low-noise satcom RF for Viasat, Lockheed Martin, and L3Harris. And it sells managed Communications-as-a-Service through a subsidiary, True G Speed Services, that exists to convert the radio hardware into a recurring revenue stream a Tier 1 carrier can actually buy.

The market cap at $5.04 against 25.34 million shares outstanding is roughly $127.7 million. Enterprise value, after backing out cash, sits near $109 million. Trailing twelve-month revenue is somewhere between $14 million and $16 million depending on how you handle the 2025 reset quarter, with Q1 2026 running at a $21.4 million annualized pace. Management has not retracted full-year 2026 guidance. CEO Fawad Maqbool said on the May 13 call that the guidance remains achievable, with the revenue ramp weighted to the second half. He owns 2,663,864 common shares directly. He owns more than ten percent of the company he founded. He has never sold a share since the IPO. He bought into the rights offering this January at his pro-rata allocation.

I am not telling you Maqbool has a clean execution record. He does not. The 2024 fiscal year was a hiring miss against a $30 million guide that landed near $10 million. The company lost $3.5 million in a custodial Bitcoin treasury arrangement to fraud the FBI is still investigating. The bear case from the original short writeups, the ones that traded the stock down to $2.75 in late 2024, is grounded in CEO credibility problems that are real and recent. I will not paper over them in this piece. They are in Section X.

What changed between $2.75 and $5.04 is the unit economics. Forty-eight percent gross margin is what a 5G radio OEM looks like at scale. It is not what a custom-microwave parts house looks like, ever. The mix shift has happened. The print proves it.

Let's go.

I. The Print

The Q1 2026 release dated May 13, 2026 disclosed four numbers that matter and three the tape did not parse correctly.

The four that matter:

The three the tape did not parse:

The manufacturing and engineering segment, the operating core that produces revenue, generated $3.28 million of Q1 revenue against $0.99 million in Q1 2025. That is a 231 percent year-over-year acceleration. The remaining $2.07 million ran through the True G Speed Services subsidiary and adjacent services lines. The mix is real. The pivot to integrated 5G radios shows up cleanest in the segment breakdown, not in the headline.

Working capital rose from $10.2 million on December 31, 2025 to $25.4 million on March 31, 2026. The current ratio went from 1.68 to 4.25. That is not a working-capital ramp from operations. It is the $16 million of fresh equity the company raised through a combined rights offering plus registered direct, both pre-funded before the quarter ended. Cash and marketable securities ended the quarter at $18.4 million. Long-term debt remained zero, where it has been for the entire post-IPO history. The balance sheet is the cleanest in the company's filing history.

Full-year revenue guidance is intact but weighted to the second half. Maqbool said on the May 13 call: "The Company continues to believe its full-year revenue guidance remains achievable; however, the Company expects revenue recognition to be more heavily weighted toward the second half of the year." There is no retracted number. There is no missed guide. There is a calendar slip.

The structural read: a parts-house revenue base at 33 percent gross margins is not the same business as an integrated-systems revenue base at 48 percent gross margins. The Q1 print is the first quarter where the systems-revenue mix carried the margin line. The rest of the year is the test of whether that mix keeps.

II. The Company

AmpliTech Group was founded in 2002. Twenty-four years old. Headquartered in Hauppauge, New York, on Long Island. Public on Nasdaq since 2021. Approximately one hundred fifty employees across engineering, manufacturing, and sales.

The founder is Fawad Maqbool. He is CEO, chairman, and beneficial owner of more than ten percent of the company. His background is the back-half of the New York RF and microwave industry: Hazeltine Corporation from 1983 to 1986 (acquired by Emerson, now BAE Systems), Miteq from 1987 to 1999 where he ran a department of thirty-two engineers (acquired by L3Harris in 2015), and AmpliComm from 2000 to 2001, a fiber-optic and cryogenic amplifier shop he founded that Aeroflex acquired within a year. He founded AmpliTech in 2002 with an explicit RF and microwave focus. He earned two bachelor's degrees from CUNY and a master's in electrical engineering from NYU. He has never sold a share of AMPG since the IPO. He has bought multiple times on the open market. He voluntarily cut his salary 20 percent in 2024.

The company organizes around four product lines:

The Hauppauge facility manufactures the components, integrates the radios, and ships the systems. The vertical integration claim Maqbool makes on every earnings call is that no other US-domiciled vendor produces ORAN 5G radios end-to-end on US soil. That is a Buy America claim, and it is the most important fact in the entire setup because the BEAD program rolling out through 2027 requires 65 percent US-sourced components today, rising to 75 percent by 2029. A vertically integrated US OEM does not have to assemble a coalition of foreign suppliers to satisfy the requirement. It already does.

The four-line product strategy is not new. The integrated systems revenue mix is.

III. The Pivot

For twenty years, AmpliTech sold parts. It built ultra-low-noise amplifiers for satellite ground stations, custom MMIC packaging for defense customers, and one-off design wins on radar and electronic-warfare programs. The revenue base was lumpy, capped by the size of the largest single contract, and structurally limited to component-level margins. Net income for the decade preceding the IPO oscillated around zero.

The pivot started after the 2021 IPO. Two events drove it. First, the Open RAN telecom standard reached commercial deployment readiness in 2023 and 2024, anchored by the O-RAN Alliance specifications that decoupled the radio unit, the distributed unit, and the centralized unit into separately specifiable hardware. That decoupling created room for a non-Ericsson, non-Nokia, non-Samsung vendor to sell an ORU specifically. AmpliTech already manufactured the RF front-end that goes into an ORU. The leap from front-end to full ORU was an integration step, not a clean-sheet design problem.

Second, the rollout of the federal BEAD program, the $42.45 billion broadband expansion authorized under the Infrastructure Investment and Jobs Act, codified domestic-content requirements that effectively excluded the European and Asian incumbents from the rural-broadband buildout. A US-domiciled radio vendor with a finished ORU and a managed-services wrapper was the exact category the BEAD funds were structured to favor.

The True G Speed Services subsidiary, branded TGSS, exists to package the radio into something a Tier 1 telco can buy as a service. Press releases from 2024 describe it as managed Communications-as-a-Service for ORAN networks and base stations. The model is the same model a hyperscaler runs: AmpliTech owns the hardware, the customer pays a recurring fee per radio per year for managed operations. The recurring revenue line is what converts the historical lumpy parts business into the kind of cash-flow profile a sell-side analyst can model.

The 64T/64R massive MIMO ORAN radio is the flagship product. The published throughput is one gigabit per second. The form factor and the power envelope target small-cell deployments, the segment of the 5G network that the Tier 1 carriers consistently undershoot because the incumbent vendor stacks were not designed for granular site rollout. The radio is the SKU that lets a telco bid a BEAD project, drop one box per site, and meet domestic-content rules out of the gate.

The pivot from parts house to integrated radio OEM is the thesis. The Q1 print is the first quarter where the integrated-systems mix actually moved the margin line. Three more quarters at 48 percent gross margin and the rerate is no longer optional. It is the first thing a generalist tech analyst initiates coverage on.

IV. The Cap Table And The Tape

Shares outstanding sit at 25,340,000 as of the most recent disclosure. Float is approximately 21.0 million, derived from the 5.82 percent short-of-float ratio against the 1,222,778 shares reported short on the April 30, 2026 settlement date. Insider ownership totals roughly 30 percent across Maqbool plus the board and named officers. Maqbool's direct stake is 2.66 million shares, 10.5 percent of outstanding.

Two things stand out on that table.

The first is the average daily volume against the float. 3.45 million shares trade per day against a 21 million share float. That is 16.4 percent of the float turning over every session, which is unusually high. The name is liquid for its market cap. A position can be built and exited inside a single day's volume. That cuts both ways: every news event prints fast in either direction. The DCA structure I am using this week is built around that liquidity profile. I am not chasing.

The second is the short interest. 5.82 percent of float is a modest short. 3.1 days to cover is below the four-day threshold that historically anchors the borrow market. The shorts here are not running a thesis trade. They are running a hedge or a pair against the broader ORAN cohort. There is no squeeze setup, and I am not pitching one. The float math says you make money on this name through fundamentals, not through a borrow event.

The capital structure is clean. There is no convertible preferred. There is no ratchet provision sitting on a PIPE. The most recent capital raise was the $16 million combined rights offering plus registered direct that closed before the Q1 print. The rights offering was structured pro-rata to existing holders, with Maqbool participating to maintain his percentage. That structure is the cheapest possible dilution for an existing shareholder. It is what a CEO who knows he is going to own the company in five years agrees to.

There is an at-the-market offering facility still outstanding. Management has historically used it for housekeeping, not for material capital raises. The ATM exists. It can be tapped. Anyone reading this should track the post-effective amendment filings on EDGAR.

V. The 5G ORAN Lane

The Open Radio Access Network standard is the single largest commercial inflection in the telecom infrastructure cycle since long-term evolution.

The O-RAN Alliance, founded in 2018, published specifications that decompose a base station into separately purchaseable hardware. A radio unit, a distributed unit, and a centralized unit. Each can be sourced from a different vendor. The integration runs over open interfaces. The economics are the economics of any open standard: incumbent margins compress, new entrants get pricing room they did not have under the closed stacks.

The total addressable market for ORAN small-cell radios from 2026 through 2030 is somewhere between $30 billion and $50 billion depending on which research house you trust, with North American deployment running ahead of the global average because the FCC C-band and CBRS auctions front-loaded the spectrum allocation.

AmpliTech's flagship 64T/64R massive MIMO ORU is the SKU that competes in the 5G mid-band small-cell slot. Throughput targets one gigabit per second. Power envelope and physical footprint target the dense-urban and rural-coverage deployments where BEAD funding is flowing. The vendor list for that slot is short: AmpliTech, a handful of Asian ORU vendors who do not qualify for Buy America preferences, and the incumbent Ericsson and Nokia stacks that are not architecturally optimized for granular ORU sales. The Tier 1 carriers in the US run formal ORAN procurement programs at AT&T, Verizon, T-Mobile, and the regional incumbents. The procurement cycles run on twelve-to-eighteen-month qualification timelines. AmpliTech is past the qualification phase with the two named LOI customers from the 2024 cycle.

The 2024 LOI bench was disclosed at $118 million across two Tier 1 customers. The named one is Telus, the Canadian carrier, at $40 million. The unnamed second customer signed at $78 million. As of the Q1 2026 print, the company had recognized $12 million of firm purchase orders against the combined LOI base. The conversion math runs slowly because the qualification gates fire one site or one regional cluster at a time. The bear point that the LOI conversion has been slower than the 2023 framing implied is correct. The bull point that the conversion is moving, that gross margins on the converted revenue have stepped up to 48 percent, and that the second-half loading of the 2026 guide reflects the next named tranche, is also correct. Both can be true.

The BEAD program funding flow is the second mechanic. The $42.45 billion is being allocated by state Broadband Equity, Access, and Deployment Grant offices over the 2025 to 2028 window. The first major procurement awards landed in late 2025 and continue through 2026. The 65 percent domestic content rule applied to every radio purchased under BEAD funds, rising to 75 percent in 2029, creates a structural moat that no foreign-headquartered vendor can satisfy without restructuring its supply chain. AmpliTech's Hauppauge manufacturing site is one of the few US-domiciled facilities that integrates an ORU end-to-end. The Buy America claim is not a marketing line. It is a procurement-rule moat.

VI. The Quantum Lane

Every superconducting qubit on every quantum computer in the IBM Quantum, Rigetti, and IonQ roadmaps requires a cryogenic low-noise amplifier to read out its state without destroying the coherence. The amplifier sits inside the dilution refrigerator stack, operates at temperatures below one Kelvin, and has to deliver a signal-to-noise floor low enough not to swamp the qubit's quantum signal.

AmpliTech is the only US-based commercial vendor of cryogenic LNAs at production volumes. The per-unit pricing is between $5,000 and $7,000 depending on the band coverage and the noise figure. Maqbool has described the customer relationship publicly: every qubit on the IBM roadmap requires an LNA, and IBM has published a target of 100,000 qubits by 2033. The unit-economics math, even at the bottom of the price range and the optimistic end of the qubit count, is in the hundreds of millions of dollars of cumulative cryo-LNA revenue over the 2026 to 2033 window, before any contribution from Rigetti, IonQ, or the academic and national-lab market.

I will not put a base case on the quantum line. The IBM roadmap could slip. The qubit-count target is a 2033 number, not a 2027 number. The competing read-out architectures, the parametric amplifiers and the traveling-wave amplifiers that European and Japanese labs are developing, could displace cryo LNAs at the high-end qubit configurations.

What the quantum line provides is optionality. It is a multi-year structural tailwind on a product AmpliTech already manufactures, against a customer set that has no obvious alternative US supplier. Even modest penetration of the IBM target produces a revenue line that exceeds 2025 total company revenue. The quantum line is not the thesis. It is the tail.

VII. The Defense And Satcom Lane

Viasat is the largest single named customer in the historical filings. The relationship covers satcom ground-station LNAs and custom RF integration. The revenue contribution is approximately $5 million annually, structured as a multi-year framework against site rollouts on the Viasat ground station network.

Lockheed Martin and L3Harris consume custom microwave components for military satcom and radar systems. The revenue is project-driven. The 10-K language is generic: "custom microwave components for military satellite communications and radar systems." Concentration metrics on these customers are not disclosed at the contract level.

The defense and satcom lane is the historical base business. It is profitable on the component-level economics that the parts-house era ran on. It is not where the growth comes from. It is the floor under the quarterly revenue line that keeps the gross margin print stable while the ORAN and quantum lanes ramp.

VIII. The Customer Bench

The customer bench, named where disclosed:

That is the named bench. The unnamed customer pool, particularly in the cryo-LNA line where every superconducting quantum hardware vendor is a potential buyer, is wider than the disclosed list. Confidentiality agreements limit what AmpliTech can name in public filings. The Tier 1 unnamed ORAN customer is the most-asked question on every earnings call. Management has declined to name them because the carrier procurement cycle does not allow vendor announcements until the BEAD or commercial deployment is publicly committed. Track the FCC and state BEAD office award announcements through 2026 for the disclosure.

IX. The Competitor Map

The competitor map is different for each lane.

For the 5G ORAN small-cell ORU lane, the named competitors are Anokiwave (private, focused on phased-array antennas, recently acquired by Qorvo), Mavenir (private, software-led ORAN stack with hardware partners), Parallel Wireless (private, end-to-end ORAN), and the Asian ORU vendors who do not qualify for Buy America. The incumbent Ericsson and Nokia stacks compete only at the high-end macro-cell tier, not the small-cell ORU slot AmpliTech targets. AmpliTech's differentiation is the vertical integration plus the US manufacturing, not the radio specification per se.

For the cryogenic LNA lane, the competitors are Low Noise Factory (Swedish, the historical incumbent in academic and national-lab cryo LNA supply), Caltech-Microsemi (research-grade, not at production volume), and a handful of in-house programs at the major quantum hardware vendors. AmpliTech's differentiation is the US-based production volume and the commercial pricing point that scales with qubit count. The Swedish incumbent does not scale to the hundred-thousand-qubit target. AmpliTech can.

For the defense and satcom RF lane, the competitors are Qorvo, MACOM, Analog Devices Wireless, MaxLinear, Skyworks, and the captive RF programs at the defense primes. The lane is crowded. AmpliTech wins on small-program flexibility and on legacy customer relationships, not on volume economics. The lane is a margin floor, not a growth engine.

The cleanest comp for valuation purposes is the public small-cap RF and microwave cohort: Qorvo and Skyworks are large-cap reference points, MaxLinear and MACOM are mid-cap reference points, and Anokiwave (pre-acquisition) priced at a strategic multiple to Qorvo. The public 5G ORAN pure-plays are nearly non-existent in the small-cap segment: Mavenir is private, Parallel Wireless is private, and the listed ORAN names are component-level rather than full-stack. AmpliTech is the only US-listed integrated ORAN ORU vendor at a small-cap valuation. The comp set is more "compound semi growth" than "ORAN pure-play," because there is no clean ORAN pure-play in the small-cap public universe.

The named competitor map, simplified:

X. The Bear Case

I owe the bear case more space than the bull case in any Part 1. Here it is.

Maqbool has a credibility problem. The 2024 fiscal year was a $30 million revenue guide that landed at approximately $10 million. The miss was attributed to delayed FCC and CE certifications on the ORU product, which is a real and recurring risk in any radio-hardware ramp. The miss happened anyway. Anyone who entered the position in 2024 based on the $30 million number got a $2.75 print twelve months later. The 2024 guide miss is the single load-bearing concern on this name.

Layered on top of that, the company lost $3.5 million in 2024 to a custodial Bitcoin treasury arrangement that the FBI is investigating as fraud. The custodian failed. AmpliTech was a victim, not a perpetrator. But the decision to place corporate cash into a digital-asset custodian without sufficient operational due diligence is a capital-allocation judgment that any allocator has to weigh against everything else. It is a small dollar amount relative to the current cash balance. It is a large signal about the management team's risk-assessment process at the time.

The 2025 revenue guide of $24 million was implied rather than formally issued, and the year almost certainly came in below that number, with the bulk of the shortfall sitting in the second customer's order delays. The published Q1 2026 number annualizes to $21.4 million, with management explicitly guiding to a second-half load that needs to deliver an additional $20 million-plus of revenue to hit any version of the originally-implied $50 million 2026 target. The math is not impossible. It requires the second customer to convert at least $15 million of LOI to firm POs in the next six months, which the company has not publicly confirmed.

The ATM facility remains outstanding. If the second-half ramp slips, the company will need to raise additional capital before mid-2027. The clean balance sheet today does not mean the company is free of dilution risk through the back of 2026 and into 2027. The dilution risk is not the $250 million catastrophe scenario the 2024 bears warned about, because the post-rights-offering share count is already at 25 million and the equity base is $48 million. The dilution risk is the $5 million to $10 million capital raise at a depressed price if the ramp slips. That is the bear case range, not the apocalypse.

Customer concentration is the third risk. Two named LOI customers account for the bulk of the 2025 and 2026 revenue ramp. If either customer pauses the rollout, the revenue line slips by the same magnitude. Diversification is occurring slowly through the BEAD-funded regional carriers, but the named Tier 1 pair carries the revenue scenario in the base case.

Competitive risk is the fourth. The Asian ORU vendors will continue to bid on US deployments where the Buy America rules do not strictly apply, and the incumbent Ericsson and Nokia stacks could decide to ramp dedicated small-cell ORU products if the BEAD funding allocation forces them to. AmpliTech's small-cap advantage on speed and customization could erode at the volume tiers.

XI. Scenario Math

I do not run probability-weighted bear, base, bull scenarios with hand-tuned weights for a Part 1. I run scenario math that lets the reader weight it themselves.

Three revenue paths from a 2026 starting point:

Three sensitivity prices using a 4x EV-to-forward-sales multiple, which is below the historical RF small-cap median and roughly half the multiple Anokiwave commanded when Qorvo acquired it:

The bear-case multiple-and-revenue combination produces a stock 22 percent below today's $5.04 print. The base case is 57 percent above. The bull case is 182 percent above. The expected-value math, even on a flat 33-33-33 weighting across the three scenarios with no quantum optionality included, lands at roughly $8.70 against $5.04 today. The probability-weighted upside is 73 percent on a one-to-two-year horizon, before any contribution from the cryo-LNA line.

I am not pricing the quantum line. The quantum line is a tail. If the IBM 100,000-qubit timeline pulls into 2030 from 2033, the cryo-LNA revenue line alone exceeds today's total company revenue. That is not in the scenario math because I cannot probability-weight an IBM roadmap. It is in the optionality.

XII. The Catalyst Calendar

Eight named catalysts on the calendar through Q1 2027:

The first read is in mid-August. Q2 2026 results will show whether the second-half ramp is starting, whether gross margin is holding above 45 percent, and whether the LOI conversion has accelerated. That single print resets the multiple either direction.

The second read, harder to date, is the named-customer disclosure. The Tier 1 carrier the unnamed $78 million LOI sits with will eventually publish a BEAD or commercial deployment announcement that names AmpliTech as the radio vendor. That disclosure is what pulls a generalist tech analyst into the name and tightens the small-cap discount.

XIII. Kill Criteria

I run every position against an explicit kill-criteria list. The position exits if any of these trips:

One. The Q2 2026 print delivers revenue below $4 million. That would indicate the back-half loading is a stall, not a ramp.

Two. Gross margin reverts below 35 percent on a sequential basis. That would indicate the integrated-systems mix shift is reversing.

Three. The company files an 8-K disclosing a second material capital raise above $10 million at a price below $4.00. That would signal the ATM facility is being tapped under duress.

Four. Either named Tier 1 LOI customer issues a public statement pausing, descoping, or withdrawing from the AmpliTech procurement relationship.

Five. Maqbool sells more than 100,000 shares on the open market in any 90-day window. The historical pattern is zero open-market sales since IPO; a break in that pattern is a signal in itself.

Six. The IBM Quantum 100,000-qubit roadmap is publicly walked back, replaced with a non-superconducting architecture, or replaced with a different LNA vendor at the readout layer.

Seven. A new short writeup from a credible source publishes a thesis that documents specific financial misstatements, customer-relationship misrepresentations, or undisclosed contract cancellations. Existing bear writeups based on the 2024 guide miss and the BTC custody loss are already priced in.

None of those seven criteria are met today. None of them are imminent. The first testable one is Q2 2026 results in mid-August. Three months from today.

XIV. Position Plan

I have no AMPG position as of today, May 25, 2026.

I will open a starter position via daily DCA buys on the following schedule:

The four-tranche structure averages me into a name with 16 percent daily float turnover. Any single-session shock prints fast in this float, in either direction. DCA into the close on each day reduces single-print risk to one-quarter of the position size.

Target starter weight is small relative to the core book. I am not framing AMPG as a top-five position. It sits behind MRLN, IQE, BMM, LPKF, and the rest of the defense and chokepoint book in conviction order. AMPG is a tradable thesis, not a multi-year compounder anchor, until the Q2 print confirms or breaks the gross-margin step-up.

Conditional add line: a print below $4.25 on a non-news session adds one additional tranche from stable holdings. A print above $7.50 before Q2 earnings holds the position size flat. I am not adding above $7.50 pre-print.

XV. Bottom Line

AmpliTech Group printed a quarter on May 13, 2026 that the tape did not parse. Revenue up 48.6 percent year over year. Gross margin up 1,500 basis points. Cash $18.4 million. Debt zero. Stockholders' equity $48.4 million. Working capital ratio 4.25 against 1.68 three months earlier. The mix shift from parts-house to integrated 5G radio OEM showed up in the segment line and the margin line for the first time.

The pivot is real. The customer bench is real. The quantum optionality is real. The CEO is founder-aligned at more than 10 percent of the company, has never sold since IPO, and bought into the rights offering at the pro-rata. The bear case from the 2024 cycle is real and recent. The $2.75 print of late 2024 happened. The credibility problem is documented.

What the tape has not yet absorbed is that the unit economics of an integrated ORAN ORU at 48 percent gross margins is a fundamentally different business than the unit economics of a custom microwave parts house at 33 percent gross margins. The Q1 print is the first quarter where the math actually shows it.

The expected-value math on a one-to-two-year horizon, weighted flat across bear, base, bull at 4x EV-to-forward-sales, lands at $8.70 against $5.04 today. The quantum tail is not in that number.

The first testable catalyst is the Q2 2026 earnings print in mid-August. The DCA accumulation window for my starter position runs Tuesday May 26 through Friday May 29, 2026.

If the Q2 print delivers revenue at $7 million or above with gross margin at or above 45 percent, this becomes a sized-up position. If it delivers revenue below $4 million or gross margin below 35 percent, this becomes a closed position and the kill-criteria framework stops the thesis where it should.

I will report the average basis after the DCA window closes Friday.

Shawarma Capital

More in this series

$AMPG: The First Conversion. FCC And ISED Cleared The Radio, And The BEAD M

AMPG Stock: AmpliTech’s 48% Gross Margins, Open-RAN Radios, and a $7 Target

A live OSINT terminal that turns every public Merlin Labs signal into one operator-grade view. Updates itself in real time. 605 flights captured, 615 airborne hours, 21 aircraft tracked across 182 days.

shawarmacapital.net

(check your spam for the login code)

What it tracks:

Fifty-two disclosed Merlin, Dynamic Aviation, and KC-135 testbed tails on a 2D map and a 3D Cesium globe with real Esri satellite imagery. Five-source ADS-B quorum: airplanes.live, <adsb.fi>, OpenSky OAuth2, adsb.lol, geo-radius. Circuit breakers and last-known-good cache fallback. The map does not blank.

Twenty live SEC EDGAR filings. Material 8-K disclosures flagged red. Direct scrape of <merlinlabs.com/news> every 15 minutes for new GlobeNewswire press releases (the $80M PIPE, the CRO and CMO appointments, the C-130J PDR completion). Thirty-one active Merlin job postings classified by department. Five upcoming dated catalysts with source URLs. Twenty-five spam-filtered Stocktwits messages with a sentiment bar. Fifty Reddit posts. Fifty academic papers citing Merlin work. METAR for every fleet airport. SBIR awards. DoD contracts. Insider net buying summary when Form 4s land. Patents.

What it shows:

The certification testbed cadence peaked March-April 2026. Three platforms carry the load. N208B Big Red is the US testbed. ZK-MLN is the New Zealand cert testbed. ZK-MLO is the NZ Super Cargomaster cargo variant. All three synchronized to a spring peak.

Cert flying alone (Merlin testbeds): 377 sorties, 316 airborne hours. USAF KC-135 military pool: 223 sorties, 290 hours. Dynamic Aviation cargo: 7 sorties opening in May 2026.

How it works:

Three views. 2D map shows every fleet aircraft right now. 3D globe shows the same fleet with motion trails, atmosphere, day/night terminator. 24-hour replay scrubber drags back to any moment in the last day. Click any flight in the log, both surfaces lockstep to its takeoff point at 60x playback.

Server-Sent Events push channel. Every new flight, every new filing, every new tail-alert fires into the browser within seconds. No polling.

Access:

All paid Shawarma Capital subscribers get full dashboard access. Sign in at shawarmacapital.net with the same email as your Substack subscription. A six-digit code arrives in your inbox. Free readers see the public sneak with live fleet count, today’s airborne activity, and the latest Shawarma coverage.

Founding members get the same dashboard access plus the standard founding benefits.

More in this series

MRLN: The Complete Thesis

The Gate Cleared and the Tape Forgot

$MRLN: I built Merlin World

MRLN Part 10: The Deep Update

I Tracked Every Merlin Aircraft for 180 Days. The Testbeds Just Peaked

Alyeska Is Underwater. None of the Thesis-Breakers Have Hit

MRLN's $105M Ceiling Is Not the Revenue, It Is the Authorization

Alyeska Wrote Both Checks and the Market Missed It

The UAE Deal Changes Everything Merlin Was Supposed To Be

MRLN: Part 4 - The Board, the Preferred Supernova, and the Catalyst Nobody

$MRLN Part 3: The DAWG, the $54.6B, and Why Merlin Is Positioned to Win the

Almost everything written about a pre-revenue defense company comes from the company. The press release. The investor deck. The CEO on a podcast. All of it useful, none of it evidence, because a company describing itself is not the same thing as a fact.

This is a different exercise.

For six months I tracked every aircraft I could connect to Merlin Labs using nothing but public ADS-B data. ADS-B is the position signal that aircraft broadcast continuously, several times a second, and that anyone with an antenna or an archive can receive. Seventeen airframes. One hundred and eighty days. Roughly six hundred individual flights, reconstructed one at a time from raw position data. No company input. No NDA. No paid feed. Just the aircraft, telling on themselves.

Here is the finding in one sentence. Merlin's certification fleet flew the busiest stretch of the entire window in March and April 2026, and three separate testbeds, owned by two different Merlin entities, flying in two different countries, all peaked in the same eight weeks.

That is an independent, non-company-sourced read on whether the program is executing. It is the exact signal the thesis in Parts 1 through 8 is priced on, and until now nobody outside the company had measured it. What follows is the full roster, the method, the complete ledger, the testbed identification, and an honest accounting of what the data does and does not prove.

It is long. It is meant to be. Every number is checkable, and the last section tells you how to check it.

Let's go.

I. The One Rule For Reading Flight Data

Before any numbers, the rule that governs all of them.

Flight tempo is direction. It is not a date, and it is not a verdict.

A flight-test program flies hard when it is winning. It also flies hard when it is losing. It flies hard repeating test points a regulator rejected. It flies hard rebuilding a flight-hours case after a setback. Tempo rises for re-tests and for problems exactly as readily as it rises for progress. Position data cannot tell a milestone apart from a do-over.

So this is monitoring intelligence. It is not a trade and it is not a secret. It tells you the testbeds are busy, and it tells you precisely when that changed. It does not tell you the certification will pass. I will hold it that honestly for the entire post, and you should too. Anyone who hands you ADS-B data as a buy signal is selling you something.

II. What I Actually Did

The project had five steps.

One, I assembled the full list of aircraft connected to Merlin across five tiers of relationship, from Merlin-owned to entirely unrelated. Forty-eight airframes entered the net. Seventeen came out the other side as aircraft worth reconstructing in full.

Two, I resolved the ICAO hex code for each aircraft. The hex is the only identifier that matters for tracking, and Section V explains why getting it exactly right is the whole game.

Three, I reconstructed one hundred and eighty days of flight history from the public adsb.lol whole-planet archive, one calendar day at a time, and segmented the raw position traces into discrete flights.

Four, I analyzed the result: monthly tempo, sortie length, intensity, and the synchronized spring ramp.

Five, I cross-checked the one genuinely new public fact the project produced, the identity of the USAF KC-135 autonomy testbed, against the Air Force's own published imagery.

Every number below is reproducible from public sources. Nothing is modeled. Nothing is estimated. Where data is missing, it is reported as missing rather than filled in.

III. The Airframes Merlin Chose To Fly

Before the roster, the hardware, because the choice of airframe is itself a piece of the thesis.

Merlin's autonomy product, the Merlin Pilot, is software that flies an aircraft from takeoff to touchdown. Software does not certify in the abstract. It certifies on a specific airframe, with a specific regulator, on a specific type certificate. So the aircraft Merlin flies are not incidental. They are the certification.

The Cessna 208 Caravan is the civil workhorse of the program. It is a single-engine turboprop, rugged, simple, flown all over the world for cargo and short-haul passenger work. It is the airframe behind Merlin's FAA Supplemental Type Certificate effort and the concurrent New Zealand Civil Aviation Authority path. Five of the seventeen aircraft in this study are Caravans or close Caravan variants. When you see a Caravan flying short repeated sorties, you are watching the civil certification campaign happen.

The C-130J Super Hercules is the defense anchor. It is the four-engine military transport at the center of the $105 million USSOCOM IDIQ. Its autonomy program is pre-first-flight, targeted for late 2026, which means it has no identifiable airframe in this dataset yet. Anyone naming a C-130J tail number today is guessing.

The KC-135 Stratotanker is the aerial-refueling platform behind the Merlin and Sierra Nevada autonomy data-collection program hosted by the 171st Air Refueling Wing. It is a 1950s-design jet tanker, and the program that flew Merlin instrumentation aboard one is the subject of Section XI.

Three airframe classes. One civil, two defense. The civil one is the one you can actually watch certifying in the flight data, and it is where this study has the most to say.

IV. What "Merlin's Fleet" Actually Is

The first hard problem is definitional. "Merlin's aircraft" is not one list. It is five concentric rings, and only the innermost ring is Merlin-owned. Collapse the rings and you get the company badly wrong, which is exactly what the public commentary does.

Ring 1. Merlin Labs owned. Five aircraft.

N208B, "Big Red," a Cessna 208B Grand Caravan, is the primary US FAA certification testbed, based at Quonset State Airport in Rhode Island. N506DB is a Burton Long-EZ experimental, a small R&D platform out of Mojave, California. ZK-MLN is a Cessna 208B Grand Caravan registered to Merlin Labs NZ Ltd, the New Zealand certification testbed. ZK-MLO is a Cessna 208B Super Cargomaster, also Merlin Labs NZ, a testbed that also flies paid charter and freight. ZK-MLP is a Cessna TU206F Turbo Stationair, lightly used, which never broadcast a clean identifier I could lock onto. It is named here for completeness and excluded from the flight history. I will not report numbers I did not measure.

Five airframes. That is the owned fleet.

I flag the count only because the figure that circulates publicly is fifty-five. Fifty-five is a 2021 press figure for the King Air fleet of a different company, Dynamic Aviation, that Merlin once announced an intent to retrofit with autonomy. Those aircraft are Ring 5 below. They are not Merlin's fleet and never were. A company with five owned aircraft and two more under operation is not a smaller version of a fifty-five-aircraft company. It is a different company, and the data should describe the real one.

Ring 2. Operated, not owned. Two aircraft.

VH-WZY is a Cessna 208B Grand Caravan, Australian-registered, operated by Merlin since January 2026. ZK-JMP is a Cessna 208B Supervan, owned by a skydiving operator and flown under lease by Merlin under the callsign MERLIN 1. As the ledger shows, it is the single busiest aircraft in the entire dataset.

Ring 3. Partner testbed. One aircraft.

N437VN is a Scaled Composites Model 437 "Vanguard," the airframe behind Northrop Grumman's Beacon autonomy testbed program. Merlin is one of six integration partners on Beacon. This aircraft is partly Merlin and mostly not, and it is weighted that way throughout.

Ring 4. The military host pool. Nine KC-135 tankers.

The 171st Air Refueling Wing, Pennsylvania Air National Guard, based at Pittsburgh, is the host unit for the Merlin and Sierra Nevada KC-135 autonomy program. Eight 171st KC-135T jets sit in the host pool, plus one separate KC-135R candidate at MacDill. Read Section X before you let this ring into the Merlin story, because it does not belong there.

Ring 5. The Dynamic Aviation King Airs. Thirty-five aircraft.

These belong to Dynamic Aviation Group of Bridgewater, Virginia, not to Merlin. They are the "fifty-five" press figure. There is no public per-tail confirmation of which were ever converted to autonomy or how many. They fly Dynamic's own ISR and survey business. Their activity is noise. I tracked them only to rule them out, and ruled them out. They appear in no number presented here as a Merlin signal.

The universe is five owned, two operated, one partner, nine military host, thirty-five unrelated. That is the shape of the thing before a single flight is counted.

V. The Unit Of Truth Is The ICAO Hex

You cannot track an aircraft by its tail number, because the tail number is not broadcast. What an aircraft transmits is its ICAO hex, a 24-bit address baked into the transponder. Every flight-tracking service in the world keys on that hex. Track the wrong hex and you have silently tracked the wrong aircraft for six months and learned nothing, and you would never know.

So the hex codes had to be exact. Sixteen of the seventeen I read straight from the raw archive, matched against known registration pairs. The seventeenth, the Northrop Model 437, never showed a clean registration-to-hex pair anywhere in the raw data.

For that one I computed the hex. The FAA assigns hex codes to US-registered aircraft by a fixed, reversible algorithm, so an N-number converts to its hex deterministically. I implemented that algorithm, and before trusting it I ran it against nine aircraft whose hex codes I had already confirmed from raw data. The first pass failed. A per-digit block offset was wrong, set to plus one when it should have been plus six hundred and one. After the fix, the algorithm reproduced all nine confirmed pairs exactly. Only then did I let it produce the Model 437 code, a53f07. That is why the Model 437 line is allowed to exist in the ledger. It was validated nine times before it was used once.

That is the standard the rest of this analysis is built to. If a number cannot survive that level of check, it is not in here.

VI. The Method: 400 Gigabytes To Surface A Few Megabytes

The data source is the adsb.lol globe_history archive. It is public and openly licensed. It is also brutally inconvenient, and the inconvenience is the reason almost nobody does this.

There is no per-aircraft endpoint. The archive does not let you ask for one plane. It publishes one whole-planet file per day, every aircraft on Earth, roughly two gigabytes compressed and split into four parts. To get one aircraft's track for one day, you download the entire planet, extract the trace files for the seventeen hex codes that matter, and delete the rest.

Then you do that one hundred and seventy-four more times.

On the order of four hundred gigabytes of archive was moved to surface a few megabytes of actual signal. Peak disk footprint stayed near three gigabytes, because each day's planet was deleted before the next was pulled. The run was driven in fifteen-day chunks, with results committed after every chunk, so an interrupted machine never lost more than a few days of progress.

Turning a raw position trace into discrete flights needs one definition: when does a sortie end. The rule here is ground time. More than twenty-five minutes stationary closes one sortie and opens the next. Any airborne segment under two minutes is discarded as noise, a stray ping rather than a flight. Airborne hours are the plain sum of segment durations. Six days in the window had no archive published at all, and those six are reported as gaps, never as zeros, because inventing a zero on a day with no data would quietly corrupt every monthly count downstream.

How the 180-day dataset was built. 181 days in window, 175 with data, 6 gaps, 17 aircraft, 400 GB processed, 597 sorties, 600 airborne hours, 9 of 9 hex pairs validated.

VII. The Finding: The Spring Ramp

Merlin's certification flying was not steady. It ramped, hard, into spring 2026. And it is not one aircraft doing it. Three separate testbeds, owned by two different Merlin entities, flying in two different countries, all posted their heaviest months of the entire window in March and April 2026.

ZK-MLN, the New Zealand certification Caravan, ran 4, 0, 0, 4, 11, 14, 8 sorties from November to May. Dead through December and January. February wakes it. March and April are its two busiest months of the window.

ZK-MLO, the Super Cargomaster, ran 0, 0, 0, 0, 9, 22, 11. Four straight months of zero. Then March takes it from nothing to nine sorties, and April to twenty-two, its single most active month.

N208B, "Big Red," the US certification testbed at Quonset, ran 12, 9, 5, 11, 1, 24, 11. Choppy all winter. Near-grounded in March at a single sortie. Then twenty-four sorties in April, its biggest month by a wide margin.

Three aircraft. Two countries. One eight-week window. Three independent flight-test programs do not all peak in the same eight weeks by accident. Something coordinated the calendar. That is the finding, and it is the kind of result you can only reach with data the company did not hand you.

Three testbeds, one spring ramp. Monthly sorties for N208B, ZK-MLN, and ZK-MLO, with the excluded 171st KC-135 fleet line for contrast. All three testbeds peak in March and April 2026.

VIII. Sortie Length Tells You What Kind Of Flying It Is

A sortie count tells you how often. The average sortie length tells you what they were doing. There is a short group and a long group, and the split is clean.

For a testbed, short sorties mean circuits. Takeoff, pattern, landing, repeat. That is the literal rhythm of certifying an autopilot. You fly the same profile again and again, varying one parameter, logging the result. N208B flew twenty-four April sorties averaging fifty-four minutes each. That is a flight-test schedule, not transport flying.

The long group is transit legs, charter runs, and tanker tracks. Ordinary point-to-point flying. ZK-MLO at seventy-two minutes is the testbed that also earns revenue, which is why it sits between the two groups. The KC-135s span sixty-three to one hundred and three minutes, which is simply what tanker flying looks like.

Intensity points the same way. When these aircraft fly, they tend to fly again the same day. N208B averaged 2.0 flights per active day. VH-WZY 2.5. ZK-JMP 2.9. A program that flies an aircraft two or three times on the days it flies at all is working a schedule, not running errands.

IX. The Full Ledger

No rounding, no summary, no aircraft hidden. Every airframe in the dataset, all 175 days of usable data. The figure below is the complete ledger: sorties, airborne hours, and active days for all seventeen aircraft.

The full ledger. Every aircraft tracked, 175 days of data, November 20 2025 to May 19 2026. ZK-JMP leads with 181 sorties, fleet total 597 sorties and 600 airborne hours.

ZK-JMP is the workhorse. One hundred and eighty-one sorties, almost one hundred and thirty-seven airborne hours, sixty-two active days. It is the busiest aircraft in the fleet by a factor of more than two over the next aircraft. Its behavior in the spring is its own story, in Section XII.

The two zeros are reported as zeros, not quietly dropped. N506DB, the Long-EZ R&D platform, logged no captured flights across the whole window. 63-8020, the MacDill KC-135 candidate, was added to the run late and covers only the tail of the window, and it logged none either. A platform that did not fly is a fact, and facts that argue against an exciting narrative get the same ink here as the ones that support it.

X. The KC-135 Ramp I Am Deliberately Not Counting

Here is the part a weaker version of this post would lead with, and it would be wrong to.

The nine 171st-wing KC-135 tankers also ramped. Steeply.

The KC-135 line climbs from eight sorties a month to seventy-five. It is the most dramatic line in the entire dataset. It is also not a Merlin signal, and folding it into the Merlin story would be the single most misleading thing anyone could do with these numbers.

Tanker wings ramp on their own cycle. Deployments, exercises, currency requirements, seasonal training blocks. A KC-135 wing flying harder in late winter tells you about that wing, not about one autonomy program riding along on a handful of its jets. The two KC-135s actually tied to Merlin by photographic evidence, 59-1460 and 58-0084, sit mid-pack at thirty and nineteen sorties against a host-fleet range of eighteen to thirty. Flight tracking cannot pick the autonomy testbed out of the pack. The data does not have the resolution, and pretending otherwise would be a lie of exactly the kind this post exists to avoid.

So the KC-135 ramp is real, it is dramatic, and it is excluded from the Merlin read on purpose. I am telling you it exists, and I am telling you to ignore it. That discipline is the whole point of doing this with data instead of vibes.

XI. The Military Testbed: 59-1460, "Classic Iron"

If flight tracking cannot identify the KC-135 testbed, how is it identified at all? Photographs. And the evidence is worth walking in full, because this is the one place the project produced a genuinely new public fact.

The aircraft is USAF KC-135T serial 59-1460, nicknamed "Classic Iron," ICAO hex ae0596, a 171st Air Refueling Wing tanker out of Pittsburgh. Three independent threads converge on it.

One, the Air Force's own imagery. The official DVIDS gallery documenting the program, titled "Next-Gen Autonomy: 171st Fuels Future of Self-Flying Tankers," shot by the 171st wing on August 1, 2024, contains six images. Five show the Merlin avionics data-acquisition rig installed inside a KC-135 cabin. The sixth, DVIDS image 8630294, is an exterior ramp shot. Two people carry that same instrumentation cart directly alongside a KC-135 whose nose is painted "1460," with a boarding airstair deployed at that aircraft's forward door. The Air Force left the tail number out of every single caption. The photograph did not.

Two, the registry. "1460" resolves to USAF serial 59-1460, an independently verified 171st-wing KC-135T.

Three, the archive. Forensic analysis of fifteen days of ADS-B archives across the July-to-August 2024 test campaign window confirms 59-1460 operating out of Pittsburgh on four days inside that window.

The precise, defensible claim is this. The Air Force's own published program photo set shows the test instrumentation at the door of KC-135 59-1460. That is an evidence-based identification, not an Air Force-stated confirmation, and the wording matters. A second tail, 58-0084 "Ghost Raiders: Final Contact," appears in a separate Merlin-supplied program photo and is the second KC-135 tied to the program by imagery.

That is as far as honest analysis goes. The program ran a first data-collection flight in late July 2024, flew three sorties, and the instrumentation was removed around August 1, 2024. Sierra Nevada Corporation is the integration partner, Merlin supplies the autonomy, and GE Aerospace joined the effort in September 2025. None of that requires guessing. The C-130J program, by contrast, has no identifiable airframe yet, and that gap is reported as a gap.

XII. Two Movements Worth Watching

Two aircraft did something in the spring the monthly numbers do not explain on their own.

ZK-JMP, the leased freight Caravan and the busiest aircraft in the fleet, flew fifty-two sorties in March and then one in April. The workhorse went dark in the exact month the certification testbeds peaked. One reasonable hypothesis is that the fleet shifted weight off revenue freight flying and onto test flying. The data is consistent with that reading. It does not prove it, and an aircraft can go quiet for maintenance, scheduling, or contract reasons that have nothing to do with the test program.

VH-WZY, the Caravan Merlin took over in January 2026, flew twenty-one sorties in its first month and nine in its second, then went silent for three months. Either it returns to the line or that arrangement stalled. Right now it is an open question, and it is flagged as one rather than guessed at.

XIII. Tying The Flight Data To The Thesis

I have spent twelve sections being careful. This one too.

The Merlin thesis, laid out across Parts 1 through 8, is priced on certification. The FAA Supplemental Type Certificate for autonomy on the Cessna 208 Caravan, on its Stage of Involvement path. The concurrent New Zealand CAA effort, which is why ZK-MLN and ZK-MLO exist as a separate testbed pair on the other side of the world. The $105 million USSOCOM IDIQ on the C-130J, whose preliminary design review cleared on March 5, 2026, with a first flight targeted for late 2026. Every one of those is a milestone that has to be flown to be reached.

For context, here is the program bench the thesis rests on.

What the flight data adds to that bench is independent corroboration that the program is, in fact, executing. Not company-sourced. Not a press release. Three certification testbeds, across two regulatory jurisdictions, flew their heaviest stretch of a 180-day window in March and April 2026, in the short-sortie circuit pattern that matches autonomy certification work.

If Merlin were quietly stalled, that is not the signature you would expect in six months of position broadcasts. You would expect flat or falling tempo and idle certification testbeds. Instead the data shows a synchronized peak across two countries.

That is corroboration. It is not proof. The flight data raises or lowers your confidence that the company is doing the work it says it is doing. It does not settle the bull or bear case, and it does not put a date on anything.

It does bear, mildly, on the kill criteria the thesis is governed by. Part 8 re-checked the seven criteria that would break the position and found zero of them tripped. The first criterion is the C-130J critical design review failing or the program being descoped. A certification fleet flying its hardest stretch of the observable window is weak evidence against a quiet descope. Weak, because tempo cannot tell a milestone from a re-test, but it is evidence, and it points the right way. The second criterion is a named customer issuing a stop-work order. Nothing in six months of flight data looks like a stand-down. The testbeds did the opposite of standing down.

XIV. Read It From Every Side

A single reading of a dataset is how you fool yourself. So here are four, including the ones that hurt.

The bull read. Three independent testbeds, two countries, two operators, all flying their heaviest of the window in the same eight weeks, at sortie lengths that match circuit work. That is a certification program accelerating into its milestones.

The bear read. Tempo is not progress. A program flies hardest when it is fixing problems, repeating failed test points, or rebuilding a flight-hours case a regulator pushed back on. The same March-April spike is fully consistent with a program behind schedule and catching up. The data cannot tell a re-test from a milestone, and an honest reader has to hold that.

The noise read. Six months is one window, not a trend. Winter weather alone suppresses flying from November through January, so part of the spring climb is seasonal thaw rather than program intent. Strip the season out and the true signal is smaller than the raw monthly counts make it look.

The discipline read. The 171st KC-135 ramp is steeper than anything on the Merlin testbeds, and it means nothing here. Putting it in the Merlin story would produce the most dramatic chart in this post, and it would be false. Leaving it out is the point.

What survives all four reads: the certification testbeds are demonstrably busier than at any other point in the window, and the timing across three aircraft is synchronized in a way coincidence does not explain easily. Everything past that sentence is interpretation, and you now hold the same numbers I do.

XV. The Catalyst Clock

The flight data is one input into a thesis that has hard structural dates ahead of it. The Section 7(b) VWAP measurement window runs across August and September 2026. The insider lockup expires September 16, alongside S-1 effectiveness. The conversion-price reset window opens in mid-October. The C-130J first flight is targeted for late 2026.

Note the timing. The next flight-data window, June, lands well before those structural dates. That is the point of watching it. The certification tempo is the cleanest read available before the capital-structure events arrive, and it is the read least contaminated by them. The lockup, the VWAP reset, and the S-1 effectiveness are share-count mechanics. The flight tempo is the operating business. Watching the second one tells you whether the first set is arriving into a company that is executing or a company that has stalled.

XVI. What I Am Watching Next

What turns this from a snapshot into a signal is the next window. June is the first clean test of whether the spring ramp was a step change or a one-off test block.

If ZK-MLN, ZK-MLO, and N208B all hold double-digit sorties through June and July, the spring ramp was a genuine gear change and the program has moved to a higher baseline. If they sink back toward winter levels, March and April were a discrete test block and not a new normal.

Then the two open questions. Does VH-WZY come back from its three-month silence. Does ZK-JMP stay dark, which would confirm a shift of fleet weight from revenue toward test flying, or return to freight.

And the highest-value watch item of all, any new tail. A previously unseen aircraft at Quonset or in New Zealand would mean the fleet is physically growing. Merlin's management has publicly described a strategy of becoming a defense "neoprime" through M&A. A new airframe appearing on ADS-B is the earliest possible read on whether that strategy is moving from talk to metal, and it is exactly what this method is built to surface.

What turns a snapshot into a signal. Combined testbed sorties per month, peaking at 60 in April, with June flagged as the first clean test, and three watch items: VH-WZY, ZK-JMP, and any new tail.

XVII. How To Check My Work

This is the part that separates analysis from assertion. Everything above is reproducible, and here is the recipe.

The aircraft list is the five rings in Section IV. The hex codes are stated next to every aircraft in the ledger. The data source is the adsb.lol globe_history archive, public and openly licensed, one whole-planet file per calendar day. The window is November 20, 2025 to May 19, 2026. The segmentation rule is a twenty-five-minute ground gap to split sorties and a two-minute floor to discard noise. Run that against the archive for those seventeen hex codes and you will reconstruct the same 597 sorties and 600 airborne hours, because nothing here is modeled.

The testbed identification rests on USAF DVIDS image 8630294, which is public domain and viewable by anyone. The registries are the FAA aircraft registry and the New Zealand Civil Aviation Authority register, both public. The company financials, contract values, and capital-structure dates referenced in Sections XIII and XV trace to SEC EDGAR filings under Merlin's filing entity and to the earlier parts of this series.

If a number in this study is wrong, it is wrong in a way you can find, because the inputs are all public and the method is all stated. That is the standard. Disagree with the interpretation all you like. The interpretation is in Section XIV and it is labeled as interpretation. The numbers are not opinion.

XVIII. The Bottom Line

Six months. Seventeen aircraft. Roughly six hundred sorties and six hundred airborne hours, segmented one flight at a time from about four hundred gigabytes of public archive. One computed hex code, validated nine times before it was used once. One military testbed identified from the Air Force's own photograph. One ramp the data shows clearly and one ramp the data shows clearly and that you must throw away anyway.

The finding stands on its own. Merlin's certification testbeds flew their busiest stretch of the observable window in March and April 2026, three aircraft across two countries peaking in the same eight weeks, in the flight pattern that matches certifying an autopilot. That is independent evidence the program is executing, gathered without a single word from the company.

It is not proof the certification passes. It is not a date. It is not a trade. It is monitoring intelligence, and the next window, June, is what tells you whether the spring ramp was a gear change or a test block. Same method, same sources, next update when the June archives are complete.

Part 8 covered the Q1 print, the burn, and the seven kill criteria. This post covered the airplanes.

Disclaimer and position disclosure. This post is research synthesis, not investment advice. The author holds no disclosed position for the purposes of this post. You should not buy or sell securities based on anything written here. I am not a registered investment advisor. I do not owe you a fiduciary duty. My conclusions could be wrong in ways I have not anticipated. Every figure is reconstructed from public data and stated sources, and where data was missing it is reported as missing. Projections and interpretations are model outputs, not guarantees. Microcaps are illiquid and can lose 50% or more in a single session. Do your own due diligence. Verify every number against the primary sources named above.

More in this series

MRLN: The Complete Thesis

The Gate Cleared and the Tape Forgot

$MRLN: I built Merlin World

MRLN Part 10: The Deep Update

Merlin Intelligence - Live OSINT Tool

Alyeska Is Underwater. None of the Thesis-Breakers Have Hit

MRLN's $105M Ceiling Is Not the Revenue, It Is the Authorization

Alyeska Wrote Both Checks and the Market Missed It

The UAE Deal Changes Everything Merlin Was Supposed To Be

MRLN: Part 4 - The Board, the Preferred Supernova, and the Catalyst Nobody

$MRLN Part 3: The DAWG, the $54.6B, and Why Merlin Is Positioned to Win the

April 30 came. The company said the phrase. The stock did not wait eighteen months. It went up roughly two and a half times year to date and re-rated past every published analyst target inside three weeks.

This is the part where I tell you what I got right, what I got wrong, what the company actually disclosed line by line, and the one thing neither Part 1 nor Part 2 mentioned at all: there is an activist investor already sitting on the LPKF supervisory board, and the chief financial officer has been buying stock in the open market with his own money.

Let's go.

Read more

The tape sold 8 percent. It sold the Q1 headline loss. The Q1 headline loss is the single most misread number in the file. This post starts there.

Let's go.

Read more

The peer tape ruled out sector rotation. The drawdown was company-specific. The question is which fact the market was actually selling.

Three candidates: the GAAP loss line, the resale registration becoming effective, or the founder-tier subscribers exiting the position in the chat in real time. All three got priced together. The first two are mechanically misread by the tape. The third is the signal I have to take seriously, and it gets its own section below.

Let's go.

Subscribe now

Read more