NDAA Replicator: actual budget, contractor-to-ticker map, and public production milestones
The question
What is the NDAA Replicator program's actual FY25-FY26 obligated budget, the prime/sub contractor list with public tickers, and what production milestones are public? Context: founder named drones / DoD Replicator (iMessage 2026-05-12) as a candidate orthogonal investing thesis; vault had zero Replicator coverage, and the appeal is a named procurement vehicle + public budget + short 2025-2026 timeline — the same demand-side-anchor shape as the power-bottleneck argument in the Innermost Loop thesis.
What we already know (from the vault)
- The vault has zero prior Replicator coverage. Closest structural analog is [[2026-05-12-innermost-loop-ai-infrastructure]], which establishes the exact pattern the founder is reaching for: a thesis anchored to a named bottleneck with a public, trackable demand signal and explicit position-sizing by leverage layer (power = TLN/GEV/OKLO/BE/FLNC).
- The RDCO investing operating model ([[readme]] — investing project) is position/capital-cycle horizon, not day-trading: tranche-accumulate, no per-trade stops, phase-marker exits, R-unit sizing. Any Replicator thesis must clear that same bar — a named demand anchor with verifiable phase markers, per [[2026-05-17-power-cycle-v1]].
- Adjacent demand-anchor discipline is already codified in the power-layer backfill work ([[2026-05-17-power-anchors-backfill-notes]]): a thesis only survives if the anchor metric stays on trajectory. Replicator's analog anchor would be appropriated-dollars + delivered-units, both partly public.
- Note of caution baked into the house style ([[2026-06-08-innermost-loop-longevity-escape-velocity]]): separate the rhetorical hook ("swarm of thousands by August 2025") from the verified datum (actual deliveries). Replicator is a textbook case where the headline and the delivered count diverge.
What the web says
- Funding has no dedicated budget line. Per the CRS product DOD Replicator Initiative: Background and Issues for Congress (IF12611, congress.gov), DoD funded Tranche 1 via a $300M FY2023 reprogramming request, $200M in FY2024 appropriations, and a $500M FY2025 request. For FY25-FY26 the department shifted to aggregated requests of ~$500M/year distributed across multiple service accounts, not a single consolidated line. CRS explicitly flags that little is public on total program cost. (https://www.congress.gov/crs-product/IF12611)
- So the honest "obligated" answer: roughly $500M committed to Tranche 1 (FY24 appropriations + reprogramming), a $500M FY25 request, and a planned ~$500M FY26 ask folded into Replicator 2 (counter-sUAS) — but precise obligated (vs requested/appropriated) figures by fiscal year are not publicly broken out, because the money is scattered across service accounts. (Federal News Network; congress.gov IF12611)
- Hardware vendors (Tranche 1/1.2): AeroVironment (Switchblade 600 loitering munition), Anduril Industries (Ghost-X UAS + Altius-600), Performance Drone Works (C-100 UAS). (DefenseScoop; Aviation Today, https://www.aviationtoday.com/2025/09/04/hundreds-of-switchblades-delivered-so-far-in-first-replicator-tranche-diu-official-says/)
- Software vendors: ORIENT (C2 resilience) — Viasat, Aalyria, Higher Ground, IoT/AI. ACT (autonomous swarm coordination) — Swarm Aero, Anduril, L3Harris. (DefenseScoop, "Pentagon unveils winners of Replicator software contracts," 2024-11-20)
- Breadth: >30 hardware/software companies awarded (≈75% non-traditional defense contractors) plus >50 subcontractors; >500 firms considered across both tranches. No contract dollar values were disclosed for the software awards. (DefenseScoop)
- Production milestones (public): Original goal — field thousands of attritable autonomous systems by August 2025. Actual disclosed delivery (DIU's Sarah Pearson, Aug 28 2025): "hundreds" of Switchblade 600s delivered in Tranche 1, against an Army commitment to buy/field 1,000+ Switchblade 600s. DIU's public posture: "we consider the problem set to have been met." Switchblade 600 Block 2 enters production 2026 for Replicator + LASSO. (Aviation Today; Army Recognition)
- AeroVironment scale-up (the most ticker-relevant datum): announced (Oct 13 2025) a ramp toward ~1,200 Switchblades/month (~14,400/yr) with a new Salt Lake City "FreedomWerx" facility online late-2026/early-2027; a separate ~$288M Army Switchblade delivery order and a reported ~$1B Army Switchblade deal are public. (Defense News; TWZ; Army Recognition)
Convergences and contradictions
- Confirmed public: the ~$500M/year scale, the no-dedicated-line structure, the named awardees, and the "hundreds delivered vs thousands promised" delivery gap. These are corroborated across CRS + DefenseScoop + DIU on-record statements.
- Speculative / not public: per-company contract dollar values (DIU withheld them), exact obligated (vs appropriated/requested) figures by FY, and the FY26 Replicator-2 final appropriation. Treat any per-vendor revenue attribution as estimate, not fact.
- Contradiction to respect: DoD's "problem set met" framing vs trade-press "drone delays / still waiting" framing (Responsible Statecraft; Washington Times). The program shipped, but well short of the headline "thousands by Aug 2025." Don't trade the press release.
Synthesis for RDCO
Is this a real orthogonal thesis? Partially — and it's narrower than it looks. The Replicator structure does rhyme with Innermost Loop: a named procurement vehicle, a public-ish budget, a short timeline. But the investable surface is thin. The defining problem for an equity thesis is that the program was deliberately built to route ~75% of awards to non-traditional, mostly private vendors. The marquee names — Anduril and Performance Drone Works are privately held and not investable on public markets (Anduril is the single most-cited Replicator prime and you simply cannot buy it). That guts the cleanest exposure.
Public tickers with credible, but mostly immaterial, Replicator exposure:
- AVAV (AeroVironment) — the only genuine pure-play. Switchblade 600 is the disclosed Replicator hardware win, and the 14,400/yr capacity ramp + ~$1B Army order are real, material company-level catalysts. Needle-mover for AVAV specifically. But note the cycle stage: AVAV is already a momentum-rich defense-drone name, so this is a "trend continues, entry is late" trade, not a discovery — same caveat applied to VRT in the Innermost Loop sizing logic.
- LHX (L3Harris) — named on ACT swarm-coordination software. For a ~$50B+ defense prime, a Replicator software award is a rounding error, not a thesis driver. Exposure is real but immaterial.
- VSAT (Viasat) — named on ORIENT C2-resilience. Same story: immaterial to a company whose stock moves on satellite-broadband fundamentals.
- KTOS (Kratos) — frequently grouped with the drone-stock cohort in trade press but not confirmed as a named Tranche-1 Replicator awardee in these sources; treat as adjacent autonomous-systems exposure, not Replicator-specific.
The honest call: Replicator is a clean demand-signal narrative but a diffuse equity thesis. The one place it concentrates into a public ticker is AVAV, and even there the right framing is "Replicator is one validating data point inside a broader loitering-munition demand surge (LASSO, FMS, Ukraine-driven restock)," not "buy the Replicator program." The dollar amounts are real but small ($500M/yr) relative to the defense primes that partly capture them, and the most exposed companies are off-limits to public investors. Versus the power layer in Innermost Loop — where the bottleneck dollars ($156B stalled) dwarf the company market caps and the pure-plays are public — Replicator is the weaker-leverage cousin. Worth a watch-item and possibly a small AVAV-anchored "autonomous attritables" sleeve, not a standalone capital-cycle thesis on the Innermost Loop tier. No fabricated per-vendor dollars: if a figure isn't above, it is not publicly confirmed.
Open follow-ups
- Pull AVAV's last two 10-Qs/10-Ks: what % of revenue is loitering munitions, and how much is attributable to Replicator vs LASSO/FMS? (feeds an AVAV-specific anchor like the hyperscaler-capex EDGAR watch)
- Track the FY2026 NDAA + appropriations bill for the actual Replicator 2 (counter-sUAS) number and which service accounts it lands in — the closest thing to a real "obligated" figure.
- Map the private primes (Anduril, PDW, Swarm Aero) for any secondary-market access or eventual IPO timeline — same private-secondaries logic as Layer 1 of Innermost Loop.
- Is there a counter-sUAS public pure-play (the defense side of Replicator 2) that's less picked-over than the attack-drone names?
- Backtest question: does a "named DoD attritable-autonomy procurement vehicle announced" event have any historical signal for AVAV/KTOS, or is it already priced by announcement day?
Related
- [[2026-05-12-innermost-loop-ai-infrastructure]] — the structural template (named-bottleneck + public demand anchor + layer-sized exposure) this brief is tested against
- [[2026-05-17-power-cycle-v1]] — sibling demand-anchor thesis; the operating model (tranche-accumulate, phase-marker exits) any Replicator thesis must clear
- [[2026-05-17-power-anchors-backfill-notes]] — the anchor-metric-on-trajectory discipline a Replicator thesis would need (appropriated $ + delivered units as the anchor)
- [[2026-06-08-innermost-loop-longevity-escape-velocity]] — house rule on separating rhetorical hook from verified datum (directly applies to "thousands by Aug 2025" vs "hundreds delivered")
- [[readme]] — Investing project operating model (position/capital-cycle horizon, not day-trading)
Sources
- Vault: ~/rdco-vault/01-projects/investing/theses/2026-05-12-innermost-loop-ai-infrastructure.md
- Vault: ~/rdco-vault/01-projects/investing/theses/2026-05-17-power-cycle-v1.md
- Vault: ~/rdco-vault/01-projects/investing/anchors/power-layer/2026-05-17-power-anchors-backfill-notes.md
- Vault: ~/rdco-vault/06-reference/2026-06-08-innermost-loop-longevity-escape-velocity.md
- Vault: ~/rdco-vault/01-projects/investing/readme.md
- CRS, DOD Replicator Initiative: Background and Issues for Congress (IF12611): https://www.congress.gov/crs-product/IF12611
- DefenseScoop, "Pentagon unveils winners of Replicator software contracts" (2024-11-20): https://defensescoop.com/2024/11/20/replicator-software-contract-awards-winners-diu-pentagon/
- Aviation Today, "Hundreds Of Switchblades Delivered So Far In First Replicator Tranche" (2025-09-04): https://www.aviationtoday.com/2025/09/04/hundreds-of-switchblades-delivered-so-far-in-first-replicator-tranche-diu-official-says/
- Defense News, "AeroVironment eyes new factory, drone launches for Switchblade" (2025-10-13): https://www.defensenews.com/air/2025/10/13/aerovironment-eyes-new-factory-drone-launches-for-switchblade/
- Army Recognition, "AeroVironment to produce 14,400 Switchblades per year": https://www.armyrecognition.com/news/army-news/2025/aerovironment-to-produce-14-400-switchblades-loitering-munitions-per-year-in-the-us
- Federal News Network, "Pentagon's Replicator 2 to focus on drone defense" (2024-09): https://federalnewsnetwork.com/defense-main/2024/09/pentagons-replicator-2-to-focus-on-drone-defense/
- TheStreet / Benzinga drone-stock cohort (ticker confirmation AVAV/LHX/VSAT, Anduril private): https://www.thestreet.com/investing/drone-stocks-to-target-as-military-appetite-surges