RTX is durable under ANCHOR.

Not clean. Not frictionless. Not software-like.

Durable because the bottlenecks are physical, regulated, certified, and installed inside aircraft and defense systems that customers cannot casually replace.

ANCHOR Score: 49 / 60

Badge: ABIP ANCHOR Certified

Gates:

RTX owns real-world aerospace and defense bottlenecks: engines, avionics, MRO, sensors, missiles, interceptors, and government programs. AI can improve inspection, planning, engineering, and factory throughput. It does not replace F135 engines, GTF shop visits, Patriot interceptors, Collins flight systems, or the certification stack around them. RTX reported Q2 2026 sales of $24.7 billion, free cash flow of $2.9 billion, and a $289 billion backlog split between $170 billion commercial and $119 billion defense.

The market is focused on backlog, defense restocking, commercial aftermarket, and the recovery path at Pratt & Whitney.

RTX raised 2026 adjusted sales guidance to $95–$96 billion and adjusted EPS guidance to $7.10–$7.25 after Q2. Reuters framed the setup as sustained aircraft repair demand plus governments replenishing weapons stockpiles. Pratt & Whitney sales rose 16% in Q2 to $8.89 billion. Raytheon sales rose 18% to $8.27 billion, helped by demand for Patriot, Standard, and AMRAAM systems.

That is the right headline.

But the real question is not “does AI help RTX?”

It does.

The real question is whether AI can compress the parts of RTX that matter.

Mostly, no.

RTX is not protected because it is big.

It is protected because replacement is slow, regulated, expensive, and operationally painful.

The company has three main segments: Collins Aerospace, Pratt & Whitney, and Raytheon. Collins sits inside aircraft systems and aftermarket support. Pratt designs, manufactures, and services commercial and military engines. Raytheon sells advanced defense systems, missiles, sensors, and air-and-missile defense capabilities.

This is not a SaaS renewal.

This is installed hardware, safety certification, government procurement, spare parts, maintenance capacity, export controls, classified work, and decades-long programs.

RTX also has customer and execution risk. U.S. government sales were $33.3 billion in 2025, or 38% of total net sales. International sales were $41.3 billion, or 47% of total net sales. Pratt’s largest commercial customer by sales is Airbus, at 29% of Pratt & Whitney segment sales in 2025 before discounts and incentives.

The moat is real.

So are the scars.

The GTF powder-metal issue is the proof. Pratt expects elevated aircraft-on-ground levels for the PW1100-powered A320neo fleet through 2026. RTX recorded a $2.9 billion pre-tax operating profit charge in 2023 and still had $0.7 billion accrued at year-end 2025 for expected customer compensation. Reuters also reported in June 2026 that ITA Airways was weighing legal action over Pratt engine problems that had grounded almost 20% of its fleet.

That is the trade.

Hard to replace does not mean easy to operate.

A — Asset-Embedded: 8/10

RTX is deeply embedded in physical systems. Collins components sit across aircraft architecture. Pratt engines sit on wings and inside defense platforms. Raytheon systems sit inside national-defense procurement stacks. The company’s $289 billion backlog is not a click-based demand signal. It is a production, certification, and delivery queue.

N — Non-Discretionary: 8/10

Aircraft maintenance is not optional. Flight safety is not optional. Defense readiness is not optional. Commercial OEM demand can cycle, and airlines can delay orders, but engines still need inspections, parts, overhauls, and support. Governments can shift budgets, but missile defense, propulsion, sensors, and sustainment are not “nice to have” categories in a hot geopolitical tape.

C — Capital-Intensive: 9/10

RTX is heavy. In 2025, it spent $2.627 billion on capital expenditures and reported $2.807 billion of company-funded R&D plus $4.886 billion of customer-funded R&D. That is not app spend. That is factories, test cells, engineering, tooling, program development, and production capacity.

H — Hard to Replace: 8/10

Replacing RTX means re-certifying systems, qualifying suppliers, reworking programs, rebuilding trust, and accepting years of execution risk. The defense side is even harder: approvals, budgets, classified work, interoperability, production slots, and government contracting rules. The weakness is that government contracts carry funding, timing, termination, and cost-control risk. Hard to replace is not the same as invincible.

O — Obsolescence-Resistant: 7/10

AI does not obsolete engines, aircraft systems, missile defense, or MRO. It improves the workflow around them. Pratt is already integrating AI-assisted borescope software to improve inspection consistency and efficiency across commercial and military engine MRO. That is a tool, not a replacement.

The risk is battlefield change. Drones, autonomy, directed energy, hypersonics, and software-defined warfare can shift budgets and product mix. RTX must keep earning the right to stay inside the stack.

R — Real-World Demand: 9/10

The demand is real. Planes fly. Engines wear. Missiles get consumed. Governments replenish. Airlines maintain fleets. RTX’s Q2 2026 defense bookings were about $23 billion, with Raytheon recording $20 billion in the quarter and $27 billion in the first half. Raytheon’s Q2 bookings included Patriot GEM-T interceptors for Ukraine, AMRAAM, AIM-9X, ESSM, NASAMS, SM-3, and classified contracts.

The GTF issue drags longer, costs more, and keeps damaging airline trust.

Defense budgets shift, appropriations stall, programs get delayed, or contract awards arrive later than expected.

Execution breaks. RTX has thousands of contracts with estimates at completion. Small errors across complex programs can become real margin pressure.

Supply chains stay tight. The market wants more engines, more MRO, more missiles, and more air-defense systems. Wanting capacity is not the same as having it.

Pratt remains too exposed to Airbus and A320neo-family engine dynamics.

Legacy Raytheon legal and compliance matters stay in the background. RTX disclosed deferred-prosecution and settlement obligations tied to legacy Raytheon matters, including DOJ and SEC-related monitoring obligations.

If I’m right:

RTX keeps converting backlog into revenue, Pratt’s GTF pain becomes more manageable, commercial aftermarket stays strong, and Raytheon benefits from sustained air-defense and missile demand.

If I’m wrong:

The GTF issue is not contained, defense bookings pull forward demand, margins get eaten by execution, and the company looks more like a stressed industrial contractor than a durable aerospace-defense compounder.

What would change my mind:

A material deterioration in backlog quality.

A major loss of trust with airlines on Pratt engines.

Evidence that missile-defense demand is peaking instead of normalizing higher.

Persistent negative EAC adjustments.

Or proof that RTX cannot expand capacity without sacrificing margin and quality.

AI Tailwind

AI helps RTX where it should: inspection, quality control, design support, predictive maintenance, supply-chain planning, and engineering speed. Pratt’s AI-assisted borescope push is exactly the right kind of AI — embedded inside a physical workflow with real consequences.

But AI is not the moat.

The moat is certified hardware in unforgiving environments.

AI can speed up the inspection. It still can’t build the engine, certify the blade, or reload the Patriot battery.

— Connor
Alpha Before It Prints

🔗 Click to Connect with me

© 2026 Alpha Before It Prints
Unsubscribe