Military Stocks: Five Companies, Five Different Tests

Aria Vantage | Research dated September 9, 2026

The investment case for defense companies has become a test of execution. Orders can fill factories for years, but shareholders receive the cash left after production costs, investment, financing and mistakes. A large backlog establishes demand visibility; it does not establish the return investors will earn at today's share price.

Lockheed Martin, Northrop Grumman, RTX, L3Harris and Palantir illustrate the distinction. They share exposure to government procurement, yet their economics differ considerably. Aircraft and interceptors involve long manufacturing schedules. Propulsion and communications suppliers depend on specific production bottlenecks. Commercial aerospace brings an additional cycle. Software offers faster expansion and different competitive risks.

Our research view is that the operating evidence supports continued attention to all five. The valuation evidence demands selectivity: Lockheed and Northrop require less ambitious earnings assumptions than RTX, while Palantir requires a substantially longer growth runway. L3Harris adds a separate question about how its missile business will be financed and owned.

From spending commitments to company cash

NATO's June 2025 Hague declaration established a 2035 commitment comprising at least 3.5% of GDP for core defense and up to 1.5% for related infrastructure, resilience and security. Those categories matter: the entire headline percentage cannot be treated as a weapons procurement budget. The July 2026 Ankara summit continued the implementation agenda, including industrial production and new capabilities. Hague declaration; NATO's July 2026 summit account.

NATO's July 15 assessment identified Russia as a continuing security threat and also discussed China's military buildup, North Korea and Middle East developments. These are the alliance's stated assessments, rather than an earnings forecast for any contractor. NATO, July 15, 2026.

For investors, the relevant sequence is an appropriation, a funded award, production, customer acceptance and collection. Each stage can create delays. Contract ceilings, framework agreements and recognized backlog also describe different commitments. Adding them together can double-count the same demand.

Lockheed's June 24 THAAD announcement provides a useful example: an award for up to $35 billion over seven years, structured as an undefinitized contract action, followed January's production framework. That is substantial company-specific evidence, but neither a single year's sales nor a guarantee that every dollar produces the same profit. Lockheed's THAAD award announcement.

Lockheed Martin: convert the backlog, protect profitability

Lockheed reported second-quarter revenue of $20.1 billion, up 11%, EPS of $7.94 and free cash flow of $2.9 billion. Backlog reached approximately $230 billion, including the multiyear THAAD contract. The earnings comparison requires care: the year-earlier quarter included $1.6 billion of program losses and other charges. The rebound should not be extrapolated as a recurring growth rate. July 23 results.

Management's updated 2026 outlook calls for $79.75–$81.75 billion of sales, $29.95–$30.65 of diluted EPS and $7.0–$7.2 billion of free cash flow. However, the cash improvement also accompanies a reduction in expected capital expenditure to $2.0–$2.4 billion. Investors should distinguish better operating cash generation from investment moving between periods. July earnings presentation, slide 4.

Aircraft remain material: Aeronautics generated $8.1 billion of the quarter's sales, compared with $4.1 billion in Missiles and Fire Control. Lockheed therefore cannot be evaluated solely through interceptor orders. Its filing also describes revenue recognition based on contract progress and estimated costs, making revisions to those estimates an essential earnings-quality check. Second-quarter Form 10-Q.

The thesis strengthens if deliveries and cash rise while costly program revisions remain contained. It weakens if the production ramp consumes working capital faster than customers pay, or if another major loss provision offsets the benefit of higher volume.

Northrop Grumman: strategic programs, uneven profitability

Northrop's second-quarter sales rose 5% to $10.9 billion, and net awards of $20 billion lifted backlog to $104.7 billion. B-21 and other restricted programs supported Aeronautics growth; Sentinel contributed to Defense Systems demand. Yet segment operating profit as a percentage of sales declined to 10.6% from 11.8%. The quarter included unfavorable adjustments of $68 million on SiAW and $91 million on GEM 63XL. Higher revenue did not translate uniformly into higher profitability. July 21 results, segment discussion.

An August development broadens the case. Northrop announced framework agreements exceeding $3 billion for PAC-3 MSE propulsion and THAAD components. Its participation alongside Lockheed demonstrates a portfolio overlap: owning both can provide exposure to different parts of the same production chain. It does not create two independent sources of end demand. August 3 announcement.

Northrop's 2026 outlook is $43.75–$44.25 billion of sales, $28.60–$29.10 of mark-to-market-adjusted EPS and $3.1–$3.5 billion of adjusted free cash flow. First-half adjusted free cash flow was negative $845 million, making the second-half collection schedule consequential. July results, guidance and cash flow.

The investment question is whether large programs become dependable cash generators. A growing backlog accompanied by repeated cost revisions would weaken that conclusion. Improved profitability and delivery milestones would provide more useful confirmation than another large headline award alone.

RTX: defense plus the aircraft maintenance economy

RTX's second-quarter sales reached $24.7 billion, up 14%, with adjusted EPS of $1.89 versus GAAP EPS of $1.57. Its $289 billion backlog included $170 billion of commercial business and $119 billion of defense. Commercial aftermarket sales rose 10% at Collins Aerospace and 25% at Pratt & Whitney. Raytheon's sales increased 18%. Management raised annual adjusted sales guidance to $95–$96 billion, adjusted EPS to $7.10–$7.25 and free cash flow to $8.50–$8.75 billion. July 23 results.

That combination offers several routes to growth, but exposes the stock to airline activity, engine maintenance and aircraft production as well as military procurement. Strong aftermarket demand is economically different from the early delivery of a new engine, where product mix and initial manufacturing costs can weigh on profitability.

The cash statement warrants close reading. RTX's filing says increased factoring activity contributed $1.5 billion to the year-over-year improvement in first-half operating cash flow. It also estimates a roughly $700 million 2026 cash impact from the Pratt & Whitney powder-metal matter. Selling receivables accelerates collections; it does not establish a permanently higher cash return. Second-quarter Form 10-Q, liquidity discussion.

For the stock to justify its premium, growth should continue alongside improving production efficiency and durable cash conversion. A weaker commercial cycle or renewed engine remediation costs could pressure both earnings expectations and the multiple investors pay.

L3Harris: attractive demand, a complicated ownership transition

L3Harris reported $5.9 billion of second-quarter revenue, up 8%, a $42 billion backlog and $771 million of free cash flow. Its annual outlook is $23.2–$23.7 billion of revenue, $11.80–$12.00 of GAAP EPS and $3 billion of free cash flow. Communications & Spectrum Dominance produced segment profit equal to 26.9% of sales; the corresponding figure for Missile Solutions was 12.3%. These are distinct businesses with different economics. July 29 results.

The proposed missile-business IPO requires an updated timeline. On the July call, management said it expected to revisit the IPO in mid-2027, superseding the earlier second-half-2026 expectation. Revisit is not a guaranteed offering date. Company-hosted July call transcript, pages 4–5.

The government's $1 billion investment is a convertible preferred security with warrants. The filing describes cumulative dividends, conversion conditions and a redemption mechanism. It is financing with claims on future value, rather than revenue or an unrestricted grant. Valuing the subsidiary requires accounting for those claims and the interest L3Harris ultimately retains. April investment announcement; July Form 10-Q, Note B.

There is also a leadership transition. Sam Mehta became CEO on August 17. The company said the predecessor's conduct issue was unrelated to financial reporting, controls, customer relationships or operations, and reaffirmed its consolidated 2026 guidance. Those are company statements; subsequent performance will provide the operating test. August 17 company announcement.

Palantir: software growth with a much higher valuation hurdle

Palantir belongs in this discussion because of government software demand, but government revenue includes more than defense. Government customers represented 52% of first-half revenue and commercial customers 48%. Second-quarter U.S. government revenue was $809 million; U.S. commercial revenue was $764 million. The company combines public-sector exposure with a substantial commercial software business. Second-quarter Form 10-Q.

The latest results demonstrate rapid expansion: quarterly revenue was $1.935 billion, up 93%, GAAP operating profit as a percentage of revenue was 47% and operating cash flow was $1.216 billion. Management projects $8.150–$8.158 billion of 2026 revenue and $4.5–$4.7 billion of adjusted free cash flow. These are management forecasts, not assured outcomes. August results and outlook.

At the observed $170.09 share price, approximately 2.403 billion reported shares imply $408.7 billion of equity value. Subtracting $9.409 billion of cash and marketable securities, with no reported debt, gives roughly $399.3 billion before lease adjustments: about 49 times 2026 revenue guidance. Using the quarter's diluted average share count instead produces approximately 52 times. This is illustrative valuation arithmetic, not a precise current fully diluted capitalization. Share count, liquidity and dilution inputs.

That valuation can fall even while the business grows. Hypothetically, 40% revenue growth combined with a decline from 49 to 30 times sales reduces enterprise value about 14%. Strong adoption must therefore persist for years. Customer expansion, GAAP profitability, collection quality and dilution deserve as much attention as contract announcements.

What today's prices require

The following calculations use September 9 intraday observations, approximately 3:20 p.m. Eastern for the four manufacturers. They divide price by the midpoint of each company's cited 2026 EPS guidance; they are not consensus multiples or target prices.

Accounting differences limit direct comparison. A lower multiple is useful only if the earnings supporting it are durable. For any of these manufacturers, an illustrative 10% EPS decline combined with a 15% multiple contraction would reduce share value by 23.5%, before dividends. Conversely, 10% EPS growth with an unchanged multiple would produce approximately 10% price appreciation. These are sensitivities, not forecasts.

The next evidence window is the third-quarter reporting cycle. Lockheed needs profitable delivery growth; Northrop needs stronger cash conversion and fewer adverse cost revisions; RTX needs sustainable aftermarket and defense profits; L3Harris needs execution under new leadership and clearer subsidiary economics; Palantir needs growth strong enough to support its valuation.

The common investment opportunity is the conversion of demand into cash. The common mistake would be assuming that strategic relevance, a large contract and an attractive stock price are the same finding.

Research note: This article was prepared with AI assistance. It is informational research, not personalized investment advice. Prices are intraday snapshots; financial information reflects the latest cited disclosures available to this research as of September 9, 2026. Valuation comparisons use company guidance on the accounting bases shown and supply no price targets.