General Dynamics Corporation (NYSE: GD) — Every Test Passed, and the Multiple Took the Winnings

Independent equity research note. Prepared as of 2026-08-15. As-of price $395.78 (close 2026-08-14). An update to my previous note on this company dated 2026-06-14.

⚡ Claude’s Take

This block is the author’s own subjective opinion. It is general information, not investment advice, and it is the single place in this article where a position is taken: the analysis that follows it takes no position, sets no price target, and carries no recommendation — by design.

Verdict: HOLD — and do not chase. The June call was “accumulate below ~$300–310.” That was too low, and the stock never came within 14% of it. I am raising the zone to reflect a genuinely better business and a higher earnings base, not to chase price: accumulate on weakness into roughly $330–360 (≈18–19.5× FY2027E EPS of ~$18.75); hold through roughly $360–415; avoid adding above roughly $415. Not a short — this is the best-executing prime in the group. Conviction: medium.

Since June, General Dynamics has done everything the bull case asked of it and one thing it did not dare ask. Q2 revenue rose 8.1%, operating margin expanded 40bp to 10.4%, EPS rose 13.4% to $4.24, and management raised full-year guidance for the third time this year, to $16.80–$16.90. Backlog hit a record $136.5B, up 32%, with a record in every segment and book-to-bill above 1.0× in all four. Then, on the afternoon of 29 July — after the earnings call, after the quarter closed — the Navy awarded $76.6B of submarine work, $71.6B of it to Electric Boat: $29.5B for five more Columbia-class boats, $42.1B for nine more Virginia-class, plus shipyard infrastructure. That is not another backlog headline. It converts the central uncertainty of the GD thesis — will the Navy actually fund the cadence it keeps describing? — from a forecast into a contract. It is also, mechanically, not in the $136.5B figure, because the quarter had already closed. Two additional facts genuinely improve the June picture: net debt has fallen to $3.2B and the August note maturity is being retired with cash, not refinanced; and my June note’s ROE figure of ~9.7% was simply wrong — the correct TTM return on equity is 17.8%, on 13.2% ROIC. GD is a mid-to-high-teens-ROE business, not a sub-10% one, and that removes a load-bearing plank from June’s bear case.

So why still HOLD? Because the entire reward went to the multiple. Forward P/E moved from 21.8× to 23.5× while the earnings guide rose only 2.1% — the +10.4% move was re-rating, not earnings. On the stock’s own ten-year history the P/E now sits at the 99.94th percentile and P/S at the 99.94th: this is, within a rounding error, the most expensive General Dynamics has ever been on earnings or sales, and the dividend yield of 1.61% is the thinnest the stock has offered in a decade. The factor tape reframes what that risk actually is. FactorsToday loads GD with essentially zero Momentum (−0.01) and essentially zero Quality (+0.01); the live loadings are Aerospace & Defense industry (+0.52), Market (+0.51), DividendYield (+0.49) and Industrials (+0.42), with R² of 0.50. Half this stock’s variance is the defense sector and the dividend-yield factor. That means the dominant risk to the price is not GD failing — it is the sector de-rating around a company that keeps performing, and a dividend-buyer base being priced out at a 1.6% yield. Underneath, the profit lever that actually matters still has not moved: Marine margin was 7.28% in Q1 and 7.34% in Q2 — flat sequentially, +40bp year-over-year, still guided to only 7.4% for the year — while Combat revenue was flat and Technologies decelerated to +4.1% with margin down 20bp. And the people who know most sold into it: across May–August, the CEO exercised 166,660 options at $165.47 and disposed of 100% of them at $379–384, with the EVP, a VP and two directors doing the same to net-zero retention. Zero open-market purchases by anyone. Framing: a quality compounder that has now been fully, fairly, and perhaps generously paid for — the risk has migrated from the business to the multiple. Tag: the Navy removed the doubt, and the market took the change.

Conviction: medium. The single fact that would flip me more bullish: Marine operating margin clearing 8% on the enlarged post-award revenue base while capex stays inside 4% of sales — that is the moment the submarine franchise stops consuming the growth and starts compounding it. The single fact that would flip me bearish: a de-rating of the defense complex (the factor read says this is the live risk, not execution), or a Marine reach-forward charge / a 2027 Aerospace air-pocket from the disclosed G280-to-G300 production gap landing while the stock carries a 99th-percentile multiple.

🔄 Changes Since the 2026-06-14 Report

Nine weeks, one reported quarter, and one very large contract. The thesis architecture is unchanged; three of its inputs are materially different.

Adjudicating the prior note’s falsification tests.

My June note’s bear test was explicit: “GD posts a mid-year guidance raise, Marine margin expands sequentially, and backlog/book-to-bill stay above 1.3× — if all three, the bear de-rating thesis is broken.” All three fired. The guidance raise came on 29 July. Backlog hit a record and book-to-bill was 1.4×. Marine margin did expand sequentially — by six basis points, 7.28% to 7.34%, which satisfies the letter of the test while barely honouring its spirit. The bear de-rating thesis, as written, was falsified. The narrower and more useful conclusion is that the test was designed to ask whether the business would validate the premium, and it did; the test was silent on how much further the multiple could travel, which is the thing that actually happened.

The June bull test — “Marine margin fails to exceed ~7.5% and cadence stays below target through FY2027 while Gulfstream book-to-bill drops below 1.0× for two-plus quarters” — was not triggered, and is not close. Gulfstream book-to-bill is 1.5× in the quarter and 1.3× trailing-twelve-months, the inverse of the failure condition. The Marine leg is still live: margin remains below 7.5% and is guided to only 7.4%. Management also clarified on the call that the two-Virginia-plus-one-Columbia cadence is an early-2030s objective — which means that leg of the June test cannot be settled for several years and should be re-specified as a question of directional progress, not a binary.

The honest scorecard: the business tracked the bull case and the stock tracked it faster. June’s HOLD was right about quality and wrong about near-term direction; its accumulation zone was set too low to be actionable. This note raises the zone on the strength of a higher earnings base and a corrected return profile — and simultaneously flags that the valuation risk is now larger, not smaller, than it was in June.

📈 Stock Price Action — Five-Year Event Map

General Dynamics has made a straight-line, low-drama round trip from wartime re-rating to record. Five years ago the stock traded at $180; it bottomed at $171.17 on 30 November 2021, and closed at $395.78 on 14 August 2026 — a +131% five-year move — after setting an all-time high of $395.97 on 10 August 2026. The stock sits 0.0% below its five-year and all-time high (FactorsToday rs_peak = −0.05%), inside a 52-week range of $308.14–$395.97. Notably, the deepest drawdown of the past twelve months was only −14.5%, and of the past three years only −22.5%: this has been an unusually orderly ascent.

Cycle narrative. (1) The stock’s modern re-rating begins with the invasion of Ukraine — GD rose 4.3% on 25 February 2022 alone, and the whole complex repriced on the recognition that European rearmament was structural, not episodic. (2–3) The 2022 drawdown was market-wide, not company-specific, and reversed sharply in October on a strong Q3 print. (4) 2023’s decline was genuinely idiosyncratic: Gulfstream’s G700 certification kept slipping, holding back the single highest-margin growth driver in the portfolio. (5) The recovery began violently — the stock rose 8.4% on 9 October 2023, the first trading day after the Hamas attack on Israel — and then compounded through 2024 as the G700 finally certified and entered service and munitions demand accelerated. (6) The one clean company-specific scare of the period was the post-election federal-efficiency drive: GD fell 6.9% on 14 November 2024 on fears that GDIT’s federal-IT contracts would be cut, and the stock did not recover its September 2024 level until mid-2025. (7) From the February 2025 low the ascent has been almost uninterrupted, punctuated by an 8.0% single-day gain on the Q1 2026 print (29 April 2026) and a 5.2% gain on 11 June 2026. (8) The final leg is the subject of this update — and contains a detail worth keeping: on Q2 earnings day itself (29 July) the stock fell 3.1%, from $393.19 to $380.96, despite a beat and a guidance raise, having already run 8.7% during July. It then made a new all-time high within eight sessions as the $76.6B Navy award — announced that afternoon, after the call — was absorbed. A stock that sells off on a beat-and-raise and then rallies on the contract is a stock where expectations are running slightly ahead of prints.

Price moves are Fact; the attributed causes are Interpretation. This section carries no recommendation and no price target.

1. Executive Summary

General Dynamics is a four-segment aerospace and defense group headquartered in Reston, Virginia, running at approximately $55.7B of revenue and $16.85 of EPS on FY2026 guidance, with a market capitalisation of roughly $108B and an enterprise value of roughly $111.5B. Roughly a quarter of revenue and a third of operating profit come from a civilian product — Gulfstream business jets — making GD a hybrid of a defense prime and a premium capital-goods OEM. The remaining three segments are defense franchises of unequal quality: Marine Systems (nuclear submarines and surface ships, ~33% of revenue), Technologies (federal IT plus defense electronics, ~26%), and Combat Systems (armoured vehicles and munitions, ~17%).

The nine weeks since my June note have resolved most of the operational questions that note left open, and sharpened the one it did not. Operationally, Q2 2026 was the second consecutive statement quarter: revenue $14.09B (+8.1%), operating margin 10.4% (+40bp), EPS $4.24 (+13.4%), operating cash flow $1.88B at 162% of net earnings, and a full-year guidance raise to $16.80–$16.90 — the third of the year. Backlog reached a record $136.5B, up 32%, with a record in every segment and book-to-bill above 1.0× in all four. Then, on 29 July, the U.S. Navy awarded $76.6B of submarine work, of which $71.6B went to Electric Boat for five additional Columbia-class and nine additional Virginia-class boats plus shipyard infrastructure. Because the award landed after the 5 July quarter-end, none of it is in the $136.5B figure; Q3 backlog will step up very materially, though GD has not yet disclosed how much of the award is incremental versus a conversion of previously-reported potential contract value.

Two corrections to the June analysis matter. First, return on equity is 17.8% TTM, not the ~9.7% printed in June — an error worth stating plainly, because it was used there to support a “structurally low return profile” characterisation that the corrected figure does not support. On 13.2% ROIC and 17.8% ROE, GD is a solidly good-returns business whose margins are monopsony-capped but whose equity returns are not poor. Second, the factor read has been revised: GD carries essentially no Momentum and no Quality style loading; it is a defense-industry-beta, dividend-yield, low-volatility instrument with an R² of 0.50 against the factor complex. That relocates the principal risk to the price from company execution to sector multiple compression.

What has not improved is the profit lever the whole Marine story depends on. Marine operating margin was 7.28% in Q1 and 7.34% in Q2 — flat sequentially, up 40bp year-over-year, and guided to only 7.4% for the full year, against revenue guided to ~$18B. Combat revenue was flat (+0.3%) with earnings down 1.9%, carried entirely by a 2.1× book-to-bill that will show up later. Technologies decelerated to +4.1% with margin down 20bp. And management disclosed a new, specific 2027 issue: a planned production gap between the last G280 (approximately Q2 2027) and the G300’s entry into service (late 2027 or early 2028).

The valuation is where the tension now sits. At $395.78 the stock trades at 23.5× the FY2026 guidance midpoint, 24.9× trailing earnings, ~16.9× EV/EBITDA, 4.16× book, and a 1.61% dividend yield, on a 99.94th-percentile P/E and 99.94th-percentile P/S against its own ten-year history. Since June the forward multiple expanded from 21.8× to 23.5× while the earnings guide rose 2.1% — the move was re-rating, not earnings. Against the rest of the large-prime peer group GD is now more expensive than Northrop (~20.3× forward) and roughly nine turns of core earnings above Huntington Ingalls, its own submarine co-builder. Insiders monetised into it: across May–August the CEO exercised 166,660 options at $165.47 and disposed of 100% of them at $379–384, with an EVP, a VP and two directors doing the same to net-zero retention, and not a single open-market purchase anywhere in the group. The analysis below takes no position and sets no price target; the valuation section frames what the price requires.

2. Business Overview

The segment architecture is unchanged from the June report; the mix within it has shifted meaningfully in six months. Below are the first-half 2026 actuals and the guided full-year shape given on the 29 July call.

Sources: GD Q2 2026 10-Q (filed 2026-07-29); Q2 2026 earnings call, 2026-07-29. Segment H1 figures derived from 10-Q segment tables; totals include corporate.

Marine Systems is now unambiguously the largest segment and, after 29 July, the most contractually secure. Electric Boat is prime and lead design yard on every U.S. Navy submarine programme — the twelve-boat Columbia-class ballistic-missile programme and the Virginia-class attack-submarine programme. Bath Iron Works is one of two builders of the Arleigh Burke (DDG-51) destroyer; NASSCO builds the John Lewis-class fleet oiler and Jones Act commercial ships. Q2 revenue grew 10.4% on, per the 10-Q, “increased material and labor volume on Columbia-class submarine construction and higher throughput on the John Lewis-class (T-AO-205) fleet replenishment oiler.” Two operating details from the call are worth carrying forward: Columbia hours earned rose 37% in the first half versus the prior year, and sequence-critical material receipts rose 65% year-over-year in Q2 — the two leading indicators of throughput. Bath delivered its most recent DDG-51 nearly three months ahead of plan. Segment backlog is $65.2B, before the July award.

Aerospace (Gulfstream plus Jet Aviation) delivered the quarter’s standout financial result: revenue +15.1% and operating earnings +26.6% on a 130bp margin expansion to 14.5%, with 41 aircraft delivered (three more than the year-ago quarter and three more than Q1). Management attributed the improvement not to a one-off but to “durable productivity improvements on all new aircraft types” and better service performance at both Gulfstream and Jet Aviation. The order book improved sharply: book-to-bill of 1.5× on 16 more aircraft orders than the year-ago quarter and 20 more than Q1 — the strongest first half for Aerospace orders since 2022. Backlog is $23.98B. The product line is mid-transition: G700 in service, G800 delivering, G280 ending, G300 and G400 coming.

Combat Systems was the quarter’s weak financial performer and its strongest order performer. Revenue was flat (+0.3%) and operating earnings fell 1.9% as Land Systems declined — the U.S. Army is in a platform-transition trough between current Abrams/Stryker volumes and the next-generation M1E3 and XM30 — while Ordnance & Tactical Systems (munitions) and European Land Systems grew. Against that, book-to-bill was 2.1×, including production of armoured combat support vehicles for the Canadian Armed Forces, taking segment backlog to $29.35B. Novakovic guided “double-digit growth continuing at European Land Systems… [and] OTS… ought to drive high-single-digit growth for the group going forward.”

Technologies (GDIT federal IT plus Mission Systems defense electronics) grew 4.1% with margin down 20bp to 9.4% — still the portfolio’s laggard, though the internal mix is improving. Mission Systems led, with its international portfolio up more than 35% since 2024. GDIT continues to face what management calls “elongated procurement cycles,” but has adapted: it won more Other Transaction Authority awards in the first half of 2026 than in all of 2025, and carries a qualified pipeline management sizes at “$120-plus billion.” Post-quarter, GDIT won a $1.3B Army National Guard enterprise network and cybersecurity contract (announced 7 August).

Business model and customer mix. GD earns on long-cycle contracts — a mix of cost-reimbursement and fixed-price government work — plus business-jet sales recognised largely at delivery and a growing high-margin services stream. The defining structural feature is that roughly two-thirds of revenue comes from the U.S. government, which caps margins by design while making the revenue extraordinarily visible: $136.5B of backlog and $186.9B of total estimated contract value against $55.7B of annual revenue is roughly 2.5 years of contracted work and 3.4 years including options.

3. Industry Dynamics

The industry framing established in the June note holds and does not need re-deriving; what has changed is the resolution of two specific uncertainties that note flagged, and one that has emerged since.

Defense: the monopsony is now paying, but the appropriation is smaller than the request. GD’s three defense segments sit inside the textbook high-barrier, low-mean-reversion structure that Marathon’s Capital Returns describes: a single dominant buyer facing a handful of qualified primes, with entry blocked by clearances, programme-specific IP, certified industrial base and political necessity. The flip side is that the same monopsony caps the returns — cost-based and negotiated fixed-price contracting with disclosed cost data holds structural margins near 10–11% and ROIC in the low teens across the entire group. This is why GD, Lockheed, Northrop and Huntington Ingalls all cluster in a narrow band of operating margin and returns despite very different portfolios: the ceiling is a feature of the customer, not of the companies.

My June note framed the demand backdrop around a “~$1.5T FY2027 request,” with the explicit caveat that a request is not law. That caveat has now partially resolved, and not in the bulls’ favour. Congressional appropriations reporting now puts the FY2027 request as having resolved into a ~$1.1T congressional base mark, up ~10%. A 10% increase on a $1T base is a genuinely strong budget by any historical standard — but it is not the 44% headline step-up that anchored sector sentiment earlier in the year. The demand environment is excellent and slightly less excellent than the market was told in the spring. For GD specifically the damage is limited, because its exposure is concentrated in the line items with the strongest bipartisan protection — submarines, munitions, and shipbuilding — and because the 29 July award removes the appropriation question for the single largest one. Novakovic’s own framing on the call was careful: “a lot of our programs are funded in the base budget. But the reconciliation is important, because… weapons production has been on very low-rate production… for quite some time.”

A structural shift worth naming: the customer is pushing toward fixed price. Asked directly about the administration’s stated preference for fixed-price contracting, Novakovic said GD has “always encouraged fixed-price contracting when it’s appropriate and the customer is interested. So we see additional interest in fixed price, which we welcome.” This is an industry-level risk transfer that deserves more attention than it gets. Fixed-price development is precisely what produced Lockheed’s ~$3.6B of 2024–25 charges, Boeing’s $5.3B FY2025 hit and RTX’s ~$3B GTF charge. GD is currently the one large prime with no reach-forward-loss history this cycle; a structural migration toward fixed price is the mechanism by which that distinction could erode. It has not eroded, and GD’s contract mix remains roughly balanced, but the direction of industry travel raises the tail risk for everyone.

Business aviation: still a supply-constrained oligopoly, now with a visible 2027 hole in GD’s own line. Gulfstream competes with Bombardier, Dassault and Textron in a consolidated large-cabin market with high barriers (certification, brand, global service networks, multi-year backlogs) and genuine cyclicality. My June note’s central worry — that Middle East order softness was the leading edge of a downturn — has not materialised. Book-to-bill was 1.5× and the first half was the strongest for orders since 2022, with management describing “very active interest across all models in the U.S. and Asia, with some cautious concern from customers in the Middle East, but they’re still active in the pipeline.” The demand question is, for now, answered. The supply question has a new wrinkle: the disclosed production gap between the last G280 (~Q2 2027) and G300 entry into service (late 2027/early 2028). Novakovic asserted this creates no absorption pressure “given the agreement we have with our partner,” but a planned break in a production line is a revenue and fixed-cost-absorption event that the current guidance horizon does not reach.

Verdict: structurally attractive, with the asterisk now more precisely drawn than in June. Entry is blocked, demand is the best in a generation, and GD is levered to the most protected line items. But the returns are capped by the buyer, the FY2027 appropriation is materially smaller than the request that fuelled the spring re-rating, and the industry is drifting toward a contracting structure that transfers risk to the contractor. Good industry; not a great one; and the gap between the demand narrative and the appropriated number is the single most under-discussed sector fact of this reporting season.

4. Competitive Position

GD’s moat is not one moat but four different competitive situations, and the July award changed the economics of the strongest one.

Marine Systems — a protected duopoly that just received demand certainty. There are exactly two qualified nuclear-submarine yards in the United States: GD’s Electric Boat and HII’s Newport News, and they partner on Virginia-class and subcontract on Columbia rather than genuinely compete. GD’s own 10-K describes “one primary competitor with which it also partners.” The barriers here are about as deep as any in industry — nuclear propulsion IP, a workforce that takes years to train, single-source critical suppliers, and a national commitment the Navy structurally cannot abandon. In Greenwald’s taxonomy this is economies of scale combined with customer captivity in close to its purest form.

What my June note could not say, and this one can, is that the demand side of that moat is now contracted rather than forecast. The 29 July award — $29.5B for five Columbia boats and $42.1B for nine Virginia boats — was characterised by Electric Boat’s president in precisely the terms that matter economically: it “provide[s] Electric Boat and our suppliers with the demand certainty we need to continue investing in capacity.” That sentence is the whole point. The binding constraint on this franchise has never been competition; it has been GD’s own throughput, and throughput requires capital and supplier commitment that neither GD nor its single-source suppliers will fund against an uncontracted forecast. A multi-boat, multi-decade contract converts a capacity-investment question into a capacity-investment plan. This is the most consequential competitive development at GD in several years.

The caveat is equally important: a contract is not a margin. Marine’s operating margin is 7.3%, the lowest in the portfolio, and the segment is simultaneously absorbing the heaviest capital investment in the company. A bigger contracted book at 7.4% margins raises absolute profit and does not, by itself, prove the franchise can earn a normal return.

Combat Systems — sole-source franchises in a platform trough, with the order book already turning. Land Systems remains the sole-source producer of the Abrams main battle tank and the Stryker for the U.S. Army, with development positions on the next-generation M1E3 and XM30. OTS is the world’s largest producer of Ball Powder propellant and a leader in large- and medium-calibre ammunition. The current-quarter financials look weak — flat revenue, earnings down 1.9% — precisely because the U.S. vehicle programmes are in a transition trough. The 2.1× book-to-bill and $29.35B backlog say the trough is a timing phenomenon rather than a franchise problem, and the growth is coming from allies rather than the U.S. Army: European Land Systems and the Canadian ACSV award. The moat is real; the current platform cycle is not helping it.

Aerospace — a brand-and-scale moat, freshly validated on the supply side. Gulfstream’s moat is brand prestige, an installed base of more than 3,000 aircraft, a global service network and the engineering to hold the ultra-long-range performance crown. This quarter produced an unusually clean piece of evidence for the scale-and-relationship component. Asked why Gulfstream’s supply chain is “humming” while peers struggle, Deep answered that “Gulfstream has a very clear relationship that has really given the supply chain visibility into our production plans for whatever period is appropriate. And so they are able to keep up now.” Supply-chain execution as a relative advantage in an industry where Boeing, Airbus and Bombardier have all been supply-constrained is a genuine, financially-visible edge — it showed up as three extra deliveries and 130bp of margin.

Novakovic also pushed back sharply on the framing that Gulfstream is a luxury business, and the pushback is analytically relevant rather than defensive: “I’m going to quarrel with you on one word, luxury… for all of our Fortune 500, Fortune 100 companies… they are business tools… These are not luxury yachts that sail around the Mediterranean.” If she is right, the demand function is corporate-capex-like rather than discretionary-wealth-like, which implies a different (and probably shallower) cyclical drawdown than the luxury analogy suggests. It also implies the 20%+ margins a luxury comparison would promise are not available — she declined to endorse them.

Technologies — still the weakest moat, but the argument has improved at the margin. Federal IT is structurally competitive: Leidos, Booz Allen, SAIC, CACI, Peraton and Accenture Federal contest the same recompetes, and switching costs are low. Mission Systems is genuinely better — encryption, space ground systems, submarine fire control and undersea vehicles are differentiated and sticky — and it is now the segment’s growth engine, with international up more than 35% since 2024. The GDIT datapoint that matters is the OTA win rate: more OTA awards in the first half of 2026 than in all of 2025 suggests GDIT is adapting to the procurement regime rather than being displaced by it. This is a scale-and-relationships business, not a moat, and the $21B of goodwill sitting on it from the 2018 CSRA acquisition remains the balance sheet’s one soft spot.

The clean-books advantage, restated and slightly qualified. GD remains the one large prime with no reach-forward losses this cycle, against Lockheed’s ~$3.6B, Boeing’s $5.3B, Northrop’s B-21 charges and RTX’s ~$3B GTF charge. It is also free of the distortion that makes Huntington Ingalls hard to value: roughly $3.7 per share of HII’s reported EPS is non-cash pension income, so that its apparent 18.6× P/E is nearer 25× on core operating earnings. GD carries no equivalent adjustment — its EPS is what it appears to be. The qualification is that this advantage is partly a contract-mix advantage, and the industry is drifting toward the contract type that created everyone else’s charges.

Verdict: durable and, in Marine, materially strengthened. The submarine duopoly is the jewel and has just been handed the demand certainty that unlocks capacity investment; Combat’s sole-source positions are intact through a platform trough that the order book says is temporary; Gulfstream has demonstrated a real supply-chain edge; Technologies remains the drag. The composite is “good, and better than it was in June” — with the honest qualifier that a strengthened franchise at 7.3% margins is a strengthened franchise that still has to prove it can earn.

5. Growth History and Forward Opportunities

The recent record. GD compounded revenue from $37.9B (FY2020) to $52.6B (FY2025), and FY2026 is guided to ~$55.7B — a five-year CAGR of roughly 8%, accelerating rather than decaying: +7.2% (FY2023), +12.9% (FY2024), +10.1% (FY2025), and +9.1% in the first half of 2026. Diluted EPS ran $11.00 (FY2020) → $15.46 (FY2025) → guided $16.85 (FY2026), a ~9% CAGR that has likewise accelerated to +13% year-over-year in the most recent quarter. The growth is overwhelmingly organic — there has been no material acquisition since CSRA in 2018.

The first-half 2026 composition tells you where the growth is actually coming from, and it is not evenly distributed:

Two-thirds of GD’s revenue growth and essentially all of its incremental operating profit came from Marine and Aerospace. That is a narrower base of growth than the “four franchises inflecting simultaneously” framing my June note used, and it is worth being precise about: two segments are inflecting, one is in a funded trough, and one is grinding.

Forward opportunities, re-ranked after the award.

1. Submarine throughput — now contracted, still the binding constraint. The $71.6B Electric Boat award covers fourteen additional boats. The operating indicators are moving in the right direction: Columbia earned hours +37% in the first half, sequence-critical material +65% year-over-year. But management was explicit about the timeline: the two-Virginia-plus-one-Columbia cadence is targeted for “the early 2030s time frame,” and Deep said “we’re on that path, and we’re actually where we expect to be at this point in the process.” This is the single most important recalibration in this note relative to June. The cadence is not a 2026–27 event; it is a decade-long ramp. Marine will grow, and the margin will improve slowly — Deep’s own phrase was “a slow, steady drumbeat.”
2. Gulfstream — demand answered, with a 2027 hole to navigate. Book-to-bill of 1.5× and the best order half since 2022 resolve the demand question. FY2026 deliveries remain guided at ~160. Beyond that, the G300 and G400 arrive, but the G280 production gap (last delivery ~Q2 2027, G300 entry into service late 2027/early 2028) is a disclosed discontinuity. Novakovic explicitly signalled that GD will “talk more about large cabins next year,” which reads as an acknowledgement that the 2027 shape is not yet settled.
3. Munitions and international vehicles — the reliable compounder. Novakovic’s own forward framing for Combat: “double-digit growth continuing at European Land Systems given the threat environment… OTS because of artillery missile components, 155. And then within Land Systems, this is a transition period… but the double-digit growth at OTS and ELS ought to drive high-single-digit growth for the group going forward.” A 2.1× book-to-bill supports it.
4. Mission Systems international — the quiet accelerator. Up more than 35% since 2024 and specifically called out as “a key driver of growth for the year and beyond.” This is the highest-quality growth inside the weakest segment.
5. Missile supply-chain content — real but secondary. Asked about second-source solid-rocket-motor opportunities, Deep was careful to place GD accurately: “we are a subcontractor, both at OTS and Mission Systems, to a number of the missile primes… So we’re, again, a part of the supply chain, but certainly not a prime there.” Investors extrapolating the missile-replenishment theme onto GD should size it as content, not franchise.

Quality of growth. This remains high-quality growth: organic, backlog-backed, and now — in Marine — contractually secured for more than a decade. The $136.5B backlog before the July award, against $55.7B of revenue, is visibility very few industrials can match. The honest qualifications are that the growth base is narrower than it appears (Marine and Aerospace are carrying it), the highest-value cadence milestone is an early-2030s event rather than a near-term one, and Aerospace faces a known 2027 production discontinuity.

Verdict: high-quality, accelerating, better-secured growth than in June — but with a narrower base and a longer runway to the prize than the market’s enthusiasm implies. The submarine ramp is a decade-long compounding story, not a two-year re-rating catalyst. It is being priced closer to the latter.

6. Financial Quality

Margins. Consolidated operating margin was 10.4% in Q2 (+40bp) and 10.4% for the first half, guided to 10.5% for the year — a gentle, genuine improvement from 10.2% in FY2025 and 10.0% in FY2023. The mix explanation is unchanged: 7.3%-margin shipbuilding (a third of revenue) and 9.4%-margin technology services drag against 14.5% Aerospace and 13.9% Combat. The improvement this year is coming almost entirely from Aerospace (+130bp in Q2) with a 40bp assist from Marine. Gross margin is ~15.4% — structurally low for the sector because shipbuilding and services carry heavy cost of sales.

Returns — and a correction. My June note stated ROE of approximately 9.7%. That figure was wrong. Trailing-twelve-month net earnings of roughly $4.49B against $26.8B of shareholders’ equity gives a return on equity of 17.8%, independently corroborated by ROIC.ai at 17.80% TTM and 17.66% at FY2025 year-end. The full corrected return profile:

This matters to the thesis. A ~10% operating margin married to a 17.8% ROE is not a “structurally low return profile”; it is a capital-efficient business operating inside a margin ceiling imposed by its customer. The correct statement is that GD’s margins are monopsony-capped while its equity returns are good and improving — helped by modest leverage, a 17.6% effective tax rate, and asset turns that a services-and-assembly mix supports. My June note’s use of a sub-10% ROE as evidence that GD did not deserve a premium multiple was, on the corrected number, unsupported. Note that ROIC of 13.2% remains the more binding constraint and is the metric management is actually compensated on.

Cash flow — outstanding, and explicitly front-loaded. First-half operating cash flow was $4,035M against $1,450M in the prior-year first half, with Q2 alone at $1,880M or 162% of net earnings. Free cash flow was $1,646M in the quarter (142% conversion) and roughly $3.6B for the half (over 150%). Working capital fell by more than $1B from year-end. This is genuine and it is also, by management’s own account, not the run-rate: Kuryea guided the full-year conversion to “around 105%” and named four specific reasons the second half will be lighter — capital expenditure stepping up sharply (guided 3.5–4.0% of sales for the year against 1.7% in Q2), an approximately $500M discretionary pension contribution to de-risk plans now near full funding, over $500M of back-half cash taxes, and the unwind of advance payments at European Land Systems. Investors extrapolating first-half cash conversion into a run-rate will be wrong, and management said so in advance — which is itself a quality signal.

Balance sheet — the cleanest it has been in years. At 5 July 2026: cash $4,333M, total debt $7,516M, net debt $3,183M — down $1.2B in the quarter and roughly $2.5B from year-end 2025. Net debt to TTM EBITDA is under 0.5×. GD repaid $500M of notes in June and, critically, resolved my June note’s open question about the 2026 note wall: the August maturity will be repaid with cash on hand, and Kuryea stated “we don’t anticipate refinancing these maturities during the year.” The company has not been in the commercial paper market at all in 2026. The consequence is visible in the income statement: net interest expense fell to $49M in Q2 from $88M, and to $118M for the half from $177M, with the full year guided to ~$270M. That interest reduction is worth roughly $0.16 of annual EPS versus the prior-year run-rate — a real, if unglamorous, contributor to the guidance raise.

The one balance-sheet watch item is unchanged: roughly $21B of goodwill from the 2018 CSRA acquisition, sitting in the weakest-moat segment, at about a third of total assets. It has never been impaired and GDIT is cash-generative, but it is the asset most exposed to a structural change in federal IT procurement.

Quality of earnings. Best in the peer group, and the peer contrast is now sharper than in June. GD has no reach-forward charges, no material pension income inflating EPS (unlike HII, where roughly $3.7 per share of reported EPS is non-cash pension income), no commercial-paper dependence, and cash flow that exceeds net income by a wide margin. Stock-based compensation is immaterial at roughly 0.4% of revenue, and the diluted share count is falling (277.7M in Q2 2024 → 273.5M in Q2 2026) even with buybacks restricted to anti-dilution.

Verdict: high-quality, conservative, improving — and materially better characterised than in my June note. Economics improve gently with scale, returns are good rather than exceptional, and the earnings are more trustworthy than any peer’s. The honest constraint is that the improvement is gradual: 30bp of consolidated margin over three years, 26bp of ROIC in two quarters. This is a grinder, not an inflection.

7. Capital Allocation

Capital allocation is the area where GD’s behaviour is most conservative and where the gap between what management could do and what it is doing is widest.

The dividend is the priority and it is well-covered. The quarterly dividend is $1.59 ($6.36 annualised), raised from $1.50 in March 2026, extending an unbroken streak of annual increases spanning three decades. First-half dividends paid were $834M. At a 36.6% payout on TTM earnings and roughly 26% of FY2026E free cash flow, coverage is comfortable and the raise cadence is sustainable. The yield, at 1.61%, is the thinnest it has been in a decade — a direct consequence of the share price, not of policy, and analytically relevant given the stock’s heavy DividendYield factor loading (see Variant Perception).

The buyback remains deliberately, and now conspicuously, restrained. GD repurchased approximately $100M in Q2 and $319M in the first half — explicitly, in Kuryea’s words, “to cover dilution.” Against roughly $4.8B of expected full-year free cash flow, a $3.2B net-debt position and a company that has just been handed $71.6B of contracted work, this is a company choosing not to use its balance sheet. Myles Walton of Wolfe Research put the question directly on the call, framing it as “obvious balance sheet strength, the DoD’s relatively negative view on share repurchase overall” — and asking whether M&A might therefore re-engage. Novakovic declined entirely: “Nice try. Look, this is something we always look at and something we never talk about.”

That exchange is the most informative capital-allocation moment of the quarter, and it cuts two ways. The charitable reading — which my June note adopted — is that restraint is politically prudent while the company is negotiating tens of billions of dollars of sole-source work with a customer that dislikes seeing its contractors repurchase stock, and that the buyback is preserved dry powder. The less charitable reading is that GD is accumulating cash it will not return and will not deploy, at a moment when its own shares trade at the 99.94th percentile of their historical valuation — which is, in fairness, exactly when a disciplined allocator should prefer not to repurchase. On the specific question of whether not buying back stock at an all-time-high multiple is good capital allocation, the answer is yes. The open question is what the cash is for. The 10-Q lists priorities as “investments in our business infrastructure, products and services… a predictable dividend, strategic acquisitions and opportunistic share repurchases primarily to address dilution.”

Capital expenditure is the real deployment, and it is rising fast. Capex was $437M in the first half, up nearly 30% year-over-year, and is guided to 3.5–4.0% of sales for the full year — roughly $1.95–2.25B against $1.16B in FY2025, an increase of 70–95%. Kuryea was explicit: “You should expect the profile of our investment to grow significantly in the back half of the year as we continue to invest, especially in our shipyards, to accelerate production and meet future demand.” This is the correct use of the cash and the correct response to the July award, and it is also the mechanism by which ROIC could stall: a franchise that requires a near-doubling of capital intensity to grow is a franchise where incremental returns need watching. Capex at 4% of sales against a 13.2% ROIC is sustainable; capex at 4.5%+ of sales with flat ROIC — which is precisely what has happened at Northrop, where ROIC fell from ~15% in 2019 to ~11.4% in 2025 — would not be.

M&A: absent, and deliberately so. No material acquisition since CSRA in 2018. In a defense M&A market where Lockheed has just paid $3.5B for Ultra Maritime, GD’s abstention is a positive by omission — it has not paid up in a seller’s market.

Pension: proactive de-risking. The planned ~$500M discretionary contribution to plans “near full funding” in order to “derisk those plans and eliminate significant volatility going forward” is unglamorous, shareholder-friendly, and the opposite of the earnings-management approach some peers take with pension assumptions. It costs cash this year and removes a source of future noise.

Insider behaviour: uniformly negative as a conviction signal, and larger than in June. Across the reviewed Form 4 corpus from 11 May to 13 August 2026 there is not a single open-market purchase (code P) by any officer or director. What there is, in volume, is option exercise paired with immediate disposal at all-time-high prices:

Source: SEC Form 4 filings, CIK 0000040533, raw XML. The 10-Q’s Item 5 confirms no director or officer adopted or terminated a Rule 10b5-1 arrangement during the quarter, and none of these filings carries a 10b5-1 footnote.

This is legally unremarkable and behaviourally routine — long-dated options were exercised and monetised, and executives are not obliged to hold. But as evidence it is one-directional: at a share price of $380–396, with the largest contract award in the company’s history freshly in hand, the eight most informed individuals collectively retained fewer than 10,000 shares of roughly 336,000 exercised, and none of them bought a share on the open market. Novakovic in particular disposed of the entirety of a 166,660-share exercise, leaving her net position unchanged at 766,457 shares. She remains very substantially invested — her stake is worth roughly $300M — so this is monetisation, not exit. It nonetheless says nothing supportive about the price.

Governance note: a control lapse. GD disclosed in the Q2 10-Q, and formalised in an 4 August 424B5, that up to 1,010,925 shares acquired by employees through the 401(k) Stock Fund between 1 July 2025 and 30 June 2026 were sold without an effective registration statement. GD filed a curative Form S-8 on 1 July 2026 and is making a rescission offer expiring 8 September 2026. The company states it does not expect a material impact on financial condition, results or liquidity, and the dollar amounts are indeed small — roughly 0.37% of shares outstanding. It is nonetheless a securities-compliance failure at a company whose governance is otherwise exemplary, and it belongs in the record.

Verdict: conservative, disciplined, well-aligned — and increasingly capital-intensive. Management is not repurchasing stock at a record multiple, is not overpaying for acquisitions in a hot market, is de-risking the pension, is retiring debt rather than refinancing it, and is redirecting the cash into the shipyard capacity that the July award justifies. That is a defensible sequence. The two things to watch are whether the near-doubling of capex holds ROIC flat or better, and what the accumulating cash is ultimately for. Insider behaviour offers nothing to lean on.

8. Changes and Headwinds — Last Two Years

The award that changed the franchise (29 July 2026). The $76.6B Navy submarine announcement, with $71.6B to Electric Boat covering five Columbia-class and nine Virginia-class boats plus shipyard infrastructure, is the most consequential single event in this window. Three analytical points. First, it landed after the 5 July quarter-end, so none of it appears in the $136.5B backlog — Q3’s backlog disclosure will be a step-change. Second, GD described the transaction as “contract modifications,” and the defense segments’ disclosed estimated potential contract value at quarter-end was only $49.2B — less than the award — so a substantial portion must be genuinely incremental to previously-disclosed contract value, but GD has not published the split and it should not be assumed. Third, the sequencing is notable: on the morning of 29 July, Novakovic was still telling analysts only that “we’re told that these contracts will be coming soon”; the award was announced that afternoon. Management was not front-running it.

Three consecutive guidance raises. January’s $16.10–$16.20 became April’s $16.45–$16.55 and July’s $16.80–$16.90. A 4.3% cumulative raise inside seven months on a $55.7B revenue base is a meaningful upgrade in a business with this much contracted visibility, and it was accompanied by segment-level guidance raises in Marine (revenue to ~$18B) and margin discipline elsewhere.

A budget request that resolved lower than advertised. The ~$1.5T FY2027 request that anchored spring sector sentiment has resolved into a ~$1.1T congressional base mark, up ~10%. Strong in absolute terms, materially below the headline the sector re-rated on. GD’s exposure to the most-protected line items limits the direct damage, but this is a sector-wide sentiment headwind that has not yet been reflected in defense multiples.

The Aerospace demand scare did not happen. My June note’s most specific worry — Middle East order softness as the leading edge of a Gulfstream downturn — was not borne out. Book-to-bill of 1.5×, 16 more orders than the year-ago quarter, and the strongest order half since 2022. The Middle East caution persists (“some cautious concern from customers in the Middle East, but they’re still active in the pipeline”) but has been more than offset by the U.S. and Asia.

A new, disclosed 2027 discontinuity. The G280 line ends with a final delivery around Q2 2027 and the successor G300 enters service “late '27, early '28.” Novakovic asserted no absorption pressure “given the agreement we have with our partner,” and signalled GD will “talk more about large cabins next year.” This is a known, planned, and currently un-guided gap in the second-largest profit segment.

Combat’s U.S. trough, and where growth is instead coming from. Land Systems revenue is declining as the Army transitions between current Abrams/Stryker volumes and the M1E3/XM30 generation; the M10 Booker was terminated in 2025. Against that, a 2.1× book-to-bill driven by allies — the Canadian ACSV award, European Land Systems demand, and OTS munitions.

Technologies: adapting, not accelerating. Growth of 4.1% with margin down 20bp. The positives are Mission Systems’ international portfolio (+35% since 2024) and GDIT’s OTA win rate (more in H1 2026 than in all of 2025). The negative is unchanged: “elongated procurement cycles” in federal IT, and no evidence yet that GDIT’s growth rate is inflecting.

Interest expense collapsed. Net interest of $49M in Q2 against $88M a year earlier, with no commercial paper outstanding and higher cash balances generating income. FY2026 net interest guided to ~$270M.

Management transition advanced. Danny Deep, named President in December 2025, was elected to the Board of Directors on 6 August 2026. He now handles the majority of operating commentary on earnings calls. This does not confirm a succession timeline, but the direction is unmistakable.

Verdict: these changes strengthen the thesis on balance, and by a clear margin. The submarine franchise received contracted demand certainty; the Gulfstream demand scare resolved favourably; the balance sheet de-levered; guidance rose three times. The offsetting negatives — a smaller-than-advertised FY2027 appropriation, a known 2027 Aerospace production gap, a still-grinding Technologies segment, and a compliance lapse — are real but second-order. The business is demonstrably better than it was in June. The question this note cannot answer favourably is whether it is 10.4% better in nine weeks.

9. Risk Analysis

The risk that has changed most. In June the top risk was framed as execution — would the submarine ramp and Gulfstream demand hold? Nine weeks later, both are demonstrably holding, and the top risk has migrated to the multiple. This is not a rhetorical flourish; it is what the factor decomposition says. With R² of 0.50 against the factor complex, a +0.52 loading on the Aerospace & Defense industry factor and a +0.49 loading on DividendYield — and effectively zero loading on Momentum or Quality — roughly half of GD’s price variance is explained by the sector and the yield factor rather than by GD. A company can execute perfectly and still lose a third of its market value if the sector multiple normalises. That is now the dominant risk, and it is the one over which management has no control.

The risk that is most under-weighted. Risk 3 — capex-driven ROIC stall — deserves more attention than it receives. GD is about to roughly double its capital intensity to build the capacity the July award justifies. The exact same sequence at Northrop (heavy capex to serve a large contracted programme) coincided with ROIC falling from ~15% to ~11.4% over six years. GD’s ROIC is currently rising (12.97% → 13.23% over two quarters), so there is no evidence of the problem yet — but the mechanism is live and the investment is only beginning.

10. Valuation Discussion (Embedded Expectations)

No price target and no recommendation. This section frames what the current price requires.

Where the multiple sits. At $395.78 with approximately 273.5M diluted shares, diluted market capitalisation is roughly $108.2B and enterprise value roughly $111.5B (adding $7,516M of debt, deducting $4,333M of cash).

The own-history percentiles are the headline, and they are now extreme. On AZI’s ten-year own-history ranks: P/E at the 99.94th percentile, P/S at the 99.94th, P/B at the 85.2nd, composite 95.0. In June these read 94.5 / 95.9 / 78.2 / 89.5. General Dynamics has, within a rounding error, never been more expensive on earnings or on sales. The P/B reading is the only one with room, and it is 85th percentile.

The decomposition of the nine-week move is the analytically important number. The stock rose 10.4%. The FY2026 earnings guide rose 2.1% ($16.50 midpoint → $16.85). Approximately four-fifths of the move was multiple expansion. That is the fact that governs everything below.

Peer context. Every large prime has been reviewed within the past five weeks on comparable terms:

Two observations. First, GD has moved above Northrop on forward earnings (23.5× vs 20.3×), reversing the June ordering in which GD sat between NOC and RTX; only RTX is now more expensive on earnings, and RTX carries a commercial-aerospace growth profile GD does not. On EV/EBITDA, GD at 16.9× is the most expensive of the group. Second, GD now trades at roughly nine turns of core earnings above Huntington Ingalls, its own submarine co-builder and partner on Virginia-class. Some of that gap is deserved — GD has clean books, a diversified portfolio, positive free cash flow and Gulfstream, while HII is a single-industry, sub-WACC, pension-flattered shipbuilder mid-turnaround. Nine turns is a large number for two companies building the same boats.

Embedded expectations — what the price requires. Reverse-engineering 23.5× forward earnings on a business guided to 10.5% operating margin, running 13.2% ROIC and about to roughly double its capital intensity, the market is underwriting four things simultaneously:

1. Sustained high-single-digit revenue growth for several years. Well supported. $136.5B of backlog before the July award, plus $71.6B more, against $55.7B of annual revenue. This is the assumption the market is most clearly right about.
2. Marine margin converting from 7.4% toward the high single digits. Not yet in evidence. Margin was flat sequentially in Q2 and is guided to 7.4%. Every 100bp of Marine margin on ~$18B of revenue is roughly $180M of operating profit, or about $0.55 of EPS — so the difference between 7.4% and 9% is worth roughly $0.90 of EPS, or 5–6% of the current guide. Material, but not transformative, and management describes the path as “a slow, steady drumbeat.”
3. Aerospace holding ~14.7% margins through the G280-to-G300 production gap. Currently tracking above guide (14.5% in Q2 against a 14.7% full-year guide that implies a stronger Q4), but 2027 has a disclosed discontinuity.
4. No compression of the sector multiple. This is doing more work than the other three combined, and more work than it was doing in June, because the multiple is now at the 99.94th percentile rather than the 94.5th. If GD simply reverted to its own five-year average forward multiple of ~19× on the FY2026 guide, the share price implied would be roughly $320 — a 19% decline on unchanged, delivered earnings.

Scenario analysis (illustrative, not a target). FY2027E EPS is taken as ~$18.75, implying 11% growth on the FY2026 guide — broadly consistent with the segment guides, the backlog, and the July award beginning to convert.

• Bull: Marine margin reaches 8.5%+ on ~$19.5B of revenue as the award-funded capacity comes online, Aerospace bridges the G300 gap without an air-pocket, Combat’s 2.1× book-to-bill converts, and Technologies stabilises. EPS reaches ~$19.50 in FY2027 and ~$22 by FY2028. If the ~23× multiple holds, the equity compounds low-teens. Requires the multiple to hold at an all-time-record percentile.
• Base: ~8% revenue growth, Marine grinding to 7.8–8.0%, EPS ~$18.75 (FY2027) and ~$20.50 (FY2028). Multiple drifts from 23.5× toward the ~20–21× that would still be above GD’s own five-year average. Total return is roughly the dividend plus low-single digits — earnings growth largely offset by de-rating.
• Bear: The defense complex de-rates on the smaller-than-advertised FY2027 appropriation and a rotation out of the sector; or Marine margin stalls at 7.4% while capex runs at 4%+ of sales and ROIC flattens; or the G300 gap produces a 2027 Aerospace shortfall. The multiple reverts toward the 17–19× that covers most of GD’s own history, producing a 20–30% drawdown on flat-to-modestly-growing earnings.

What the market has right, and what it may have wrong. Correctly priced: the backlog visibility, which is now genuinely exceptional; the demand certainty from the July award; the clean accounting; the de-levered balance sheet; and — a correction from my June note — the return profile, which at 17.8% ROE is better than that note credited. Potentially mispriced in two directions. The market may be over-extrapolating first-half cash conversion (over 150%) that management explicitly guided down to ~105% for the year, and under-pricing the sector-beta character of the stock — with essentially zero Quality and zero Momentum factor loading and a +0.52 A&D industry loading, GD is not the idiosyncratic quality compounder the narrative describes; it is a high-quality operator inside a re-rated sector, and it will de-rate with that sector regardless of how well it executes. Conversely, the market may be under-pricing the duration of the submarine annuity: fourteen contracted boats plus the existing programme of record is a two-decade revenue stream with essentially no competitive risk, and a stream of that duration arguably warrants a higher multiple than a defense average.

The valuation is full, the business quality is real and better-established than in June, and the asymmetry from here is unfavourable absent a pullback — with the specific note that the dominant downside mechanism is now sector multiple compression rather than company execution.

11. Variant Perception

Consensus view. GD is a high-quality, defensively-positioned defense prime with a uniquely diversified four-franchise portfolio, a record backlog freshly augmented by the largest submarine award in its history, three consecutive guidance raises, the cleanest accounting in the peer group, and a de-levered balance sheet — a “sleep-well-at-night” compounder riding a generational defense cycle. The sell side reflects this: the stock beat consensus by $0.28 in Q2 and made an all-time high within eight sessions of the print. Even the downgrades in the news flow are valuation downgrades — two Seeking Alpha pieces published 30 July are titled “General Dynamics Has Run Its Course (Downgrade)” and “Strong Tailwinds, But The Valuation Is Pricey (Downgrade)” — which is to say consensus has stopped arguing about the business and now argues only about the price. That is itself a signal about where we are in the cycle.

Strongest bull case. The July award transforms the character of the Marine franchise. For a decade the submarine story was a promise — the Navy said it wanted two Virginias and one Columbia a year, but would not contract for the boats, so GD and its single-source suppliers could not justify the capacity investment to deliver them, so the cadence never arrived. The $71.6B award breaks that loop, and Electric Boat’s president said so explicitly: it provides “the demand certainty we need to continue investing in capacity.” Layer on: Gulfstream’s best order half since 2022 at 1.5× book-to-bill with 130bp of margin expansion; Combat’s 2.1× book-to-bill from allied rearmament; Mission Systems’ international portfolio up 35%; net debt at $3.2B with the note wall being retired rather than refinanced; a corrected 17.8% ROE that is rising; and the cleanest GAAP in a peer group where every other prime has taken nine- or ten-figure charges. A business with 2.5 years of contracted backlog, no accounting distortions, a rising return profile, and a two-decade annuity in its largest segment is worth a premium to the defense average — and GD is arguably the only prime that deserves one.

Strongest bear case. You are paying the highest multiple in the company’s measurable history — the 99.94th percentile on both P/E and P/S — for a business whose consolidated operating margin is 10.4%, whose ROIC is 13.2%, and which is about to roughly double its capital intensity. Four-fifths of the last nine weeks’ return was multiple expansion, not earnings. The single profit lever the whole Marine thesis rests on has not moved: margin was flat sequentially at 7.3% and is guided to 7.4%, with management describing the cadence prize as an early-2030s event, not a near-term one. Combat’s revenue is flat and Technologies is decelerating with eroding margin. The FY2027 budget that fuelled the sector’s spring re-rating has resolved at ~$1.1T rather than ~$1.5T. A known Aerospace production gap arrives in 2027. And the factor decomposition says the thing the narrative denies: with zero Momentum and zero Quality loading and a +0.49 DividendYield loading at a 1.61% yield — the thinnest in a decade — GD’s price is half-explained by a defense-sector factor that has already re-rated and a dividend-yield factor whose natural buyer has been priced out. Meanwhile the eight most-informed individuals at the company exercised roughly 336,000 options into the high and retained fewer than 10,000 shares between them, with zero open-market purchases. A reversion to GD’s own five-year average multiple, with earnings exactly as guided, is a ~19% decline.

The 3–5 assumptions that matter most, with falsifiers:

1. The defense-sector multiple holds. This is now the dominant assumption, ahead of any company-specific one. Falsifier: three-to-four turns of sector multiple compression, or GD’s own P/E percentile retreating below the 80th while earnings track guidance.
2. Marine margin converts from 7.4% toward 8.5%+ as award-funded capacity comes online. Falsifier: two more quarters of flat-to-down sequential Marine margin, or FY2027 Marine margin guided at or below 7.5%.
3. Rising capex does not stall ROIC. Falsifier: capex sustained above 4.0% of sales with ROIC flat or falling for four consecutive quarters — the Northrop pattern.
4. Aerospace bridges the G280-to-G300 gap without a 2027 revenue or margin air-pocket. Falsifier: FY2027 Aerospace guided below ~$13.5B of revenue or below ~14% margin.
5. The July award is substantially incremental to backlog rather than a conversion of previously-disclosed contract value. Falsifier: Q3 backlog rising by materially less than ~$60B.

The factor-positioning read — and a correction to June. My June note characterised GD as “a low-volatility quality+momentum name that has been rewarded.” The current loadings do not support the quality or momentum half of that. In FactorsToday’s “All Factors” model GD’s live loadings are Aerospace & Defense industry +0.519, Market +0.508, DividendYield +0.494, Industrials +0.417, LowVolatility +0.162, Value +0.067 — with Quality at +0.006 and Momentum at −0.011, i.e. statistically zero. The risk-adjusted record is genuinely excellent: beta 0.494, alpha 0.129, one-year return +27.5% at a Sharpe of 1.17 with a maximum drawdown of only −14.5%; the most recent quarter annualises to +99.3% at a Sharpe of 4.19 (approximately +18.8% raw). Relative strength is +26.9% over twelve months, and rs_peak of −0.05% puts the stock at its all-time high. Idiosyncratic volatility is 16.5% annualised against a model R² of 0.50, and the factor-similar peer list is dominated by defense ETFs — PPA (0.91), DFEN (0.89), ITA (0.89), SHLD (0.87), XAR (0.85) — with NOC (0.85) the nearest single stock.

The variant-perception implication is sharper than June’s. This is not a crowded momentum trade and it is not a quality-factor trade; it is a defense-sector-beta-plus-dividend instrument that has been carried by a sector re-rating, wearing an operational record good enough to make the re-rating feel earned. Consensus is correct that the business is excellent. Where consensus may be offsides is in attributing GD’s return to GD. Half of it belonged to the sector and to the dividend-yield factor, and both of those are now less supportive than they were: the sector’s anchoring budget number came in a third lower than the request, and the yield has compressed to 1.61%. The place to disagree with consensus is not on business quality — that argument is over and consensus won it — but on the attribution of the return and therefore the durability of the multiple.

12. Fact vs. Interpretation Table

13. Open Questions

1. How much of the $71.6B Electric Boat award is incremental to backlog? GD called them “contract modifications,” and defense-segment potential contract value at quarter-end ($49.2B) was smaller than the award — so a substantial portion must be new, but the split is undisclosed. The Q3 backlog print (call scheduled 30 October) settles it.
2. Can Marine margin clear 7.5%, and on what timetable? It has been flat sequentially for two quarters at ~7.3% and is guided to 7.4%. This is the single most important operating number in the company.
3. Does the near-doubling of capex hold ROIC flat or better? Capex moves from 2.2% to 3.5–4.0% of sales. Northrop’s precedent is unflattering.
4. What is the FY2027 shape of Aerospace given the G280-to-G300 gap? Novakovic deferred it explicitly to next year. A one-to-three-quarter production break in the highest-margin segment is not currently in any published guidance.
5. What is the accumulating cash actually for? Buybacks are anti-dilution only, M&A is refused as a topic, the dividend is 37% of earnings, and net debt is already under 0.5× EBITDA.
6. Can Technologies inflect, or is 4% the structural rate? Mission Systems is growing well and GDIT’s OTA win rate is improving, but the segment has now decelerated with eroding margin for several quarters, and $21B of goodwill sits on it.
7. What contract structure did Block VI Virginia-class carry? Asked directly on the call, Novakovic deferred: “let’s give you the clarity that you’re seeking here… after the awards are granted.” Fixed-price versus cost-type terms on the new boats materially affects the risk profile of GD’s cleanest advantage.
8. Is the succession timeline set? Deep is President and now a Director; Novakovic has not signalled a date.

14. What Must Be True

Bull case — what must be true:

• The $71.6B award proves substantially incremental, taking Q3 backlog toward ~$200B and confirming a two-decade contracted submarine annuity.
• Marine operating margin converts from 7.4% toward 8.5%+ as award-funded capacity comes online, without capex above 4% of sales stalling ROIC below 13%.
• Aerospace sustains ~14.7% margins and bridges the G280-to-G300 gap without a 2027 air-pocket; Combat’s 2.1× book-to-bill converts to high-single-digit growth as guided.
• The defense-sector multiple holds near current levels, allowing low-teens EPS compounding (~$18.75 FY2027, ~$22 FY2028) to translate into shareholder return.
• Falsification test: Marine margin fails to exceed 7.5% in any quarter through FY2027 while FY2027 Aerospace is guided below ~$13.5B of revenue — a stalled profit lever plus a production-gap air-pocket in the two segments carrying all the growth would break the bull case outright.

Bear case — what must be true:

• The defense complex de-rates as the FY2027 appropriation settles a third below the request that fuelled the spring re-rating, dragging GD’s 99.94th-percentile multiple toward its own historical mean.
• Marine’s margin stays near 7.4% while capital intensity roughly doubles, so backlog growth converts into revenue but not into returns — ROIC flattens or falls.
• Technologies stays at ~4% growth with eroding margin; Combat’s U.S. trough runs longer than the order book implies; the G300 gap costs 2027 Aerospace.
• Multiple reversion toward the ~19× five-year average — on earnings exactly as guided — delivers roughly a 19% decline; toward the 17× that covers much of GD’s history, closer to 28%.
• Falsification test: Marine margin exceeds 8.0% in two consecutive quarters and Q3 backlog rises by more than ~$60B on the award and GD’s P/E percentile falls below the 90th while the share price is flat-or-higher — i.e. earnings growing into the multiple rather than the multiple growing into the earnings. If all three occur, the de-rating case is dead and the premium is structurally earned rather than merely cyclically granted.

The honest synthesis. In June this note’s predecessor wrote that “the business is tracking the bull case, but the stock prices it in,” and set a bear test that the business promptly passed on all three legs. The business has now done more than track the bull case — the July award materially strengthened the franchise, the return profile was better than we credited, and the balance sheet de-levered. And yet the investment conclusion has not improved, because the multiple absorbed the entire improvement and then some: the P/E percentile went from 94.5 to 99.94 while the earnings guide rose 2.1%. The bear case has migrated from “the business will disappoint” to “the business will deliver and the multiple will not hold” — which is a weaker argument about the company and a stronger one about the price. That is why this remains a quality name to own on weakness rather than to chase at the record, and why the accumulation zone in Claude’s Take has been raised on the strength of the business rather than the strength of the tape.

15. Source Appendix

The full citation list follows in Appendix B. Primary sources: GD Q2 2026 Form 10-Q for the quarter ended 2026-07-05 (SEC EDGAR, CIK 0000040533, filed 2026-07-29); GD FY2025 Form 10-K; GD Q2 2026 earnings release and 8-K (2026-07-29); GD Form 424B5 rescission-offer prospectus supplement (2026-08-04) and Form S-3ASR (2026-07-30); GD Form 4 corpus (raw XML), 2026-05-11 through 2026-08-13; GD Q2 2026 earnings call transcript (2026-07-29, via ROIC.ai MCP). Quantitative data: ROIC.ai MCP (income statement, profitability ratios, enterprise value); AZI valuation-index own-history percentiles and five-year price CSV; FactorsToday factor model (stock loadings, leaderboard, stock info, specific volatility, related stocks). Corporate announcements via PR Newswire: Electric Boat $76.6B Navy award (2026-07-29), GDIT ENOCS award (2026-08-07), Danny Deep board election (2026-08-06), dividend declaration (2026-08-05). Industry and peer framing: the author’s prior published notes on GD (2026-06-14), NOC (2026-08-15), LMT (2026-08-14), RTX (2026-08-12), BA (2026-08-07) and HII (2026-07-10), each built from those companies’ own public filings. All non-obvious facts are cited with form and date in the appendix.

APPENDIX A — Standard Diligence Questionnaire

General Dynamics Corporation (NYSE: GD) — as of 2026-08-15

Supplemental to the analysis above. Fact / Interpretation / Assumption labels applied where it matters. This is an update note; answers that are unchanged from the 2026-06-14 edition are stated concisely, and the emphasis falls on what the Q2 2026 quarter and the 29 July Navy award changed.

General

What thoughtful questions have other investors asked about this company? The Q2 2026 call Q&A is the cleanest available proxy, and the questions have shifted noticeably since Q1. The Street is no longer asking whether the business works; it is asking about pace, margin ceilings and capital. The live questions were: (1) how far out does the Aerospace backlog now extend, and how much can production rise? (Strauss, Wells Fargo — Novakovic declined, citing competitive sensitivity); (2) is there a reason Marine growth decelerates from here? (Strauss — Deep: second-half moderation is material-receipt timing, “we don’t have any expectation that it will slow down much”); (3) where is the Virginia-class build rate now? (Epstein, BofA — Deep declined the number but confirmed the 2+1 target is an “early 2030s” objective); (4) is M&A re-engaging given balance-sheet strength and the DoD’s negative view of buybacks? (Walton, Wolfe — refused outright); (5) what is the Aerospace margin ceiling, and can Gulfstream reach luxury-goods margins? (Kahyaoglu, Jefferies; Khanna, TD Cowen; Liwag, Morgan Stanley — Novakovic declined out-year margin guidance three separate times and rejected the “luxury” framing); (6) is Marine’s 7.3% a good baseline? (Kahyaoglu — Deep: “a slow, steady drumbeat”); (7) what forces the Navy to place the submarine orders? (Khanna — answered that afternoon by the $76.6B award); and (8) how sensitive is Technologies to extended continuing resolutions? (Godyn, Citi). The density of margin-ceiling questions is itself the tell: sophisticated investors are trying to work out what this company can earn at maturity, because at 23.5× forward that is the only question left.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: Earnings are mid-cycle and rising; the valuation is at an all-time high. Defense revenue is early-to-middle in a multi-year up-cycle now underwritten by contract rather than forecast — the $71.6B Electric Boat award covers fourteen boats stretching into the 2030s. Consolidated operating margin at 10.4% is still below GD’s own 2020 level of ~10.9%, so there is recovery runway rather than peak-margin risk. Gulfstream is closer to a cyclical high on deliveries (~160 guided for 2026, up from 111 in FY2023), though it remains supply- rather than demand-constrained. Net: earnings are not at a cyclical peak; the multiple is at a historical one.

Driven by the external environment or internal actions? Both, and the split is roughly even. External: the post-Ukraine rearmament cycle, NATO spending commitments, and a ~$1.1T FY2027 congressional base mark (up ~10%). Internal: Gulfstream’s G700/G800 learning-curve progress (130bp of Q2 margin expansion on “durable productivity improvements”), Marine throughput gains (Columbia hours earned +37% in H1; a DDG-51 delivered three months early), and a deliberate de-levering that cut net interest expense from $88M to $49M year-over-year in the quarter.

How stable are revenues? Fact: Exceptionally stable by industrial standards. Backlog was $136.5B at 2026-07-05 — approximately 2.5 years of revenue — with $186.9B of total estimated contract value, or roughly 3.4 years including options and IDIQ. Book-to-bill was above 1.0× in all four segments. The 29 July award adds materially on top. Aerospace (~25% of revenue) introduces the only genuine new-order cyclicality, and it is currently running a 1.5× book-to-bill.

Outlook for products and services? Strong and narrowing in dispersion. Strongest: submarines (contracted for a decade-plus), munitions and international armoured vehicles (OTS and European Land Systems both guided to double-digit growth), and Mission Systems international (+35% since 2024). Solid: Gulfstream, with a known 2027 production gap between the G280 and G300. Weakest: GDIT federal IT, still constrained by “elongated procurement cycles,” and U.S. Land Systems, in a platform-transition trough between current Abrams/Stryker volumes and the M1E3/XM30 generation.

How big will this market be — growing, shrinking, domestic or international? Growing, and internationalising. U.S. defense sits at roughly $1.0T with an FY2027 congressional base mark near $1.1T — a strong ~10% increase, though materially below the ~$1.5T request that drove sector sentiment earlier in 2026 (Fact). This gap is the sector’s principal under-discussed risk. Internationally, NATO commitments toward 3.5–5% of GDP are driving European Land Systems and Combat exports, and Gulfstream reports “very active interest” in the U.S. and Asia. Roughly 17% of GD revenue is non-U.S. and rising.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Interpretation: Structurally less, with one important qualification. Entry into the prime franchises is effectively blocked — the nuclear-submarine business has exactly two qualified yards, and GD’s 10-K describes HII as “one primary competitor with which it also partners.” The qualification is that the customer is changing the terms: Novakovic confirmed on the Q2 call that GD sees “additional interest in fixed price, which we welcome.” A structural migration toward fixed-price contracting transfers risk from the buyer to the contractor and is precisely the mechanism that produced Lockheed’s ~$3.6B, Boeing’s $5.3B and RTX’s ~$3B of charges. Competition is not increasing; contractual risk is.

How profitable is the business (ROIC, ROE)? Fact, and a correction to the June edition: TTM ROIC 13.2%, ROE 17.8%, ROA 7.7%, operating margin 10.3%. The June appendix stated ROE of ~9.7%; that figure was wrong. TTM net earnings of roughly $4.49B against $26.8B of equity gives 17.8%, independently confirmed by ROIC.ai. All four return metrics are on a rising two-year trend. The accurate characterisation is that GD’s margins are capped by its monopsony customer while its equity returns are good and improving — not that its returns are structurally poor.

How profitable is the industry — how many competitors, what barriers to entry? Moderately profitable and tightly clustered: prime operating margins run 7–14% and ROIC low-teens across GD, LMT, NOC and HII, because cost-based and negotiated fixed-price contracting with disclosed cost data caps the surplus the contractor may retain. Barriers are extreme — security clearances, decades of programme-specific IP, a certified industrial base, and in the submarine case a political necessity that the yards not fail. The result is the classic Marathon capital-cycle position: capital cannot enter to compete the returns away, and the incumbent cannot earn extraordinary returns on them.

Can the business be easily understood? At segment level, yes. In aggregate it requires genuine sum-of-the-parts thinking, because a submarine yard, a business-jet OEM, a tank plant and a federal IT integrator have almost nothing in common operationally, and they carry margins from 7.3% to 14.5%.

Can it be undermined by foreign low-cost labour? No. Cleared U.S. defense work and high-end aircraft completion are not offshorable. The single relevant exposure is the Israel-based G280 production partner — and that line is being retired by ~Q2 2027 in any case.

Do brands matter? For Gulfstream, decisively — brand, installed base of 3,000-plus aircraft and a global service network are core to pricing and share. Novakovic explicitly rejected the “luxury” framing, arguing these are “business tools,” which if correct implies a corporate-capex demand function rather than a discretionary-wealth one. For the defense segments, programme incumbency and IP matter far more than brand.

What is the nature of competition? Sole- and dual-source incumbency in defense (Abrams, Stryker, submarines); brand, range performance and service in business jets; price-and-recompete in federal IT. Gulfstream demonstrated a genuine supply-chain competitive edge this quarter — Deep attributed the ability to hold schedule while peers struggle to giving suppliers “visibility into our production plans for whatever period is appropriate.”

What are customers’ switching costs? Very high in defense — the Army cannot re-source a tank line, and the Navy cannot replace a submarine yard. Moderate for Gulfstream (fleet commonality, multi-year delivery slots, residual values). Low for GDIT, where contracts are routinely re-competed.

Financial Condition & Balance Sheet

Assets not fully recognised on the balance sheet? Interpretation: The submarine franchise, the sole-source Army vehicle positions and the Gulfstream brand carry no balance-sheet value but represent most of the enterprise’s economic worth. The 29 July award arguably created a large off-balance-sheet asset — a fourteen-boat contracted revenue stream — that will appear only as backlog disclosure.

Off-balance-sheet liabilities? Standard and modest. Operating leases are capitalised; the pension is being de-risked rather than left as an exposure, with a ~$500M discretionary contribution planned in the second half to plans already near full funding. Nothing unusual is flagged in the Q2 10-Q. The one new contingent item is the 401(k) rescission offer covering up to 1,010,925 unregistered plan shares, which the company states will not have a material impact.

How conservative is the accounting? Fact: The most conservative in the peer group, and the gap has widened. GD carries no reach-forward charges this cycle, against Lockheed’s ~$3.6B (2024–25), Boeing’s $5.3B (FY2025), Northrop’s B-21 charges and RTX’s ~$3B GTF charge. It carries no material non-operating pension income inflating EPS — a distinction that matters against Huntington Ingalls, where roughly $3.7 per share of reported EPS is non-cash pension income, turning an apparent 18.6× P/E into roughly 24.7× on core earnings. Cash from operations consistently exceeds net income (162% in Q2). Stock-based compensation is ~0.4% of revenue. GD’s reported EPS requires fewer normalising adjustments than any peer’s.

How CapEx-hungry is the business? Moderate historically and rising sharply. Capex was 2.2% of sales in FY2025 and is guided to 3.5–4.0% for FY2026 — an increase of roughly 70–95% in absolute dollars — with Kuryea stating investment will “grow significantly in the back half of the year… especially in our shipyards, to accelerate production and meet future demand.” Interpretation: this is the correct response to the July award and the single most important thing to monitor for return quality. The directly comparable precedent is unflattering: Northrop’s ROIC fell from ~15% (2019) to ~11.4% (2025) through a heavy capex programme serving a large contracted franchise. GD’s ROIC is currently rising (12.97% → 13.23% over two quarters), so there is no evidence of the problem — only the mechanism.

Capital Allocation & Management

How much free cash flow does the business generate, how does management use it, and what is the philosophy? First-half 2026 operating cash flow was $4,035M (against $1,450M a year earlier) and free cash flow roughly $3.6B, a conversion rate above 150%. This is explicitly not the run-rate: Kuryea guided full-year conversion to “around 105%,” naming four second-half drags — a sharp capex step-up, a ~$500M discretionary pension contribution, over $500M of back-half cash taxes, and the unwind of European Land Systems advance payments. FY2026E free cash flow is therefore approximately $4.8B. Uses, in priority order: rising capital investment, a predictable and growing dividend ($834M in the half), debt retirement ($500M repaid in June, a further $500M in August to be repaid with cash, not refinanced), and anti-dilution-only buybacks. The philosophy is conservative, per-share-oriented, and currently biased toward reinvestment.

Any significant acquisitions recently? None material since CSRA in 2018. Asked directly on the Q2 call whether balance-sheet strength and the DoD’s dislike of buybacks might re-engage M&A, Novakovic refused the topic entirely: “Nice try… we just never talk about it.” Interpretation: abstention in a seller’s market — Lockheed has just paid $3.5B for Ultra Maritime — is a positive by omission.

Is the company buying back shares? Fact: Only to offset dilution — approximately $100M in Q2 and $319M in the first half, against ~$4.8B of expected annual free cash flow. The 10-Q language is explicit: “opportunistic share repurchases primarily to address dilution.” Interpretation: declining to repurchase stock at the 99.94th percentile of its own historical valuation is defensible capital allocation, arguably good capital allocation. The unanswered question is what the accumulating cash is for, given net debt is already under 0.5× EBITDA.

Is the company issuing large amounts of new shares to insiders? No. Stock-based compensation runs ~0.4% of revenue and the diluted share count is falling — 277.7M in Q2 2024 to 273.5M in Q2 2026 — despite buybacks being restricted to anti-dilution.

What is the compensation policy for directors and management? Unchanged and well-designed: the annual incentive weights EPS, free cash flow, operating margin and strategic objectives; long-term incentives are performance share units on three-year ROIC with a relative-TSR modifier, with ROIC named the “most important financial performance measure.” Ownership requirements are among the most stringent in the Fortune 100 (CEO at 15× salary). Say-on-pay support was 96%.

What are the motivations of management? Interpretation: Returns on capital and per-share value, consistent with the incentive design and with observed behaviour — no overpriced M&A, no buybacks at a record multiple, pension de-risking, debt retirement rather than refinancing. The one datapoint that cuts the other way is personal: across 11 May to 13 August 2026 the eight most senior insiders exercised roughly 336,000 options and retained fewer than 10,000 shares between them, with zero open-market purchases. Novakovic alone exercised 166,660 options at $165.47 and disposed of all of them at $379–384, leaving her net position unchanged at 766,457 shares (still roughly $300M of stock, so this is monetisation rather than exit). Legally routine; as a conviction signal, unhelpful.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No. Standard U.S. common stock, NYSE-listed, ordinary 1099 dividend reporting.

What is the dividend policy? $1.59 per quarter ($6.36 annualised), raised from $1.50 in March 2026, extending an unbroken run of annual increases spanning three decades. The payout ratio is 36.6% of TTM earnings and roughly 26% of expected FY2026 free cash flow — comfortably covered with room to keep raising. The yield of 1.61% is the thinnest in a decade, which is analytically relevant because the stock carries a +0.49 factor loading on DividendYield: its natural income buyer is being priced out.

How profitable is the business? TTM ROE 17.8%, ROIC 13.2%, operating margin 10.3%, all rising. See the correction noted above.

Is net income diverging from cash from operations? Fact: Yes, favourably — cash from operations exceeded net income by 62% in Q2 and by roughly 50%+ across the first half. This is a quality-of-earnings positive, with the caveat that management has explicitly guided the full-year figure back toward 105% and named the four reasons.

Risks & Downside

What factors would cause the stock to decline? In descending order of probability-weighted impact: (1) compression of the defense-sector multiple — with the stock at a 99.94th-percentile P/E, a +0.52 loading on the A&D industry factor and a model R² of 0.50, roughly half the price variance is sector-driven, and the FY2027 appropriation has resolved a third below the request that drove the re-rating; (2) Marine margin failing to convert from 7.4% while capex roughly doubles, stalling ROIC; (3) a 2027 Aerospace air-pocket from the disclosed G280-to-G300 production gap; (4) a prolonged continuing resolution; (5) a first reach-forward charge, which would remove GD’s cleanest competitive distinction. Reversion to GD’s own five-year average forward multiple of ~19×, on earnings exactly as guided, implies roughly a 19% decline; reversion to 17× implies closer to 28%.

Is there a risk of catastrophic loss? Interpretation: Low. The most plausible large negative surprise remains a reach-forward loss on a major fixed-price programme, and the industry’s drift toward fixed-price contracting raises that tail. But GD’s contract mix remains balanced, its execution record is unblemished this cycle, and the submarine work is contracted with a customer that structurally cannot let the yard fail.

What is the chance of a total loss? Negligible. Diversified across four franchises, investment-grade, net debt under 0.5× EBITDA, with multi-decade national-security programmes and no single point of failure.

Recent News & Events

Has the business environment changed recently? Yes, and materially — on balance favourably. The defining event is the $76.6B U.S. Navy submarine award of 29 July 2026, $71.6B of it to Electric Boat, covering five additional Columbia-class and nine additional Virginia-class boats plus shipyard infrastructure. Electric Boat’s president framed the economic significance precisely: it provides “the demand certainty we need to continue investing in capacity.” Because the award landed after the 5 July quarter-end it is not in the reported $136.5B backlog. Alongside it: a third 2026 guidance raise to $16.80–$16.90; a record backlog in every segment; Gulfstream’s strongest order half since 2022; and net debt down to $3.2B. Against it: the FY2027 budget request resolving into a ~$1.1T congressional base mark rather than ~$1.5T, and a newly disclosed 2027 Gulfstream production gap.

Any significant acquisitions? None.

Any change in accounting policies? None. Accounting remains conservative and unusually clean.

Recent changes — new markets, facilities, management? Facilities: a substantial shipyard capacity expansion is underway, funded by capex rising from 2.2% to 3.5–4.0% of sales. Markets: Combat is pivoting toward allied demand (a $1.4B Canadian armoured-vehicle contract in July, plus European Land Systems growth); Mission Systems’ international portfolio is up more than 35% since 2024; GDIT won a $1.3B Army National Guard enterprise network and cybersecurity contract announced 7 August. Products: G800 delivering, G280 ending ~Q2 2027, G300 entering service late 2027/early 2028, G400 in development; M1E3 and XM30 in development for the Army. Management: Danny Deep, named President in December 2025, was elected to the Board of Directors on 6 August 2026 — an advancing but undated CEO succession. Governance: GD disclosed that up to 1,010,925 shares acquired through the 401(k) Stock Fund between July 2025 and June 2026 were sold without an effective registration statement; a curative Form S-8 was filed 1 July 2026 and a rescission offer expires 8 September 2026. Small in dollars, but a compliance lapse worth recording.

APPENDIX B — Source Appendix

General Dynamics Corporation (NYSE: GD) — as of 2026-08-15

All sources accessed 2026-08-15 unless otherwise noted. Primary sources are listed first. This is an update note; the durable scaffolding (FY2021–FY2025 Forms 10-K, the DEF 14A, and the pre-2026 8-K corpus) was established in the 2026-06-14 note and is not re-listed exhaustively here.

A. Primary — SEC filings (EDGAR, CIK 0000040533)

Enumerated via scripts/edgar.sh since GD 2026-05-01. Form types 424B*, FWP and 144 excluded from the material-event review as structured-note/administrative noise, with the exception of the 424B5 above, which is substantive.

B. Primary — Company communications

C. Quantitative data services

Authority note: ROIC.ai, AZI and FactorsToday are third-party aggregated or statistically-estimated data, not primary. Every material figure driving a verdict in this note has been reconciled to the Q2 2026 10-Q or the earnings release. Where the 10-Q and an aggregator differ, the filing governs.

D. Insider-transaction sources (SEC Form 4, raw XML)

Reviewed directly from EDGAR accession XML rather than rendered HTML, so that transactionCode sweeps run against the filed data.

Finding: zero transactions with code P (open-market purchase) across the entire reviewed window. Approximately 336,000 options exercised; fewer than 10,000 shares net retained in aggregate. No Rule 10b5-1 footnotes appear on any of these filings, consistent with the 10-Q’s Item 5 statement that no director or officer adopted or terminated such an arrangement during the quarter.

E. Third-party press (used for triage and event dating; all material claims verified against primary sources above)

News triage conducted via the ROIC.ai MCP get_company_news tool (50 items, window from 2026-06-01). The tape over the period is heavily weighted toward Zacks-style screening content; the genuinely material items are the Q2 print, the $76.6B Navy award, the GDIT ENOCS award, the board election and the dividend declaration, all of which are sourced to primary releases above.

F. Author’s prior work on the sector

G. Analytical frameworks

H. Data limitations and caveats

1. The $76.6B award’s backlog treatment is not yet disclosed. GD described the transaction as “contract modifications”; defense-segment estimated potential contract value at quarter-end was $49.2B, less than the award, implying a substantial incremental component — but the split is unpublished and is flagged as Interpretation, not as fact. The Q3 disclosure (call scheduled 2026-10-30) resolves it.
2. Marine’s Q1 2026 segment margin is derived, not directly reported — computed from the 10-Q’s six-month and three-month segment tables as ($658M − $342M) / ($9,003M − $4,660M) = 7.28%.
3. Diluted market capitalisation and enterprise value are computed at the 2026-08-14 close using the Q2 diluted share count of 273.5M and the 5 July balance sheet; ROIC.ai’s reported figures are as of 30 June and therefore reflect a lower share price.
4. FactorsToday loadings are in-sample statistical estimates, L1-sparse (an absent factor is zeroed, not missing) and level-dependent across nested models; betas are compared only within a single model, per the service’s methodology. Reported loadings, returns and drawdowns are facts; any statement about persistence is Interpretation and is labelled as such.
5. AZI valuation percentiles are own-history only and are never used cross-sectionally. GD’s GAAP EPS carries no distortion (no reach-forward charges, no material pension income), so the P/E percentile is readable here without the adjustment required for REITs, deep cyclicals or pension-flattered issuers.