The Iran conflict has reached a point where even professionals say they 'can't model the endgame' — that kind of unpriceable risk doesn't resolve quickly. Unlike oil, where a ceasefire could crater prices overnight, defense contractors get multi-year back

Verdict: the defense-cash story is real, but the trend is broken — wait for confirmation

The verdict: wait. The idea's strongest support is fundamental — Lockheed grew revenue 3.4% year over year (96th percentile of Industrials peers), delivered $6.9B of free cash flow, and its Q2 2026 margins improved across the board, while Northrop's net margin rose 1.2 points to 10.1%. The strongest point against is that the trade itself is not live: all three names sit below their EMA (50) — LMT $23.52 below, NOC $22.40 below, RTX $11.54 below — and the most recent ownership filings (period ended June 30, 2026) show net open-market insider selling at all three, about -$2.8M at RTX and nobody buying across 45 filers. The completed 9-month backtest is a caution, not a vindication: 37 trades, a 48.6% win rate, a -1.5% return, and a 13.2% maximum drawdown, with the 60-, 24-, and 12-month windows unevaluable. What would flip this to a buy is a dual confirmation: LMT reclaiming its EMA (50) at $550.53 with RSI (14) above 45, and the next insider filing cycle (due by early November) showing selling stop rather than extend.

Trade now: all three legs still short of their entry triggers

Nothing is actionable today. The strategy wants each name trading above its 50-day EMA with RSI (14) above 45 and ADX (14) above 20 before a long entry arms. Right now none of the three qualifies. LMT closed at $527.01, about $23.52 below its EMA (50) of $550.53, with RSI (14) at 36.6 — it needs a recovery of roughly 4.5% in price plus an RSI gain of 8.4 points. NOC is closer on price ($520.2 versus an EMA (50) of $542.6, a $22.40 gap) but its ADX (14) at 13.3 is well short of the 20 threshold. RTX is the farthest away: at $192.16 it sits $11.54 below its EMA (50) of $203.7, and RSI (14) is 18.8, more than 26 points from the 45 trigger, though its ADX (14) of 81.4 already signals a strong trend. What "wait" means concretely: no position, no partial entry, no scaling. The risk framing once a trade triggers is fixed — a stop at 2.3% below entry and a take-profit at 4.6% above, an effective reward-to-risk of about 2:1, with no single position larger than 25% of the book and sizing set by a 2.3% risk-per-trade rule. Entries also require the day's low to touch the 61.8% retracement while the close holds above it (the simpler second rule set drops the retracement condition), so the trigger is a pullback-recovery day, not a breakout chase. On evidence: the completed 9-month backtest of this exact rule set on LMT produced 37 trades, a 48.6% win rate, a -1.5% return and a 13.2% maximum drawdown — that is the realized bar this setup has to beat from here, and it argues for discipline at the triggers rather than anticipation. On the longer evaluation windows no result could be computed, and no robust parameter setup was established, so the frozen rule set above is what you trade as written. The thesis itself — the idea argues defense names are a durable way to own war-risk versus oil — does not change the mechanical answer: the market is currently pricing these names below their trend baselines, and forcing a long before the RSI and price conditions repair would mean taking the thesis risk without the entry edge the rules were built to provide.

Backlogs, Cash Flow, and a Signal That Actually Traded

The idea's core claim — that defense contractors are a more durable way to own geopolitical risk than oil — is supported by what these companies actually earn and return. Lockheed Martin generated $7.5B of revenue growth of 3.4% year over year in fiscal 2025, a figure in the 96th percentile of the Industrials sector, and delivered $6.9B of free cash flow (99.8th percentile). RTX grew revenue 1.3% year over year on $88.6B of sales while producing $7.9B of free cash flow, also in the 99.8th percentile. These are not defenseless cyclicals; they are cash machines whose demand is contractually anchored to government budgets, which is exactly the durability the thesis is betting on. Lockheed's fiscal-year return on equity of 74.6% reflects a heavily leveraged but highly productive equity base — the operating model monetizes escalation without needing new capital. Momentum within the panel is consistent with the backlog-beneficiary framing. Lockheed's Q2 fiscal 2026 quarter (ended June 28, 2026) showed gross margin improving 0.66 points quarter over quarter to 12.2%, operating margin up 0.91 points to 12.4%, and net margin up 0.89 points to 9.2%. Northrop's net margin rose 1.2 points to 10.1% in the same quarter. Margin expansion at the group level, in a period when the Iran conflict dominates the news cycle, is what you'd expect if multi-year procurement demand is flowing through to pricing and mix. The dividend track record reinforces the cash-generation argument. Lockheed has paid $13.35 per share in 2025 versus $10.60 in 2021, roughly 4.6% annual growth, with a current quarterly payment of $3.45 per share. Northrop grew its payout at a 9.4% annual rate to $7.25 in 2026, and RTX at 7.6% to $2.14. These are board commitments paid from operating cash, not promises contingent on a war continuing — they exist today and rise even in calm periods. On the quant side, this is a backtested, completed evaluation rather than a hypothetical: over a 9-month window on daily bars, the strategy ran 37 trades across the cluster with a 48.6% win rate, an equity curve that peaked above 13.6% before fading, and a worst drawdown of 13.2%. The only cited catalyst — per the Yahoo Finance piece dated September 17, 2026, "Energy markets grapple with Iran war uncertainty: 'We simply don't know how to model the endgame'" — is precisely the kind of unpriceable geopolitical tail the thesis argues sticks to defense budgets rather than evaporating like an oil price. In short: percentile-leading growth and cash flow, group-wide margin expansion in Q2 2026, and rising dividends all argue the fundamental substrate beneath this trade is real, not just headline-driven.

Insiders Are Selling, Margins Are Thin, and the Backtest Lost Money

Start with what the people closest to the numbers are doing. As of the June 30, 2026 filing period, all three companies show net open-market insider selling: Lockheed's insiders net sold about $117.6K, Northrop's about $106.3K, and RTX's about $2.79M. Across 45 filers, nobody is buying. If the…

Scores

• Conviction score breakdown: 47
• Thesis support: 72
• Trade readiness: 25
• Risk quality: 42
• Backtest evidence: 30
• Fundamentals trend: 65

Watch items

• LMT — RSI (14)
• LMT — Price vs EMA (50)
• NOC — ADX (14)
• NOC — RSI (14)
• RTX — RSI (14)
• RTX — Price vs EMA (50)
• RTX — Insider net open-market value
• LMT — Net margin (Q3 earnings)