Lockheed Martin (LMT): Backlog Record, Margin Recovery
Lockheed Martin‘s backlog hit an all-time high of $230.4 billion at the end of the June quarter, according to its second-quarter 10-Q, and the stock still trades 21.9% below the $683 high it set earlier this year. A backlog that big and a stock that beaten down do not usually sit together for long.
The gap comes down to one thing: fixed-price development programs that turned profitable-looking contracts into loss-making ones once the engineering ran into trouble. Lockheed booked roughly $1.6 billion of combined program losses in the second quarter of 2025, mostly from an ongoing classified program at its Aeronautics division plus two troubled helicopter contracts, the Canadian Maritime Helicopter Program and the Turkish Utility Helicopter Program, per the company’s own second-quarter release. A year later, in the quarter ended June 28, 2026, none of those programs added a new charge, and net earnings jumped to about $1.8 billion from $342 million in the year-ago period. My take: the demand side of this business was never in doubt, and the margin side just had its first clean quarter in a while, so last year’s multiple compression is now pricing in more bad news than the company is currently producing.
What the record backlog is made of
Ending backlog was $230.4 billion as of June 28, 2026, up roughly $64 billion from a year earlier. About $36.8 billion of that six-month increase came from a single item: an undefinitized contractual action for the THAAD missile defense program at Lockheed’s Missiles and Fire Control segment, layered on top of a separate $35 billion THAAD interceptor contract and a $3 billion Guided Multiple Launch Rocket System order. That is signed government funding, not hopes about an unwon bid, and Lockheed says it expects to convert about 30% of the backlog to revenue over the next 12 months and roughly half over the next 24.
Missile defense demand did not appear from nowhere. Rising conflict risk across several regions and allied stockpile replenishment pushed the THAAD program specifically toward this kind of order size, a pattern playing out across a few of Lockheed’s missile lines at once rather than one isolated contract.
Backlog did not jump overnight, either. It was $176.0 billion at the end of 2024, $193.6 billion at the end of 2025, and $230.4 billion by the end of June 2026, a climb that ran through two separate program-charge cycles without slowing. New orders in the second quarter alone were $65 billion, more than the entire backlog most defense contractors carry.
That is the demand side of the story. It has not been the problem.
One program is more than a quarter of sales
The F-35 is not just the biggest program at Lockheed, it is close to being the company by itself. F-35 sales were approximately 28% of total consolidated revenue in the second quarter of 2026 and 27% for the first six months, roughly where the split has sat for the last few years. A program that large is not going away on its own; allied air forces do not have another fifth-generation fighter built at this scale to switch to. But it also means a single production hiccup, a software delay, a parts shortage, an allied budget fight, moves the whole company’s numbers more than it would at a more diversified prime. I would treat F-35 news, deliveries, unit costs, any new country signing on, as close to the single most important recurring data point for this stock, alongside the quarterly earnings headline itself.
Fixed-price contracts are not unique to Lockheed. The Pentagon has pushed harder toward fixed-price development deals across the defense industry over the past decade, a way to force contractors to control cost overruns instead of passing them to taxpayers on a cost-plus basis. The tradeoff is that the contractor eats the difference when a new radar, software stack, or airframe design does not work the first time, and primes with heavy development exposure book the loss the moment they can reasonably estimate it, not once the program eventually gets fixed. That is a structural risk difference between a mature, high-rate program like the F-35 and a newer, still-developing effort, and it is why not every dollar of Lockheed’s backlog carries the same risk.
A clean quarter after a year of charges
Go back to the second quarter of 2025 and net earnings were $342 million, or $1.46 a share, dragged down by that roughly $1.6 billion in combined program losses, about $950 million of it on the classified Aeronautics program alone. A year later, in the quarter that ended June 28, 2026, none of those programs added a new charge, and net earnings came in around $1.8 billion, or $7.94 a share. Revenue for that quarter was $20.1 billion, 11% above a year earlier and 11% above the prior quarter, which is the kind of backlog-to-revenue conversion that only shows up once a program stops absorbing losses.
| Metric | Value | Context |
|---|---|---|
| Price | $533.38 | 21.9% below 52-week high of $683 |
| Backlog | $230.4 billion | record, as of June 28, 2026 |
| Q2 2026 net earnings | about $1.8 billion | vs $342 million a year earlier |
| Revenue, FY2025 | $75.0 billion | up 6% from $71.0 billion in FY2024 |
| Operating margin | 10% | EBIT margin 9.4% |
| Forward P/E | 17.4 | five-year average 20.7 |
| Analyst target | $630 | 18% above current price, 36% buy ratings |
| Dividend yield | 2.56% | $13.65 trailing twelve months |
Margins are recovering slower than demand
Full-year operating margin was 10%, with an EBIT margin of 9.4%, both still below where a defense prime with this kind of backlog should eventually land once the fixed-price overhang clears. Gross margin ticked up to 10.2% from 9.8% the year before. A small move, but the right direction.
Net margin for the year was 7%, thin for a company generating $75.0 billion in revenue. That gap between backlog quality and margin quality is exactly why the stock trades at 17.4 times forward earnings against a five-year average of 20.7, a discount rather than the premium a record order book alone would usually justify.
Compare that discount to where the stock sat before the charges started piling up. A five-year average multiple of 20.7 times forward earnings reflects a period when fixed-price losses were rarer news, not the norm investors had to price around every quarter. Getting back to that average, or above it, requires more than one clean print; it requires enough consecutive quarters without a new reach-forward loss that the market stops discounting the backlog for execution risk. That is a slower repair than a single earnings beat can deliver, and I would rather underestimate the pace of that repair than assume the June quarter alone settled it.
Why only a third of analysts say buy
14 analysts cover the stock and only 36% rate it a buy, a low ratio for a company that just posted a +10.5% earnings-day move and a record backlog in the same quarter. The average target is $630, 18% above today’s price, but the low estimate in that range sits -6% below where the stock trades now. The full range, $503 to $700, is unusually wide for a company this size, a sign that even coverage this deep cannot agree yet on how fast the margin recovery shows up in reported numbers. That spread tells me the analyst community is more divided on this name than the headline buy-rating count on a growth story like Palantir, where the disagreement is mostly about the size of the number, not its direction.
My own quant score on Lockheed slipped from a C to a D over the past few months, even as the fundamentals improved. I read that as the model penalizing the multiple relative to its own history rather than penalizing the business itself.
The dividend is the easier reason to hold
At a 2.56% yield and $13.65 paid over the trailing twelve months, Lockheed offers something a pure growth story does not: income large enough to matter on its own while the margin story plays out. Short interest sits at just 1.2% of the float, thin for a stock trading 21.9% below its high, which suggests the market is not betting heavily against the turnaround so much as waiting for it to keep proving itself one quarter at a time. Trading well off a high while the underlying numbers firm up is not unusual; JPMorgan did something similar on its way back toward new highs after a guidance raise, though that comparison only holds if Lockheed’s margin trajectory keeps improving instead of stalling out again.
Where I land: the backlog removes the demand question for years, so this is not a broken franchise trading at a discount for a good reason. It is a business still proving it can convert fixed-price risk into predictable earnings, and another clean quarter would close more of the gap between 17.4 times forward earnings and the 20.7 times investors used to pay for this name. The next earnings report is the number to hold against this thesis directly: another quarter with no new reach-forward loss, on top of this one, would be the first real two-quarter streak in a while and the stronger signal that the margin problem, not just one comp, is actually behind the company.
Where I would be wrong: if a new charge shows up on a different fixed-price program, even a small one, it would suggest the June quarter was a comparison against an unusually bad prior year rather than a real turn, and the stock’s discount multiple would start looking earned instead of cheap.
Analysis and opinion only, not investment advice. Figures come from Lockheed Martin’s filings on SEC EDGAR and its investor site; valuation multiples are approximate and were checked on September 23, 2026.
Sources: Dividends (SEC Investor.gov glossary) (https://www.investor.gov/introduction-investing/investing-basics/glossary/dividend) · Dividends tax topic (IRS) (https://www.irs.gov/taxtopics/tc404)