Lockheed Martin

IThe business

Lockheed Martin makes weapons and the machines that carry them, and sells most of them to the United States government. In 2025 it had sales of $75.0B, up 5.6%, operating profit of $7.7B, net earnings of $5.0B and about 123,000 employees, 72,000 of them engineers, scientists and IT staff.³ It is headquartered in Bethesda, Maryland, and its shares trade on the New York Stock Exchange under LMT.⁶ The Department of War, which most people still call the Pentagon, took 61% of nine-month 2026 sales, and the U.S. government as a whole 70%.⁶

It has four businesses, and the table shows them. They are not equals. Aeronautics is 40% of sales and the F-35 alone is 27% of the whole company and 67% of Aeronautics.³ Missiles and Fire Control is the third largest by sales and the best by margin, and it is where the current boom is.³⁴ Rotary and Mission Systems is Sikorsky plus naval and radar systems. Space builds satellites and the submarine-launched Trident missile, and holds a 50% stake in United Launch Alliance, which flies rockets for the government.³⁶

Who pays, and how

I want the customer concentration said plainly. In the first nine months of 2026, 70% of sales came from the U.S. government, either as prime contractor or as a subcontractor, and 61% from the Department of War.⁶ International customers were $17.9B of the $58.8B, or 30%, and that figure includes foreign military sales, in which a foreign government buys through the U.S. government, so the U.S. government is in the contract chain for part of it.⁶ Commercial sales in the U.S. are close to nothing, $59M in the third quarter.⁶ There is no second customer to fall back on. If Washington changes what it buys, or when it pays, the company has no other buyer for a fighter or an interceptor. The 10-K lists dependence on U.S. government funding among its risk factors, and I put it first.³

The second question is the contract type. Of 2025 sales, $45.2B or 60% were fixed-price and $29.8B or 40% cost-reimbursable.³ That mix matters more than any other line on the page. On a cost-plus contract, the government pays the cost and a fee, so a cost overrun is the customer’s problem and the company’s fee shrinks at worst. On a fixed-price contract, the company agrees the price first and keeps the difference, so a cost overrun is the company’s loss in full. In the third quarter of 2026, fixed-price work was 60% of sales again, $12.5B of $20.7B.⁶ Development work with new technology is where the company has been burned, and the filings say so directly: competitively bid programs with fixed-price development work or fixed-price production options “increase the risk of a reach-forward loss.”⁶ A reach-forward loss is the accounting charge taken for the whole expected loss on a contract as soon as the company sees that costs will exceed the price. It is booked at once and paid out over years, which I come back to in Valuation.

The programs that matter

F-35. A single-engine stealth fighter bought by the U.S. services, international partners and foreign military sales customers.³ Lockheed has delivered 1,387 production aircraft since the program began, 989 F-35A, 255 F-35B and 143 F-35C, including 43 in the third quarter of 2026.⁶ Its backlog was 274 aircraft at 27 September 2026, against 368 at the end of 2025.⁶³ The fall reflects Lots 18 and 19, 296 aircraft, agreed in 2025, with Lots 20 and above not yet under contract.³⁶ The company “remains in negotiations” and says its costs on the Lot 20 advance contract “began to exceed the current contractual authorization and funding” in the third quarter; without more funding, it says, “our results of operations, cash flows, and financial condition will be negatively impacted.”⁶ It also has a framework agreement with the government for multi-year production and sustainment of the F-35.⁶ That sentence is the biggest open question on the page. I do not know the price, quantity or terms of Lots 20 to 22, and neither does anyone outside the negotiation.

The F-35 also has a history that explains why I treat its schedule with suspicion. In 2024 the company had no customer deliveries in the first quarter because the aircraft’s Technology Refresh 3 (TR-3) software, the upgraded computer and software package, was not finished, and it delivered the first TR-3 aircraft in July 2024 while it worked with the government on when final payments on those aircraft would be made.¹¹¹² The 2023 annual report had put customer acceptance of the software at the second quarter of 2024 and warned that the third quarter might be likelier.¹³ The company has said it continues to advance TR-3 and the later Block 4 capabilities.¹² I did not find a Block 4 date in the 2025 annual report or the third-quarter 2026 filing, so I do not state one.

Missiles. This is the growth. In the third quarter of 2026, Missiles and Fire Control sales rose 24% to $4.5B, with $480M more on air and missile defense (PAC-3 and THAAD) and $330M more on tactical and strike missiles (GMLRS and PrSM).⁴ During 2026 the company signed multi-year contracts with the Department of War for PAC-3 MSE and THAAD, and munitions framework agreements for PrSM and the Joint Advanced Tactical Missile.⁶ The PAC-3 award alone was $52.5B and took the backlog from $230.4B to $292.0B in a quarter.⁶ JASSM and LRASM orders drove the backlog at this segment in 2025.³

Helicopters. Sikorsky builds the Black Hawk, the Seahawk and the CH-53K King Stallion. It is also the source of two of the loss-making programs, the Canadian Maritime Helicopter Program, where Canada is the customer, and the Turkish Utility Helicopter Program.⁶ In the nine months of 2026, Rotary and Mission Systems booked $125M of unfavorable adjustments on the Heavy Lift program and $95M on Seahawk.⁶

Space. Lockheed builds the Next Gen OPIR missile-warning satellites, the GPS III navigation satellites, the Orion crew capsule for NASA and the Next Generation Interceptor for homeland missile defense.³ In the third quarter, Space sales rose 4%, with $185M more on strategic and missile defense (Trident and the Next Generation Interceptor) and $115M less on national security space from the OPIR program moving through its life cycle.⁴ Space has the lowest share of fixed-price work, 25% of its sales, and the most cost-plus.⁶ Its operating profit for nine months, $983M, was lower than the $1,072M of 2025.⁴

Golden Dome and the budget

The company’s filing cites “Golden Dome missile defense” and an increase in F-35 purchases among the priorities of the Administration’s FY 2027 request, which asks for $1.5 trillion for defense, including $350 billion of mandatory funding through reconciliation and $760 billion for weapons procurement and modernization.⁶ Congress has not passed the FY 2027 appropriation, and on 2 September 2026 a continuing resolution holding FY 2026 funding levels was signed, to 11 December 2026.⁶ The FY 2026 law provided $839.2B of discretionary funding for national defense, and the 2025 tax act added more than $150B of mandatory defense funding available to September 2029.⁶ I could not open a primary document that puts a dollar figure on Golden Dome or on Lockheed’s share of it, and the company does not give one, so I do not include any in my values. That is an honest hole in the case, and it cuts both ways: the bull case leans on it, and I have not paid for it.

The state of play now

The third quarter of 2026 was strong. Sales were $20.7B, up 11%, net earnings $1.8B, or $7.97 a share, up 14%, and all four segments grew.⁴ Free cash flow was $3.0B in the quarter and $5.6B for nine months, against $4.2B a year before.⁴ On 22 October the company raised its 2026 outlook to sales of $80.0B to $82.0B, earnings of $30.35 to $30.75 a share and free cash flow of $7.8B to $8.0B.⁴ The board authorized a fourth-quarter dividend of $3.60 a share, up $0.15, the 24th increase in a row, as the company puts it.⁶⁴ The shares closed at $589.91 that day, up 7.0%, and at $643.15 on 2 November.¹ A month earlier, on 30 September, they were $509.25.¹

Two things in the quarter are less pleasant. The company has not bought back any shares in 2026, against $2.25B in the same period of 2025.⁶ And its contract assets, the revenue it has recognized but not yet billed, rose from $13.0B at the end of 2025 to $16.9B in September, a $3.9B build in nine months.⁶ I return to both below.

IIThe moat

The test of a lasting business is what it earns when the customer holds the power. Lockheed Martin has one buyer that matters, and that buyer decides price, quantity and the contract terms. So I ask two questions. Does it earn good returns on the money it puts in? And does anyone else do what it does?

On the second, the answer is close to no. The F-35 is the only fifth-generation fighter the U.S. and most of its allies buy in quantity; 1,387 are built, 274 are on order and the 10-K treats the program as the company’s largest.⁶ The PAC-3 MSE is the interceptor behind the Patriot system, and the Department of War just committed $52.5B to its multi-year production.⁶ No competitor can enter either market in a decade. A satellite maker, a missile maker and a helicopter maker each need certifications, facilities and cleared staff that take years to assemble. That is the first source of the advantage, and it is genuine.

On the first question, the numbers say less. Operating margin was 13.6% in 2021 and 10.3% in 2025 C (calculated).⁵ Return on the money invested shows a good business rather than a great one: operating profit of $7.7B on total assets of $59.8B at the end of 2025 is 12.9% before tax, and the equity was $6.7B because the company has spent more than it earned on buybacks and dividends over the years C (calculated).⁵³ A business that really cannot be matched usually earns more. The explanation is the customer. When one buyer purchases most of your output, it pays cost plus a fee set by negotiation, and it watches your margins.

What protects it

• A franchise on the F-35. 67% of Aeronautics sales and 27% of the company come from one fighter, which is also its sustainment business for decades. The service sales of Aeronautics were $4.5B in nine months of 2026, rising with the fleet.⁶ Once an air force flies an F-35, it buys spares, software and support from Lockheed for the aircraft’s life.
• A backlog that does not cancel easily. $292.0B, or 3.6 times 2026 sales, and 40% of it converts to sales in 24 months.⁶⁴ Governments can cancel programs, and the company’s own release says its contracts can be terminated for convenience, which makes the backlog a protection and a risk in one number.
• Cost-plus work that is hard to lose. About 40% of sales are cost-reimbursable, mostly in Space and Rotary and Mission Systems, where the government bears the cost risk.⁶
• Missile demand that has outrun supply. The government is paying to expand capacity: it signed multi-year contracts to raise production of PAC-3 and THAAD and wrote munitions framework agreements for PrSM.⁶

Where it is thin

The protection is thin in the places where the work is new and the price was set before the cost was known. The company has said it has “experienced significant performance issues” on an Aeronautics classified program with a fixed-price incentive-fee first phase. It booked $950M of additional losses there in the second quarter of 2025, and the cumulative loss recognized stands at about $1.8B.⁶ A Missiles and Fire Control classified contract, with a cost-reimbursable base and fixed-price options that the company disclosed would each be performed at a loss, has produced cumulative losses of about $1.5B.⁶ The Canadian Maritime Helicopter Program has cumulative losses of about $670M, and the company says restructuring talks with Canada “may be prolonged or unsuccessful” and could end in a termination.⁶ These three alone are about $4.0B of losses recognized, in a company that earns $5B to $7B of net earnings a year.

The same year-by-year pattern shows in the segment results. Net profit adjustments on all programs added $1,585M to segment operating profit in 2023, took away $180M in 2024 and added $75M in 2025, but the 2025 figure is after $1.6B of new reach-forward losses.³ Missiles and Fire Control operating profit fell from $1.5B in 2023 to $0.4B in 2024 and came back to $2.0B in 2025, a margin of 13.7%, then 3.3%, then 13.8%. The company attributes the 2025 rise of $1.6B mainly to about $1.4B of reach-forward losses on a classified program in 2024.³ That is what a fixed-price option on a development program does. A business with a true price advantage does not have swings like that. A business with a strong position but an adverse contract structure does.

Finally, the customer itself is changing the terms. The 10-K reports that the President has issued executive orders to overhaul the federal acquisition rules and move to longer contracts. It also reports a recent order that could limit contractors on critical defense contracts from paying large dividends, buying back shares and paying executives freely. The company expects it to be put into effect through a new DFARS clause and to apply only after the Department of War decides a contractor has failed to meet identified performance requirements.³ The government has also said that it would consider progress payments, not performance-based payments, as the baseline for fixed-price contracts, which could slow the cash.⁶ So the buyer is writing rules that touch what the company can do with its own profits, and I weigh that in the capital-return part of the valuation.

IIIThe competitive advantage

The question here is what Lockheed can do that a rival cannot. The rivals are the other primes: Northrop Grumman, General Dynamics, RTX, L3Harris and Boeing’s defense arm. They are not substitutes for one another on any single program, but they compete for the same dollars and, increasingly, for the same new programs.

Scale is the plain difference. Lockheed has $75.0B of sales against $52.6B for General Dynamics and $42.0B for Northrop, and a backlog of $193.6B against $118.0B and $95.7B at the end of 2025.⁵³⁸⁹ It is 2.6 times sales, against 2.2 and 2.3.³ Margins are in the same band, 10.2% to 10.8%, and Lockheed’s was dragged down in 2025 by the program losses. In 2026, the guidance is for segment margin of about 10.65% and consolidated operating margin of about 11.8%, using the middle of the ranges C (calculated).⁴ So its edge is volume and depth, not a higher return on each dollar.

I can find four places where the edge is real, and one where it is being tested.

The F-35 installed base. A fleet of more than a thousand aircraft needs parts, software updates and depot work for 30 years or more, and Lockheed holds the data and the design authority. Aeronautics service sales were $4.5B for nine months of 2026.⁶

Missile production at scale. The Department of War is paying to expand the capacity of this one supplier. The multi-year PAC-3 and THAAD contracts, the munitions frameworks and the revised acquisition strategy all point to a customer who wants fewer, larger, longer contracts, which suits the largest prime.⁶³ Missiles and Fire Control sales rose 17% in nine months at a steady 14% margin.⁴

Integration of many systems. The company says its four businesses work as one company offering integrated solutions across every domain of warfare.⁶ I can’t measure this from outside, and the claims every large prime makes about it are the same. The one visible sign is that it builds both the Next Generation Interceptor and the Aegis combat system, two missile-defense programs that connect to others.³

Financial strength. It has $5.96B of cash, $20.5B of debt, a new $2.25B 364-day credit facility, and its $3.0B five-year facility was extended to August 2031.⁶¹⁴ That is more than most rivals could put behind a fixed-price development bid.

Where rivals can catch up

The test is in new programs, where the government now picks a vendor by competition and pays fixed prices for development. The company’s own filings say the customer is “increasingly” using this structure and that it increases the chance of a reach-forward loss at the start of a contract.⁶ The chief executive’s own comment in the results release names start-ups, new entrants and commercial technology leaders as sources of the technology the company now combines with its own.⁴ I read that as an admission that the old model, where the prime does everything, is under pressure. If the Department of War spreads Golden Dome and other new work across several vendors, Lockheed’s share of the incremental dollar could be smaller than its share of the existing budget. If it concentrates the work, the opposite. I cannot tell which, and I do not know anyone who can.

IVManagement

James Taiclet has been President and Chief Executive Officer since June 2020, Chairman since March 2021, and a director since January 2018.³ Frank St. John is Chief Operating Officer and has been since June 2020. Evan Scott became Chief Financial Officer in April 2025, after a year as finance chief of the Missiles and Fire Control segment and before that as Treasurer.³ Greg Ulmer runs Aeronautics, and ran the F-35 program before that.³ It is a team that grew up inside the company. The chief executive’s tenure covers the whole period in this page’s recent history: the pandemic, the F-35 software delay, the buyback surge of 2021 to 2023 and the program losses of 2024 and 2025.

How they are paid, and what that pays them to do

Mr. Taiclet’s total pay for 2025 was $23.45M, against $23.75M in 2024 and $22.81M in 2023. The 2025 figure is $1.75M of salary, $13.67M of share awards, $5.50M of cash incentives and $2.54M of other pay.¹⁰ The median employee earned $130,614, so the ratio is about 180 to 1 C (calculated).¹⁰ I don’t think the size is the issue. What the pay rewards is.

The 2025 annual bonus counted sales (20%), segment operating profit (40%) and free cash flow (40%). Sales paid 112% of target and segment profit paid 0%, after losses on classified programs cut segment profit; free cash flow paid the maximum 200%, after the compensation committee adjusted it for unplanned pension contributions, tariffs and tax benefits from the 2025 tax act. The overall financial payout was 102.4% of target, and Mr. Taiclet’s annual bonus was $3.99M, 114% of his $3.50M target.¹⁰ Look at what happened. Profit missed badly and the bonus came out slightly above target, because cash held up. That is cash that comes partly from billing customers early and paying suppliers late, and partly from losses that are booked now and paid later.

For long-term pay, 50% of the share units depend on total shareholder return against peers, 25% on return on invested capital and 25% on free cash flow, over three years.¹⁰ For 2023 to 2025 the relative return was below threshold and paid nothing, free cash flow was above target and the return on capital was below target.¹⁰ That part of the system worked: when the share price lagged, the pay lagged. But the committee has changed the design for 2026, and I read it with care. The annual bonus will count financial goals at 50% rather than 70%, with the other half on strategic and operating goals such as on-time delivery and production speed; free cash flow becomes cash from operations; and the long-term units will measure relative return (30%), sales (30%) and cash from operations (40%), so return on invested capital drops out and the long-term awards no longer measure profit.¹⁰ The committee says the aim is to put weight on backlog conversion and customers. The proxy does not tell me what the bonus weights are within its financial half. Sales and operating cash can both be improved by taking on work at poor margins, particularly work that is paid for in advance. I make no charge of bad faith; this is simply what the long-term pay rewards.

Ownership is more reassuring. The chief executive must own stock worth six times his salary, about $10.5M, and counts as his own 78,052 shares plus 32,298 units and unvested awards, 110,349 in all, worth about $71M at $643.15, about 40 times his salary C (calculated).¹⁰¹ All 19 directors, nominees and executive officers together held 307,730 shares and units, about $198M.¹⁰ That is 0.13% of the company. His own money is in, at a size well above the requirement.

Who owns the shares

The proxy lists three holders above 5%, each in 13G/A filings dated January and February 2024, so the figures are old: State Street 37.1M shares (14.9%), Vanguard 22.1M (8.9%) and BlackRock 18.3M (7.4%).¹⁰ Most of the State Street figure, 27.7M shares, is held as trustee for Lockheed Martin’s own benefit plans, not as an investment of State Street’s clients.¹⁰ These are index and trustee holdings. The company has no controlling holder and no founder, and what the owners can really do is limited to the vote on pay and directors. I did not find a named active investor with a verified, dated position in the filings.

What they did with the money

The record over a full cycle is the best evidence. From 2016 to 2025, free cash flow was $55.8B and the company paid $27.1B in dividends and $32.6B in buybacks, $59.7B in all, which is $3.9B more than it earned. Diluted shares fell from 303.1M to 233.5M, a drop of 23%.⁵ Debt rose from $14.3B to $21.7B.⁵ The last five years are in the table.

From 2021 to 2025 the company bought back 55.2M shares for $24.7B, an average of about $447 a share: 9.4M in 2021 at about $435, 18.3M in 2022 at $432, 13.4M in 2023 at $448, 7.5M in 2024 at $493 and 6.6M in 2025 at $455.⁵¹³³ The price is now $643.15, so those repurchases have gained in value. The test is whether the money had a better use. The company paid $24.7B to retire about a fifth of its shares when free cash flow did not cover that and its dividend, while debt rose $9.5B, and in 2024 and 2025 it booked about $3.0B of program losses.⁵³ In 2026 it has made no purchases in nine months, against $2.25B in the same period of 2025, and held cash of $5.96B at 27 September.⁶ The 10-Q does not say why, and I will not guess. The authorization has $8.3B left from year-end 2025, and the company says its capital deployment will continue to be “disciplined and dynamic.”³ One possible reason is that its customer is now writing rules about dividends and buybacks, but the filing does not say so.

The dividend is different. It has risen for 24 years in a row, from $6.77 a share in 2016 to $13.35 in 2025 and $14.40 annualized now, and it cost $3.1B in 2025, 45% of free cash flow.⁵⁶⁴ I count it as a commitment. The other use of capital is the business: capital spending of $1.65B in 2025 and $2.1B to $2.3B guided for 2026, a rise of 33% at the midpoint, for munitions capacity.³⁴ I want to see it return a profit, and the test is the Missiles and Fire Control margin, 14.1% over nine months.⁴ The one large purchase in this period was Sikorsky, which cost $9.0B in 2015, and I did not find another in the cash flow data I opened.¹⁶

VFinancials

Four things in Lockheed Martin’s statements don’t mean what they first seem to. Some years carry one-time items: 2017 net earnings had a $1.9B non-cash tax charge from the 2017 tax act, and 2018 cash from operations was cut by a $5.0B pension contribution.¹⁶¹⁵ Reported profit is helped each year by a pension adjustment of about $1.7B, which I explain below. Losses on programs are expensed when the company sees them and paid in cash over later years, so earnings and cash move apart. And the cash flow in 2026 is lifted by items that will reverse.⁶

With those cautions, the picture is a steady business that grew, then lost margin. Sales rose from $47.3B in 2016 to $75.0B in 2025, 5.3% a year, and in 2026 the guidance is $80.0B to $82.0B.⁵⁴ Operating profit rose from $5.9B to $9.1B in 2021, then fell to $7.0B in 2024 and $7.7B in 2025. The operating margin went from 13.6% in 2021 to 9.9% in 2024 and 10.3% in 2025 C (calculated).⁵ Net earnings went from $6.3B in 2021 to $5.0B in 2025, and earnings a share from $22.76 to $21.49, even after the share count fell 16% in that period.⁵ So the company has grown sales by 12% over four years and earned less. That is the main thing to carry forward.

Why did the margin fall? Two reasons stand out. First, the losses: roughly $1.4B at Missiles and Fire Control in 2024 and $1.6B in 2025 at Aeronautics and Rotary and Mission Systems, about 2% of sales each year.³⁶ Second, cost pressure, with the filings naming inflation, supplier performance and tariffs.⁶ One thing did not cause it: the pension adjustment that adds to operating profit was $1.0B in 2018 and is about $1.7B in the 2026 guidance, so the underlying business has lost more margin than the headline shows.¹⁵⁴ For 2026 the guidance has segment margin at about 10.65%, up from 9.0% in 2025, with no new program losses so far this year C (calculated).⁴³⁶

The backlog is the growth story. It was $96.2B at the end of 2016, $147.1B in 2020, $193.6B at the end of 2025 and $292.0B at 27 September 2026.¹⁶⁵⁶ It rose $61.6B in the third quarter alone, $52.5B of it the PAC-3 award.⁶ Backlog is not sales, and only 23% of the September 2026 figure converts within a year, against 37% at the end of 2025.⁶³ The new award lengthens the book more than it lifts the next four quarters. The company says that about 40% converts in 24 months.⁶ If backlog is a promise of work, this one is a promise stretched over the late 2020s and early 2030s.

What an owner can really take out

Free cash flow, as the company defines it, is cash from operations less capital spending. Stock-based pay is added back in operating cash flow, because it is not paid in cash, but the shares it issues have to be bought back or they dilute the owners, so I take it off. After that, owner cash was $6.6B in 2025 and $30.9B over the five years 2021 to 2025, which is 106% of net earnings C (calculated).⁵

The cash is higher than the earnings for reasons I can name. The company books a program loss as an expense at once and pays for it later, so in 2025 earnings carried $1.6B of losses that had not yet been paid in cash.⁶ Pension recoveries come in as cash through contract prices while the matching expense is lower. And the company receives money in advance: contract liabilities rose $1.2B in nine months of 2026.⁶ Cash will fall below earnings when those turn. The accrued loss reserves tell me how large that is: $394M on the Aeronautics classified program, $1.0B on the Missiles and Fire Control one and $360M on the Canadian helicopter program, $1.75B to pay out.⁶ That is about a quarter of one year’s owner cash, and I take it off in my model.

Now the 2026 number, which is the one the market is looking at. For nine months, cash from operations was $6.9B against $5.3B in 2025. Within that, deferred income taxes added $1.1B (against $72M in 2025), contract liabilities $1.2B (against $38M) and accounts payable $1.5B.⁶ Offsetting them, contract assets, revenue recognized and not yet billed, rose $3.9B.⁶ The quarter is good, and a 2026 free cash flow of $7.9B still cannot be the new normal. The October update raised free cash flow guidance from $7.1B to $7.9B at the middle, and cash from operations guidance by $0.8B.⁴ My base uses a normal year, below it.

The pension

Lockheed has a large defined-benefit pension plan, and the accounting is odd enough that it needs explaining. The government pays for pensions through contract prices under cost accounting standards (CAS). Those recoveries were $1.6B in 2025, and the guidance for 2026 puts CAS pension cost at about $1.7B.³⁴ The company also reports a pension expense for financial statements (FAS), which is much lower. The difference, about $1.685B in the 2026 guidance, shows up as extra operating profit.⁴ That is cash the company collects through prices, and it goes back out when the company contributes to the plan: $860M in 2025 and $990M in 2024, none in the first nine months of 2026, and no required contribution in 2026.³⁶ The plans were $3,885M underfunded at the end of 2025, down from $4,785M, and prepayment credits, contributions above the CAS cost recovered so far, were $1.9B.³ I count the pension through the cash flow, not by subtracting the deficit again, and in my base case the profit from the adjustment shrinks to $1.0B by 2033 as the plans are de-risked.

The balance sheet

Debt was $11.7B at the end of 2021, $15.5B in 2022, $17.5B in 2023, $20.3B in 2024 and $21.7B at the end of 2025, and $20.5B at 27 September 2026 after a $1.2B repayment.⁵⁶ Cash is $5.96B, so net debt is $14.5B.⁶ Interest expense was $1.1B in 2025.³ In August 2026 the company signed a new $2.25B 364-day facility and extended its $3.0B five-year facility to 2031; the 364-day facility has no financial maintenance covenant, and nothing was drawn at signing.¹⁴ I did not open the rating agencies’ documents, so I state no rating.

The balance sheet also carries items that are not debt but behave like it. The 50% stake in United Launch Alliance is carried at $578M, and ULA’s Vulcan rocket “experienced performance challenges” in the first quarter of 2026; in the third quarter the company agreed to guarantee up to $750M of ULA borrowings and says it and Boeing expect to provide more support.⁶ I include the full guarantee in my claims and nothing for the stake. Equity is $10.1B, up from $6.7B at the end of 2025.⁶

The price in numbers

On those numbers the shares trade at 21.1 times 2026 guided earnings, 17 times enterprise value to operating profit, and a free cash flow yield of 5.3% on guidance, with the caveat above that guidance has working-capital help in it. The shares closed at $483.67 at the end of 2025, reached a high close of $676.70 on 2 March 2026, were $509.25 on 30 September and rose 26% in October.¹ The last close is 5% under the March high. Over ten years the price went from $245.11 on 2 November 2016 to $643.15, 10.1% a year for the price alone C (calculated).¹ The shares fell 18% in 2018 and rose 49% in 2019, measured from year-end closes, and in 2025 they were flat while sales rose 5.6%.¹⁵

VIValuation

First, how this goes badly wrong

Before any numbers, I ask what would hurt an owner most. I count three ways.

The fixed-price risk keeps costing real money. The company has lost about $3.0B on programs in two years: roughly $1.4B at Missiles and Fire Control in 2024 and $1.6B in 2025 across Aeronautics and Rotary and Mission Systems.³⁶ Three programs carry $1.75B of accrued losses still to be paid in cash, and the filings say more could come on each.⁶ The F-35 adds a fourth question: its costs on the Lot 20 advance contract began to exceed funding in the third quarter, and the terms for Lots 20 and above are not agreed.⁶ The F-35 is 27% of the company. If the government forces a price that does not cover cost on an aircraft that size, a one-point margin loss on the F-35 is about $0.2B a year.³ This is the scenario where the stock is most wrong. The company cannot walk away from the customer and the customer knows it.

The buyer changes the terms. The government can slow progress payments, delay a budget with a continuing resolution, pass rules on payouts, or shift work to new entrants. It is already doing three of those: the continuing resolution runs to 11 December 2026, the progress-payments policy is under discussion, and an executive order on payouts is being implemented.⁶³ None of them is a disaster alone. The danger is the combination with a $16.9B pile of contract assets, revenue recognized and not yet paid.⁶ If the government pays slowly, working capital eats the cash while earnings look fine.

I pay a full price for a good result. The shares rose 26% in October on a strong quarter, a $52.5B award and a raised outlook.¹⁴⁶ At $643.15 the price already pays for margins that are back above 10.5% and sales that grow 7% to 8% a year. A business with one customer and fixed-price risk, bought after a month like that, is exposed to two things at once: the earnings missing and the multiple falling. The risk lies in what an owner pays.

What an owner can really take out

In 2025 owner cash was $6.6B, which is 4.4% of today’s market value C (calculated).⁵ Guidance for 2026 is $7.8B to $8.0B of free cash flow, or about $7.5B after stock pay, but it has $1.1B of deferred tax and $1.2B of customer advances in it.⁴⁶ In my base case, cash before interest is $5.7B in 2027, $6.8B in 2028 and $8.2B in 2030, after tax, capital spending, $0.5B a year of pension contributions, working capital, $0.4B a year for new program losses and the cash still to go out on the $1.75B already booked. Take off net interest of about $0.8B before tax and owner cash in 2028 is about $6.2B, which is 4.2% of market value E (estimate). That is 12% under the $7.1B of earnings the 2026 guidance implies, and I want to say why.

Earnings carry the pension adjustment of $1.7B, which is cash now but shrinks in my cases. They do not carry the $0.5B of pension contributions, which the company controls and paid at $860M in 2025 and $990M in 2024. And they carry no standing charge for losses on programs, which I cannot leave out when each of the last two years had them. I do not claim that my number is right. I claim that it lies somewhere between $5.5B and $7.5B, and that at this price the market is paying for the top of that range.

What I can’t judge

I can’t judge the F-35 negotiation. The price, quantity and risk-sharing for Lots 20 to 22 are unknown outside the room, and the company’s framework agreement does not give them.⁶ I can’t judge the classified programs. The company discloses cumulative losses, not the designs, the remaining work or the odds of more. I can’t put a number on Golden Dome. The Administration’s request names it, and I could not open a primary document giving its cost or the share that goes to Lockheed, so I have left it out of every case.⁶ I can’t judge how the executive order on payouts will be written. I can’t say what ULA will need. And I can’t say what the Congress will do with the FY 2027 budget, which asks for $1.5 trillion but has not been passed.⁶ Where an input can’t be known, I show a range, and the bear, base and bull differ most on margin and the amount of cash that stays in the business.

How the cases are built

I build each case year by year in US dollars. Sales start at the $81.0B middle of the 2026 guidance and grow at the rate in the table. Segment operating profit is sales times a margin; I add the pension adjustment (FAS/CAS) and take off unallocated costs of $650M a year, which include amortization of acquired intangibles. I charge tax on that operating profit, add back depreciation at 2.05% of sales and take off capital spending, pension contributions, the cash that working capital absorbs as sales grow, a standing allowance for new program losses and the cash still to go out on losses already booked. Stock pay is a cost, so I do not add it back. The value counts the fourth quarter of 2026 (the guidance-implied free cash flow of $7.9B less $5.6B already reported, adjusted for interest and stock pay) and then each full year from 2027 to 2036, with growth after that. Net claims are those at 27 September 2026, so I do not count the cash earned before that date.⁶⁴

I discount at 8.5%: the 10-year Treasury yield of 4.91% on 2 November 2026 plus 3.6 points, because 70% of sales depend on one customer, the program losses are real and I can’t see the next ones, and because the customer can change the terms.⁷ It is lower than the 9.5% I use for Bombardier, which is smaller and more cyclical, and it reflects a larger, longer-lived business.¹⁷

Three comparisons help. My base cash before interest in 2030 is $8.2B, against the $8.4B or so that 2026 guidance implies before interest and after stock pay C (calculated), because 2026 has no pension contributions, no new program losses and the working-capital help I described; my base cash is lower in 2027 by design.⁴ My base sales grow 6.5% a year to 2030 and then 4.5% falling to 3%, which is faster than the 5.6% of 2025 and slower than the 7.4% of the first nine months of 2026 and the 11% of the third quarter, because I haven’t seen the F-35 lots or the budget.⁵⁴ And my base segment margin of 11.0% is 2 points above 2025 and 0.35 above the 2026 guidance, which I think the missile ramp can support if the losses stop.⁴³

Put plainly, the price needs sales of about $116B in 2030, 11% more than my base and $1B more than my bull, or a segment margin of 12.7% held every year from 2027, which is 3.7 points above 2025 and 1.7 above my base. It also corresponds to a discount rate of 7.6%. Any one of these is possible. All of them are above my base, and the price asks for one of them in full.

The margin of safety, and the price against the range

My margin-of-safety price is $379. It is 70% of my base value. At that price an owner would have a 30% cushion against being wrong on margins, program losses and the F-35, and it would be 42% above my bear. It would be about 12 times the 2026 guided earnings of $7.1B and a free cash flow yield near 9% on guidance. The shares were $509.25 on 30 September, 21% below today’s price, and $439.19 at the low close of the last 12 months on 1 December 2025.¹ Neither was a margin of safety by my numbers, but the comparison shows how far the price has travelled in a month.

The price reflects a real debate, and the case for the market being right needs naming. It says that the company is a monopoly supplier on its two largest products, that governments are rebuilding stockpiles of the weapons it makes, that the backlog is the highest ever and the customer wants it to grow capacity, that the program losses are mostly behind it and the new multi-year contracts bring volume, and that the buyer is a government whose FY 2027 request is $1.5 trillion. If all of that holds, a business with a $292B backlog would be priced above a one-customer industrial. My base case grants part of this and my bull case most of it. The case against says the company has lost $3.0B in two years, is arguing with its customer over the F-35 price, has paid out more than it earned and is the one the customer is writing new rules for. The price is 19% above my base and 14% below my bull. The price sits 48% of the way from my base to my bull.

The next best use of the same dollar

The Treasury 10-year pays 4.91% with no risk to the dollars promised.⁷ My base return to October 2036 on Lockheed is 6.6% a year E (estimate). That is 1.7 points more than the bond, with all the risk of a share, and 1.9 points under the 8.5% I require. On this site, Berkshire Hathaway returns 6.6% a year on its base from a price 16% above that base value, and Bombardier 4.5% a year.¹⁸¹⁷ So Lockheed sits level with Berkshire and ahead of Bombardier on those terms, and under what I ask. It differs from the 7.6% above because that is the rate at which my base cash flows equal the price, while this return assumes the value I put on the business in October 2036. Lockheed is a good business that has stopped converting growth into profit, at a price that assumes it will start again.

The bull case needs many things at once, so I multiply them out. A segment margin of 11.8% held for years, above anything in the guidance (about 20% likely, by my guess), no new program loss above $0.5B in four years (35%), the F-35 Lots 20 to 22 settled at a profit (60%), sales growth of 6% to 11% a year through 2030 as the backlog converts (40%), and pension contributions that stay low (70%). Together that is about 1% if they are independent. They are not, since a good budget helps most of them, so I call it 5% to 8% E (estimate). The price sits about halfway between my base and my bull, so it asks for a good part of a case that I give a small chance.

Where this could be in 5 and 10 years

For the base to be right at five years, sales have to reach $109B in 2031 with operating profit of $12.5B, and the value is then $639 a share, about level with today’s $643.15, because five years of cash have gone out and the rest has grown. At ten years the base reaches $735. In the bear, sales growth stays at 2% to 3% and the margin at 10% before losses, and the value at ten years is $331. In the bull the value at ten years is $1074, a return of 11.0% a year. The range is wide because margin and cash conversion do all the work.

VIIThe case for the price

What most people miss

Three things cut in Lockheed’s favour and get little attention. First, part of the bad news is already booked. Three programs carry $4.0B of cumulative losses, and $1.75B of that is accrued but not yet paid, so it leaves as cash in the next three years without touching earnings unless the estimates worsen.⁶ Second, contract assets of $16.9B, up $3.9B in nine months, are work done and not yet billed; when the government pays, the cash arrives, and nine-month cash from operations of $6.9B already absorbed that build.⁶ Third, 40% of sales are cost-reimbursable, including 75% of Space, and 30% are international, where allied governments are rebuilding stocks of missiles Lockheed makes.⁶

The bull case

Lockheed is the prime contractor on the F-35 and builds the PAC-3 interceptor, its backlog is the largest it has had, the customer is paying for more capacity, and the margin hit of 2024 and 2025 may be behind it. The third-quarter release says the quarter gave management confidence to raise its outlook, and the chief executive says pulling existing platforms through with new technology can “drive elevated growth in our financial performance.”⁴ The FY 2027 request is for $1.5 trillion, with $760 billion for procurement and modernization, and it names munitions, Golden Dome and more F-35s; the Department of War has signed multi-year contracts for PAC-3 and THAAD.⁶ If that is real, my bull sales path, 11% growth in 2027 and about 9% a year to 2030, is not extreme. I found no named investor or analyst view on this company that I could open and date, so the bull case here is management’s. The proxy lists State Street, Vanguard and BlackRock as holders in 13G/A filings dated January and February 2024, which records a holding and says nothing of their views.¹⁰

Against reality

If Lockheed converts the backlog on schedule, holds an 11% segment margin from 2028 and books no new losses, the model gives my base, $541. The price of $643.15 is 19% above it. To be worth $643.15, the business needs a segment margin near 12.7% from 2027, or about $116B of sales in 2030 on my base margins, against 10.6% over nine months and $104B in my base. Only my bull, $752, clears the price, and it assumes both: $114B of sales and 11.8% by 2030. The shares trade at 21.1 times guided 2026 earnings and about 22 times 2025 owner cash, and I give my bull about a 5% to 8% chance E (estimate).

What a 10x would take. Ten times $643.15 is $6,431.50, a market value of about $1.49 trillion. At 20 times earnings that needs about $74B of net earnings a year, 10.5 times the $7.1B in 2026 guidance and above the best year in my ten-year record, $6.9B in 2023 C (calculated).⁴⁵ At a 9% net margin it means about $825B of sales, ten times the 2026 guide and close to the $839.2B FY 2026 national defense appropriation.⁶ It needs 26% a year for ten years. The shares rose 10.1% a year in the last ten, from $245.11 to $643.15.¹ I have no sourced base rate for companies this size, and I see no path to this, so I do not model one.

Where I come down

What moved me was the demand: a $292.0B backlog, a $52.5B award and a raised outlook. The price did not move me, because I cannot build a case for $643.15 without a segment margin well above the recent record, and the 2026 cash flow has $2.3B of deferred tax and customer advances in it. What would raise my values is the F-35 lots signed at a price that covers cost, no new program loss for four quarters and a segment margin held near 11.5%.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest, and a test of each.

• The budget. Congress has not passed FY 2027 appropriations. A continuing resolution runs to 11 December 2026, and the company lists the debt ceiling, shutdowns and priority shifts among the factors that could change results.⁶⁴ A stall of two years costs about $41 a share in the model.
• Contract terms. Fixed-price work was 60% of 2025 sales, and the filings say customers continue to move risk to contractors.³⁶ The government may use progress payments rather than performance-based payments as the baseline, which would slow the cash.⁶
• Classified programs. The company cannot describe two of the three loss programs. It says they carry fixed-price phases and that further losses could be material.⁶
• Payout rules. The executive order on contractors’ dividends, buybacks and pay is expected to take effect through a new DFARS clause and to apply only after a contractor fails to meet performance requirements.³ The dividend was 45% of 2025 free cash flow, and my base assumes it continues; I can’t say what the final clause will say.
• Supply chain and materials. The filing names supplier performance, tariffs and rare earth minerals; the company received about $225M of tariff refunds in 2026 and says it does not expect existing tariffs to have a material long-term effect.⁶
• International demand. 30% of nine-month sales were international, and the company is in talks with Canada to restructure the helicopter contract, with trade relations named as a risk.⁶
• Other claims. ULA may need more support than the $750M guarantee, and the pension needs cash if markets fall; the plans were $3.9B underfunded at the end of 2025.⁶³
• People. About 19% of the workforce is covered by collective bargaining agreements, and the company hired nearly 10,800 people in 2025 to grow.³

What a bad year looks like in the model

The surprise is how little a stalled budget costs if it ends: two slow years take $41 off the base. What matters more is anything permanent. Margins stuck at 9.0% cost $124 a share, and a new $1.5B loss every year costs $94. A bad decade of fixed-price work is the risk that matters.