KTOS - Kratos Defense: Market Reads Prepaid Revenue as Cash Burn

What Kratos Actually Builds

Start with the company, because most write-ups collapse it into one airframe and then argue about that airframe. Kratos was created on January 1st 2008 out of the wreckage of a telecom-services business, with no revenue and some cash from asset sales. Eric DeMarco has run it since 2004 and Deanna Lund has been chief financial officer for exactly as long, which by defence standards is a geological era. Today it employs more than 5,200 people across 13 countries and reports in two segments.

Kratos Government Solutions is the larger segment and where growth lives: $379.7 million of June-quarter revenue, 82.8% of the total, growing 22.0% organically. Inside it sit six businesses that have almost nothing to do with each other operationally and everything to do with each other strategically. Defense rocket and hypersonic support grew 50.2% organically. Turbine technologies grew 43.3%. Microwave electronics, headquartered in Israel and working with Israel Aerospace Industries, RAFAEL and Elbit, grew 29.5%. Space, training and cyber grew 8.7%. Add radar sustainment, C5ISR mobility hardware and directed-energy integration and you have a portfolio of about thirty programs.

Unmanned Systems is the famous half and the smaller one: $79.1 million, 17.2% of revenue, up 8.1% organically. XQ-58A Valkyrie lives here, alongside Mako, Firejet and those BQM-167 and BQM-177 target drones nobody writes about, which steadily bankroll the tactical work. Kratos produces 165 to 170 jet drones a year, which DeMarco believes makes it the largest jet-drone producer in the world outside Ukraine.

That sentence is the entire operating philosophy, and it explains a financial statement that otherwise looks broken. Kratos deliberately does not chase the exquisite end of the market. It designs for mass manufacture at a price point the large primes structurally cannot match, funds the development itself, and sells either as a prime where it thinks it wins or as a merchant supplier underneath Lockheed Martin, Northrop Grumman and Boeing where it does not. DeMarco described the future force structure on August's call as a barbell: enormous quantities of attritable munitions and drones on one side, a handful of exquisite systems on the other, very little in between. Kratos sells into both ends and owns neither middle.

Be concrete about how a dollar actually reaches this company, because contract structure explains both the thin margin and the ugly cash flow. Some 67% of second-quarter revenue came from fixed-price contracts, 29% from cost-plus and 4% from time and materials. By customer it was 69% United States federal, 20% foreign and 11% commercial, state and local. Fixed price on physical hardware caps gross margin at 21.8%. A software-led defence name would post 60% or 70%. Kratos is a manufacturer, it carries the overrun risk itself, and no amount of narrative changes that arithmetic.

What it buys with that risk is the design-in. Hold on to that mechanism for the rest of this study. Kratos funds development out of its own pocket, gets its engine or its subsystem specified into somebody else's missile or aircraft, and then earns on every unit that platform ever ships. Development is the entry fee. Production is the annuity. A company running that model will always look worst in the years when it is paying entry fees for programs that have not started shipping. That is precisely the year you are looking at. Once you see the business this way, the cash-flow statement stops being a verdict and becomes a calendar.

Green channel narrows to a thread before it ever reaches net income. On trailing twelve-month figures $1.3 billion of revenue produces $307.9 million of gross profit, and then an operating-expense block of $280.2 million consumes almost all of it, leaving $27.7 million of operating income and $22.0 million of net income. Selling and administrative cost is $240.2 million of that $280.2 million, so the binding constraint is overhead. You are looking at a company carrying a corporate structure sized for a business two or three times its current revenue. Read as a static picture it looks like failure; read as a leading indicator it tells you where operating leverage comes from if revenue keeps compounding into an expense base that is already built. Source: GF, Company filings

One more structural point, because it decides how the next three years look. DeMarco runs six autonomous divisions, each with a long-tenured president holding full profit-and-loss responsibility and its own business-development function. None of those presidents came from a traditional prime. He is explicit about why: "At Kratos, if there are ten decisions to make, we make ten decisions. Eight are good, two aren't. But at least we made ten decisions and didn't miss an opportunity waiting for fifteen committees." In an industry where threat evolution outruns procurement, that is the actual moat. Not a patent. A decision cycle.

Metrics at a Glance

Numbers below are as reported for the quarter ended June 28th 2026, with market data to the August 17th close. Read them as a pair of opposing columns, because that is what they are.

Sit with the last figure for a second, because it governs everything about how this position should be sized. A stock that moves 10% in an average week does not respect a tight stop, and any risk framework built for a 3% mover will be taken out by noise long before the thesis is tested. High risk on the header row is not a formality here.

Anatomy of a 30% Beat

Kratos told the market in May to expect $400 to $410 million of revenue, an operating loss of $6 to $8 million and adjusted EBITDA of $30 to $35 million. It delivered $458.8 million, an operating loss of $1.6 million and $38.2 million. Clearing the top end of your own revenue guide by $48.8 million goes well past a rounding surprise. A company that misjudges its own conversion rate by almost twelve percent in one quarter is telling you how fast the backlog has started to turn. Shares rose 30.4% over the month that followed the print.

Widening green bars on the right are the story. Four consecutive quarters where actual earnings cleared consensus, each gap larger than the one before, ending with $0.21 against a $0.14 estimate in June. What makes that more than a good run is the blue line underneath it: analysts left their June-quarter number at $0.14, barely above the $0.135 they had carried for March, while the actual went from $0.16 to $0.21. Estimates are anchored and the company keeps stepping over them. Left panel shows why that matters. Consensus already models earnings near tripling from $0.55 in 2025 to $1.47 in 2028, yet the error bar on that final year runs from $1.25 to $2.16. Direction is agreed. Magnitude is not, and a 73% spread on the terminal year is precisely where this valuation argument lives. Source: Barron's, consensus estimate data

Now do the uncomfortable arithmetic. Reported net income of $4.4 million exists because $10.2 million of interest income on the February cash pile more than covered the operating loss. Strip that and Kratos lost money at the operating line on record revenue. Adjusted earnings per share of $0.21 rests on adding back $16.3 million of stock compensation and $24.5 million of depreciation and amortisation, the latter inflated by the Nomad and Orbit purchases. Both numbers are legitimate under their own definitions. Neither is evidence of a profitable business yet.

Most coverage stopped here. What follows is where it gets interesting, because guidance moved in two directions at once and almost nobody separated them.

Two bars point the same way and tell the whole story: capital expenditure fell $30 million and operating cash flow fell $25 million in one revision. Spending did not shrink by a dollar. It migrated from the investing section of the cash-flow statement into inventory and vendor prepayments in the operating section, which is exactly why free-cash-flow use came out unchanged at $85-105 million. Any bear reading only the operating-cash-flow line saw deterioration where Kratos had merely changed the address of one cheque. Source: SimplyWallSt, Company guidance

Two negatives survive that reframing. Full-year adjusted EBITDA guidance of $174.5 million at midpoint landed below the $177.7 million analysts carried, so that raise was revenue-only. And a strengthening Israeli shekel cost $2.5 million of EBITDA in the quarter alone, with $5 to $7 million expected for the year, because Kratos bills more than 700 Israeli employees' work in dollars and pays them in shekels. Lund called it "the single most significant headwind from a margin perspective." That is real money and it is not going away this year.