Chapter 9

Backlog Durability

Mission Technology Solutions is the larger company after the spin and the source of 57% of KBR's revenue, so how sticky its government order book is weighs heavily on the case. The book is more layered than "stable government annuity": roughly two-fifths is long-dated U.K. infrastructure work that never recompetes, and 91% of 2026 revenue is already under contract — but most U.S. work is cost-reimbursable that recompetes every few years, and HomeSafe's abrupt termination shows even won government work can vanish. The read here is a narrow, incumbency-based moat.

The pricing of the government half is set out separately in The Government Half; this chapter asks the prior question — whether that revenue holds together when contracts come up for renewal.

What the order book is made of

KBR's total backlog was $16.9 billion at the start of fiscal 2026, of which about 17% sits inside equity-method joint ventures, and the company expects to convert roughly 36% of it into revenue within fiscal 2026 [1]. A book that turns over roughly a third each year implies an average life close to three years for the work that recognizes as revenue — short enough that renewal risk is a live, recurring event, not a distant one.

The order book splits into two very different kinds of contract. Private finance initiatives (PFIs) — the Aspire Defence and U.K. Military Flying Training System programs — are long-term, predominantly fixed-price U.K. contracts with terms of 15 to 35 years covering the construction and lifecycle management of government assets [2]. These do not recompete during their term; they are the annuity. The rest — cost-reimbursable and time-and-materials work performed mostly for U.S. defense, intelligence and civilian agencies — is the part that periodically goes back out to bid.

Source: contract-type mix from the FY2025 10-K [3] and FY2024 10-K [4]; June 2022 (then-Government Solutions segment) from the Q2 2022 earnings release [5].

At the start of fiscal 2026, 39% of backlog was PFIs, 24% cost-reimbursable, 20% time-and-materials and 17% fixed-price [3]. The PFI share has been stable near 36–39% for three years. Because every one of KBR's PFIs sits inside MTS, and because the fixed-price bucket is largely STS engineering work, the annuity share within the government segment is higher than the 39% headline — comfortably over half of MTS's own backlog. The remaining cost-reimbursable and time-and-materials work — the recompete-exposed core — is roughly 44% of the consolidated book.

Two qualifiers keep this from being over-read. First, the reported "backlog" excludes $6.3 billion of priced option periods the customer has not yet exercised, and only 40% of MTS backlog is actually funded once PFIs are stripped out [6]. The growth the company advertises in its $19 billion "backlog and options" figure is unexercised optionality, not committed work. Second, the company's headline book-to-bill deliberately excludes the U.K. PFIs, the Plaquemines LNG project and the HomeSafe joint venture [17] — so the ~1.0x order intake investors watch is the recompeteable core, and MTS's fell to 0.5x in the fourth quarter of fiscal 2025 [18].

Diversification is the first line of defense

Within the government segment, no single program dominates. KBR's largest customer disclosure names the U.S. Army, Navy, Air Force, Space Force, Missile Defense Agency, the National Geospatial-Intelligence Agency, the National Reconnaissance Office, NASA and the U.K. government — and states that beyond the U.S. government (57%) and U.K. government (9%) in aggregate, no other customer represented 10% or more of consolidated revenue in any period shown [8]. A recompete loss on any one vehicle is survivable; the concentration is at the level of the customer (the U.S. government), not the contract.

The internal mix has also been shifting toward the more durable work. Defense and Intel revenue rose from $2.50 billion in fiscal 2023 to $3.18 billion in fiscal 2025, while the Readiness and Sustainment line — the contingency, logistics and base-support work most exposed to troop levels and the European drawdown — fell from $1.50 billion to $1.28 billion; Science and Space held flat near $1.1 billion [9].

Source: KBR segment revenue disclosures, FY2023–FY2025 10-K [9]. Fiscal 2025 is the year ended Jan 2, 2026.

Sources: backlog mix and 2026 revenue coverage from the FY2025 10-K and Q1 FY2026 call [3] [7]; protest figure from the Q3 FY2025 call [10].

The recompete mechanism, and what it costs

KBR states the risk in its own words: "Many of our existing contracts must be recompeted when their original period of performance ends, representing opportunities for competitors to take market share away from us." The same passage warns that even a won award "may be subject to a lengthy protest process" that can "delay the recognition of sales and defer underlying cash flows" [11]. This is not boilerplate for KBR; it is what fiscal 2025 actually looked like.

By the third quarter of fiscal 2025, KBR disclosed $3.0 billion of contracts it had won still frozen in protest — up 50% in a single quarter, with the largest addition a classified program in the Indo-Pacific command — and management said conversion "remains uncertain" in the government-shutdown environment [10]. Winning the bid is not the same as booking the revenue; protests sit between the two, and the gap was widening.

The sharper illustration is HomeSafe. KBR's 72%-owned joint venture had been selected as the exclusive provider of household-goods moves for the U.S. armed forces — precisely the kind of long-duration, sole-award win that is supposed to anchor a backlog. In June 2025, U.S. Transportation Command "unexpectedly terminated" that role; by the start of fiscal 2026 all of HomeSafe's operations had ceased and the business was carried as discontinued operations [12]. KBR's own risk factors now cite HomeSafe as the working example that "following contract award, we may also encounter significant expense, delay, contract modifications or even contract loss or termination" [13]. A won, exclusive, multi-year government contract went to zero inside a year.

A quieter version of the same risk runs through NASA, where KBR has worked for more than 60 years. Management flagged in early 2026 that the agency's administrator had signaled interest in in-sourcing certain core workforce competencies, an outcome already reflected in the 2026 outlook as a change in the mix of work [7]. Recompete is not the only way government revenue leaves; the customer can also decide to do the work itself.

The incumbency that holds it together

Against those losses sits real evidence of stickiness. The clearest is that KBR keeps winning its recompetes: its single largest recompete of fiscal 2025 was the NASA astronaut health-and-human-performance contract, re-awarded to KBR at a $2.5 billion ceiling plus $1 billion of options — though management was candid that the booking value was below $1 billion, "more consistent with the current run rate" [14]. That single award captures the whole tension: KBR retains the mission, but the government sets the price, and the run-rate can reset lower even on a win.

The incumbent's advantages are genuine but ordinary for the industry: cleared personnel, decades of past performance, and cost-reimbursable structures under which the customer bears most of the execution risk and rarely switches a functioning provider mid-mission. Management describes the bulk of the MTS U.S. portfolio as "mission essential operational work" on "well-funded multiyear programs," which is why 91% of fiscal 2026 revenue was already under contract entering the year and why recent quarters added a LOGCAP extension and a Department of Transportation recompete that "extended a long-standing partnership" [7].

The advantage is not a wall, and KBR's own disclosures show why. It performs mostly under cost-reimbursable contracts — the same low-risk, recompete-exposed model a pure government-services peer such as Booz Allen Hamilton runs, where 59% of revenue is cost-reimbursable [15]. And KBR is perfectly capable of taking share the same way: it recently won a seven-year, $176 million space-surveillance program by displacing the incumbent [16]. Incumbency slows competition; it does not stop it, in either direction.

The read

MTS is a narrow moat, not a fortress and not a house of cards. The durable core is real and quantifiable: an annuity of 15-to-35-year U.K. PFIs that is more than half of the segment's backlog, a customer base diversified across every major defense and civil agency with no single program above 10% of revenue, high past-performance win rates on recompetes, and 91% revenue visibility a year out. The offsetting evidence is equally concrete: most U.S. work must be re-won every few years, $3 billion of awards can sit stranded in protest, a customer can in-source or — as HomeSafe showed — simply terminate, and even retained work can reprice lower.

On balance the segment should be valued as a stable-but-slow government-services franchise carrying a genuine recompete discount, which is consistent with the low end of the peer multiple range applied in The Government Half rather than a premium. What would move the read toward the durable end: firm backlog — not options — resuming growth, the recompete-core book-to-bill holding above 1.0x for several quarters, the $3 billion of protested awards converting to revenue, and the NASA in-sourcing staying contained to mix. What would move it the other way: a second HomeSafe-style termination, a lost anchor recompete, or the funded share of backlog continuing to erode. The Form 10 that splits the company will be the first place to see MTS's standalone backlog laid out on its own — and the first clean test of how much of this order book the market should pay for.