Recompete Durability

Recompete Durability

Maximus owns no contracts outright. Every dollar of revenue sits inside a government agreement that runs for a fixed term and is then reopened to competitors, so the franchise is a retention business, not a set of annuities. The evidence supports a durable, retention-based moat: management models — and has held for five years — a rebid win rate near 90% plus or minus three points [1], the work is anchored to mandatory entitlement programs that survive administrations, and the contracting model gains volume when eligibility rules tighten [2]. The binding constraints are concentration and the veterans-exam contract that closes at the end of December 2026 [3].

Sources: rebid win rate [4]; concentration [5]; DOGE impact [6].

The moat is retention, not lock-in

Maximus's contracts run for a fixed number of years, may be extended if the agency elects to, and are then opened to competing bidders with no guarantee of renewal [7]. Almost all of them also carry a "termination for convenience" clause that lets the customer end the work at any time [8]. There is no contractual monopoly here; the advantages the company itself lists are experience-based — decades of program-operations expertise, the ability to convert program rules into compliant technology, financial strength, and a familiarity with government bidding "which can be difficult for new or inexperienced competitors" [9].

What turns that into a moat is the win rate on rebids. Management describes the target as a 90% plus-or-minus-three rebid win rate, notes it "has been the case over the last five years," and reports it holding "at historic levels near 90%" through the FY2024 slow patch [10] [11]. An incumbent that keeps roughly nine of ten recompetes it defends compounds an installed base rather than treading water — and the switching cost a customer bears is real: in one state, Maximus staff are trained across five separate legacy systems to deliver a single program, the kind of integration a challenger cannot replicate on a bid timeline [12].

The metric investors watch for renewal health — book-to-bill — is deceptively noisy, and the noise is the first thing to understand about this business. Signings cluster in whatever year the large multi-year contracts happen to come up for rebid, so book-to-bill swings widely without the underlying franchise changing. Four quarters in FY2022–FY2023 ran above 2.0x on a wave of rebids; FY2024 then fell to 0.4x because so little was up for renewal, before recovering toward 0.9x in FY2025 [13] [14].

Sources: TTM book-to-bill at each period end — Sep-2023 1.2x [15]; Sep-2024 0.4x [16]; Mar-2025 0.8x [17]; Sep-2025 0.9x [18].

A book-to-bill below 1.0x for two straight years is not, in this case, a signal that the franchise is shrinking; a normal year with rebids evenly distributed would produce close to 1.0x from renewals alone, and the low readings coincide with a thin rebid calendar rather than lost work [19]. The genuine caution is that this same lumpiness front-loads risk: the years with the heaviest rebid volume are the years the win rate matters most, and the next such year is now approaching.

Why the programs themselves persist

Retention only matters if the underlying programs keep running, and Maximus's portfolio is deliberately built around work that does. Its largest lines administer Medicaid eligibility, Medicare and Affordable Care Act enrollment, and veterans' disability benefits — mandatory-spending entitlement programs with bipartisan support that "stand the test of various presidential administrations" [20]. A significant majority of the work is classified "essential" by the government, so it continues even through a federal shutdown [21].

The more important structural feature is how Maximus is paid. Most state Medicaid contracts pay for the volume of activity performed — eligibility checks, redeterminations, enrollments — not a flat fee per enrolled member. That inverts the usual political risk: a policy that cuts the number of recipients does not necessarily cut Maximus's volume, because verifying who still qualifies is itself billable activity [22]. The 2025 One Big Beautiful Bill Act sharpens that dynamic. It raises the redetermination frequency for the Medicaid adult-expansion population from annual to every six months and adds an 80-hour monthly work-and-community-engagement requirement, both effective December 31, 2026 — and it explicitly bars health insurers from administering the engagement requirement on behalf of states, leaving the work to third-party operators [23]. It is a plausible FY2027-and-beyond volume tailwind, though the same filing cautions that states may request delays of up to two years and that final rules are still pending [24].

The DOGE stress test

The 2025 federal efficiency drive was the sharpest real-world test of this moat, because it hit the whole government-services sector at once. Maximus's measured exposure was small: management put the direct impact of Department of Government Efficiency actions at roughly $4 million of FY2025 revenue — a rounding error on a base above $5 billion — confined to a handful of small contracts, some already scheduled to end [25].

The contrast with a same-sector peer is the cleanest evidence that the insulation is structural, not luck. ICF International — screened as a Maximus peer but running a discretionary advisory, IT-modernization, and consulting model rather than mandatory-benefit operations — disclosed that it "received contract terminations and temporary stop-work orders" from DOGE in early 2025, and its share of revenue from the U.S. federal government fell from 54% to 43% in a single year [26]. Discretionary consulting was cuttable; running the eligibility engine for a mandatory program was not.

Sources: Maximus federal revenue ~55% and ~$4M DOGE impact (0.07% of revenue) [27] [28]; ICF federal revenue 54%→43% (an ~11-point / ~20% relative decline) [29].

The same episode also exposed the moat's vulnerability, and the clearest case study is the CMS contact-center contract — the 1-800-Medicare and federal-marketplace operation Maximus has run since 2018, worth about $6.6 billion over its full option life and roughly $600–700 million of annual revenue, under 15% of the company [30]. In 2024, CMS moved to recompete it early, attaching a labor-harmony requirement Maximus argued was unprecedented for a services contract of this kind. Maximus contested the move at the Government Accountability Office and then the Court of Federal Claims; after the election, the government canceled the procurement, and the existing contract is now expected to run through 2031 on its option periods [31]. The outcome protected a fifth of federal revenue — but it turned on a legal fight and a change of administration, not on incumbency alone. A recompete can be used as a policy lever, and the defense is not automatic.

Concentration and the December 2026 recompete

The reason these single events matter is concentration. About 60% of revenue comes from the ten largest contracts, and roughly one-fifth from contracts with a single federal agency [32]. With the CMS contact center resolved and the VA medical-disability-exam successor contracts reawarded from January 2025 [33], the largest identified recompete now sits ahead: the veterans-exam work — anchor of the 2021 Veterans Evaluation Services acquisition — runs through December 31, 2026 for all vendors, and the VA has not yet published a rebid timeline. Management calls everything else "normal recompete cadence" [34].

Maximus enters that contest as a strong incumbent — it staffs the work substantially with veterans, has invested in AI to cut case-preparation time, and expresses optimism about the outcome — but "optimism" is not a 90% win rate on a specific, large, single award, and the company concedes it cannot yet say whether the VA will extend the contract while it recompetes [35]. This is the identifiable event most capable of dislodging the durability thesis in the next two years, and it is worth watching for a disclosed timeline and any bridge-contract extension.

Revenue a decade out

For a reader whose test is whether cash flows still stand a decade out, the durability question comes down to whether revenue will be higher in year ten than today. The historical base rate is encouraging — revenue has risen from about $2.45 billion in FY2017 to $5.43 billion in FY2025, roughly doubling across two administrations and a pandemic, with only one modest down-year in FY2018 [36]. The near-term reads flatter: FY2026 revenue is guided to $5.2–5.35 billion, level-to-slightly-below FY2025, even as adjusted EPS is guided up 14% on margin gains [37].

Source: derived from reported financials, FY2017–FY2025 10-Ks (FY2025 revenue $5,431.3M on the Consolidated Statements of Operations [38]); FY2026E is the midpoint of management guidance [39].

The weight of the evidence points to revenue being higher in a decade than today: the programs are mandatory and bipartisan, the activity-based model turns eligibility tightening into volume, OBBBA adds work from FY2027, and a ~90% rebid win rate compounds the base. The strongest facts against that read are the FY2026 flat-to-down guide, the concentration of a fifth of revenue in one agency, and the reminder — from the CMS episode — that a determined customer can force a recompete and win it takes a fight. What would change the read: the veterans-exam recompete lost or materially de-scoped, the rebid win rate slipping below the high-80s for more than one cycle, or a structural shift from activity-based to per-member pricing that removes the volume hedge. On the durability at the center of the investment question, the moat is real and measured, but it is a retention moat that has to be re-won on schedule — not an annuity.