Science Applications International Corporation (NASDAQ: SAIC) — No Moat, a New Sheriff, and a 12% Free-Cash-Flow Yield

Independent Research Note Report date: July 21, 2026 | Price (July 20, 2026 close): $112.66 | Sector: Technology / Government IT Services | HQ: Reston, Virginia Fresh initiation. (Price per AZI/FactorsToday; a third-party feed’s $115.66 print for the same session was internally inconsistent and is rejected.) Fiscal year ends ~January 31 — “FY2026” ended January 30, 2026; the company is currently in FY2027.

The analytical body of this report carries no position: no BUY/SELL recommendation and no price target appears anywhere outside the clearly-labeled Kimi’s Take block immediately below. Valuation in the body is discussed only as embedded expectations and scenarios. This article is the author’s independent analysis and general information, not investment advice.

⚡ Kimi’s Take

Kimi’s own subjective opinion — the author’s independent view, offered as general information only; it is not investment advice. The analytical body (Sections 1–15) below carries no position and no price target — the sole exception is this block.

Verdict: speculative BUY — accumulate on weakness toward ~$100–105, hold what you buy through the Vanguard decision. Conviction: medium. Valuation zone: fair value ~$105–140 per share (10–12x EV/adjusted EBITDA on the ~$725M FY27 guided base, cross-checked by a 10–12% FCF-yield band on >$600M of guided free cash flow). At $112.66 the stock sits just above the base case (~$108) — the margin of safety is decent, not fat, so stage the entry; the bear case is ~$75 (8x) and the bull case ~$141 (12x).

Let me be honest about what this is, because the two halves of the story pull hard in opposite directions. The business is genuinely mediocre: SAIC is the smallest, lowest-margin, weakest-positioned of the Big-5 cleared government-services primes, in a low-moat industry that re-competes everything every 3–5 years — and it proved the point by losing a ~$1.4B Army Corps recompete as incumbent to Accenture. Revenue has gone nowhere for four years and is guided to shrink again in FY27 (organic −2% to −4%). ROIC of ~12% is roughly zero economic profit. If you stopped there, you would pass. But the equity is a different animal. The market is pricing ~$112 for a company guiding to >$600M of free cash flow — a ~12% FCF yield that underwrites roughly −2% to 0% perpetual growth — while a new CEO (Jim Reagan, the ex-Leidos CFO the board installed after firing the growth-strategy CEO in October 2025) has delivered two consecutive beat-and-raise quarters, a record 11.6% EBITDA margin, a raised FY27 guide, and a deliberate shrink-to-quality pivot out of the commodity enterprise-IT book where every big recompete loss occurred. The cash-return machine is real and accelerating: ~28% of the share count retired since FY2020, ~$400M/yr of buybacks (~9% of the market cap annually at these prices), and — the signal I weight most — a one-sided insider tape: 37 open-market buys versus zero sales in 18 months, including Reagan himself buying at ~$100 in December, seven weeks after taking the interim job. The February guidance massacre (−16% in a day, to $80.50) looks, with two clean prints behind it, like the capitulation reset that marks a floor.

The framing is contrarian/value with early tape confirmation — a falling knife that has stopped falling, not a quality compounder and not a one-way street. The factor data supports that: SAIC still carries a negative momentum loading (quant screens haven’t caught the turn), zero Quality loading, 36.6% idiosyncratic vol — an event-driven value stock rallying inside a friendly value regime, already the strongest name in the services complex over 3/6/12 months. What keeps this “speculative” rather than high-conviction: the entire EPS story is financial engineering (buybacks + a collapsing tax rate) on a flat pre-tax profit base; the Vanguard/State Department recompete (~$250M/yr, above-average margin, bidding as the $10B EVOLVE vehicle) is a pending coin-flip that could repeat the February playbook; leverage re-ticked to ~3.7x EBITDA; and peers LDOS/BAH are statistically even cheaper against their own histories (1st–15th percentile vs SAIC’s 39th) if you prefer the complex without the turnaround-specific risk. What would flip me more bullish: Vanguard retained (or EVOLVE won) plus a third consecutive beat-and-raise at ≥10.3% margins — that forces the re-rate toward ~$141. What would flip me bearish: a Vanguard loss, or any quarter where the margin gains prove to be the one-time items (the $12M venture gain, the ~$70M tax benefit) rather than the cost program — that re-opens the ~$75 bear case. Tag: “The worst house on a cheap street — with a new contractor, cash in the walls, and the price of a teardown.”

📈 Stock Price Action — Five-Year Event Map

SAIC’s five years are a round trip: ~$81 in mid-2021, a defense-cycle recovery to a $152.64 intraday peak the day after the November 2024 election, a −47% collapse through the DOGE era to $80.50 in February 2026, and a ~+40% repair to $112.66 at the July 20 close — −25% off the high, +40% off the low, above all three EMAs (21/50/200: $111.65/$107.44/$102.35) for the first time in this cycle. Beta is 0.40, but realized vol runs 30–40% with repeated ±10–16% earnings-day gaps: this is an idiosyncratic, event-driven stock (~85% of variance is stock-specific), not a market proxy.

Cycle narrative. The pre-2024 arc (rows 1–4) is a cyclical defense-services name riding the post-invasion budget upcycle and a new-CEO honeymoon to a peak that, in hindsight, marked the exact top of both the cycle and the multiple. Row 5 is the regime break: the November 2024 announcement of the Department of Government Efficiency triggered an indiscriminate de-rating of every government-services name — Leidos, Booz Allen, KBR, CACI and SAIC all fell, while hardware prime GD never de-rated at all; the services-vs-hardware split is the defining sector fact of the period. Rows 6–7 are where SAIC’s company-specific problems compounded the regime: three consecutive problem prints, a guidance cut, a fired CEO, and finally the February 11, 2026 pre-announcement — FY27 revenue cut to $7.0–7.2B on two lost recompetes and procurement delays — that capitulated the stock to a four-year low of $80.50. Row 8 is the repair: the board made the ex-Leidos CFO permanent, the December and June prints both beat-and-raised (record margins), insiders bought the lows, and the stock reclaimed all its moving averages. The honest read of the map: the crash was roughly half regime (sector-wide, arguably overdone) and half self-inflicted (revenue erosion that was SAIC’s alone) — the recovery to date has priced the regime repair; the self-inflicted half is what the rest of this memo adjudicates.

1. Executive Summary

Science Applications International Corporation is a ~$7.3B-revenue U.S. government technology-and-mission-services contractor — systems engineering, IT modernization, cybersecurity, space and intelligence support, training and simulation, logistics — deriving 98% of revenue from the U.S. government (~52% Department of War, ~46% intelligence community and civilian agencies), with ~23,000 employees (a majority holding clearances), ~1,700 active contracts, and a $22.9B backlog. It is the smallest and lowest-margin of the Big-5 cleared primes (behind Leidos, GDIT, Booz Allen, and CACI), with FY2026 EBITDA margin of 9.7% and ROIC of ~12% — a structurally thin-margin, recompete-exposed, single-customer business. This memo’s industry verdict is that federal services is a low-moat, competed market; its business-quality verdict is that SAIC holds at best a narrow, eroding-in-places advantage — the weakest of the Big-5, with demonstrated incumbent losses on marquee recompetes.

That is the bad news, and it is fully priced. The investment case rests on four harder facts. First, cash: FY2026 free cash flow was $577M (1.6x net income) and FY2027 guidance is >$600M — a ~12% yield on the ~$4.9B market cap, which embeds roughly zero perpetual growth. Second, the reset: the board fired CEO Townes-Whitley in October 2025 after two years of decline and installed director Jim Reagan — the former Leidos CFO — who cut five business groups to three, launched $100M of cost actions, and has since produced two consecutive beat-and-raise quarters, a record 11.6% Q1 FY27 EBITDA margin, and a raised FY27 guide (adj. EBITDA $720–730M, adj. EPS $9.90–10.10). Third, the capital-return engine: ~28% of shares retired since FY2020, ~$400M/yr of buybacks (including ~$175M in Q1 FY27 alone at ~$92), a 1.3% dividend, and incentives that pay on EBITDA, free cash flow, and relative TSR — with zero payouts on missed legs. Fourth, the insider tape: 37 open-market purchases versus zero sales over 18 months, clustered at $95–117, including the incoming CEO buying at ~$100 in December 2025.

The balance of the memo argues this is a cheap, cash-generative, no-moat business mid-turnaround — where the tension is not “good company vs. bad company” but “eroding revenue base vs. per-share compounding machine.” The three numbers that decide it: whether the revenue decline stops after the ~$400M FY27 recompete headwind lapses; whether margins hold ≥10% after one-time items wash out; and the outcome of the State Department Vanguard recompete (~$250M/yr, above-average margin, pending as the $10B EVOLVE vehicle) — the single largest swing item in the story. Valuation is ~10.2x TTM EV/EBITDA, ~11.3x FY27 guided EPS, and a 39th-percentile own-history composite (9th percentile on P/E) — cheap against history, though notably not the cheapest name in its own de-rated complex (Leidos and Booz Allen screen cheaper still). This memo takes no position and sets no price target outside Kimi’s Take above.

2. Business Overview

What SAIC does. SAIC is a pure-play U.S. federal services contractor. Its offer is cleared people plus mission knowledge: it designs, integrates, modernizes, and operates the technology and engineering systems its government customers run on — enterprise and mission IT, cybersecurity, cloud and data, software and digital engineering, space systems support, training and simulation, and lifecycle logistics. It does not build platforms; it makes platforms, networks, and agencies work. Representative franchises: the $1.8B Army CCDC AvMC software-lifecycle program, repeated seven-year space/intelligence SETA (systems engineering and technical assistance) recompetes, the $928M HOPE 2.0 Air Force TENCAP contract, the $1.4B COBRA task order, Navy propulsion-test and C4ISR work, FAA task orders, Treasury T-Cloud, and the State Department Vanguard enterprise-IT program.

How it makes money — contract economics. Revenue is ~62% cost-reimbursable, ~22% time-and-materials, and ~16% firm-fixed-price. That mix is the single most important structural fact about the P&L: cost-plus work caps margins (the government pays cost plus a negotiated fee) and, as CFO Natarajan put it, is “very hard to differentiate on” — it is where SAIC’s big recompete losses occurred. Fixed-price work is where margin expansion lives: on FFP contracts SAIC keeps the productivity gains from automation and AI, which is precisely why the Civilian segment (more FFP/T&M) earns 15.5% adjusted operating margins versus 10.0% in Defense & Intelligence. The business is people-economics: ~$316K of revenue per employee, gross margins of ~12%, and SG&A discipline determining whether EBIT margins land at 7% or 9%.

Segments and customers. Two reportable segments — Defense & Intelligence (D&I: $5,581M FY2026 revenue, −3%, 8.6% adj. operating margin) and Civilian ($1,681M, −4%, 13.6%) — over an internal structure that was just consolidated from five customer-facing business groups to three (effective the first day of FY2027, part of the Reagan reset). Customer concentration is total: 98% U.S. government, about half Department of War (Army-heavy) and half IC/civilian (NASA, DHS, State, Treasury, FAA, GAO). The company cites ~1,700 active contracts/task orders — granular revenue, no single-program death risk, but constant recompete churn: the typical vehicle is a 3–5-year task order under a multi-award IDIQ/GWAC, after which the work is re-bid.

Revenue visibility. Backlog is $22.9B (~3.2x forward revenue), of which only $3.7B is funded — the unfunded $19B is negotiated-but-not-yet-appropriated, which is normal for the industry but means “backlog” overstates certainty. The more honest near-term metric is book-to-bill: it bottomed at 0.8x TTM in mid-2025 (the bleed) and recovered to 1.1x in Q1 FY27 (1.0x TTM) — stabilization, not yet growth. A qualified pipeline of ~$85B (deliberately shrunk ~25% year-over-year as management no-bids commodity work) supports the pivot.

The workforce and delivery model. The company’s ~23,000 employees — a majority holding active security clearances, and a large veteran contingent — are simultaneously its product, its cost base, and its scarcest resource. Cleared-labor supply is the industry’s binding constraint (peers report 50–77% cleared workforces and multi-month onboarding pipelines), which cuts two ways: it protects incumbents’ staffing on existing programs, and it makes organic growth partly a recruiting exercise. SAIC primes most of its revenue but also appears as a subcontractor to the OEMs and to fellow primes (the 98% U.S.-government figure combines both; the prime/sub split is not disclosed — a genuine disclosure gap). Around ~1,700 active contracts and task orders, no single program exceeds a low-single-digit share of revenue — granularity that prevents any one loss from being fatal, while guaranteeing that some recompete is always being decided somewhere in the book.

The products ambition — SilverEdge and the pivot’s sharp end. The October 2025 SilverEdge acquisition ($203M net) is small but strategically pointed: it brings intelligence-community mission-technology — a SOAR (security orchestration, automation and response) platform and the MynAI AI stack — sold more as product than as labor. The strategic logic is the lesson of the comp set: CACI trades at ~14x EBITDA precisely because its technology-products content escapes the labor-arbitrage economics of pure services, and Leidos’s premium franchises (QTC, FAA) are owned, differentiated workflows rather than billed hours. SAIC’s version of this is nascent — products are a rounding error today — but the intent is declared, the first asset is in hand, and the December 2026 portfolio-review readout is the venue where further reshaping (more commodity exits, more product bolt-ons) would surface.

The portfolio pivot. The defining strategic fact of 2026 is “addition by subtraction”: management is deliberately exiting commoditized cost-plus enterprise IT (EIT) — shrinking from 17% of revenue (FY2025) toward ~10% (FY2027E) — because that is where it cannot differentiate and where it lost the CASTLE and Cloud One Next recompetes. In its place: mission IT, space/IC SETA, engineering, and products — the latter seeded by the October 2025 SilverEdge acquisition ($203M, intel/mission-tech SaaS). The FY2027 guide (revenue $7.0–7.2B, organic −4% to −2%) explicitly embeds this shrink; the margin guide (10.1–10.3% adjusted EBITDA, a first-ever double-digit full-year guide) is the payoff.

Verdict: A clear-eyed, if unglamorous, business model — granular, appropriations-backed revenue with real visibility, structurally capped margins on two-thirds of the book, and economics that only improve where contract structure (FFP) or niche position (space SETA, Army software) lets SAIC keep productivity gains. The pivot attacks exactly the right problem; it is also an admission that a sixth of the company was moatless commodity work.

3. Industry Dynamics

Structure — a fragmented oligopoly selling to a monopsony. The U.S. federal IT and mission-services market is a “fragmented oligopoly” of cleared primes: Leidos (~$17.2B revenue), GDIT (~$13B within GD), Booz Allen (~$11.2B), CACI (~$8.6B), SAIC (~$7.3B), Parsons (~$6.4B), V2X (~$4.3B), plus Peraton, Amentum, Accenture Federal, the services arms of the defense OEMs, and thousands of small businesses fed by set-aside programs. Work is competed as task orders under multi-award IDIQ/GWAC/GSA vehicles on past performance, technical rating, and price. The industry is consolidated at the prime tier but brutally competitive at the task-order level, where everything re-bids every 3–5 years.

Barriers to entry are real but shared — the Greenwald test fails at the firm level. Personnel and facility clearances (a majority of SAIC’s workforce is cleared), DCAA-compliant cost accounting, and qualifying past-performance records keep new entrants out. But these are class-level table stakes possessed by every incumbent — they do nothing to stop incumbents from bidding against each other, which is what actually determines returns. SAIC’s own 10-K concedes the structural drift against incumbents: the shift to multi-award vehicles “has led to greater competition and increased pricing pressure,” small-business award initiatives “could result in fewer new opportunities for our industry as a whole,” and FAR/OTA reform is widening the funnel to non-traditional players (Palantir, Anduril). CACI notes LPTA (lowest-price-technically-acceptable) pressure “has moderated, though price still remains an important factor.” In Greenwald’s taxonomy this is a low-to-no-moat industry: demand is stable and appropriations-backed, but the customer is a monopsony whose procurement philosophy — LPTA vs. best-value, set-asides, insourcing, efficiency reviews — sets pricing for everyone. The profit pool stays thin (services EBITDA margins ~9–12%) and ROIC clusters at or modestly above the cost of capital, except in genuinely differentiated niches: Leidos’s QTC health-exams and FAA franchises, CACI’s technology-products tilt. Scale alone does not pay; differentiated content does.

The budget backdrop — record topline, services winter. The dollar environment is, on its face, the most favorable in history: FY2026 defense funding (enacted late, on 2026-02-03, after a full-year-CR start and the longest government shutdown on record, October 1 – November 12, 2025) plus the OBBBA reconciliation’s ~$150B of non-border defense funding pushed total FY2026 defense spending above $1 trillion for the first time; the FY2027 request (April 2026) asks for ~$1.5T including a presumed second reconciliation tranche, and the House’s June 2026 bill exceeds $1T (+$234B vs. 2026). Yet the services slice spent 2025–26 in a procurement winter: the CR and shutdown froze awards, procurement staffs turned over, award cycles elongated, and the DOGE efficiency drive went directly at the contractors — GSA’s February 2025 memo ordered agencies to review and terminate contracts with the ten highest-paid consulting firms (SAIC explicitly named, part of a group “set to receive over $65B in fees”), ultimately claiming ~$23B of multi-year savings. The damage was highly mix-dependent, which is the analytical point: Booz Allen’s consulting-heavy Civil book drove a −6.4% FY26 revenue decline (Civil −22%); SAIC says its direct DOGE impact was <1% of revenue; CACI’s defense-and-technology mix grew double digits and raised guidance twice. The forward setup is a barbell: defense, space, border, and mission-technology content funded generously; commodity IT and consulting structurally scrutinized. That barbell happens to point exactly where SAIC is trying to pivot — the strategic question is whether SAIC can get there before the shrinking legacy book drags the total lower.

Where the industry sits in the capital cycle (Marathon lens). The services complex spent 2025–26 being divested by investors, not invested in: one-year returns of LDOS −33%, BAH −39%, KBR −24%, CACI −6.5%, SAIC +0.8%, while hardware prime GD returned +26%. That is a capital-cycle signal in reverse — the market withdrew capital from the services labor model en masse (multiple compression, not capacity addition), which historically is how the next upcycle’s returns are seeded. Whether the discount is secular (procurement permanently hostile, insourcing, Palantir-ization) or cyclical (a DOGE shock that fades into the largest defense topline ever) is the industry’s central debate, and the first calendar-Q2 peer prints (BAH July 24; CACI and Leidos late July/early August) are the next evidence.

Verdict: NOT a structurally attractive industry — a low-moat, competed market with defensible niches. Demand is large, growing in dollars, and visible; but the seller side sells substantially undifferentiated cleared labor to a single buyer that re-competes everything, pushes set-asides and efficiency reviews, and is actively widening competition. The industry’s thin profit pool and WACC-level returns are features of that structure, not of any temporary dislocation. The caveats that keep this from outright “bad industry”: demand durability is exceptional (appropriations-backed, mission-critical), the niches are real, and the 2025–26 capital exit may have priced the structure more than once.

4. Competitive Position

The moat — named, then stress-tested. SAIC’s candidate advantages, in Greenwald’s taxonomy, are demand-side switching costs rooted in incumbency: mission knowledge embedded in the customer’s operations, cleared staff already in seat, past-performance credentials, and the customer’s rational reluctance to risk a transition on mission-critical work. That is a real but modest captivity — and the last eighteen months stress-tested it to destruction on the exact contracts where it should have been strongest:

• CASTLE (the smoking gun). SAIC was the incumbent on the ~$1.4B Army Corps of Engineers CASTLE enterprise-IT task order — and lost it to Accenture Federal; its GAO protest was denied in May 2026. An incumbent losing its marquee recompete to a non-traditional-in-defense rival is direct evidence that switching costs did not bind.
• Cloud One Next (Air Force). Lost (and partially no-bid) — the second of the two “unfavorable recompete award decisions” behind the February 2026 guidance cut; together the losses equaled ~4% of revenue.
• RITS (Army). Lost after protest — a ~$200M FY2027 headwind rolling off around Q3.
• NASA (FY2025). An unnamed program loss cited by management.

CFO Natarajan’s own autopsy: “the one common thread across most of [our recompete losses] has been large EIT, cost-plus and (DoD).” If the moat were real, incumbent win rates on marquee cost-plus EIT recompetes would hold; they did not. Management’s claimed 85–90% recompete win rates “ex-commoditized EIT” is, note, a claim that excludes exactly the category where the losses occurred.

Where SAIC demonstrably wins — the pockets of competence. The same record shows positions that have held repeatedly: space/IC SETA advisory (~$970M of seven-year recompetes retained in Q1 FY27 alone, including $540M and $330M awards — the one place incumbency has visibly bound, because mission knowledge in intelligence space operations is genuinely scarce); Army aviation/missile software engineering (the $1.8B AvMC franchise); new mission-tech wins ($928M HOPE 2.0, $1.4B COBRA, $192M ABMS digital infrastructure, $242M Navy propulsion test); and civilian FFP enterprise IT, where fixed-price structure lets SAIC bank automation gains (Civilian 15.5% adj. operating margin vs. 10.0% D&I) — it took the $95M GAO TIS contract from GovCIO by bidding $95.4M against a $134.3M rival bid and still making money. This is a pockets-of-competence profile, not a franchise.

Against the Big-5. SAIC is the smallest (~$7.3B vs. Leidos ~$17B), the lowest-margin (FY26 EBITDA 9.7% vs. LDOS ~14%, CACI ~12%, BAH ~11%), the slowest-growing (the only Big-5 prime shrinking through the largest defense topline in history), and earns ROIC of ~12% versus mid-teens at Leidos and Booz Allen. It has no crown-jewel analog to Leidos’s QTC/FAA and no products engine like CACI’s. Its relative-value offset is price: it has historically traded at a discount to the complex (though after its 2026 rally and its peers’ continued declines, it currently trades slightly above the LDOS/KBR/BAH cluster on EV/EBITDA — unusual, and examined in the valuation section).

Verdict: weak/narrow advantage — the weakest-positioned of the Big-5 cleared primes. The moat candidates are shared table stakes that empirically failed on the largest recent tests; the metric that would prove a moat — organic growth at or above budget growth with stable margins — shows the opposite (−3.3% organic FY26, guided −2% to −4% FY27). What keeps this above “no advantage at all”: the repeatedly re-won space/IC SETA book, sticky Army engineering franchises, attractive civilian FFP economics, a $22.9B backlog, and >$600M of guided free cash flow. Durability depends on whether the Reagan pivot converts the portfolio toward the pockets where incumbency actually holds before the shrinking EIT base drags consolidated revenue further.

5. Growth History and Forward Opportunities

History: four years of stall, then contraction. Revenue: $7,394M (FY22) → $7,704M (FY23) → $7,444M (FY24) → $7,479M (FY25) → $7,262M (FY26) — a −2.9% reported / −3.3% organic decline in FY2026 (adjusting for $27M of SilverEdge contribution), with management attributing ~$200M of the FY26 decline to the deliberate Cloud One no-bid and ~26M to the record shutdown. FY27 is guided to $7.0–7.2B, organic −4% to −2% — a fourth consecutive year without organic growth, making SAIC the only Big-5 prime shrinking through a record defense budget. Growth quality, where it exists, is mixed: the FY21–22 growth was substantially acquired (Unisys Federal, Halfaker); the FY25–26 decline is concentrated in the commodity EIT book the company is deliberately exiting, plus recompete losses and procurement friction. Q1 FY27 printed +1.5% reported (+0.5% organic) — stabilization, with management now expecting revenue “at or slightly above the midpoint” of guidance.

The bookings engine underneath the revenue line. The demand indicators turned before the P&L did: FY26 net bookings of ~$7.8B (book-to-bill ~1.07x) against the shrinking revenue line, the TTM ratio bottoming at 0.8x in mid-2025 and recovering to 1.1x in Q1 FY27 ($2.1B of Q1 bookings; 1.0x TTM). The composition of recent bookings matters as much as the level: ~$970M of Q1 FY27 space/IC recompetes (seven-year durations — long-visibility annuity work), the $1.4B COBRA task order, $928M HOPE 2.0, $242M Navy propulsion test, $192M ABMS digital infrastructure, $123M NAVWAR C4ISR, $100M of FAA task orders, and a $200M DHS recompete. Management cites win rates above 30–50% on non-EIT new business and 85–90% on recompenses outside the commodity EIT book it is exiting — claims that are hypotheses, but consistent with the bookings mix. The deliberately smaller qualified pipeline (~$85B, down ~25% y/y) is bid discipline, not demand weakness: fewer, better shots.

The forward growth algorithm — shrink, then (maybe) grow. The explicit plan: absorb the ~$400M FY27 recompete headwind (RITS ~$200M rolling off ~Q3, plus CASTLE/Cloud One annualization) while ramping FY25–26 wins — ~$350M of actual FY26 new-business revenue against a >$800M potential annual run rate if the won book ramps fully — and targeting FY28 as the re-acceleration year (consensus: +1.1% revenue, +8% EPS). The swing items, in order of size: Vanguard/State EVOLVE (~$250M/yr at above-average margins; SAIC is pursuing 4 of 5 workstreams on the $10B vehicle — retention stabilizes the Civilian book, loss re-opens the revenue decline in FY28); the second reconciliation bill (~$333B assumed in the FY27 request — upside to defense/space/border content); and the ramp of the won book (COBRA $1.4B, HOPE 2.0 $928M, ABMS $192M, the ~$970M Q1 space/IC recompetes). Management’s bid discipline (pipeline deliberately cut ~25% to ~$85B; >30–50% win rates on non-EIT new business) means reported growth will stay muted even if the pivot works — this is a shrink-to-grow transition, and FY28 is the first clean test.

Verdict: low-quality current growth, credible but unproven forward algorithm. The track record is poor (no organic growth in four years, guidance for another decline), and a deliberate-shrink strategy is only half a growth story. The credible pieces: the headwind is quantified and lapsing, the won-book ramp is arithmetic rather than hope, the budget barbell favors the target portfolio, and the first clean evidence (book-to-bill back to 1.0x TTM, Q1 organic +0.5%) has appeared. What would falsify it: a Vanguard loss, or FY27 organic landing below the −4% floor — evidence the erosion is secular, not transitional.

6. Financial Quality

Margins: a genuine, slow climb — with one-time scaffolding. Gross margin has ground from 11.6% (FY22) to 12.0% (FY26); adjusted EBITDA margin reached 9.7% in FY26 and is guided to 10.1–10.3% in FY27 — the first double-digit full-year guide in company history. Q1 FY27 printed a record 11.6% (+320bp y/y) — but ~60bp of that is a one-time $12M gain on an investment sale, so the clean run-rate is closer to 11.0%, still a real step up. Two normalizations matter for the trend: FY2024’s GAAP operating income of $741M included a $247M gain on the supply-chain divestiture (ex-gain, FY24 operating margin was 6.6% — FY24 was never a margin peak; third-party feeds that show a margin “decline” from FY24 are reading the gain), and the FY26 effective tax rate of 7.5% (an IRS audit settlement worth $47M, plus a −$19M Q2 credit) flattered EPS — normalized at a ~20–23% rate, FY26 EPS would be roughly $1 lower.

The five-year picture, condensed (fiscal years ending ~January 31; FY24 operating income and EPS include the $247M divestiture gain):

*FY2024 includes a one-time $247M gain on the supply-chain divestiture (plus a $7M deconsolidation gain); ex-gain FY24 operating income was ~$494M (6.6% margin). Adjusted (non-GAAP) EBITDA, the metric management guides and is paid on, was $711M (9.7%) in FY26 and is guided to $720–730M (10.1–10.3%) in FY27.

The table tells the whole financial story in one glance: a revenue line that goes nowhere, a profit line that is flat once FY24’s one-time gain is stripped, a cash line that trends up, and a share count that falls ~5% a year. Everything per-share is the fourth line working on the first three.

Earnings quality: the engine is financial, not operational. GAAP diluted EPS grew $4.77 → $7.70 (FY22→FY26, +61%) while pre-tax operating profit ex-items has been roughly flat at $500–560M for four years. The bridge: diluted shares fell 58.1M → 46.5M (−20%, buybacks), and the tax rate fell 23.1% → 7.5%. Neither is a sin — buybacks at these multiples are intelligent allocation — but an analyst must see that consolidated EPS growth is per-share engineering on a flat operating base, and the FY27 guide repeats the recipe (the >$600M FCF guide embeds a ~$70M nonrecurring cash tax benefit; FY28 FCF guidance of ≥$530M strips it). Underlying trajectory, honestly told: revenue −3%, margins +40–60bp/yr, pre-tax profit flat-to-slightly-up, per-share metrics +10–18%.

Cash flow: the genuine strength, and the reason the equity case exists. FY26 operating cash flow $609M, capex a trivial $32M (~0.4% of revenue — services are asset-light), FCF $577M = 1.6x net income and ~10% above guidance; Q1 FY27 FCF $118M. Conversion has been aided by a durable working-capital harvest (receivables $1,000M → $853M; DSO 61 → 46.6 days since FY20) — a real but finite tailwind. Balance-sheet quality flags are modest: no material impairments or restructuring in FY25–26, trivial contract liabilities ($41M — no deferred-revenue games), SBC of $64M (0.9% of revenue).

Balance sheet: levered for a services company, goodwill-heavy. Total debt $2,685M (including $198M finance leases), cash $182M, net debt $2,503M = ~3.7x EBITDA (3.4x ex-leases; company-adjusted basis ~3.1x) — re-ticked up from 3.1x at FYE25 after funding SilverEdge ($203M) plus $445M of FY26 buybacks. Interest coverage ~5.2x. Goodwill + intangibles of $3,705M are 69% of assets and tangible book is deeply negative (−$47.6/share) — the Engility/Unisys/Halfaker acquisition legacy; ROE (23%) is therefore meaningless and ROIC (~11.8% FY26, up from ~9.3% in FY22–24) is the right returns metric. The leverage is manageable for a business with 98% government revenue and 62% cost-plus mix, but it caps M&A capacity and makes the buyback-vs-deleverage tension a live question (management calls ~3x its target ceiling and describes buybacks as “opportunistic”).

Verdict: improving economics at the margin, flat economics at the core — and a cash machine either way. The margin climb is real (mix shift + cost actions), the cash conversion is excellent and the balance sheet is serviceable. But normalize the tax rate and one-time gains and the honest picture is a flat pre-tax profit base whose per-share metrics are being compounded by buybacks. That is a perfectly good equity story at 11x earnings and a 12% FCF yield; it is not operating leverage, and it should not be underwritten as such.

7. Capital Allocation

The record — genuinely good, and the strongest section of this report. Over five fiscal years (FY22–26) SAIC returned ~$2.3B to shareholders — $1.878B of buybacks plus dividends — against cumulative FCF of roughly $2.5B, retiring ~28% of the share count (59.0M diluted FY20 → 46.5M FY26; 42.3M outstanding at May 2026). Crucially, the buying has been counter-cyclical and price-disciplined: $445M in FY26 as the stock bled, then $188M in Q1 FY27 alone at an average of ~$92 — ~20% below today’s price — explicitly framed by management as buying a “dislocation to the stock price.” A December 2024 authorization of $1.2B (no expiry) leaves ample room. The dividend, by contrast, has been frozen at $0.37/quarter since 2022 (~1.3% yield, ~20% payout) — a minor critique: buybacks have absorbed everything, which is defensible while the stock trades at 11x but signals no dividend-growth commitment.

M&A: disciplined after a leveraged start. The transformative deals were Engility (2019, ~$2.5B all-stock — dilutive in count, transformative in scale) and Unisys Federal (2020, $1.2B at 13x adjusted EBITDA, debt-funded — credited with the FY21–22 growth). Since 2021 the record improves markedly: Halfaker (~$250M, VA health entry), the small Koverse AI deal, a well-timed divestiture of the logistics/supply-chain business to ASRC Federal for $350M (May 2023) — selling a low-multiple commodity book at a good price — and SilverEdge (October 2025, $203M), a bolt-on in intel/mission-tech SaaS that is the first tangible piece of the “products” pivot. Nothing oversized in five years; leverage has oscillated around ~3x rather than ratcheting.

Incentives: the board pays on the right things and lets the math bite. Short-term incentives pay equally on revenue, adjusted EBITDA dollars, and adjusted operating cash flow; FY26 results scored revenue and EBITDA legs at 0% (actual misses — no discretionary softening) and OCF at 200%, for a 66.7% total. Long-term incentives are 60% PSU / 40% RSU, with PSUs on three-year cumulative adjusted EBITDA, cumulative adjusted free cash flow (newly added, replacing OCF — directionally right), and relative TSR versus peers; the FY24–26 cycle paid 113.1% with the rTSR leg at the 25th percentile paying zero. Ownership guidelines (CEO 5x, NEOs 3x base), no employment agreements, no gross-ups, hedging/pledging banned. The gap: no ROIC metric — mitigated by the FCF and rTSR legs. Insider ownership is thin (~0.65% all-in), but the behavior offsets it: 37 open-market buys versus zero sales over 18 months, including the CFO, GC, multiple directors, and the incoming CEO at $95–112.

The CEO transition as a capital-allocation signal. The board fired Toni Townes-Whitley (“without cause,” October 23, 2025 — seven weeks after the Q2 guidance cut, with the stock at bottom-quartile TSR) and installed a director, Jim Reagan — the former CFO of Leidos (2015–2021), Vencore, and PAE — who was made permanent in February 2026 and immediately consolidated five business groups into three (two EVPs out on severance), launched $100M of cost reallocation, and reset guidance to a credible floor. The read: the board swapped a transformation narrative for a capital-discipline operator, and the two prints since (December beat-and-raise, June beat-and-raise) are early validation. The optics flags are real but secondary: the ouster cost ~$16.7M all-in ($8.25M severance plus unprorated vesting of $8.4M of equity — the board paying for its own 2023 hiring error), and Reagan’s $2.0M interim-CEO RSU vested upon his own permanent appointment. CFO Prabu Natarajan (ex-Northrop Grumman, in seat since November 2020) provides continuity across both regimes.

Verdict: has management allocated capital intelligently? Yes — with this section carrying the bull case. Counter-cyclical buybacks retiring 28% of the count, a well-priced divestiture, bolt-on-only recent M&A, cash-flow-weighted incentives with demonstrated zero payouts, and a board willing to fire a CEO and hire a discipline operator. The watch-items: a frozen dividend, thin insider ownership (though heavy insider buying), leverage back at the ~3x target ceiling, and the buyback’s role as the primary EPS engine while revenue declines — if Vanguard is lost, watch whether the board keeps buying back stock at the ceiling or preserves the balance sheet.

8. Changes and Headwinds — Last Two Years

• The DOGE regime (Nov 2024 → ). The Department of Government Efficiency’s announcement de-rated the entire services complex; GSA’s February 2025 ten-firm consulting review named SAIC explicitly. Direct revenue damage was modest (<1% per management) but the procurement friction — CR, the record 43-day shutdown (Oct–Nov 2025, ~$26M SAIC revenue impact), elongated awards, a second shutdown at DHS (Feb–Apr 2026) — was pervasive.
• Three problem prints and a guidance collapse (Jun 2025 – Feb 2026). Q1 FY26 miss (−13.3%), Q2 FY26 guidance cut (−6.9%), and the February 11, 2026 pre-announcement cutting FY27 revenue guidance to $7.0–7.2B on two lost recompetes and procurement delays (−16.0%, four-year low $80.50).
• The CEO change (Oct 2025 – Feb 2026). Townes-Whitley fired; Reagan interim (Oct 23) then permanent (Feb 17); 5→3 business-group consolidation (effective Jan 31, 2026); two business-group EVPs exited; a new Chief Growth Officer (Dec 2025); board refreshed (Urban, then Eremenko and Adm. Mike Rogers, taking it to 13).
• Recompete losses (the self-inflicted wounds). CASTLE (~$1.4B, lost as incumbent to Accenture Federal; protest denied May 2026), Cloud One Next, RITS (~$200M FY27 headwind), an unnamed FY25 NASA loss — ~4% of revenue in announced losses.
• The won book (the offset). $928M HOPE 2.0, $1.4B COBRA, $242M Navy propulsion test, ~$970M of Q1 FY27 space/IC recompetes, $192M ABMS digital-infrastructure leadership, $123M NAVWAR C4ISR, $100M FAA, and the $203M SilverEdge acquisition.
• The financial turn. FY26 FCF $577M (~10% above guide); December 2025 beat-and-raise (+16.3%); Q1 FY27 record 11.6% margin and raised FY27 EBITDA/EPS guidance (+10.4%); FY27 guide now 10.1–10.3% margins — first double-digit full-year guide.
• Pending: the State Department Vanguard recompete (EVOLVE, $10B ceiling) — the largest open swing item; results of management’s portfolio review, due on the December 2026 call.

Verdict: the last two years weakened the revenue thesis (demonstrated recompete vulnerability, four years of no growth) but materially strengthened the execution thesis (new operator, margin records, cost discipline, insider buying). The net effect on the thesis is refocusing: this is no longer underwritten as a growth story at all — it is underwritten as a cash-flow floor with a margin-expansion and per-share-compounding kicker, and the events of the last two years are exactly what forced that framing.

9. Risk Analysis

Catastrophic-loss / total-loss risk: very low. A 98%-government-revenue, cost-plus-weighted, cash-generative contractor with a $22.9B backlog does not go to zero; the realistic bear path is a slow grind (another lost recompete, margin stall, multiple stuck at 8x → ~$75, ~−34%), not impairment. The leverage and negative tangible book amplify equity moves but do not threaten the enterprise.

10. Valuation Discussion (embedded expectations)

Where it trades. At $112.66 and ~43.4M diluted shares: ~$4.9B market cap; ~$2.57B net debt (Q1 FY27, including finance leases); ~$7.5B enterprise value. Against TTM results (revenue $7,291M, EBITDA ~$732M) and FY27 guidance (adj. EBITDA $720–730M, adj. EPS $9.90–10.10, FCF >$600M):

Own history. SAIC’s FY-end EV/EBITDA ran 10.8–13.9x across FY17–FY26 (average ~12.4x), with within-year troughs at ~8.3–8.5x (printed in FY21 and again at the February 2026 low). The current ~10.2x is the bottom quartile of its own ten-year range; the AZI own-history composite percentile is the 39th — but the split tells the real story: P/E 9th percentile, P/B 54th, P/S 54th. On earnings power, SAIC is near its cheapest ever; on sales and book, merely average — consistent with a market that doubts the earnings base, not the revenue.

Peer context (TTM, re-struck at July 20 prices — approximations). The entire services complex de-rated through 2025–26:

Two honest observations. First, the group prices growth direction: CACI (+8.5% organic) holds a ~14x premium while every shrinker sits at 8–10x — the market pays for revenue trajectory in this space, not for cash flow alone. Second, after SAIC’s +40% repair rally and its peers’ continued declines, SAIC now trades slightly above the LDOS/BAH/KBR cluster — historically unusual for the lowest-margin prime. The relative-value purist’s argument for LDOS (8.4x, 15th percentile) or BAH (9.5x, 1st percentile) over SAIC is real; the counter is that SAIC is the only one of the three with a delivered margin inflection, two beat-and-raises, and insider buying — the market is paying a small premium for the only visible turn in the complex.

Scenario analysis (FY27E adj. EBITDA $725M, net debt $2.57B, ~43.4M shares; implied equity = multiple × EBITDA − net debt):

Embedded expectations. The arithmetic is stark: a >12% guided FCF yield at a 9–10% cost of equity implies the market is underwriting roughly −2% to 0% perpetual FCF growth — i.e., the flat base case, with the FY28 re-acceleration, the buyback’s ~9–10%-of-cap annual compounding, and any multiple normalization all as unpriced optionality. What the market is pricing correctly: the structural low multiple (monopsony customer, 3.5x+ leverage, negative tangible book, budget cyclicality — none of these argues for 15x), and genuine doubt about the revenue base (the group prices growth direction, and SAIC’s is negative). What it may be underpricing: the probability that February 2026 was the reset floor (two beat-and-raises since), the insider signal, and the per-share math of retiring ~9% of the equity annually at an 11x multiple — which compounds per-share value at ~10%/yr even if the enterprise never grows. The asymmetry the tape offers: base ≈ current price, bull +25%, bear −34% — negative skew if probabilities were equal; the whole debate is whether the pivot evidence shifts the odds toward the bull. Sensitivity: ±0.5x multiple ≈ ±$8.35; ±$25M EBITDA ≈ ±$5.80.

No price target, no recommendation (see Kimi’s Take for the fenced exception). The embedded-expectations read: priced for continued slow erosion, with the turnaround and the compounding machine as the free options — and the cheapest-in-complex claim no longer available as a crutch after the rally.

11. Variant Perception

Consensus. The Street is at Hold (11 analysts: ~2 Strong Buy / 8 Hold / 1 Strong Sell — Goldman at Sell, $96; Citi at Buy, $132; TD Cowen $130; Jefferies $115), mean target $120, median $125 — barely ~6% above the tape. The consensus narrative: a no-growth, recompete-exposed services name whose margin story is real but whose revenue base keeps eroding; respect the cash flow and the buyback, but don’t pay for a pivot until revenue stabilizes. In short, the Street sees the cheapness and the turnaround and refuses to underwrite either.

Strongest bull case. A ~12% FCF-yield annuity whose revenue decline is quantified, lapsing, and deliberately self-inflicted in the worst segment — run by a new operator with a Leidos-honed capital-discipline playbook, two consecutive beat-and-raises, record margins, a raised guide, and a friendly budget barbell (>$1T defense, FY27 request +$234B). The per-share engine compounds ~10%/yr mechanically (buyback of ~9% of cap at 11x), insiders are buying with their own money at these levels, and the factor regime (value in favor, momentum screens still carrying SAIC as negative-momentum) means the turn is uncrowded. Bull’s falsification: FY27 organic revenue below the −4% floor, or adjusted EBITDA margin back under 10% ex-one-times — either breaks “the decline is transitional.”

Strongest bear case. A melting ice cube in a structurally hostile industry: the only Big-5 prime shrinking through the largest defense budget in history, with demonstrated incumbent recompete losses, a pending Vanguard decision that could remove ~$250M/yr of its best-margin revenue, EPS “growth” that is entirely buybacks and a 7.5% tax rate on a flat pre-tax base, 3.7x leverage, 69% goodwill — and a stock that has already re-rated 40% off the low to a multiple above its cheaper, arguably better-positioned peers (LDOS at 8.4x/15th percentile, BAH at 9.5x/1st percentile). The February pattern — guidance reset, relief rally, next disappointment — has already played once under this strategy. Bear’s falsification: Vanguard retained plus a third consecutive beat-and-raise with organic revenue at/above the guide midpoint — evidence the base has stopped eroding.

The 3–5 assumptions that matter most:

1. The revenue floor. FY27 organic lands inside −4% to −2% and FY28 turns positive on the won-book ramp (COBRA, HOPE 2.0, ABMS, space/IC recompetes). The single most important number in the story.
2. Vanguard/EVOLVE. Retention keeps the civilian FFP margin engine intact; loss is ~$250M/yr of above-average-margin revenue and a second credibility hit.
3. Margin durability ≥10%. Whether the 10.1–10.3% guide survives the wash-out of the $12M gain and depends on cost actions rather than mix luck.
4. The buyback’s continuation. ~$400M/yr at 11x is the EPS engine; it requires leverage staying ~3x and no Vanguard shock forcing cash preservation.
5. The services discount is cyclical, not secular. If the DOGE-era de-rating of services labor is permanent (insourcing, Palantir-ization), there is no multiple mean-reversion to harvest.

Factor-positioning read (as evidence, not a call). SAIC’s factor signature is a crash-era hangover: negative Momentum loading (−0.22), zero Quality, small-size tilt, low beta, 36.6% idiosyncratic vol — quantitatively, screens still see “weak tape, unproven quality,” even as the actual tape has turned (above all three EMAs, strongest 3/6/12-month relative performance in the services complex, +19% m3). That gap — the model pricing the crash while the company prints beat-and-raise — is the variant perception in its purest form: the turn is real but unacknowledged by systematic money. The regime is a tailwind (Value z +1.4, DividendYield in favor, high-beta out of favor). Two positioning caveats: factor-similar names are generic small-cap value cash-flow stocks, not the gov-services comp set (so cross-sectional “cheap vs peers” arguments carry less weight than own-history ones), and short-interest data was unavailable — a positioning blind spot worth closing.

12. Fact vs. Interpretation

13. Open Questions

1. Vanguard/EVOLVE outcome and timing. The single largest swing item (~$250M/yr, above-average margin). What are realistic odds, when is it decided, and what is plan B for the Civilian book if lost?
2. The revenue floor. Does FY27 organic land inside guidance (−4% to −2%) and does FY28 consensus (+1.1%) hold as the recompete headwind lapses? The won-book ramp (~$350M actual vs. >$800M potential run rate) is the mechanism — how fast does it convert?
3. Clean margin run-rate. What is Q1 FY27’s 11.6% ex the $12M gain (~11.0%), and does ≥10% hold through the year as RITS rolls off?
4. Normalized earnings power. At a standard ~21–23% tax rate and without the ~$70M FY27 cash tax benefit, what is sustainable EPS and FCF? (FY28 FCF guide of ≥$530M is the honest baseline.)
5. Buyback durability at the leverage ceiling. With net debt ~3.7x EBITDA (incl. leases) and a ~3x stated target, does the ~$400M/yr pace survive a Vanguard loss — or does the board choose the balance sheet?
6. December 2026 portfolio review. Management has flagged portfolio-review conclusions for the December call — further divestitures (an Engility-era logistics remnant? more commodity EIT exits?) could reshape both revenue and margin.
7. 10b5-1 vs. discretionary selling history and short interest. The 18-month zero-sale streak is striking; current short interest (unavailable from our feeds) would complete the positioning picture.
8. Palantir-ization risk. How exposed is SAIC’s mission-IT and data book to product-led entrants (Palantir, Anduril) as FAR/OTA reform widens the funnel — is the SETA/advisory pocket defensible against software?

14. What Must Be True

For the bull case (cheap annuity + successful pivot → re-rate to ~$140) to be right:

• The revenue decline stops on schedule: FY27 organic within −4% to −2%, FY28 positive as ~$400M of recompete headwind lapses and the won book ramps.
• Vanguard is retained (or EVOLVE workstreams won), preserving the civilian FFP margin engine.
• Margins hold ≥10% ex-one-times, proving the cost program and mix shift are structural.
• The buyback continues at ~$400M/yr, compounding per-share value ~10%/yr while the base stabilizes.
• Falsification test: FY27 organic revenue below −4%, or adjusted EBITDA margin back under 10% excluding one-time items, or a Vanguard loss without an offsetting plan — any one breaks “transitional, not secular.”

For the bear case (melting ice cube at a full-for-the-complex multiple → ~$75) to be right:

• The erosion is secular: procurement hostility, set-asides, and product-led entrants keep shrinking the addressable book faster than the pivot can replace it.
• Vanguard is lost, removing ~$250M/yr of above-average-margin revenue and repeating the February credibility break.
• Margin gains prove one-time (venture gain, tax benefit) rather than structural, and the flat pre-tax base starts to decline.
• The sector multiple stays pinned at ~8x, and SAIC’s premium-to-peers closes downward.
• Falsification test: three consecutive beat-and-raise quarters (two are in hand) with organic revenue at/above the guide midpoint and a Vanguard retention — evidence the base has stopped eroding.

The two numbers to watch quarterly: organic revenue growth against the −4%/−2% guide rails, and adjusted EBITDA margin excluding one-time items — with the Vanguard decision as the single binary event that can settle the debate early.

15. Source Appendix

See the accompanying Appendix B — Source Appendix in the combined report for the full source list. Primary sources: SAIC FY2026 Form 10-K (filed March 16, 2026) and Q1 FY2027 Form 10-Q (filed June 1, 2026); FY2026 DEF 14A (April 22, 2026); the CEO-transition 8-Ks (October 23, 2025; February 17, 2026); earnings releases and 8-K exhibits for Q2 FY26 through Q1 FY27 (including the February 11, 2026 pre-announcement); the full 60-month EDGAR corpus (451 filings enumerated, ~385 mirrored locally, including 261 parsed Forms 3/4/5); Q4 FY26 and Q1 FY27 earnings-call transcripts. Market/factor data: AZI price series and valuation_index percentiles; FactorsToday factor model; ROIC.ai (cross-check; its SAIC FCF field and 7/20 price print were rejected as erroneous). Analyst/consensus data via S&P Global poll (stockanalysis.com) and trade press — attributed, never adopted. Peer context: the author’s prior published coverage of LDOS, CACI, BAH, GD, KBR. Management commentary is treated as hypothesis and validated against filings throughout.

APPENDIX A — Standard Diligence Questionnaire

Science Applications International Corporation (NASDAQ: SAIC) — July 21, 2026

Answers are grounded in the FY2026 10-K, Q1 FY2027 10-Q, FY2026 proxy, earnings calls, and supporting analysis. Labels: F = Fact, I = Interpretation, A = Assumption.

General

What thoughtful questions have other investors asked about this company? (1) Is the revenue decline transitional (deliberate shrink + quantified recompete headwinds) or secular (structural share loss)? (2) Is the record margin real or one-time (venture gain, tax benefit)? (3) What happens on the State Department Vanguard recompete (EVOLVE)? (4) Can EPS growth built on buybacks and a 7.5% tax rate be underwritten? (5) Is the services-complex de-rating cyclical (DOGE shock) or permanent (procurement reform, Palantir-ization)? (6) Does the ex-Leidos CFO as CEO change the trajectory, or is a low-moat business a low-moat business regardless of operator? The Street’s answer so far: Hold, mean target ~$120.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? (I) Revenue is at a cyclical low (procurement winter: CR, record shutdown, efficiency reviews, elongated awards); margins are at a cyclical high (record 11.6% Q1 EBITDA margin, first double-digit full-year guide). Earnings are neither cleanly high nor low — the mix makes this a transition year.

Driven by external environment or internal actions? (I) Both, unusually clearly separated: the revenue problem is roughly half external (DOGE/procurement friction, sector-wide) and half internal (recompete losses, deliberate no-bids); the margin story is internal (cost actions, mix shift, FFP discipline).

How stable are revenues? (F/I) Very stable in structure — 98% U.S. government, 62% cost-reimbursable, $22.9B backlog, ~1,700 contracts — but not immune: four years without organic growth and ~4% of revenue lost to recompetes in 18 months.

Outlook for products/services? (I) Barbell: mission tech, space/IC SETA, engineering, and products (SilverEdge) funded generously by the >$1T defense budget; commodity enterprise IT structurally scrutinized. SAIC is pivoting toward the funded half.

How big is this market — growing, shrinking, domestic or international? (F/I) Federal IT/mission services is a large, growing-in-dollars pool inside a >$1T FY26 defense topline and a ~$1.5T FY27 request; essentially 100% domestic/U.S.-government by definition.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? (I) More: multi-award vehicles, set-asides, efficiency reviews, and FAR/OTA reform widening entry to non-traditional players (Palantir, Anduril). SAIC’s 10-K itself flags the pricing pressure.

How profitable is the business (ROIC, ROE)? (F/I) ROIC ~11.8% FY26 (up from ~9.3% FY22–24) — roughly WACC-level, i.e., near-zero economic profit. ROE (23%) is meaningless (negative tangible book; goodwill 69% of assets).

How profitable is the industry — competitors, barriers? (F/I) Thin: services EBITDA margins 9–12%, returns clustering near cost of capital; barriers (clearances, DCAA, past performance) are class-level and shared. Five big cleared primes plus Peraton/Amentum/Parsons and set-aside smalls.

Can the business be easily understood? (I) Yes — cleared people billed to the government under re-competed contracts; the nuances (contract-type mix, funded vs. unfunded backlog, fiscal-year offset) are manageable.

Can it be undermined by foreign low-cost labor? (I) No — clearances and ITAR-ish sensitivity insulate it; the threat is domestic: product-led entrants and set-aside smalls.

Do brands matter? (I) Past performance matters (it is scored in bids); “brand” as pricing power does not.

What is the nature of competition? (I) Task-order bake-offs every 3–5 years on price/technical/past-performance; incumbency helps (industry recompete win rates ~70–90%) but did not save SAIC’s marquee cost-plus EIT contracts.

Customers’ switching costs? (F/I) Real but modest — transition risk on mission work buys incumbents a benefit of the doubt in differentiated niches (space/IC SETA, where SAIC keeps winning), and nothing in commodity EIT (CASTLE lost as incumbent).

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? (F/I) The cleared workforce, past-performance record, and $19B of negotiated-unfunded backlog are economic assets not on the books; conversely 69% of the balance sheet is goodwill/intangibles from acquisitions.

Off-balance-sheet liabilities? (F) Finance leases ($198M, included in our leverage math); operating leases; standard government-contract termination/audit exposures (DCAA incurred-cost audits can claw back years of billings).

How conservative is the accounting? (I) Reasonably: trivial contract liabilities ($41M — no deferred-revenue games), no material impairments/restructuring recently, capex honest at ~0.4% of revenue. Watch-items: a 7.5% FY26 tax rate and ~$70M FY27 nonrecurring cash tax benefit flatter EPS and FCF — normalize; non-GAAP “adjusted EBITDA” is the metric management is paid on, so read its reconciliation each quarter.

How CapEx-hungry is the business? (F) Barely: capex $27–36M/yr (~0.4% of revenue). Capital intensity sits in working capital (receivables — DSO already harvested from 61 to 47 days) and in cleared-people recruiting.

Capital Allocation & Management

How much FCF, and how is it used? (F/I) FY26 FCF $577M (1.6x net income); FY27 guided >$600M; FY28 guided ≥$530M ex-tax-benefit. Uses: ~$400M/yr buybacks (28% of shares retired since FY20), a flat $0.37/quarter dividend (~20% payout), bolt-on M&A (SilverEdge $203M), leverage held ~3x.

Significant acquisitions recently? (F) SilverEdge (Oct 2025, $203M, intel/mission-tech SaaS) — first “products” bolt-on of the pivot. History: Engility (2019, ~$2.5B stock), Unisys Federal (2020, $1.2B, 13x), Halfaker (2021, ~$250M); divestiture of logistics/supply-chain ($350M, 2023).

Buying back shares? (F) Aggressively: $445M FY26; $188M in Q1 FY27 alone at ~$92 average; $1.2B authorization (Dec 2024, no expiry).

Issuing large amounts of new shares to insiders? (F/I) No — SBC ~$64M (0.9% of revenue), modest against ~$400M/yr of buybacks; net dilution strongly negative.

Compensation policy? (F) STI: revenue/adj EBITDA/adj OCF equally weighted (FY26 scored 66.7%, with zero on the missed legs); LTI: 60% PSU on cumulative adj EBITDA, cumulative adj FCF, rTSR (FY24–26 paid 113.1%, rTSR leg zero); no employment agreements, no gross-ups; guidelines 5x/3x.

Motivations of management? (I) The board fired a growth-narrative CEO and installed a capital-discipline operator (ex-Leidos CFO Reagan) who bought stock at ~$100 with his own money; CFO Natarajan (since 2020) bought repeatedly at $100–110. Incentives and behavior both point to per-share value creation over empire-building. Watch: thin absolute ownership (~0.65%), rich exit economics for the fired CEO.

Valuation & Market Data

ADR, MLP, or K-1 issuer? (F) No — U.S. Delaware C-corp, standard common stock.

Dividend policy? (F) $0.37/quarter, flat since 2022 (~1.3% yield, ~20% payout); no growth commitment — buybacks are the chosen return channel.

How profitable is the business? (F/I) FY26 adj. EBITDA margin 9.7% guided to 10.1–10.3% FY27; ROIC ~12%; FCF margin ~8%.

Is net income diverging from cash from operations? (F) Favorably — FY26 FCF of $577M is 1.6x net income (D&A, working-capital harvest); the divergence to watch is the other direction: net income flattered by the 7.5% tax rate.

Risks & Downside

What would cause the stock to decline? (I) A Vanguard loss; another marquee recompete failure; FY27 organic below the −4% floor; margin ex-one-times back under 10%; a sector tape that keeps de-rating services; or normalization of the tax-flattered EPS by the Street. Bear case ~$75 (8x EV/EBITDA).

Risk of a catastrophic loss? (I) Very low — 98% government revenue, 62% cost-plus, $22.9B backlog, ~$600M FCF; the enterprise is not at risk, only the multiple and the growth narrative.

Chance of a total loss? (I) Negligible.

Recent News & Events

Has the business environment changed recently? (F) Yes — the DOGE/procurement winter of 2025–26 (CR, record shutdown, GSA consulting review naming SAIC), then a record >$1T FY26 defense topline and a ~$1.5T FY27 request; SAIC-specific: CEO fired (Oct 2025), ex-Leidos CFO installed, 5→3 reorg, FY27 guidance cut (Feb 2026) then raised (Jun 2026) after two beat-and-raises.

Significant acquisitions? (F) SilverEdge ($203M, Oct 2025).

Change in accounting policies? (F) None material; watch the non-GAAP definitions management is paid on.

Recent changes — new markets, facilities, management? (F) New CEO (Reagan, permanent Feb 2026), new Chief Growth Officer, two EVP exits, board additions (Urban; Eremenko + Adm. Rogers), 5→3 business-group consolidation, and a deliberate portfolio pivot out of commodity enterprise IT with a portfolio-review readout due December 2026.

APPENDIX B — Source Appendix

Science Applications International Corporation (NASDAQ: SAIC) — July 21, 2026

All non-obvious facts in this report are cited to the sources below. Primary sources (SEC filings, company releases) take precedence; third-party data is used for cross-check and market context and is labeled as such. Management commentary is treated as a hypothesis and validated against filings. Fiscal-year convention: FY2026 ended January 30, 2026.

Primary sources — SEC filings (EDGAR, CIK 0001571123; 60-month corpus reviewed in full)

1. Form 10-K, FY2026 (year ended 2026-01-30), filed 2026-03-16 — segments (D&I $5,581M / Civilian $1,681M); contract mix (62% cost-reimb / 22% T&M / 16% FFP); 98% U.S. government revenue (~52% Department of War); backlog $22,622M ($3,572M funded); SilverEdge ($203M net, 2025-10-15); ~$26M shutdown impact; $16M executive-transition costs; $47M IRS-audit tax benefit; buybacks ($422M FY26) and the Dec-2024 $1.2B authorization; risk factors (budget dependence, IDIQ pricing pressure, set-asides).
   • URL: https://www.sec.gov/Archives/edgar/data/1571123/000157112326000029/saic-20260130.htm
2. Form 10-Q, Q1 FY2027 (quarter ended 2026-05-01), filed 2026-06-01 — revenue $1,906M (+2%, +0.5% organic); adj. EBITDA $222M (11.6%, incl. $12M investment-sale gain); book-to-bill 1.1x; backlog $22,860M; 5→3 business-group consolidation; Q1 buyback ~$175M.
   • URL: https://www.sec.gov/Archives/edgar/data/1571123/000157112326000079/saic-20260501.htm
3. CEO transition 8-Ks — 2025-10-23 (Townes-Whitley out “without cause”; Reagan interim CEO; acc. 0001193125-25-248919) and 2026-02-17 (Reagan permanent CEO; acc. 0001193125-26-053003).
   • URLs: https://www.sec.gov/Archives/edgar/data/1571123/000119312525248919/d59585d8k.htm ; https://www.sec.gov/Archives/edgar/data/1571123/000119312526053003/d93127d8k.htm
4. DEF 14A (FY2026), filed 2026-04-22 — STI metrics and FY26 scores (66.7%; zero on revenue/EBITDA legs); PSU metrics (cumulative adj. EBITDA, cumulative adj. FCF, rTSR) and FY24–26 certification (113.1%, rTSR leg 0%); separation economics (Townes-Whitley $8.25M + continued vesting); ownership (~0.65%).
   • URL: https://www.sec.gov/Archives/edgar/data/1571123/000157112326000069/saic-20260422.htm
5. Earnings 8-Ks / exhibits — Q2 FY26 guidance cut (2025-09-04, acc. 0001571123-25-000164); Q3 FY26 beat-and-raise (2025-12-04, acc. 0001571123-25-000187); FY26/Q4 + FY27 initial guidance (2026-03-16, acc. 0001571123-26-000025); Q1 FY27 + raised FY27 guidance (2026-06-01, acc. 0001571123-26-000077); EVP departures (2025-11-13, acc. 0001193125-25-280143; 2026-06-01, acc. 0001193125-26-249675); board additions (2026-04-09, acc. 0001193125-26-148919); SilverEdge (2025-10-06, acc. 0001571123-25-000181); $500M 5.875% notes due 2033 (2025-09-25, acc. 0001571123-25-000176).
   • Example URL: https://www.sec.gov/Archives/edgar/data/1571123/000157112326000077/saic06012026ex991earningsr.htm
6. Forms 3/4/5 (60-month corpus; 352 non-derivative transactions parsed) — 37 open-market purchases vs. 18 sales; zero open-market sales since 2025-01-13; December 2025 cluster at ~$100 including Interim CEO Reagan (1,000 sh @ $100.17, acc. 0001571123-25-000198), CFO Natarajan (2,000 @ $100.49, acc. 0001571123-25-000196); March–April 2025 cluster at $108–117 including then-director Reagan (acc. 0001571123-25-000048).
7. Q4 FY2026 earnings call (2026-03-16) and Q1 FY2027 earnings call (2026-06-01) transcripts (via ROIC.ai) — FY26 “year of multiple disruptions”; recompete losses named (NASA Aegis, Cloud One, CENTCOM, Army RITS); “very hard to differentiate on predominantly cost-plus work”; “addition by subtraction”; $100M cost cuts; recompete win rates “stabilizing”; ~$400M/yr buyback plan; portfolio-review readout due December 2026.

Company releases and trade/primary press

8. SAIC Q1 FY2027 results release (GlobeNewswire, 2026-06-01): https://www.globenewswire.com/news-release/2026/06/01/3304209/0/en/saic-announces-first-quarter-of-fiscal-year-2027-results.html
9. SAIC Q4/FY2026 results release (GlobeNewswire, 2026-03-16): https://www.globenewswire.com/news-release/2026/03/16/3256144/0/en/SAIC-Announces-Fourth-Quarter-and-Full-Fiscal-Year-2026-Results.html
10. SAIC preliminary Q4/FY26 + FY27 guidance cut (2026-02-11; two unfavorable recompete decisions, procurement delays): https://www.nasdaq.com/press-release/saic-announces-preliminary-unaudited-fourth-quarter-and-full-fiscal-year-2026
11. Washington Technology — “SAIC plans partial pivot away from enterprise IT” (2026-02-12; CASTLE + Cloud One Next ~4% of revenue; EIT ~10% of revenue; Natarajan quotes): https://www.washingtontechnology.com/companies/2026/02/saic-plans-partial-pivot-away-enterprise-it/411404/ ; “SAIC loses protest fight over $1.4B Army contract it once held” (2026-05-20; CASTLE to Accenture Federal): https://www.washingtontechnology.com/contracts/2026/05/saic-loses-protest-fight-over-14b-army-contract-it-once-held/413667/ ; CEO coverage (2025-10-23; 2026-02-17): https://www.washingtontechnology.com/companies/2026/02/saics-board-stays-reagan-names-him-full-time-ceo/411456/
12. GSA ten-firm consulting review — GovExec (2025-02-27): https://www.govexec.com/management/2025/02/trump-administration-asks-agencies-cull-consultants/403355/ ; FedScoop (2025-03-01): https://fedscoop.com/gsa-tells-agencies-to-terminate-contracts-with-top-10-consulting-firms/ ; Washington Technology (2025-06-27, ~$23.3B claimed savings): https://www.washingtontechnology.com/contracts/2025/06/gsa-adds-third-set-companies-consulting-contract-review/406370/
13. TBR, “DOGE Federal IT Vendor Impact Series: SAIC” (2025-06-11; DOGE impact <1% of revenue per management; bid goals; civil-account concentration): https://tbri.com/blog/doge-federal-it-vendor-impact-series-saic/
14. Budget documents — White House FY2027 budget (April 2026): https://www.whitehouse.gov/wp-content/uploads/2026/04/budget_fy2027.pdf ; House Appropriations Democrats (June 2026, >$1T bill): http://democrats-appropriations.house.gov/news/press-releases/republicans-unveil-1-trillion-defense-funding-bill-largest-history-amid-steep ; DoD Comptroller FY2026 overview: https://comptroller.war.gov/Portals/45/Documents/defbudget/FY2026/FY2026_Budget_Request_Overview_Book.pdf ; Leonardo DRS 10-Q (FY26 approps enacted 2026-02-03; OBBBA ~$150B through GFY2029): https://www.sec.gov/Archives/edgar/data/1833756/000183375626000025/drs-20260331.htm
15. Awards — COBRA $1.4B (2025-11-20): https://www.globenewswire.com/news-release/2025/11/20/3191911/0/en/ ; Navy propulsion $242M (2025-11-24): https://www.globenewswire.com/news-release/2025/11/24/3193491/0/en/ ; HOPE 2.0 $928M (2025-06-25): https://www.globenewswire.com/news-release/2025/06/25/3105083/0/en/ ; ABMS $192M (2026-06-09): https://www.globenewswire.com/news-release/2026/06/09/3308832/0/en/ ; SilverEdge (Reuters, 2025-10-06): https://www.reuters.com/technology/saic-acquire-silveredge-government-solutions-205-million-2025-10-06/
16. M&A history — Unisys Federal $1.2B at 13x adj. EBITDA (Unisys release, 2020): https://www.unisys.com/news-release/unisys-federal-business-saic-company-achieves-guidance-on-metrics/ ; Halfaker (Washington Technology, 2021): https://www.washingtontechnology.com/2021/06/revenue-growth-new-acquisition-fuel-saics-optimism/359332/ ; LSCM divestiture $350M (ExecutiveBiz, 2023): https://www.executivebiz.com/articles/asrc-finalizes-acquisition-of-saics-logistics-and-supply-chain-management-arm
17. Peer prints — Leidos Q1 CY26 (PR Newswire, 2026-05-05): https://www.prnewswire.com/news-releases/leidos-posts-strong-first-quarter-results-and-raises-full-year-guidance-302761919.html ; CACI FQ3 FY26 (2026-04-22): https://investor.caci.com/2026-04-22-CACI-Reports-Results-for-Its-Fiscal-2026-Third-Quarter ; Booz Allen FQ4 FY26 (2026-05-22): https://247wallst.com/companies/bah/earnings

Analyst actions and consensus (third-party signals only; never adopted)

18. S&P Global poll via stockanalysis.com (accessed 2026-07-21): Hold consensus (11 analysts; ~2 Strong Buy / 8 Hold / 1 Strong Sell); mean target $120, median $125, range $93–137; FY27/FY28 EPS and revenue consensus: https://stockanalysis.com/stocks/SAIC/forecast/
19. Post-Q1 FY27 actions — TD Cowen Hold $115→$130 (2026-06-05); Jefferies Hold $100→$115 (2026-06-07); Goldman Sell $85→$96 (2026-06-11); Citi Buy reiterated $132 (2026-07-01); Zacks earnings coverage (2025-06-03, 2025-12-05, 2026-03-17, 2026-06-01).

Market and third-party data (cross-check only; not primary)

• AZI price series (SAIC daily OHLCV 2013–2026; 21/50/200-EMAs; beta) and AZI valuation_index own-history percentiles (SAIC composite 39th: P/E 9th, P/B 54th, P/S 54th; LDOS 15th; BAH 1st; KBR 34th; CACI 51st; GD 93rd) — pulled 2026-07-21; third-party aggregates.
• FactorsToday factor model — SAIC loadings (Base model: SmallSize +0.52, Market +0.39, Momentum −0.22; R² 0.111), leaderboard, stock-info, specific-vol (36.6%), related-stocks, factor-returns/historic (Value z +1.44; A&D soft; Technology extreme leadership) for SAIC and peers (LDOS, CACI, BAH, GD, KBR) — pulled 2026-07-21; third-party statistical estimates.
• ROIC.ai MCP — statements, ratios, enterprise value, multiples (SAIC + peers), transcripts, profile. Rejected data points: ROIC’s SAIC “FCF” field equals CFO (capex misclassified — filing values used); ROIC’s 2026-07-20 SAIC close ($115.66, internally inconsistent bar — AZI/FactorsToday’s $112.66 used); FY24 operating income per the filing ($741M) includes the $247M divestiture gain that ROIC strips.
• Prior coverage by the author — Leidos (July 17, 2026), CACI (July 2, 2026), Booz Allen Hamilton (June 14, 2026), General Dynamics (June 14, 2026) — used for peer sizing, ROIC context, industry framing, and the DOGE regime cross-read.