CACI International Inc (NYSE: CACI) — From Billable Hours to Hyperspectral Sensors: A Roll-Up That Won the Regime Shift, Priced to Keep Winning

Report date: 2026-07-02 · Price: $502.83 (2026-07-02 close, +7.7% on the day) · Shares: ~22.1M · Market cap: ~$11.1B · Enterprise value: ~$16.1B · Fiscal year-end: June 30 (FY2026 Q3 reported 2026-04-23)

⚡ Claude’s Take

This block is the author’s own independent opinion. It is general information, not investment advice. The analysis in the sections below is presented without a recommendation or price target; the single directional view is confined to this block.

Verdict: HOLD at ~$503 / accumulate-on-weakness below ~$430–450. Not a short, not a table-pounder here. Framing: a quality-improving government-technology roll-up that just de-rated ~30%, now trading at a full-but-fair multiple with the balance sheet levered for a big space/AI bet.

CACI is the government-services name that won the regime shift that broke its peers. While Booz Allen’s civil-consulting book collapsed ~22% under DOGE and the “consulting-is-waste” purge, CACI — 75% Department of Defense, heavy in electronic warfare, SIGINT, space and C4ISR, and increasingly selling technology rather than hours — grew revenue ~10% and kept raising guidance. That is not luck; it is a decade of deliberate acquisitions moving the company up the value chain from staff augmentation toward differentiated hardware and software. The reward is a business compounding revenue ~8.5%/yr and adjusted EPS at a low-double-digit clip, with a $33B backlog and a 1.2x book-to-bill.

So why only a HOLD? Three things temper the “quality winner” reflex. First, price. After a sharp fall from a ~$662 January-2026 all-time high to a $443 June low, the stock has bounced to ~$503 — but that is still the ~60th percentile of its own decade-long valuation range (composite), ~18x forward adjusted EPS and ~14x EV/EBITDA. This is a fair multiple, not a cheap one; the easy money was the recovery off $443, and the sell-side’s $673 average target already underwrites the story continuing. Second, the returns. For all the growth, CACI earns only ~10% ROIC and ~11% ROE — roughly its cost of capital — because it has paid up for M&A: ~$6.5B of goodwill and ~$2.2B of intangibles now sit against $4.3B of equity, and tangible book is negative. The roll-up works for the seller of capital (per-share FCF grows) but only modestly for the return on capital. Third, the bet. The $2.6B all-cash ARKA acquisition (space sensors + “agentic AI,” closed March 2026) took net-debt/EBITDA from ~2.4x to ~4.7x. If ARKA scales and the budget stays friendly, it’s brilliant; if space budgets wobble or integration slips, CACI has bought a full-priced asset at the top of its own leverage range.

I land on a genuine HOLD: own it through the compounding, but the risk/reward is only decent at ~$503, not screaming. My fair-value zone is roughly $470–560 (17–20x forward adjusted EPS of ~$28, ≈13–14x EV/EBITDA); the risk/reward turns genuinely attractive back toward the $430–450 June-low zone (~15–16x). Conviction: medium. The single fact that flips me more bullish: ARKA-and-organic revenue re-accelerating with net-debt/EBITDA back below ~3.5x within 18 months — proof the bet delevers itself. The single fact that flips me bearish: organic growth decelerating to low-single-digits into a shutdown/CR air pocket, or a goodwill write-down on ARKA/Azure Summit — proof the roll-up was buying multiple, not moat. Tag: it climbed the value chain the right way — just don’t confuse a great operator with a cheap stock.

📈 Stock Price Action — Five-Year Event Map

Factual price history and the events that most plausibly drove each move. Price moves are Fact; attributed causes are Interpretation. No recommendation, no price target, no chart-pattern reading.

The arc. Over five years CACI roughly doubled and then some — from ~$255 (mid-2021) to an all-time high of ~$662 on 2026-01-23, before a ~33% draw-down to a 52-week low of ~$443 on 2026-06-25 and a sharp bounce to ~$503 by 2026-07-02. The stock now sits ~24% below its January-2026 peak, near the middle of a wide 52-week range ($443–$662). Unlike its peer Booz Allen (which round-tripped to a multi-year low), CACI is a long-term compounder (~+16.7%/yr over ten years) that suffered a normal-sized cyclical de-rating, not a franchise break.

Cycle narrative. (1) 2021–23 was a quiet, steady climb on organic growth and small capability deals. (2) The 2023–24 run reflected accelerating growth (FY24 revenue +14.4%), the transformative Azure Summit electronic-warfare acquisition, and a post-election bid for defense names. (3) The Feb-2025 crash was the DOGE shock — the market sold all government contractors on fears of mass federal spending cuts, without discriminating between civil-consulting exposure and defense/intel. (4) CACI then more than doubled off the low as it became clear the cuts fell on civil consulting (crushing Booz Allen) and not on defense/intelligence/space, where CACI is concentrated; the ARKA announcement in December 2025 added a growth narrative and carried the stock to its $662 ATH on the Q2-FY26 print. (5) The H1-2026 de-rating combined generic defense-sector profit-taking, appropriations/continuing-resolution uncertainty, the leverage taken on for ARKA, and a Middle-East de-escalation that trimmed the sector’s risk premium. (6) The recent bounce is a sector move off an oversold low; there is no confirmed company-specific catalyst on 2026-07-02, and the day’s +7.7% should be read as tape, not news.

1. Executive Summary

CACI International is a ~$9.5–9.6B-revenue (FY2026E) prime contractor that sells technology and expertise almost entirely (~95%+) to the U.S. federal government — and, within that, overwhelmingly to national security: 75% of revenue comes from the Department of Defense, with most of the balance from the Intelligence Community and a modest federal-civilian and international (UK/Europe) tail. It is a member of the cleared-services oligopoly (with Booz Allen, Leidos, SAIC, GDIT and Parsons), delivering ~96% of revenue as a prime contractor, with ~25,000 employees, a majority of them security-cleared. Founded in 1962 and headquartered in Reston, Virginia, CACI has spent the last decade deliberately repositioning from a labor-based services firm toward what it calls “differentiated Technology” — electronic warfare, signals intelligence, space sensing, secure communications, photonics, counter-UAS and, most recently, “agentic AI” — bolted on through serial acquisition.

That strategy has produced enviable growth for the category: revenue compounded ~8.5%/yr from FY20–FY25 and accelerated into the recent period (+14.4% FY24, +12.6% FY25, ~10% guided FY26), while EBITDA margins expanded to ~11.8–12.3% — the high end of the services peer set. Crucially, CACI grew through the 2025–26 government-spending upheaval that broke its civil-consulting-heavy peers. Where the DOGE / “consulting-is-waste” purge cut Booz Allen’s civil book ~22% and drove that stock to a multi-year low, CACI’s DoD/intelligence/technology concentration proved defensive: the cuts largely missed it, and management raised FY26 guidance twice.

The complications are three. First, capital returns are only ordinary. Despite the growth, CACI earns ~10% ROIC and ~11% ROE — roughly its cost of capital — because it has paid full prices for acquisitions: the balance sheet now carries ~$6.5B of goodwill and ~$2.2B of intangibles against $4.3B of equity, and tangible book value is negative. The underlying labor/technology business earns far more on tangible capital; the reported return is diluted by what CACI paid to buy growth. Second, the balance sheet is stretched. The $2.6B all-cash ARKA acquisition (space-based EO/IR and hyperspectral imaging sensors plus agentic-AI GEOINT software, closed March 9, 2026) lifted net-debt/EBITDA from ~2.4x to ~4.7x — the top of CACI’s historical leverage range — and widened the gap between GAAP and adjusted EPS (Q3-FY26 GAAP diluted $5.90 vs. adjusted $7.27) via intangible amortization. Third, valuation is full, not cheap. After a bounce off its June low, the stock trades at ~18x forward adjusted EPS and ~14x EV/EBITDA — the ~60th percentile of its own decade-long range, and the sell-side’s ~$673 average target already assumes the story continues.

The investment debate reduces to whether CACI is (a) a genuine quality-compounder whose technology mix-shift and $33B backlog justify a premium GovCon multiple and support durable low-double-digit per-share growth, or (b) a serial acquirer whose ~10% ROIC and negative tangible book reveal that it has been buying multiple and growth rather than moat, now levered near its limit into an uncertain-appropriations, potentially AI-deflated future. This memo presents the evidence on both sides. It offers no recommendation and no price target outside Claude’s Take above.

2. Business Overview

What it does. CACI provides “Expertise and Technology” in support of national-security missions across the intelligence, defense and federal-civilian markets, primarily in the United States and secondarily in the United Kingdom and continental Europe. In plain terms, it (1) staffs and runs mission and IT programs for government agencies (the “Expertise” business — analysts, engineers, cleared personnel embedded in agency missions), and (2) designs and sells differentiated hardware and software — electronic warfare and SIGINT systems, space sensors, secure/tactical communications, photonics and optical communications, signals processing, counter-UAS and, increasingly, AI-enabled software (the “Technology” business). Management frames the portfolio along this Expertise-vs-Technology axis rather than by end product; the technology content is what separates CACI from pure staff-augmentation peers and is the source of its margin premium and its acquisition strategy.

Reporting segments. CACI reports two geographic operating segments: Domestic Operations (the vast majority of revenue — U.S. government customers across Digital Solutions, C4ISR, Enterprise IT, Cyber, Space, Engineering Services and Mission Support) and International Operations (a small tail, conducted through CACI Limited and CACI BV in the UK/Europe, offering IT services plus proprietary data and software products to commercial and government customers). The segment structure is coarse — it does not break out the strategically important Expertise-vs-Technology or by-capability economics, which is a genuine disclosure limitation for outside analysts.

Customer mix (FY2025). The concentration on national security is the single most important fact about the business and the reason it behaves differently from Booz Allen:

• Department of Defense: 75.4% of revenue (74.4% FY24) — the dominant customer group.
• The balance is Intelligence Community, federal-civilian agencies, a modest commercial component, and international.
• ~95.7% of revenue is earned as a prime contractor (95.1% FY24) — CACI controls the customer relationship on nearly all of its work, a stronger position than a subcontractor-heavy book.
• Federal-government contracts are expected to remain “the primary source of our revenues for the foreseeable future.” Suspension/debarment from DoD or IC contracting is a named existential risk.

How it makes money — contract structure (FY2025). Revenue is predominantly labor delivered under cost-based government contracts, with a meaningful and rising layer of product/technology work:

• Cost-plus-fee: 60.5% of revenue — costs (including inflation) largely pass through; markup is fee-capped, which limits both downside and operating leverage.
• Time-and-materials: 13.2% — labor billed at negotiated rates.
• Fixed-price: ~26.3% (the remainder) — notably higher than Booz Allen’s ~19%, reflecting CACI’s greater product/technology content; fixed-price carries more execution risk but more margin upside and is where differentiated technology is sold.

Recurring vs. non-recurring / backlog. Revenue is contractual and visible but recompete-exposed. Total backlog was ~$33.4B at Q3-FY26 (~3.6x annual revenue), up 6% YoY; funded backlog grew 19% YoY; the trailing book-to-bill was 1.2x and the weighted-average duration of awards ~6 years. The vast majority of revenue flows through IDIQ (“indefinite-delivery / indefinite-quantity”) vehicles and GWACs (e.g., GSA schedules) as task orders that are periodically re-competed. So revenue is best understood as an incumbency annuity: highly visible and sticky (incumbents win most recompetes on mission knowledge and cleared staff), but not subscription-recurring — each vehicle must be re-won on a rolling basis, and a lost recompete can transfer both the work and the people.

Scale and people. ~25,000 full- and part-time employees (June 30, 2025), a majority holding U.S. security clearances — the workforce and its clearances are the true productive asset and a genuine barrier (the work cannot be offshored to low-cost labor).

3. Industry Dynamics

Structure — a fragmented oligopoly of cleared primes. The federal technology-services market is a large, appropriations-backed demand pool served by a handful of scaled prime contractors — CACI, Leidos (~$17.2B revenue), Booz Allen (~$11.2B), GDIT (a ~$13B segment of General Dynamics), SAIC (~$7.5B), Parsons (~$6.4B) and V2X (~$4.3B) — plus thousands of smaller and small-business-set-aside firms. Work is competed as task orders under multiple-award IDIQ and GWAC vehicles (GSA schedules and the like); award turns on past-performance scoring, technical responsiveness and price. Barriers to entry by new firms are real — security clearances, facility accreditation, DCAA-compliant cost-accounting systems and a qualifying past-performance record take years to build — but barriers among the incumbent primes are low, since they all clear their people and hold the vehicles. The result is high demand stability with persistent price competition at the commoditized-labor end.

Procurement dynamics. CACI reports that LPTA (“lowest-price, technically-acceptable”) pressure “has moderated, though price still remains an important factor,” with best-value source-selection increasingly used for high-end electronic-warfare, space, cyber and intelligence work. That is structurally favorable to CACI’s differentiation strategy: the more the government buys mission-critical capability rather than bodies, the more CACI’s technology content earns a premium and the less its high-end book is exposed to price shootouts. The low-end Expertise book remains price-competitive and recompete-exposed. Bid protests and procurement delays are a chronic friction that pushes awards to the right.

The FY2025–26 budget and DOGE backdrop — the single most important recent development. On 2025-03-15 the government was funded by a full-year continuing resolution (the first time the DoD ran a full-year CR), but at higher appropriation levels — defense raised to roughly $893B — with new-program authority and expanded transfer authority. Budget priorities are rotating toward defense, intelligence, space and border security, exactly CACI’s core (~75% defense). At the same time, the Department of Government Efficiency (DOGE) and a broad “consulting-is-waste” push targeted advisory/consulting spending. This bifurcated the industry: civil-consulting-heavy firms were hit hard (Booz Allen took >$200M of terminations, cut ~7% of staff and guided its civil book down “low double digits”; Accenture Federal and Deloitte were also affected), while defense/intelligence/technology-weighted contractors rode the budget tailwind.

Does the “consulting-is-waste” narrative hurt CACI less? Yes — materially. CACI’s federal-civilian book is only ~20% of revenue and skews to IT modernization and mission delivery, not McKinsey-style advisory; its DoD/IC concentration (~75%+) sits on the favored side of the rotation. This is the empirical reason CACI grew ~10% while Booz Allen shrank 6.4% in overlapping periods. The residual industry risk is not thesis-specific to CACI: continuing resolutions and shutdowns delay new starts and awards regardless of priority (visible in CACI’s soft 0.9x Q3 book-to-bill), and long-run insourcing / small-business set-asides can erode the recompete base.

Capital-cycle lens (Marathon). Elevated defense/space budgets and a wave of “defense-tech” enthusiasm are drawing capital into the space-sensing, EW and counter-UAS niches CACI is buying into — venture-funded newcomers (Anduril, Palantir-adjacent software, small-sat sensor startups) plus the primes all bidding up assets. That is a caution flag for acquisition prices (CACI paid a full multiple for ARKA into a hot space-tech market) even as it is a tailwind for demand. High returns in a hot niche attract capital and tend to mean-revert; the durable winners will be those whose capability is genuinely defensible, not merely acquired at the top of a funding cycle.

Verdict: a moderately attractive industry, tilting favorable for defense/IC-weighted primes. Large, stable, appropriations-backed demand with high entry barriers, offset by fragmentation, commodity-labor price competition, protest/CR friction, and a single dominant customer with the structural power to insource, terminate for convenience and reprioritize at will. CACI sits on the better side of the current rotation — but “better side of a government-budget cycle” is a weaker foundation than a private-market moat, and the cycle can turn.

4. Competitive Position

The moat, named. CACI’s competitive advantage is a narrow moat built from clearances + prime incumbency + accumulated mission knowledge, with a small but growing core of genuine acquired product IP (RF/EW/SIGINT hardware and space sensors). In Greenwald’s taxonomy this is primarily an intangibles/customer-captivity advantage (cleared workforce, past-performance qualifications, embedded program knowledge that raises switching costs) supported by modest scale (large enough to prime major integration programs). It is not a wide moat: CACI itself concedes it holds “a relatively small share of the addressable market” with “no single competitor dominant,” and it flags recompete losses as a top risk. Barriers protect the incumbent primes as a class from newcomers far more than they protect CACI from its peers.

Where the moat is real. (1) Clearances and facility accreditations are difficult and time-consuming to obtain — a genuine barrier against new entrants and offshore labor (the work cannot be sent to low-cost geographies). (2) Incumbency on multi-year mission programs creates real switching costs: incumbents win the majority of recompetes because they hold the cleared staff and the mission context, and losing an incumbent risks program disruption. (3) Acquired product IP — Azure Summit’s software-defined RF/EW/SIGINT hardware (~$440M year-one revenue at ~25% adjusted-EBITDA margin, roughly double the corporate average) and ARKA’s EO/IR and hyperspectral space sensors, photonics and optical-communications technology — is genuinely product-like, higher-barrier and margin-accretive. This is where a durable advantage could form: differentiated payloads and sensors are far harder to recompete away than staff augmentation.

Where it is weak — and the central skeptical point. For all the “differentiated Technology” branding, the economics still say services, not software. “Technology” is now 55–57% of revenue, yet consolidated operating margin is only ~8.9% and EBITDA margin ~12%. A true software/product mix would show 20–40% margins. Most of “Technology” is agile-software, ERP-modernization and cyber work delivered by cleared engineers — labor with a technology label, not licensed IP. The genuinely product-like franchises (Azure Summit, ARKA, SA Photonics) are a minority of the ~$9.5B base, and CACI is buying them rather than building them organically — stacking ~$6.5B+ of goodwill and intangibles onto a 9%-margin services core. The differentiation is real but purchased, and its durability depends on integrating these franchises into higher-margin businesses rather than absorbing them into the services average.

Direct comparison to peers. CACI is more differentiated-technology and DoD/IC-weighted than Booz Allen (advisory/civil-tilted, which is exactly why BAH’s civil book cratered and CACI’s did not) and than V2X (base operations/logistics). It has built real EW/space sensor IP that pure staff-augmentation peers lack, and its ~12% EBITDA margin is at the top of the services peer set. Against Leidos and GDIT it is subscale on the very largest full-and-open integration programs. The honest read: a good, not great, competitive position — a narrow, defensible incumbency moat with a real but still-minority product-IP core, carried on a highly concentrated (95.7% federal, 75.4% DoD, top-10 contracts = 46% of revenue) customer base.

Moat-to-financial-outcome test. Does the moat show up where it should? Partially. Retention/incumbency shows up as a ~$33B backlog and high recompete win rates; the technology tilt shows up as a top-of-peer-set EBITDA margin and above-GDP organic growth. But it does not show up in ROIC — which at ~10% is merely at the cost of capital — because the price paid for the acquired differentiation offsets the operating advantage. A moat that cannot lift returns on invested capital above the cost of capital is, at the corporate level, a moat the seller captured. Verdict: a durable but narrow advantage in a good-not-great competitive position; the differentiation is genuine at the program level but has been bought, not compounded, at the corporate level.

5. Growth History and Forward Opportunities

The record. Revenue compounded from $6,044M (FY21) to $8,628M (FY25) — a ~9.3% CAGR that accelerated into the recent period: +2.6% FY22, +8.1% FY23, +14.3% FY24, +12.6% FY25, and ~10% guided for FY26 ($9.5–9.6B). GAAP diluted EPS rose from $16.43 (FY23) to $22.32 (FY25); adjusted EPS grew faster (Q3-FY26 adjusted diluted EPS $7.27, +17% YoY). This is genuinely strong for the category — Booz Allen’s revenue fell 6.4% over an overlapping window.

But how much is organic? This is the key quality question, and the answer is sobering: the double-digit FY24–FY25 headline growth was substantially acquired, not organic. Q3-FY26 organic growth was 6.8% and nine-month-FY26 organic just 5.6%; the reported acceleration came from Azure Summit (~$440M year-one revenue), Applied Insight ($314M deal) and now ARKA ($2.64B). Sustainable organic growth is mid-to-high single digits — respectable and demand-backed, but not the low-double-digit rate the headline implies. An investor paying a premium multiple for “double-digit growth” should underwrite the organic ~6% plus a continued, debt-funded M&A cadence — a very different and riskier proposition.

What’s driving it. (1) Budget rotation toward CACI’s strengths — electronic warfare, SIGINT, space-domain awareness, counter-UAS, cyber and IT modernization — all budget-priority growth areas. (2) The technology mix-shift, deliberately engineered through acquisition, raising both growth and margin. (3) Strong recompete performance and a 1.2x trailing book-to-bill sustaining the base. (4) The ARKA platform, adding space EO/IR + hyperspectral sensing and “agentic AI” GEOINT — a new, higher-growth adjacency (~$150M contribution already in FY26 guidance).

Forward opportunities. The space and agentic-AI adjacencies are the genuine forward story: space-domain awareness and resilient sensing are among the fastest-growing national-security budget lines, and ARKA gives CACI a differentiated payload/sensor position rather than a services seat. Counter-UAS and electromagnetic-spectrum operations are similarly funded growth areas where Azure Summit gives CACI product IP. The international segment (proprietary UK software/data, ~21% margins) is structurally attractive but too small (~3% of revenue) to move the consolidated needle. The swing factor is appropriations timing — CRs and shutdowns delay new-start awards (the soft 0.9x Q3 book-to-bill is the tell), so even a well-positioned CACI can see a quarter or two of decelerated organic growth on Washington’s calendar, not its own execution.

Verdict: high-quality demand positioning, mixed-quality growth composition. CACI is aimed at the right budget lines and is mixing toward higher-margin technology — but the reported growth leans heavily on a debt-funded roll-up, and the durable organic rate is mid-single-digit, not double-digit. The growth is good; it is not as good, or as organic, as the headline.

6. Financial Quality

Revenue and margins. CACI reports revenue net of costs (direct + indirect + D&A), so its “operating margin” of ~8.9% (FY25) is the relevant profitability line, with EBITDA margin ~11.1% (FY25) rising to ~12.3% in Q3-FY26 and guided to 11.8–11.9% for FY26. Margins have expanded ~200bp over five years on the technology mix-shift and program execution — a real, if modest, improvement. The incremental EBITDA margin from acquired product franchises (Azure Summit ~25%) is well above the corporate average, which is the mechanical case for the M&A strategy: each differentiated-technology deal, if integrated without margin erosion, lifts the blended rate.

Returns on capital — the crux, and the weakest link. Despite the growth and margin expansion, CACI earns only:

• ROE ~10.8% (FY25), down from ~15.4% (FY21) as the equity base and acquired intangibles grew;
• ROIC ~10.0% (FY25) — roughly its cost of capital.

The reason is structural: CACI has paid full prices for acquisitions, stacking ~$6.5B of goodwill and ~$2.2B of intangibles (post-ARKA) against $4.3B of equity, so tangible book value is negative (~−$100/share and worsening). The underlying labor/technology business is capital-light (capex <1% of revenue) and earns very high returns on tangible capital; the corporate ROIC is dragged to ~10% by the price paid for growth. This is the defining financial signature of the business: operationally excellent, but a serial acquirer that has largely paid away the excess returns to the sellers of the companies it bought. A roll-up that earns its cost of capital creates value for shareholders only through (a) per-share compounding funded by cheap-enough capital and (b) genuine post-deal margin/synergy capture — not through a return spread.

Cash flow and its quality. Operating cash flow was $547M (FY25) against net income of $500M; with capex of only ~$60–70M, real free cash flow runs ~$480–550M/year. Two quality caveats: (1) CACI uses an accounts-receivable purchase (securitization) facility that accelerates collections and flatters the timing of operating cash flow — normalize for it when judging underlying cash generation; and (2) working-capital swings around large program ramps and the ARKA integration can move OCF materially quarter to quarter (unbilled receivables rose sharply in FY25). FCF conversion is solid but not pristine, and forward FCF carries a higher interest burden post-ARKA.

The GAAP-vs-adjusted gap. The widening spread between GAAP diluted EPS ($5.90 in Q3-FY26) and adjusted diluted EPS ($7.27) — a ~$1.37/quarter, ~19% gap — is driven mainly by amortization of acquired intangibles ($890M of ARKA customer relationships + $290M of ARKA technology, on top of the prior intangible base). “Adjusted” figures exclude this as non-cash, but it is a real economic cost of the roll-up: it represents capital spent to buy customer relationships and technology that deplete over time. An investor should weight GAAP earnings more heavily than management’s adjusted figures when judging the return on the acquisition strategy, even while using adjusted figures for run-rate comparability.

Balance sheet and leverage. Post-ARKA (2026-03-31): total debt ~$5.6B, net debt ~$5.0B, equity $4.3B. Net-debt/EBITDA jumped to ~4.7x from ~2.4x a quarter earlier — the top of CACI’s historical range — with interest coverage (EBITDA/interest) of ~5.5x, still comfortable but down from ~5.8x. The debt is the direct cost of the $2.6B all-cash ARKA deal. CACI has historically de-levered quickly after large acquisitions (it carries no dividend obligation and generates steady FCF), so the base case is a glide back toward ~3x over 18–24 months — but that glide is now the load-bearing assumption for both the credit and the equity: it requires organic growth and margins to hold while free cash flow is directed at debt rather than buybacks or new deals.

Verdict: economics improve modestly with scale, but the returns are ordinary. Margins expand on the technology mix, cash flow is steady and capex-light, and per-share metrics compound — but ROIC sits at the cost of capital, tangible book is negative, and the balance sheet is now stretched. This is a well-run, cash-generative services-plus-technology business whose reported returns have been diluted by the price of its own growth strategy. The quality is in the operations and the mix-shift; it is not in the return on invested capital.

7. Capital Allocation

The framework, in one line: reinvest everything into acquisitions, pay no dividend, buy back stock opportunistically between deals, and delever after big ones. CACI has never paid a dividend, by design; its stated financial objective is to grow free cash flow per share, and its growth engine is M&A. Judging capital allocation therefore means judging the M&A.

M&A track record. CACI’s model is to buy differentiated, clearance-heavy technology/IP businesses serving DoD/IC customers and fold them into the platform. The deals skew intangible-heavy — the two largest allocate ~99% of purchase price to goodwill + intangibles:

Has M&A created value? The evidence cuts both ways, and lands on “competent, not exceptional.” The genuinely bullish fact is that CACI has never recorded a goodwill impairment — a rare clean record for a serial acquirer of this cadence, and real evidence of disciplined pricing and effective integration. Azure Summit at ~11.6x forward EBITDA for a scarce RF/EW asset with ~25% margins is a defensible strategic price. But the skeptical read dominates the return math: consolidated ROIC (~10%) sits right on top of an ~8–9% WACC, so the value spread on acquired capital is thin; the two deals that now dominate invested capital (Azure, ARKA) are too recent to have proven their returns; and ~99% of ARKA’s price is goodwill + intangibles — CACI is paying for future contract wins and an assembled cleared workforce, not hard assets, into a hot space-tech acquisition market (a Marathon capital-cycle caution). Critically, the entire “expertise → technology” pivot is an acquisition story: CACI discloses essentially no internal R&D line (R&D and bid-&-proposal are buried in indirect costs), so the differentiation is bought, not built.

Buybacks — small, opportunistic, subordinated to deals. Repurchases were $273M (FY23, incl. a $250M ASR at ~$303.57), $161.5M (FY24 at ~$318.99) and $168.6M (FY25 at ~$344.35) — all well below the current ~$500 price, so accretive in hindsight, but trivial next to the billions spent on Azure + ARKA. Only ~$187.3M remained authorized at 6/30/25, and buybacks are effectively paused while CACI delevers post-ARKA. Share count has drifted down from ~25.1M (FY20) to ~22.1M — a ~2%/yr reduction that is a genuine but modest per-share tailwind.

Balance sheet and leverage policy. ARKA was funded with debt: principal jumped from $2,942.6M (6/30/25) to $5,210.8M (3/31/26), taking net-debt/EBITDA from ~2.4x to ~4.7x. The maturity profile is well-laddered (revolver and term loans due 2030–31, senior notes due 2033; the credit facility was upsized to $3.25B in Nov 2025), and coverage (~5.5x EBITDA/interest) is comfortable, so this is a stretched-but-manageable balance sheet, not a distressed one. The base case is a glide back toward ~3x over 18–24 months, funded by FCF directed at debt rather than deals or buybacks — which is now the load-bearing assumption for the equity.

Compensation — the empire-building flag. This is the sharpest governance concern. The annual bonus for the CEO/CFO is 50% CACI Revenue + 50% CACI EBITDA — with no ROIC, no EPS, no margin, and no relative-TSR gate — and acquired companies are folded into the bonus targets after closing. The long-term plan (65% of target pay) is better but still a growth metric: PRSUs vest on three-year free-cash-flow growth (not per-share, not return-on-capital), and that FCF is itself flattered by the receivables facility (below). In sum, the incentive plan rewards getting bigger — Revenue, EBITDA and FCF growth, all directly inflatable by debt-funded acquisition — not capital efficiency. For a serial acquirer earning ~10% ROIC, that is precisely the wrong scorecard. The meaningful offset is strong ownership guidelines (CEO 8× salary, CFO 7×, NEOs 5×) that tie executives’ own wealth to the share price. CEO John Mengucci earned $16.7M in FY25 (CFO $3.76M).

Insider behavior. Across the trailing ~2-year Form 4 corpus there is not a single open-market purchase (code P). Activity is entirely routine (RSU/PRSU vesting, tax withholding, small director grants) plus modest open-market sales — the CEO sold ~10,175 shares (~$5M) in September 2025 at ~$490–494; the GC and several directors made small sales into the early-2026 strength. This is a classic grant-and-sell pattern, not a red flag — but the complete absence of conviction buying through a period in which the stock more than doubled is a mild negative tell about how insiders view the risk/reward at recent prices.

Verdict: competent capital allocation undermined by an empire-building scorecard. The clean no-impairment record, disciplined-enough deal pricing, sub-market buybacks and laddered debt are real positives; the ~10% ROIC hugging WACC, the ~99%-intangible mega-deals, the negative tangible book, the FCF-flattering receivables facility, and a compensation plan that pays for scale rather than returns are real negatives. Management has allocated capital capably — it has not demonstrably allocated it in a way that earns a durable spread over the cost of capital.

8. Changes and Headwinds — Last Two Years

1. The DOGE / spending-reform regime shift (2025–26) — net positive for CACI. The single biggest change in CACI’s world was the federal-spending upheaval that bifurcated GovCon. The “consulting-is-waste” purge and DOGE terminations hammered civil-consulting-heavy peers (Booz Allen cut ~7% of staff and guided civil revenue down double digits), while CACI’s ~75% DoD / ~20% mission-oriented-civil mix rode the budget rotation toward defense, intelligence, space and border security. CACI raised FY26 guidance twice through the period. The residual headwind is procurement friction — the full-year continuing resolution, appropriations delays and shutdown risk push new-start awards to the right (visible in the soft 0.9x Q3 book-to-bill), even as the dollars ultimately flow.

2. The ARKA acquisition (Dec 2025 announced, Mar 2026 closed) — a strategic step up the value chain and a leverage event. CACI’s largest-ever deal ($2.6B all-cash) added space-based EO/IR and hyperspectral sensing, photonics/optical communications and “agentic AI” GEOINT software — a genuinely higher-growth, more product-like adjacency that fits the technology pivot. It also took leverage to ~4.7x and added ~$1.18B of amortizing intangibles that widen the GAAP-vs-adjusted EPS gap. This is the defining event of the last two years and the fulcrum of the forward thesis.

3. Prior transformative deal — Azure Summit (Oct 2024). The $1.3B RF/electronic-warfare acquisition (~25% EBITDA margins) was the prior step-change, materially lifting CACI’s technology content and margin mix and driving much of the FY25 growth.

4. Leadership and governance. Continuity at the top (CEO John Mengucci since 2019) with an ongoing U.S.-operations leadership transition and a new EVP/COO (Dave Young) appointed June 2026. Founder/Chairman J.P. “Jack” London died in 2021; the board is now independent-chaired with a clearance-heavy, ex-DoD/IC roster.

5. Litigation. A $42M jury judgment (November 2024) in the long-running Abu Ghraib-related civil suit is under appeal, with no amount accrued — a tail legal/reputational item, not financially material at the corporate scale but worth monitoring.

6. Tax/cash-flow mechanics. The TCJA R&D-capitalization rule reduced FY25 operating cash flow by ~$47.4M; its legislative reversal (OBBBA) is expected to benefit FY26 cash taxes — a non-recurring FCF tailwind to normalize out before extrapolating the (LTI-relevant) FCF trend.

Verdict: on balance these changes strengthen the thesis operationally (CACI is positioned on the right side of the budget rotation and has added a differentiated growth platform) while raising financial risk (leverage, integration, a wider GAAP-adjusted gap). The business is better; the balance sheet is more fragile.

9. Risk Analysis

Catastrophic-loss risk: low. CACI is investment-grade, cash-generative, diversified across thousands of task orders, with no single existential contract. The realistic downside is a de-rating plus a growth stall, not insolvency; a total loss would require a systemic collapse in U.S. defense spending — not a base case. The realistic asymmetry risks are #1 (a budget-timing air pocket that decelerates organic growth just as leverage is elevated) and #2/#3 (ARKA failing to earn its keep while the balance sheet is stretched), which could combine to compress both estimates and the multiple.

10. Valuation Discussion — Embedded Expectations

No price target, no recommendation. This section frames what the current price implies and stress-tests it against scenarios.

Where the multiple sits. At $502.83 (2026-07-02) on ~22.1M shares, market cap is ~$11.1B; adding post-ARKA net debt of ~$5.0B gives an enterprise value of ~$16.1B. Against FY2026 guidance (revenue $9.5–9.6B; EBITDA margin 11.8–11.9% → EBITDA ~$1.13B; adjusted net income $615–630M → adjusted diluted EPS ~$28), that is:

• ~18x forward adjusted EPS
• ~14.2x EV/FY26E EBITDA (and ~12.5–13x on a full-year-ARKA FY27 run-rate)
• ~1.7x EV/revenue
• ~4.5–5% free-cash-flow yield on equity (pre the FY26 OBBBA tax tailwind).

On CACI’s own history, this is middling-to-slightly-full, not cheap: the AZI own-history percentiles are composite ~60th, P/E ~49th, P/B ~66th, P/S ~65th, and EV/EBITDA of ~14x sits above the ~13x decade average (range ~10.8x–17x). This is the key contrast with Booz Allen, which trades at the ~10th percentile of its range: CACI is priced as the quality-and-growth leader of the services group, not as a distressed bargain.

Peer context. CACI’s ~14x EV/EBITDA is at the premium end of the pure GovCon services set — above Booz Allen (~10.8x, distressed), Leidos and SAIC (~10–12x) and V2X (~9x), and roughly in line with growth-tilted Parsons (~14–16x). The premium is earned by CACI’s faster growth, top-of-peer-set margins and defense/technology tilt — but it is a premium, so the multiple offers little cushion if growth decelerates.

Embedded expectations — what must be true at $503. A ~14x EV/EBITDA multiple on a business that (a) earns only ~10% ROIC and (b) generates most of its headline growth from debt-funded M&A is implicitly underwriting continued successful roll-up execution: that organic growth holds at mid-single-digits through appropriations noise, that ARKA and Azure integrate into durable higher-margin franchises, that margins keep drifting up on the technology mix, and that leverage glides back toward ~3x without a demand shock. The market is pricing the defensive DoD positioning, the margin expansion and the backlog visibility correctly; it is arguably pricing too little probability on the two-sided risk that ARKA fails to earn a spread over its cost of capital and that a CR/shutdown air pocket decelerates organic growth just as leverage is elevated.

Scenario analysis (≈3-year forward, EV-to-equity; shares ~flat ~22M).

The distribution is symmetric-to-modestly-favorable: a probability-weighted center of gravity slightly above spot, with a genuine fat left tail from the combination of elevated leverage, integration risk and budget-timing. This is not a mispriced bargain; it is a fairly priced quality-growth compounder where the reward for being right (base/bull) is decent and the penalty for a stumble (bear) is real. The sell-side’s ~$673 average target (range $513–787) essentially encodes the base-to-bull path with little weight on the bear.

A note on the “right” earnings number. Because ~$1.18B of ARKA intangibles (plus the prior base) amortize for years, GAAP EPS (Q3-FY26 diluted $5.88) understates cash earnings while adjusted EPS ($7.27) overstates the return on the acquisition capital. For run-rate comparability use adjusted; for judging whether the roll-up creates value, weight GAAP and the ~10% ROIC. The honest owner-earnings figure sits between the two.

11. Variant Perception

Consensus view. The sell-side is bullish (consensus Buy; ~$673 average target, ~34% above spot; range $513–787). The consensus narrative: CACI is the best-positioned mid-cap GovCon prime — defense/intel-weighted, riding the budget rotation, transforming into a differentiated defense-technology company via Azure Summit and ARKA, with a $33B backlog and a credible path to sustained low-double-digit adjusted-EPS growth. Recent target cuts (Citi to $519, JPM to $645) trimmed the optimism but kept the direction.

The strongest bull case. CACI has demonstrably won the regime shift that broke its peers: it grew ~10% while Booz Allen shrank, because ~75% of revenue is DoD and its book skews to funded, mission-critical work (EW, SIGINT, space, cyber) on the favored side of the budget. The technology mix-shift is real and margin-accretive (top-of-peer-set ~12% EBITDA; Azure ~25%), the space/agentic-AI adjacency (ARKA) opens a genuinely higher-growth vector, the backlog gives multi-year visibility, and management has a clean no-impairment integration record. On this reading, ~18x forward adjusted EPS for a defensively-positioned low-double-digit compounder is reasonable-to-cheap, and the base/bull scenarios (+26% to +82%) dominate.

The strongest bear case. Strip the narrative and the returns are ordinary: ~10% ROIC hugging the cost of capital, negative tangible book, and a compensation plan that pays for Revenue/EBITDA/FCF growth with no return-on-capital or per-share gate — the textbook set-up of a roll-up that buys multiple and scale rather than moat. Most of the headline growth is acquired (organic just ~5.6% in nine-month-FY26), the differentiation is bought (no internal R&D line), FCF is flattered by a $300M receivables facility that also happens to be the LTI’s payout metric, leverage is at a decade-high ~4.7x on a full-priced ($2.6B, 99%-intangible) space deal into a hot acquisition market, and insiders have bought nothing through a doubling. Layer on the structural question of whether agentic AI deflates the billable-hours labor model that is still ~74% of revenue, and the bear sees a fairly-to-fully-priced ~10%-ROIC serial acquirer one budget air-pocket or one impairment away from a de-rating.

The 3–5 assumptions that actually matter.

1. Organic growth holds at mid-single-digits through appropriations noise (bull) vs. decelerates to low-single on a CR/shutdown air pocket (bear).
2. ARKA/Azure returns — do the ~99%-intangible deals earn a spread over the ~8–9% WACC, or is ~10% ROIC the ceiling?
3. Margin trajectory — does the technology mix keep lifting EBITDA margin toward the mid-teens, or does it plateau at ~12% as acquired franchises get absorbed into the services average?
4. Deleveraging — does FCF glide net-debt/EBITDA back to ~3x within ~18–24 months, restoring optionality, or does it stick near ~4x?
5. AI — is agentic AI a product CACI sells (demand tailwind) or a force that deflates its labor base (revenue headwind)?

What would falsify each side. Bull falsified if FY27 organic growth prints low-single-digit with book-to-bill stuck <1.0x and margins flat, or a goodwill impairment lands on ARKA/Azure. Bear falsified if organic re-accelerates to high-single with ARKA visibly accretive to margin and net-debt/EBITDA back below ~3.5x within 18 months.

The factor-positioning read (what the tape is pricing). CACI is, empirically, a low-beta, highly idiosyncratic defense-technology name, not a momentum or value trade. Its market beta is only ~0.43 with an R² of ~16% (most of its movement is stock-specific, not market-driven); its factor loadings are Aerospace & Defense (+0.28), SmallSize (+0.31) and Cybersecurity (+0.12), with essentially no Momentum, Quality or Growth loading and only a whisper of Value (+0.05). The risk-adjusted record is that of a genuine long-term compounder that just had a bad year: ten-year annualized return ~+16.7% (Sharpe ~0.52), but a −3.4% trailing twelve months and a sharp recent drawdown (the last quarter was deeply negative on an annualized basis) before the current bounce. The factor-similar peer set is exactly the GovCon/defense complex (Leidos 0.96, V2X 0.92, Booz Allen 0.91, Moog 0.89). Translation for the thesis: the tape shows a de-rated compounder that momentum investors left on the drawdown and that value investors have not embraced (it never got cheap enough) — consistent with the “fair, not cheap, mid-recovery” framing, and with a consensus that is bullish on fundamentals while the factor crowd sits it out. It is neither a crowded momentum long to fade nor an abandoned deep-value name to pound the table on — which is exactly why the honest call is a HOLD.

12. Fact vs. Interpretation

13. Open Questions

1. What is the true organic growth rate ex-M&A over a full cycle? Nine-month-FY26 organic was ~5.6%; is that the durable rate, or does it decelerate further as CRs bite?
2. Will ARKA earn a return above its cost of capital, or does ~10% ROIC prove to be the structural ceiling of a full-price, 99%-intangible acquisition strategy?
3. What is the actual cleared-workforce percentage and how tight is the cleared-talent constraint on growth? (Not disclosed precisely in the 10-K.)
4. Is agentic AI a net tailwind or headwind to a business that is still ~74% cost-plus/T&M labor? CACI is betting it can sell the AI faster than the AI deflates its seat count — unproven.
5. How fast does leverage actually glide down, and will management resist a new large deal (or resume buybacks) before net-debt/EBITDA is back near 3x?
6. Does the technology mix keep lifting margins, or do acquired high-margin franchises (Azure ~25%) get diluted into the ~12% blended rate as they scale within CACI?

14. What Must Be True (Bull and Bear, with falsification tests)

Bull case — what must be true:

• CACI’s DoD/intel/technology positioning keeps it on the favored side of the budget through the appropriations cycle, sustaining mid-single-digit-or-better organic growth.
• Azure Summit and ARKA integrate into durable, higher-margin franchises that lift the blended EBITDA margin toward the mid-teens and prove accretive to returns.
• FCF deleverages the balance sheet to ~3x within ~18–24 months, restoring capital-allocation optionality.
• Agentic AI is a product CACI sells profitably, not a force that deflates its labor base.
• Falsification test: If FY27 organic revenue growth prints low-single-digit with trailing book-to-bill stuck below 1.0x and EBITDA margin flat-to-down, or a goodwill impairment is recorded on ARKA/Azure, the bull thesis is broken.

Bear case — what must be true:

• The ~10% ROIC and negative tangible book reveal a roll-up that buys multiple and scale rather than moat; ARKA fails to earn a spread over WACC.
• A CR/shutdown air pocket (or an insourcing push) decelerates organic growth just as leverage is at a decade high, forcing estimate cuts and a de-rating from the ~60th valuation percentile.
• The technology margin premium plateaus as acquired franchises are absorbed; agentic AI begins to deflate the labor book.
• Falsification test: If organic growth re-accelerates to high-single-digit with ARKA visibly accretive to margin, net-debt/EBITDA back below ~3.5x within 18 months, and insiders begin buying, the bear thesis is broken.

15. Source Appendix

See the Source Appendix below for the full list of primary and secondary sources: CACI’s FY2025 Form 10-K, FY2024/FY2023/FY2022 10-Ks, Q3-FY2026 Form 10-Q, FY2026 8-Ks and earnings materials, the DEF 14A proxy (filed 2025-09-05), the Form 4 insider-filing corpus (CIK 16058), the Q3-FY2026 earnings-call transcript, ROIC.ai fundamentals, AZI price/valuation data, the FactorsToday factor model, and trade-press and peer-company sources as cited throughout.

This article carries no investment recommendation and no price target outside the clearly-labeled Claude’s Take block at the top. The analysis discusses valuation only as embedded expectations and scenarios. It is general information, not investment advice.

APPENDIX A — Standard Diligence Questionnaire

CACI International Inc (NYSE: CACI) — as of 2026-07-02

Supplemental to the research memo. Answers are grounded in filings, the fiscal 2026 third-quarter results (quarter ended 2026-03-31), and market data as of the report date. Labels: F = Fact, I = Interpretation, A = Assumption.

General

What thoughtful questions have other investors asked about this company?

• Is CACI’s “Expertise vs. Technology” pivot real product differentiation, or is “Technology” mostly relabeled cost-plus labor plus a thin layer of acquired hardware IP? (I)
• Does the ARKA acquisition ($2.6B all-cash, closed March 9, 2026) earn its cost of capital, or is CACI paying up (~15–17x EBITDA) for growth and space-domain optics while net-debt/EBITDA jumps from ~2.4x to ~4.7x? (F/I)
• Why does a business that has compounded revenue ~8.5%/yr and adjusted EPS at a low-double-digit rate earn only ~10% ROIC and ~10–11% ROE? (Answer: goodwill/intangible-heavy serial acquisition. The incremental labor business earns far more; the reported return is diluted by the price paid for M&A.) (F/I)
• Is CACI structurally insulated from the DOGE / “consulting-is-waste” pressure that broke Booz Allen’s civil book, because ~70%+ of revenue is Defense/Intelligence and increasingly product/technology rather than advisory? (I)
• Does agentic AI deflate the billable-hours labor model that still underlies most of CACI’s revenue — a tailwind (CACI sells the AI) or a threat (fewer staff-augmentation seats)? (Open question)

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? (I) Neither extreme. CACI is riding an up-cycle in defense/intelligence/space budgets and its own technology mix-shift; margins (EBITDA ~11.8–12.3%) are at the high end of the company’s history but not obviously unsustainable. The risk is not a cyclical peak in profitability but an air pocket in the top line from continuing resolutions, shutdowns, and appropriations delays.

Driven by external environment or internal actions? (I) Both. The external tailwind is elevated national-security spending and a budget mix tilting toward CACI’s strengths (EW/SIGINT, space, C4ISR, cyber). The internal driver is a decade of acquisitions moving CACI up the value chain from staff augmentation toward differentiated technology, plus disciplined program execution.

How stable are revenues? (F) Highly contractual: total backlog $33.4B (~3.6x annual revenue), weighted-average award duration ~6 years, funded backlog up 19% YoY, trailing book-to-bill 1.2x. But “backlog” is recompete-exposed — task orders are periodically re-bid, so revenue is an incumbency annuity, not a subscription.

Outlook for products/services; how big is the market? (I) The addressable federal technology-services market is large (hundreds of billions) and growing in CACI’s priority areas (space, EW, cyber, counter-UAS, agentic AI). CACI is a mid-cap prime (~$9.5–9.6B FY26 revenue) with room to gain share and to keep buying capability. Domestic (U.S.) is ~93%+ of revenue; international (UK/Europe) is a small, slower tail.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? (I) Structurally consolidated at the prime-contractor tier (a cleared-services oligopoly: CACI, Booz Allen, Leidos, SAIC, GDIT, Parsons), but individual task orders are competitively bid and LPTA (“lowest price technically acceptable”) pressure persists on commoditized labor. The differentiated-technology tier (EW, space sensors, SIGINT hardware) is less price-competitive and higher-margin.

How profitable is the business (ROIC, ROE)? (F) Modest on a reported basis: FY25 ROE ~10.8%, ROIC ~10.0% — roughly at the cost of capital, because the balance sheet carries ~$6.5B goodwill + ~$2.2B intangibles against ~$4.3B equity (negative tangible book). The underlying operating business earns far higher returns on tangible capital (it is a low-asset labor/technology business); reported returns are depressed by acquisition prices.

How profitable is the industry? (F/I) GovCon operating margins are structurally thin (~8–13% EBITDA) because cost-plus contracts pass through cost and cap markups. CACI’s ~11.8–12% EBITDA margin is at the top of the services peer set, reflecting its higher technology/product content.

Can the business be easily understood? (F) Yes at a high level (sell expertise and technology to national-security agencies), but the segment reporting is coarse (two geographic segments) and the “Expertise vs. Technology” split is a management framing, not an audited segment — so the true product economics are partly opaque.

Undermined by foreign low-cost labor? (F) No — the work requires U.S. security clearances and on-shore/cleared-facility delivery; offshoring is largely precluded. This is a genuine barrier.

Do brands matter? Nature of competition? Switching costs? (I) “Brand” matters as past-performance record, clearances, and incumbency, not consumer brand. Switching costs are real but bounded: incumbents win recompetes at high rates because of mission knowledge and cleared staff, but a lost recompete transfers the work (and often the people). The durable moat is clearances + incumbency + accumulated mission IP, strongest in the classified/technology work and weakest in commoditized IT staff augmentation.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? (I) The cleared workforce, past-performance qualifications, and contract vehicles are valuable intangibles not capitalized. Conversely, a large share of booked assets is acquisition goodwill/intangibles of uncertain standalone value.

Off-balance-sheet liabilities? (F) An accounts-receivable purchase (securitization) facility is used to accelerate cash collection — it flatters reported operating cash flow / FCF timing and should be normalized when judging cash generation. Operating leases are modest.

How conservative is the accounting? (I) Broadly mainstream for GovCon; the main quality caveat is the widening gap between GAAP and adjusted EPS (Q3 FY26 GAAP diluted $5.90 vs. adjusted $7.27) driven by amortization of acquired intangibles — a real, if non-cash, cost of the roll-up strategy that “adjusted” figures exclude.

How CapEx-hungry? (F) Very light — capital expenditure runs ~$60–70M/yr (<1% of revenue). CACI’s “capex” is effectively M&A, not PP&E.

Capital Allocation & Management

How much FCF, and how is it used? (F/I) Real FCF ~$480–550M/yr (operating cash flow less ~$60–70M capex; note the AR facility timing benefit). CACI pays no dividend and deploys cash into (1) acquisitions (the priority), (2) debt reduction, and (3) opportunistic buybacks. The stated financial objective is to grow free cash flow per share.

Significant acquisitions recently? (F) Yes — this is the core strategy. ARKA Group ($2.6B, space/EO-IR/hyperspectral/agentic-AI, closed March 2026); Azure Summit Technology (~$1.275B, EW/SIGINT/photonics, Oct 2024); plus a long tail (Applied Insight, SA Photonics, ID Technologies, Ascent Vision, LGS Innovations, Mastodon, etc.).

Buying back shares? (F) Yes, opportunistically and at modest scale (FY25 ~$169M, FY24 ~$161M, FY23 ~$273M, FY21 ~$509M). Share count has drifted down from ~25.1M (FY20) to ~22.1M. Buybacks are subordinate to M&A.

Issuing large amounts of stock to insiders? (F/I) No large dilution; ARKA was all-cash. Stock-based compensation is modest (~$60M/yr).

Compensation & motivations of management? (See the Capital Allocation section.) Incentive metrics and their emphasis on growth/EBITDA vs. return-on-capital are the key governance question for a serial acquirer — detailed above.

Valuation & Market Data

ADR/MLP/K-1? (F) No — CACI is a Delaware C-corporation (NYSE: CACI); a standard 1099, no K-1. Dividend policy? (F) None; the company has never paid a cash dividend, by design (reinvestment strategy). How profitable? (F) FY26 guide: revenue $9.5–9.6B, adjusted net income $615–630M (~6.5% adjusted margin), EBITDA margin 11.8–11.9%. Net income diverging from cash flow? (F/I) Operating cash flow modestly exceeds net income in most years (non-cash amortization add-back), but is flattered by the AR facility; watch the working-capital swings around large program ramps and the ARKA integration.

Risks & Downside

What factors would cause the stock to decline? (I) (1) Organic revenue decelerating to low-single-digits on appropriations delays / shutdowns / insourcing; (2) ARKA integration disappointment or a goodwill impairment; (3) leverage (~4.7x) staying elevated into a budget air pocket; (4) a lost major recompete; (5) evidence that agentic AI is deflating rather than expanding the billable base; (6) multiple compression from a ~mid-range level back toward the low-teens EV/EBITDA.

Risk of catastrophic / total loss? (I) Low. CACI is an investment-grade, cash-generative prime with a diversified contract base and no single existential contract. The realistic downside is a de-rating and a growth stall, not insolvency; a total loss would require a systemic collapse in U.S. defense spending, which is not a base case.

Recent News & Events

Has the business environment changed recently? (F/I) Yes: (1) the FY2025–26 DOGE / federal-spending-reform wave reshaped GovCon demand — punishing civil-consulting-heavy peers (Booz Allen) while CACI’s defense/intel/technology tilt held up and grew; (2) CACI closed its largest-ever acquisition (ARKA) in March 2026, adding space-sensor and agentic-AI capability and levering the balance sheet; (3) continuing-resolution / shutdown dynamics and shifting appropriations remain the swing factor; (4) leadership continuity at the top (CEO John Mengucci) with ongoing U.S.-operations leadership transitions announced in 2026.

Significant acquisitions? (F) ARKA (March 2026) — see above. Change in accounting policies? (F) None material identified. Recent operational changes? (F) New EVP/COO appointment (Dave Young, June 2026); continued build-out of space, EW, and agentic-AI portfolios; ARKA integration underway.

APPENDIX B — Source Appendix

CACI International Inc (NYSE: CACI) — as of 2026-07-02

Primary sources prioritized over secondary. All SEC filings accessed via EDGAR (CIK 0000016058) and mirrored locally. Fact/Interpretation/Assumption labels appear inline in the memo; this appendix lists the evidentiary base.

Primary — SEC filings (CACI International Inc, CIK 0000016058)

Primary — earnings materials

Quantitative data services

Secondary — trade press & peer sources (public)

• CACI/ARKA acquisition coverage — GovConWire, Virginia Business (deal size, rationale, close date).
• CACI/Azure Summit acquisition — CACI investor release; deal-multiple estimates.
• Booz Allen DOGE/civil-book coverage — Washington Technology (staff cuts, civil guidance) — peer contrast.
• Peer sizing (Leidos, SAIC, GDIT, Parsons, V2X) — company FY2025/FY2026 results.
• Sell-side target-price changes (Citi, JPMorgan) — as reported via financial media; used only for consensus context, not as valuation inputs.
• Peer coverage of Booz Allen Hamilton (BAH) — used for GovCon industry structure, moat framing and the DOGE civil-vs-defense contrast; grounded in BAH’s own public SEC filings.

Notes on data reconciliation

• CACI reports revenue net of costs; ROIC.ai maps its “gross profit” to operating income, so ROIC’s “gross margin” (~8.9%) is the operating margin. Reconciled to the 10-K income statement.
• ROIC.ai’s “free cash flow” equals operating cash flow (does not subtract capex); real FCF is ~$60–70M lower. Capex and the MARPA receivables-facility timing benefit were normalized when discussing cash generation.
• Enterprise value uses the current price ($502.83) × ~22.1M shares plus post-ARKA net debt (~$5.0B, per the Q3-FY26 10-Q), not ROIC’s period-end EV.
• All valuation figures are as of the 2026-07-02 close; the stock rose ~7.7% that day on broad defense-sector strength with no confirmed CACI-specific catalyst.