Jacobs Solutions Inc. (NYSE: J) — A Focused, Well-Run Services Roll-Up Priced as a Compounder

Independent Equity Research Report date: 2026-07-17 · Price (2026-07-16): $132.31 · Market cap: ~$15.1B · EV: ~$18.3B Sector: Industrials · Engineering & Consulting / Professional Services · FY-end: late September

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — it is not investment advice. The analysis in the sections below is presented on its own terms and, aside from this block, expresses no buy/sell recommendation and no price target.

Verdict: HOLD — quality-at-a-price, not a compounder-at-a-discount. Not a short. Accumulate on weakness below ~$110–115 (~15x forward adjusted EPS). Fair-value zone ~$120–135 on the FY26 base; the risk/reward only turns genuinely attractive in the low-$100s. Conviction: medium.

Jacobs is a materially better company than it was three years ago — the Amentum spin (Sept 2024) left a focused, asset-light engineering-and-consulting franchise pointed at four real secular tailwinds (water/PFAS, semiconductor and AI-data-center facilities, life sciences, defense/grid), with a record ~$27B backlog, a 1.4x book-to-bill, and a genuinely high-quality consulting jewel in PA Consulting (~22% margins). Management has done the right strategic things: exited fixed-price-at-risk EPC (now 0% of revenue), cut federal exposure to ~8%, and returned ~150% of free cash flow in FY25. The adjusted earnings power is real: FY25 adjusted EPS $6.12 (+16%), FY26 guided to $7.10–7.35 (+18%). At $132 that is ~18x forward adjusted EPS — the widely-quoted 41x GAAP P/E and “richest-ever multiple” (AZI 93rd-percentile composite) is a tax/separation/Amentum-mark artifact, not the real price. The market is not being crazy here.

But it is not being generous to me either. You are paying a full quality-services multiple (~18x forward adjusted EPS, ~13–15x EV/adjusted-EBITDA — in line with AECOM and below Tetra Tech) for a business that earns the lowest ROIC in its entire peer set: ~6.6%, at or below its ~8–9% cost of capital, versus ~14% at both ACM and TTEK. That gap is the whole story. Jacobs bought its scale (CH2M in 2017, now a debt-funded ~$1.6B buy-in of the PA minority at 13x EBITDA) rather than compounding an advantage, and the balance sheet shows it: $5.4B of goodwill and intangibles, negative tangible book, and leverage just stepped up from 0.8x to 2.1x for a slow-growing advisory asset. The entire premium rests on a show-me bridge — adjusted EBITDA margin rising from 13.9% to a “17%+” FY29 target and ROIC finally inflecting above WACC. If that self-help lands, ~$146 base / ~$184 bull is on the table (~6–19%/yr). If growth normalizes to the mid-single-digit services rate the industry structurally supports, the multiple de-rates to 14–15x and you lose money from here (~$110). The skew at $132 is symmetric-to-negative with no margin of safety — hence HOLD, and a preference to own it cheaper.

Framing: a stalled, out-of-favor quality-income defensive-industrial (beta 0.84, no momentum/growth/value factor loading, off ~19% from its Oct-2025 high, basing after a de-rate) — the opposite of the crowded AI-capex momentum trade in the self-perform contractors (PWR/EME/MTZ), and not a falling knife. Tag: “You pay for the compounder; you own the roll-up.” · Bull-flip trigger: two-plus quarters of the 13.9%→17% margin bridge tracking and ROIC visibly inflecting toward double digits — proof the self-help is operational, not financial engineering. · Bear-flip trigger: organic net-revenue growth decelerating to mid-single-digits with margins stalling, which breaks the multiple.

📈 Stock Price Action — Five-Year Event Map

Over five years Jacobs has round-tripped a full re-rating. On a split/spin-adjusted basis the stock ran from a ~$86 trough (Sept 2022) to an all-time high of $163.10 (21 Oct 2025), then de-rated ~34% to a ~$107 low (May 2026) before recovering to $132.31 today — leaving it ~19% below the high, mid-range of a 52-week $106.93–$163.10 band. Note that pre-Sept-2024 adjusted prices are deflated for the Amentum value stripped out at the spin, so the earlier climb partly reflects that mechanical adjustment; the cleaner read is the standalone post-spin arc from ~$127 (Sept 2024) to $163 and back.

Cycle narrative. (1) Jacobs fell with every long-duration industrial in the 2022 rate shock despite no fundamental deterioration. (2–3) The 2023–2024 recovery was underwritten by IIJA-funded work and, decisively, the market’s anticipation of the Amentum government-services spin, which promised a cleaner, higher-margin, faster-growing pure-play. (4) That thesis peaked in October 2025 at $163 and the richest own-history multiple Jacobs has ever carried, as the data-center/AI-infrastructure narrative and the announced PA Consulting buy-in stacked on top of a record backlog. (5) The subsequent ~34% de-rate was the market re-pricing three things at once: a full multiple with little margin of safety, a genuinely soft environmental/water-remediation quarter (net revenue only +2%), and a $1.6B debt-funded acquisition that lifted leverage from 0.8x to 2.1x. (6) The rebound since May tracks two consecutive guidance raises and the backlog inflection to $27B. Each move is a Fact; the attributed driver is Interpretation, cross-referenced to earnings dates, 8-K events, and guidance changes.

1. Executive Summary

Jacobs Solutions is a global engineering-design and technical-consulting firm — the #1-ranked ENR design firm in seven of the last eight years — that, following the September 2024 spin of its government-services arm (Amentum, via Reverse Morris Trust), now operates two segments: Infrastructure & Advanced Facilities (I&AF) — 89.5% of FY25 revenue, 8.4% operating margin — and PA Consulting, a UK-based strategy/innovation consultancy — 10.5% of revenue but ~24% of segment operating profit at a ~22% margin. It is an asset-light (capex <1% of revenue), fee-and-cost-plus professional-services model with ~43,000 employees and deliberately zero fixed-price-at-risk backlog.

The investment tension is clean. The good: a focused post-spin franchise aimed at four durable demand pools (water/environmental, semiconductor and AI-data-center “advanced facilities,” life sciences, defense/grid); a record $27.0B backlog (+22% YoY) with a 1.4x gross / 1.2x net book-to-bill; real and growing adjusted earnings power (FY25 adjusted EPS $6.12, +16%; FY26 guided $7.10–7.35, +18%); light capital intensity; and disciplined shareholder returns (~150% of FCF returned in FY25 via $754M buyback + a serially-raised dividend). The problem: at the consolidated level there is no demonstrable moat. Jacobs’ FY25 ROIC of ~6.6% is the lowest in its peer set and sits at or below its ~8–9% cost of capital, versus ~14% for both AECOM and Tetra Tech in the same “commoditized” industry. The low return is company-specific — the product of $5.4B of goodwill and intangibles from CH2M (2017) and the just-closed PA buy-in, carried against a negative tangible book value. Jacobs bought scale; it has not yet earned excess returns on it.

Valuation embeds the optimistic resolution. At ~18x forward adjusted EPS and ~13–15x EV/adjusted-EBITDA, the stock trades in line with pure-design peers and carries its richest-ever own-history multiple (on GAAP optics), while underwriting the full FY29 plan: 7%+ net-revenue CAGR, adjusted EBITDA margin rising from 13.9% to “17%+,” and — implicitly — ROIC inflecting from ~6.6% toward double digits. That is low-double-digit EPS growth priced as durable on a base of at/below-WACC returns. Scenario work frames a symmetric-to-negative skew (bear ~$110, base ~$146, bull ~$184). This memo takes no position and sets no price target; that judgment sits solely in Claude’s Take above.

2. Business Overview

What Jacobs does. Jacobs is a provider of professional engineering, design, technical, and management-consulting services — it sells the time and expertise of its ~43,000 professionals, not manufactured products or self-performed heavy construction. It plans, designs, engineers, and program-manages complex facilities and infrastructure for public and private clients, and (through PA Consulting) advises on strategy, innovation, and technology. Founded in 1947 and now headquartered in Dallas, Texas, it operates across the US, Europe, the Middle East, Asia-Pacific, and India.

Two segments post-spin. The September 2024 Amentum separation removed the lower-margin, US-government-heavy Critical Mission Solutions and Divergent Solutions businesses, leaving:

• Infrastructure & Advanced Facilities (I&AF) — FY25 revenue $10,764M (89.5% of total) at an 8.4% operating margin. This spans (a) water & environmental — water/wastewater treatment, PFAS remediation, environmental restoration; (b) transportation — rail, transit, aviation, ports, highways (global); © advanced facilities — the fast-growing cluster of semiconductor fabs, AI/data-center campuses, life-sciences/pharma manufacturing, and advanced manufacturing; and (d) energy, power & national security — grid, cyber, and mission-critical infrastructure. Advanced facilities and the AI-ecosystem are the growth engines (data-center pipeline >100% YoY; AI ecosystem ~10–11% of the business growing >40%); water/environmental is the current soft spot (net revenue +2% in Q2 FY26).

• PA Consulting — FY25 revenue $1,266M (10.5% of total) at a 22.0% operating margin. A UK-anchored strategy/technology/innovation consultancy (financial services, government, defense, energy, consumer, health/life-sciences). It is the economic jewel: 10.5% of revenue but ~24% of segment operating profit, and the highest-return asset Jacobs owns. In March 2026 Jacobs bought in the ~29% it did not already own, taking it to 100% (see §6).

How it makes money — gross vs. net revenue matters. Jacobs’ reported GAAP revenue is gross — it includes ~$3.36B (~28% of FY25’s $12.03B) of pass-through subcontractor and procured-materials costs on which Jacobs earns little or no margin. The economically meaningful base is net revenue (revenue net of pass-throughs) — ~$8.67B in FY25 — which is what management guides on and against which the 13.9% adjusted EBITDA margin is measured. Any EV/Sales or growth framing that uses gross revenue overstates the base and understates the multiple; this memo uses net revenue where it matters. (FACT that filings report only gross revenue; net revenue and book-to-bill are IR-deck non-GAAP metrics — an OPEN QUESTION on disclosure quality, §13.)

Revenue quality and contract mix. The mix is deliberately de-risked: ~68% cost-reimbursable, ~32% fixed-price with limited risk, and 0% fixed-price at-risk (turnkey EPC) — the latter cut from 4% to 0% as management exited lump-sum construction risk. Revenue is largely reimbursable and framework-based (multi-year master service agreements, IDIQ vehicles, AMP regulatory cycles in the UK), which produces recurring, relationship-anchored work but not contractual subscription lock-in. US federal revenue is only ~8% and falling (10%→9%→8%, FY23–25), which structurally limits DOGE/shutdown exposure at the consolidated level. Geography: US ~61.6%, Europe ~23.9%.

Verdict. A clean, understandable, asset-light professional-services model with a genuinely high-quality consulting segment bolted onto a large, lower-margin engineering-services core. The economics are decent and de-risked, but the revenue is fee-for-service and re-competed, not annuity-like.

3. Industry Dynamics

Structure. The engineering, design, and technical-consulting industry is large, fragmented, and structurally below-average in profitability. Jacobs’ own 10-K names 20+ competitors — AECOM, Tetra Tech, Stantec, WSP, Arcadis, HDR, Fluor, Bechtel — and, in advisory, Accenture, McKinsey, and the strategy houses. No single firm dominates; the top players each hold low-double-digit shares of any given vertical, and the filing openly concedes the market is “highly competitive” with “downward pressure on prices and margins.” This is the textbook Greenwald signature of an industry without meaningful barriers to entry: you cannot count the leaders on one hand, and the primary input — engineering and consulting talent — is mobile and owned by the talent, not the firm.

Economics of the model. The business is asset-light (low capital intensity is a genuine positive — it converts earnings to cash and needs little reinvestment), but it is fundamentally a human-capital, cost-plus model: revenue scales with billable headcount and utilization, and margins are set by mix, utilization, and wage inflation rather than by pricing power. Returns on tangible capital can look high precisely because there is little tangible capital; returns on total invested capital (including the goodwill paid to acquire talent pools) are mediocre. Switching costs exist at the project and framework level (mid-stream engineer changes carry real error/continuity risk) but are weak at the enterprise level, where clients routinely re-compete master agreements.

Demand drivers — genuinely favorable. Where the industry is attractive is on the demand side, and Jacobs is well-positioned: US infrastructure funding (IIJA, only ~⅓–½ outlaid — a multi-year tail; OBBBA a net positive per management); water/PFAS regulation; the CHIPS-driven semiconductor reshoring wave (Micron/HBM); the AI-data-center capex super-cycle; life-sciences onshoring; and defense/grid modernization (US and European/UK). These are secular, funded, multi-year pools.

Capital-cycle read (Marathon). The supply-side caution is precisely in the hottest pocket: data-center and advanced-facility engineering margins are attracting industry-wide capacity — every design/EPCM firm and the self-perform contractors are racing into AI-infrastructure work. High returns attract capital; this is a mid-cycle inflow, not a defended niche. The broader engineering-talent pool is tight (a modest supply constraint that supports utilization and wage recovery), but the value-added margin on the marquee data-center work is more likely to be competed down than to widen. The industry is not consolidating in a way that structurally lifts returns.

Verdict: a structurally below-average industry riding an above-average demand cycle. The tailwinds are real and durable, but they accrue to a competitive, low-barrier services market. Demand strength is not the same as industry attractiveness, and it does not, by itself, justify a premium multiple.

4. Competitive Position

The moat question, answered by the returns. The decisive evidence is comparative ROIC. In the same “commoditized” industry, AECOM earns ~14.7% and Tetra Tech ~14.1% ROIC; Jacobs earns ~6.6% — the lowest in the peer set and at or below its ~8–9% cost of capital. If two direct competitors in a low-barrier industry sustainably earn ~14%, then the industry is not the reason for Jacobs’ low return — the company is. And the company-specific reason is transparent on the balance sheet: $4.78B of goodwill plus ~$0.64B of intangibles (CH2M in 2017 at ~$3.3B; earlier deals; now the PA minority buy-in) acquired at full premiums, carried against a negative tangible book value (TBVPS −$6.85). Jacobs bought its scale and reputation; the excess return that scale is supposed to generate has not materialized in invested-capital terms.

What Jacobs genuinely has — and what it is not. Jacobs has real competitive assets: the #1 ENR design ranking (seven of eight years), a deep technical reference base and delivery track record that help win mega-projects, security clearances and incumbency on multi-year framework and national-security work, and a global low-cost delivery network (India/Philippines). These are advantages in bidding. They are not, on the evidence of the returns, a moat — a durable structural barrier that would cause Jacobs’ economics to deteriorate if a competitor tried to replicate them. A genuine moat claim must tie to a financial outcome that would erode without it; Jacobs’ scale and reputation have not produced a differentiated ROIC, so at the consolidated level the moat is narrow-to-absent.

The exception: PA Consulting. The one genuinely advantaged asset is PA — a ~22%-margin strategy/innovation consultancy where reputation, senior-relationship depth, and talent create real (if human-capital-dependent) pricing power and search/switching costs of the kind Greenwald recognizes in professional services. This is why Jacobs paid 13x EBITDA to own all of it. But PA is ~10% of revenue and is not itself a wide-moat compounder — consulting is talent-owned and cyclical, and the acquired majority now sits in goodwill.

Advanced facilities: a tailwind, not a fortress. The bull case increasingly leans on “advanced facilities” (semis, data centers) as a differentiated niche. It is a high-demand niche, but the memo’s capital-cycle read (§3) is that it is a competed, mid-cycle pocket, not a defended one. Cyclical mix-tailwind ≠ durable advantage.

Verdict: a crowded market with weak enterprise-level differentiation. Jacobs is a well-run, scaled, reputable bidder in a low-barrier industry, with one genuinely advantaged segment (PA). The 6.6% ROIC — worst in its cohort — is the financial proof that scale here has not become a moat.

5. Growth History and Forward Opportunities

History — modest organic growth, restated by the spin. Reported revenue is discontinuous because of the Amentum separation and discontinued-operations restatement (continuing-ops revenue: FY22 $9.78B → FY23 $10.85B → FY24 $11.50B → FY25 $12.03B). Underlying organic growth has been moderate — roughly mid-single-digit in net-revenue terms — with the reported FY25 +4.6% and the eye-catching Q2 FY26 gross +27% both flattered by pass-through costs, FX, and easy comps (a prior-year JV reserve); gross margin actually fell ~390bps in that quarter on pass-through and lower PA utilization. The honest read: this is a mid-single-digit organic net-revenue grower, currently accelerating.

The backlog inflection — real but read it carefully. Backlog has genuinely inflected: $23.06B (FY25) → $26.97B at Q2 FY26 (+22% YoY), with a trailing book-to-bill of 1.4x gross but only 1.2x net — and management added the net metric this year, a tacit admission that the gross figure is inflated by high-pass-through data-center and semiconductor awards. The cleaner tells are gross-profit-in-backlog (+13–15%) and net-revenue-in-backlog (+12%). For conservatism, note that ASC-606 remaining performance obligations (RPO) are only $19.0B — “backlog” is a broader, management-defined metric.

Forward drivers. The pipeline is legitimately strong where it counts: AI-data-center infrastructure +~400% pipeline YoY (AI ecosystem ~10–11% of the business, growing >40%); life sciences +81%; semiconductors (HBM/Micron reshoring); transportation (global rail/aviation/ports); energy/grid; and PA cross-sell now unconstrained by the prior UK conflict-of-interest limits. The soft spot is water/environmental (+2% net revenue), hit by a regulatory pause, a FEMA federal→state funding shift, and DOGE-adjacent DoD-environmental cuts — management expects it to inflect by Q4 FY26.

Quality of growth — half self-help. Management’s FY29 framework targets ~7%+ net-revenue CAGR, adjusted EBITDA margin to “17%+” (from 13.9%), and ~mid-teens-to-18% adjusted EPS growth. Decompose that: with ~7% net-revenue growth, roughly half of the targeted EPS growth is margin expansion + buyback, not volume. That is achievable and not disreputable, but it is lower-quality than volume-led compounding — margin self-help has a ceiling, and buyback-driven EPS growth is financial, not franchise, growth.

Verdict: moderate-quality growth, tailwind-supported. Real secular demand and a real backlog inflection, but organic volume growth is high-single-digit at best on a clean net-revenue basis, and a large slice of the EPS-growth algorithm depends on a margin bridge that must still be delivered.

6. Financial Quality

The GAAP-to-adjusted wedge is enormous — and mostly explainable. FY25 GAAP diluted EPS was just $2.37 (continuing-ops $2.58) against adjusted EPS of $6.12 — a ~$3.5+ wedge. The drivers are: (a) a −$227.3M mark-to-market loss on the retained Amentum stake (largely non-deductible, which is why the GAAP effective tax rate spiked to 39.7% vs. FY24’s 16.9% and an adjusted tax rate guided at 27–28%); (b) ~$134M of separation/restructuring costs; © a $20.5M debt-extinguishment loss; and (d) amortization of acquired intangibles. Most of this is genuinely non-recurring separation noise, and the adjusted figure is a fairer picture of run-rate earnings than the GAAP optic. The 41x GAAP P/E and AZI’s 97.5th-percentile P/E rank are artifacts; the real forward multiple is ~18x adjusted.

But the adjustments are not pristine. Two add-backs deserve scrutiny. First, intangible amortization is a real (if non-cash) cost of the acquisition-led model and is added back to adjusted EPS — for a serial acquirer, excluding it flatters returns on capital that were, in fact, spent. Second, “strategic-investment/acquisition” costs are excluded from the incentive-comp metrics as well as from adjusted earnings — a recurring cost of a recurring strategy treated as one-time. SBC is modest and honest (~$61M, ~0.5% of revenue). Net: adjusted EPS is directionally right but carries perhaps $0.30–0.50 of add-back generosity.

Margins and the bridge. FY25 adjusted EBITDA was ~$1.20B at 13.9% of net revenue; FY26 is guided to 14.6–14.9%, and the FY29 target was raised to 17%+. The drivers management cites — global delivery (India/Philippines), commercial-model/mix shift toward advanced facilities and PA advisory, OpEx leverage, and PA cost synergies ($20M+) — are credible and partly in motion (~200bps of expansion FY25→FY26). This is the single most important number in the thesis: the multiple is underwritten by this bridge.

Cash generation. The model converts well: FY25 FCF ~$607M (FY24 $934M, FY23 $837M) on capex of just ~$79M (<1% of revenue). FY26 FCF conversion is guided below 100% of adjusted net income on one-time PA cash costs and a tax event, then recovers toward the 11%+ FCF-margin FY29 target ($1.2–1.3B). Working capital (unbilled receivables/DSO) is a perennial swing item in this industry and bears watching, but the multi-year trend is healthy cash conversion.

Returns and the balance sheet. ROIC ~6.6% (below WACC); ROE ~14% (flattered by leverage and the negative-tangible-equity denominator). Tangible book is negative; $5.4B of goodwill/intangibles dominate the asset base. The proxy’s “ROIC” metric uses a lean/adjusted capital base that flatters the true ~6.6% figure — a governance flag (§6/§13).

Verdict: real, cash-generative adjusted earnings power — but returns on total capital do not yet improve with scale. The economics are decent and the cash conversion is genuine; the missing ingredient is a return on invested capital that clears the cost of that capital.

7. Capital Allocation

The framework. Post-spin, management targets returning ≥60% of free cash flow to shareholders and has over-delivered: ~150% of FCF returned in FY25 ($754M buyback + $153M dividend). Buybacks scaled FY24 $355M → FY25 $754M → H1-FY26 $472M, with $746M left on a $1.5B authorization. The dividend has been raised every year (to $1.26 in FY25; recent quarterly raises of ~10–12.5%). For an asset-light business with light reinvestment needs, aggressive buyback + a growing dividend is the correct default — and management has executed it.

The defining event: the PA Consulting buy-in. In March 2026 Jacobs acquired the ~29% of PA Consulting it did not already own (it held ~71%), taking the stake to 100%, for ~£1.21B initial (£997.6M cash + 2.04M new Jacobs shares) plus a £75M deferred payment, ~$113.5M of EBT compensation, and ~$36M of costs — a headline 13x EV/EBITDA (12.3x post-synergies), ~$1.6B of cash consideration, implying a ~$5.5B value for 100% of PA. It was debt-funded: $1.3B of new senior notes (4.75% 2031 / 5.375% 2036) plus ~$1.2B of term loans and a new $1.5B revolver. This drove the redeemable-NCI redemption (minority interest +$1,097M Q1 → −$8.4M Q2), lifted net debt from $933M to $2.71B, and pushed net leverage from 0.8x to 2.1x — ~0.5 turn above the 1.0–1.5x target, with deleveraging to <2x guided by FY26-end.

Assessment (mixed). The strategic logic is sound: PA is the highest-return asset Jacobs owns, full ownership removes UK conflict-of-interest limits on joint pipeline, the deal is accretive to adjusted EPS in year one, and $20M+ of cost synergies are credible. But the price is full (13x EBITDA for a cyclical, talent-owned consultancy), it is debt-funded at ~5% into a rising leverage profile, and — per Marathon’s asset-growth anomaly — it is exactly the kind of premium-priced, debt-financed asset expansion that historically precedes muted forward returns. It concentrates capital into a good asset at a price that leaves little margin of safety.

The M&A track record. Jacobs is a serial acquirer — CH2M (~$3.27B, 2017) is the anchor of today’s $5.4B goodwill/intangible base. The 6.6% consolidated ROIC is the verdict on that record: the deals bought revenue, reputation, and scale, but not excess returns on the capital deployed. This is the central capital-allocation criticism.

Incentive alignment. Annual incentives key off Adjusted Operating Profit (50%), Adjusted EBITDA Margin (20%), and Adjusted Net Revenue (20%); the LTIP is 50% Adjusted EPS + 50% ROIC + a stock-price multiplier. Including ROIC is good — but the proxy’s ROIC definition uses a flattering adjusted capital base, so the metric that should discipline the acquisition habit is measured on a basis that obscures the ~6.6% reality. Insider ownership is modest and there is no conviction open-market buying (see §8.9).

Verdict: disciplined shareholder returns, questionable acquisition returns. Management allocates free cash flow well (buyback + dividend) but has a decade-long record of deploying acquisition capital at premiums that have not earned their cost. The debt-funded PA buy-in is strategically defensible but adds leverage and asset-growth risk for a full price.

8. Changes and Headwinds — Last Two Years

Portfolio transformation (net positive). The dominant change is the September 2024 Amentum spin (Reverse Morris Trust), which separated the lower-margin, government-services businesses and created today’s focused, higher-margin pure-play. Jacobs has since fully exited the retained Amentum stake (29.2M shares at spin → 19.5M used in an equity-for-debt exchange, 7.3M distributed in-kind to holders, 2.4M to sponsors; zero remaining) — removing the mark-to-market volatility that distorted FY25 GAAP. The March 2026 PA Consulting buy-in (§7) is the second major structural change.

Strategic de-risking (positive). Exit of fixed-price-at-risk EPC (4%→0%), federal exposure cut to ~8%, and a shift toward advanced facilities and advisory — all reduce risk and lift mix.

Segment realignment & disclosure. The move to the two-segment I&AF + PA structure, and the growing reliance on non-GAAP “net revenue” and “book-to-bill” metrics that do not appear in the filings, are a mild disclosure-quality headwind (§13).

Headwinds. (a) Environmental/water softness — regulatory pause, FEMA federal→state shift, DoD-environmental cuts; expected to inflect by Q4 FY26. (b) Leverage step-up — 0.8x→2.1x for the PA buy-in, into a ~5% cost of debt. © Federal/DOGE noise — structurally overstated at 8% of revenue, but a sentiment overhang. (d) FX — a meaningful GBP/EUR translation exposure via PA and European I&AF. (e) Wage inflation / utilization — the perennial services-margin risk. Management: Bob Pragada (Chair & CEO), Venk Nathamuni (CFO).

Verdict: the changes strengthen the business and the mix; the headwinds are mostly cyclical/transitional, with the leverage step-up the one genuinely thesis-relevant negative.

9. Risk Analysis (Risk Matrix)

Catastrophic-loss risk is low. The asset-light, 0%-fixed-price-at-risk, diversified, investment-grade profile makes a total or near-total loss highly improbable. The dominant risk is valuation — a de-rating of a full multiple on a business whose returns do not yet justify a premium.

10. Valuation Discussion (Embedded Expectations)

Use the right numerator. On GAAP the stock looks absurd (~41x P/E, richest-ever); on adjusted figures it is a full-but-defensible ~18.3x forward adjusted EPS (FY26 mid $7.22) and ~13–15x EV/adjusted-EBITDA (adjusted EBITDA ~$1.20B FY25), ~1.9–2.1x EV/net-revenue, with a ~3.9% FCF yield. The honest “elevated” tell in the AZI percentile suite is the P/S rank (84th), not the tax-distorted P/E rank (97.5th) or the meaningless P/B (negative tangible book).

Peer comparison — peer-equivalent price for worst-in-class returns.

Jacobs trades roughly in line with the pure-design cohort on EV/EBITDA and below it on adjusted P/E — but earns roughly half the ROIC of ACM and TTEK. You are paying a peer-equivalent price for materially lower returns and higher leverage. On a relative-value screen the stock is mediocre, not cheap — it is not a value opportunity, and it is not an obvious short either.

What the price embeds. ~18x forward adjusted EPS / ~13–15x EV/adjusted-EBITDA underwrites the full FY29 plan: 7%+ net-revenue CAGR, adjusted EBITDA margin rising ~75bps/yr to 17%+, 11%+ FCF margin — and, implicitly, ROIC inflecting from 6.6% toward double digits as the margin bridge lands and the acquisition base finally earns its cost. This is low-double-digit EPS growth priced as durable on a base of at/below-WACC returns. The market is underwriting the optimistic resolution of the very tension this memo identifies.

Scenarios (FY2028, off the FY26 ~$7.22 adjusted-EPS base):

The base case delivers roughly a market return; the skew is symmetric-to-negative, and there is no margin of safety at $132. The upside requires both the margin bridge and multiple maintenance; the downside needs only growth to normalize to the services-sector mid-single-digit rate for the multiple to compress. (This section renders no price target and no recommendation; the scenario prices are illustrative embedded-expectations math, not targets.)

11. Variant Perception

Consensus. A quality pure-play compounder: post-spin focus, mid-teens adjusted-EPS growth, secular demand across data-center/AI, semiconductors, water, and defense, a record $27B backlog, and disciplined capital return — worth a premium multiple.

Strongest bull case. The secular end-markets are real and funded; the margin path to 17%+ is credible and already showing ~200bps of progress; the PA buy-in adds a high-margin advisory cross-sell engine now unconstrained by conflict rules; ROIC will inflect as mix shifts and synergies land; and an asset-light model returning ≥60% of FCF compounds intrinsic value even at a mid-single-digit organic growth rate. If ROIC moves from 6.6% toward 10%+, today’s multiple is justified and the stock re-rates.

Strongest bear case. The “moat” is rented talent; consolidated ROIC (~6.6%) sits at/below WACC and is the worst in the peer set; clean organic net-revenue growth is high-single-digit (the headline gross backlog and +27% revenue prints are inflated by data-center pass-throughs); GAAP earnings are one-third of adjusted; tangible book is negative; the $1.6B debt-funded PA buy-in lifted leverage 0.8x→2.1x for a cyclical consulting asset at 13x EBITDA; and the stock carries its richest-ever own-history multiple. This is a well-run services roll-up priced as a wide-moat compounder — the classic setup for a de-rating when growth normalizes.

The five assumptions that matter most (with falsification tests):

1. ROIC inflects above WACC. Falsified if ROIC remains stuck at ~6–7% through FY27 despite the margin bridge — proving scale is not becoming a moat. (Highest weight.)
2. Adjusted EBITDA margin reaches ~17%. Falsified if margin stalls below ~15% on wage inflation/utilization/mix.
3. Organic net-revenue growth holds high-single-digit. Falsified if clean net-revenue growth decelerates to mid-single-digit as data-center/pass-through comps normalize.
4. The premium multiple holds. Falsified if the stock de-rates to the ~14–15x that at/below-WACC returns historically command. (Second-highest weight.)
5. Federal/environmental softness is transitory. Falsified if water/environmental fails to inflect by FY26-end and federal cuts deepen.

The variant view. Consensus is right about the demand and the management quality and wrong about the durability of the premium. The factor tape agrees the enthusiasm has cooled — Jacobs now screens as a low-beta (0.84), out-of-favor quality-income industrial with no momentum, growth, or value factor loading, off ~19% from its high and basing. That is neither a crowded momentum trade (unlike the self-perform contractors) nor a falling knife; it is a stalled compounder whose premium is on probation pending proof that the ROIC inflection is real.

12. Fact vs. Interpretation Table

13. Open Questions

1. Exact adjusted net revenue and adjusted EBITDA in dollars — derived here from FCF-margin and guidance because the earnings-release Ex-99.1 reconciliations are not in the local filing corpus. Confirm against the IR deck.
2. The I&AF end-market split (water/env vs transportation vs advanced facilities vs energy/national-security) in exact revenue terms — disclosed in IR materials, not the 10-K.
3. Pass-through percentage precision — ~28% is derived; the exact gross-to-net bridge is an IR-deck figure.
4. Disclosure quality — “net revenue” and “book-to-bill,” the metrics management guides on, do not appear in the 10-K/10-Q. Why not, and will they be formalized?
5. PA Consulting standalone growth and margin durability post-buy-in — is high-single-digit constant-currency growth sustainable, and how cyclical is the ~22% margin?
6. Proxy ROIC definition — the exact capital base used, and the gap to the true ~6.6% figure.
7. FY26 sub-100% FCF conversion — confirm the one-time tax event and PA cash costs fully explain it, and that FY27 reverts.
8. Water/environmental inflection — does the guided Q4-FY26 recovery actually materialize?

14. What Must Be True

For the bull case to work:

• ROIC must inflect from ~6.6% toward double digits as the margin bridge lands and the acquired base (CH2M + PA) finally earns its cost of capital. Falsification test: ROIC still ~6–7% at FY27 year-end. — This is the single load-bearing assumption; the entire premium rests on it.
• Adjusted EBITDA margin must reach ~17% (from 13.9%) via delivery-mix, PA synergies, and OpEx leverage. Falsification test: margin stalls below ~15% through FY27.
• Organic net-revenue growth must hold high-single-digit as data-center/pass-through comps normalize. Falsification test: clean organic net-revenue growth <5% for two consecutive quarters.
• The premium multiple must hold (~18x forward adjusted EPS). Falsification test: sustained de-rating below ~15x.

For the bear case to work:

• Growth normalizes to the services-sector mid-single-digit rate and margin gains stall, exposing a ~6.6%-ROIC business at a compounder multiple. Falsification test: two-plus quarters of accelerating clean net-revenue growth with margin tracking to 17%.
• The multiple compresses to 14–15x as the market re-prices at/below-WACC returns. Falsification test: the stock holds ≥17x while ROIC stays flat — i.e., the market keeps paying up on faith.
• The PA leverage/asset-growth bet weighs on returns (Marathon anomaly). Falsification test: PA synergies + deleveraging to <2x by FY26-end with visible ROIC accretion.

The pivot: Both cases resolve on one question — does ROIC inflect above the cost of capital, or does Jacobs remain a well-run bidder earning ~WACC on a goodwill-laden balance sheet? Everything else (backlog, tailwinds, buyback) is agreed; the disagreement is entirely about whether scale finally becomes a moat.

15. Source Appendix

See the Diligence Questionnaire (Appendix A) and Source Appendix (Appendix B) below. Primary sources: Jacobs Solutions FY2025 Form 10-K (filed 2025-11-20), Q1/Q2 FY2026 Forms 10-Q, FY2026 8-Ks (PA Consulting buy-in and debt issuance, Feb–Mar 2026), DEF 14A proxy, and Q3-FY2025 through Q2-FY2026 earnings-call transcripts. Quantitative data reconciled via SEC EDGAR and third-party aggregators; price and factor-positioning data from public market-data sources; peer framing via AECOM/Tetra Tech and the broader engineering & construction peer set.

This article is independent research for general information only. Aside from the labeled “Claude’s Take” block, it contains no buy/sell recommendation and no price target. It is not investment advice.

APPENDIX A — Standard Diligence Questionnaire

Jacobs Solutions Inc. (NYSE: J) — 2026-07-17

Fact / Interpretation / Assumption labels applied where material.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions are: (1) Is the post-Amentum-spin margin/ROIC uplift structural or one-time? (2) How much of the “record backlog” and revenue growth is real value-added net revenue vs. low-margin data-center/semiconductor pass-through? (3) Was the debt-funded PA Consulting buy-in (13x EBITDA, leverage 0.8x→2.1x) good capital allocation or empire-building at a full price? (4) Does the ~18x forward adjusted P/E (richest-ever own-history composite) hold when ROIC (~6.6%) sits at/below WACC? (5) Is “advanced facilities” (AI data centers/semis) a durable niche or a mid-cycle tailwind that gets competed away?

Cyclicality & Earnings Nature

Cyclical high or low? Mid-to-early-cycle on demand (IIJA only ~⅓–½ outlaid; AI/data-center and semiconductor waves early), but margins are being lifted by internal self-help toward a cyclical-high target (13.9%→17%). Interpretation: demand is not at a peak, but the margin/multiple leave little room for disappointment. External environment or internal actions? Both — secular demand (external) plus mix shift, global delivery, and the exit of fixed-price EPC (internal). Revenue stability? Reasonably stable: ~68% cost-reimbursable, 0% fixed-price-at-risk, multi-year framework/AMP/IDIQ contracts, ~$27B backlog. Not annuity-recurring, but low-volatility for a project business. Market size/outlook? Large and growing — global infrastructure, water, advanced facilities, life sciences, defense. Multi-year funded pools; international as well as domestic.

Business Quality & Competitive Moat

Industry more or less competitive? Highly competitive and fragmented (20+ named rivals); not consolidating in a returns-lifting way. Data-center engineering is attracting industry-wide capacity (capital-cycle caution). How profitable (ROIC/ROE)? ROIC ~6.6% (Fact) — below WACC and the lowest in the peer set; ROE ~14% (flattered by leverage/negative tangible equity). This is the central weakness. Industry profitability / barriers? Structurally below-average; low barriers to entry (talent is mobile and owned by the talent). Barriers exist at project/framework level (switching/continuity), not enterprise level. Easily understood? Yes — a fee-for-service engineering/consulting model. Undermined by low-cost foreign labor? Partly the opposite — Jacobs uses India/Philippines low-cost delivery as a margin lever; high-end design/advisory is less offshorable. Do brands matter? Reputation/reference base and #1 ENR ranking help win work; PA’s brand carries real advisory pricing power. Neither has produced excess consolidated ROIC. Nature of competition / switching costs? Competitive bidding on frameworks and re-competes; switching costs are moderate mid-project, weak at renewal.

Financial Condition & Balance Sheet

Assets not on the balance sheet? The reputational/reference/relationship capital and clearances — real but unbooked. Conversely, $5.4B of goodwill/intangibles are booked and dominate the asset base. Off-balance-sheet liabilities? Standard operating leases; JV exposures; performance guarantees typical of E&C. Nothing flagged as unusual. How conservative is the accounting? Mixed. Cash conversion is genuine and SBC modest (~0.5% of revenue), but heavy reliance on non-GAAP “adjusted” figures, add-back of intangible amortization for a serial acquirer, and IR-deck-only “net revenue”/“book-to-bill” metrics warrant scrutiny (§13). CapEx-hungry? No — capex <1% of revenue; asset-light.

Capital Allocation & Management

FCF generation and use? ~$607M FCF FY25 (~$1.2–1.3B target by FY29); policy ≥60% of FCF returned; ~150% returned FY25. Philosophy: buyback + growing dividend + bolt-on/strategic M&A. Significant acquisitions? Yes — CH2M (~$3.27B, 2017) and the March 2026 PA minority buy-in (~$1.6B, 13x EBITDA, debt-funded). A serial acquirer; the 6.6% ROIC is the verdict on the record. Buying back shares? Yes — $754M FY25, $472M H1-FY26; $746M remaining on $1.5B authorization. Issuing shares to insiders? Modest SBC (~$61M); 2.04M new shares issued as part of the PA consideration. No egregious dilution. Compensation policy / motivations? Annual: adj operating profit/EBITDA margin/net revenue. LTIP: 50% adj EPS + 50% ROIC + stock-price multiplier. Caveat: proxy “ROIC” uses a flattering lean capital base vs. the true ~6.6%. Management (Pragada/Nathamuni) is credible and execution-focused.

Valuation & Market Data

ADR/MLP/K-1? No — US C-corp common stock, NYSE. Dividend policy? Growing dividend ($1.26/sh FY25; ~10–12.5% annual raises); ~1.0% yield, ~45% GAAP / lower adjusted payout. How profitable? Decent adjusted profitability (13.9% adj EBITDA margin on net revenue) but low returns on total invested capital (6.6%). Net income vs. cash from operations diverging? GAAP net income is well below adjusted and below cash flow (separation/Amentum-MtM/tax distortions); cash conversion of adjusted earnings is healthy. Watch working-capital/unbilled-receivables swings.

Risks & Downside

What causes the stock to decline? A de-rating of the full multiple if ROIC fails to inflect and/or growth normalizes to mid-single-digit; margin-bridge stall; environmental softness persisting; leverage/PA integration stress. Catastrophic-loss risk? Low — asset-light, IG credit, 0% fixed-price-at-risk, diversified end markets/geographies. Total-loss risk? Negligible.

Recent News & Events

Business environment changed recently? Yes — Amentum spin (Sept 2024) reshaped the company; PA buy-in (Mar 2026) took PA to 100% and lifted leverage; guidance raised twice in FY26; backlog inflected to $27B. (News flow was quiet in this period; the material events — the spin, the buy-in, and the guidance raises — are all sourced from filings and transcripts.) Acquisitions? PA Consulting minority buy-in (Mar 2026). Accounting-policy changes? Segment realignment to I&AF + PA; increased non-GAAP metric emphasis. Other recent changes? Full exit of the Amentum stake; new senior-notes/term-loan debt; dividend raises; management stable (Pragada CEO, Nathamuni CFO).

APPENDIX B — Source Appendix

Jacobs Solutions Inc. (NYSE: J) — 2026-07-17

Primary sources prioritized over secondary; quantitative figures reconciled to filings. All figures accessed 2026-07-17 unless noted.

Primary — SEC filings (EDGAR, CIK 0000052988; local corpus in output/J/sources/)

• Form 10-K, FY2025 (period end 2025-09-26), filed 2025-11-20 — segments (I&AF, PA Consulting), revenue disaggregation, contract mix, competition, risk factors, employees (~43,000), backlog. output/J/sources/10-K/2025-11-20_jec-20250926.htm
• Form 10-K, FY2024 filed 2024-11-25; FY2023 filed 2023-11-21 — multi-year trend, Amentum-spin accounting.
• Form 10-Q, Q2 FY2026 (period end 2026-03-27), filed 2026-05-05 — PA buy-in balance-sheet impact, leverage, gross-margin/pass-through, backlog $27.0B. output/J/sources/10-Q/2026-05-05_j-20260327.htm
• Form 10-Q, Q1 FY2026 (period end 2025-12-26), filed 2026-02-03.
• 8-K, 2026-03-03 — completion of $800M 4.750% Senior Notes due 2031 + $500M 5.375% Senior Notes due 2036 to fund the PA Consulting minority acquisition. output/J/sources/8-K/2026-03-03_d16121d8k.htm
• 8-Ks, 2026-01-05 / 2026-02-26 / 2026-03-17 — PA Consulting acquisition announcement, senior notes, term-loan/revolver facilities.
• 8-Ks, 2025-11-19 / 2025-11-20 — FY2025 Q4/full-year earnings release and results.
• DEF 14A proxy (latest, in output/J/sources/DEF_14A/) — executive compensation (annual: adj operating profit / adj EBITDA margin / adj net revenue; LTIP: adj EPS / ROIC / stock-price multiplier), insider ownership.
• Form 4 corpus (241 filings, output/J/sources/4/) — insider transactions; reviewed for open-market purchases (only a de-minimis director buy, ~253 sh @ $112.56) vs. routine grants/tax-withholding.

Primary — Earnings-call transcripts (via ROIC.ai MCP)

• Q2 FY2026 (2026-05-05), Q1 FY2026 (2026-02-03), Q4 FY2025 (2025-11-20), Q3 FY2025 (2025-08-05) — guidance (FY26 adj EPS $7.10–7.35; adj EBITDA margin 14.6–14.9%; FY29 target 17%+), backlog/book-to-bill, PA buy-in rationale (13x EBITDA, accretive), Amentum stake exit, end-market demand, margin drivers, capital return.

Quantitative data services

• ROIC.ai MCP — income statement, balance sheet, cash flow, profitability ratios (ROIC ~6.6%, ROE ~14%), enterprise value (~$18.3B), valuation multiples, per-share data (TBVPS −$6.85), FY2020–2025 + Q1/Q2 FY2026. Third-party aggregated; reconciled to filings.
• SEC EDGAR XBRL (scripts/edgar.sh) — CIK resolution, filings index, corpus enumeration.
• AZI — price/OHLCV CSV (5-year event map; 52-wk $106.93–$163.10; ATH $163.10 on 2025-10-21) and valuation_index own-history percentile ranks (P/E 97.5th [GAAP-distorted], P/B 98.3rd [meaningless — negative tangible book], P/S 84.0th, composite 93.3rd). News feed returned no scored items (quiet tape).
• FactorsToday — factor loadings (beta 0.836; Market ~0.74, DividendYield ~0.50; no Momentum/Growth/Value loading), leaderboard (y1 −2.9%, m6 −8.3%, y3 +10%/yr, m3 +22% annualized), stock-info (rs_6m −4.6, off ~19% from high), related-stocks (quality-income industrials).

Peer / cross-read (public filings)

• Peer framing and comp multiples drawn from public filings of the engineering/consulting cohort: AECOM (ACM) and Tetra Tech (TTEK) as the true design/consulting comps (ROIC ~14.7% / ~14.1%); EMCOR (EME), Quanta (PWR), MasTec (MTZ), Primoris (PRIM), IES Holdings (IESC) as the self-perform construction contrast.

Frameworks

• Greenwald & Kahn, Competition Demystified (barriers-to-entry / three-advantage taxonomy / EPV); Chancellor (Marathon), Capital Returns (supply-side capital cycle; asset-growth anomaly) — applied in §3, §4, §6, §7.

Notes / limitations

• Adjusted net revenue (~$8.67B) and adjusted EBITDA ($) are derived from FCF-margin and guidance; the earnings-release Ex-99.1 GAAP-to-non-GAAP reconciliations are not in the local corpus (OPEN QUESTION §13).
• CH2M purchase price (~$3.27B, 2017) predates the 5-year local corpus.
• Management commentary is treated as a hypothesis and validated against filings and financial data.