Quanta Services, Inc. (NYSE: PWR) — The Indispensable Builder of the Grid, Priced as if It Has Already Built It

Report date: 2026-06-12 Price at analysis: ~$707.74 · Market cap: ~$106B · EV: ~$108.6B · Diluted shares: ~151M · Net debt: ~$5.5B (~1.9x adj. EBITDA) Fiscal year: December · Sector: Industrials — Engineering & Construction (Infrastructure Solutions) · CIK: 0001050915

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analysis that follows is presented without a recommendation or price target; the directional valuation view appears only in this opening block.

Verdict: HOLD / quality-compounder-at-the-wrong-price. A wonderful business, but the supercycle is in the price. Not a fresh buy at ~$708, and explicitly not a short. Constructive-accumulation zone below ~$450–500; fair-value band ~$450–560 (≈28–32x a blended forward adjusted EPS bridging FY26 ~$13.90 to FY27 ~$16). Conviction: medium.

Quanta is the best operator in one of the best multi-decade demand setups in all of industrials — a genuine grid/electrification + data-center + generation capex supercycle, throttled on the supply side by a craft-labor shortage that money cannot quickly solve and that favors the largest labor aggregator in North America. The moat is real (scale + customer captivity + a 25-year training pipeline + ~85% self-perform + a vertical supply chain), the balance sheet is investment-grade, capital allocation is disciplined, and incentives are aligned to ROIC and per-share earnings rather than size. None of that is the debate. The debate is price × durability. At ~51x forward adjusted EPS / ~30x forward EV/adjusted-EBITDA / ~100x GAAP / the 98th–99th percentile of its own ten-year history, the market is underwriting that PWR delivers the mid-to-high end of its own 15–20% 2030 EPS algorithm and keeps a richly premium multiple at the end of it. The arithmetic is unforgiving: even if PWR hits its plan’s midpoint (~$24 adjusted EPS by 2030) but the multiple matures to a still-premium ~25x, the stock compounds at roughly flat-to-negative from here. The bear case (normal capex-cycle deceleration + a de-rate toward peers) needs nothing exotic to produce ~−40–50%; the bull case needs a “double hold” — high-end execution and a persistent ~30x multiple — to deliver only a mid-single-digit IRR.

The framing is momentum / great-business-priced-for-perfection, not falling-knife and not contested short (short interest ~3.4%, consensus “Buy,” several Street targets already at or below spot — a crowded, well-owned long). I would happily own this franchise — at a price that leaves a margin of safety the current quote does not. The thing that flips me bullish: a market drawdown that resets the multiple toward ~25–28x while the backlog and book-to-bill stay intact (the business is far more durable than the multiple’s exit risk). The thing that flips me bearish on the business (not just the price): book-to-bill slipping below 1.0 for two consecutive quarters, a hyperscaler/utility capex guide-down, organic growth decelerating below ~6%, or a large debt/equity-funded deal at a stretched multiple signaling the roll-up is buying growth it can no longer grow. Tag: best builder of the grid, priced as if it already built it.

1. Executive Summary

Quanta Services is the largest specialty infrastructure-services contractor in North America — the dominant builder, upgrader, and maintainer of the electric grid (transmission, distribution, substations), and increasingly of power generation, renewables, gas/pipeline systems, communications, and the electrical/mechanical guts of data centers. FY2025 revenue was $28.48B (+20.3%), compounding at a ~20% five-year CAGR ($11.2B in 2020). It is the services/construction expression of the same secular electrification thesis underpinning grid-equipment franchises like Eaton and AMETEK — but as a ~10%-adjusted-EBITDA-margin labor business, not a 20%+ equipment franchise.

The business is genuinely high quality. PWR’s moat is an economies-of-scale-plus-customer-captivity franchise rooted in the scarcest input in the industry: skilled craft labor. It fields the largest craft workforce in North America (~69,500 employees, ~28,000 journeymen and foremen in the field daily), owns the training pipeline that replenishes it (Northwest Lineman College, 10 campuses; enrollments +120% YoY), self-performs ~85% of its work, and has verticalized into transformer/breaker manufacturing to sell “certainty” to capacity-starved utilities and hyperscalers. Roughly three-quarters of its work is negotiated rather than hard-bid, embedding it in customers’ multi-year capital plans. Backlog stood at a record $48.5B at Q1 2026 (total backlog +27% in FY2025; firm remaining performance obligations +42%).

The demand backdrop is exceptional and unusually durable. US investor-owned utility capex hit a record ~$204B in 2024 (14 straight record years) and is forecast to exceed $1.1 trillion cumulatively over 2025–2029, with data-center grid load roughly tripling toward ~130 GW by 2030. The binding constraint is not capital but craft labor — which delays the normal capital-cycle margin mean-reversion and favors the largest aggregator.

But the quality is dressed up, and the price discounts perfection. GAAP operating margin is thin (~5.7%); GAAP net margin ~3.6%; GAAP ROIC is ~8% — at or slightly below cost of capital — with the more flattering cash-ROIC at ~11%. Adjusted EPS ($10.75 FY2025 vs. GAAP $6.80) leans on a recurring $499M intangible-amortization add-back that is the structural cost of the roll-up and will never converge to GAAP. ~58% of FY2025’s growth was acquired, organic ~8.5%; M&A has exceeded free cash flow every year. Tangible equity is negative. And at ~51x forward adjusted earnings — the richest the stock has ever been on its own history — the market prices the 2030 algorithm as near-certain with the premium multiple assumed to persist.

The variant question is therefore not whether Quanta is a good company (it is) but whether the supercycle is already in the price (it largely is). The remainder of this memo argues the case, section by section, and lays out the bull/bear falsification tests. No recommendation and no price target appear below; valuation is discussed only as embedded expectations and scenarios.

2. Business Overview

Quanta Services provides specialty contracting services — engineering, procurement, construction (EPC), upgrade, repair, and maintenance — to the infrastructure that moves and delivers energy and data. It is a labor-and-equipment services business, not a product company: revenue is earned by deploying skilled craft crews, fleets, and project-management expertise against customers’ capital programs.

Two reportable segments (since Q1 2025). A first analytical correction: PWR consolidated from three segments to two beginning Q1 2025, recasting all prior periods. The current structure (FY2025):

(Source: FY2025 10-K, Note 5 “Segment Information.”) The “Electric” segment now houses the old Electric Power and Renewable Energy segments — management folded them together because the craft labor is fungible across electric, generation, and renewable work, an operating-model point that is itself a moat data point (§4). The cost is reduced transparency: the renewables-vs-grid split is no longer disclosed at the segment-margin level.

What each segment does. Electric Infrastructure Solutions: electric transmission and distribution (overhead and underground), substations, grid hardening/smart-grid, and emergency storm restoration; utility-scale renewable generation (solar, wind, storage) EPC and repowering (via Blattner, 2021); low-voltage electrical and mechanical-electrical-plumbing (MEP) inside data centers and large-load centers (via Cupertino Electric, 2024); wireline/wireless communications and fiber; and a growing in-house manufacturing of power transformers and breakers (vertical supply chain). It also includes Northwest Lineman College and the 50% LUMA Energy JV operating Puerto Rico’s grid. Underground Utility & Infrastructure Solutions: gas-utility distribution, pipeline construction/integrity/rehabilitation, civil/earthwork, midstream/downstream industrial services, and — via Dynamic Systems (DSI, July 2025) — turnkey mechanical/plumbing/process infrastructure for data centers, semiconductors, and healthcare.

Customer mix is shifting toward technology. Revenue by customer type (FY2025 / 2024 / 2023): Utility and Power 70% / 74% / 75%; Energy and Other 17% / 17% / 19%; Technology, Manufacturing & Communications 13% / 9% / 6%. The technology/data-center cohort more than doubled in two years and is the fastest-growing market — the diversification engine of the forward story, but also the pull toward larger, lumpier, more fixed-price EPC work.

Revenue model. Three contract types, with a meaningful and important mix shift (FY2025 / 2024 / 2023): Fixed-price 60.6% / 56.2% / 49.1%; Unit-price 23.8% / 27.4% / 31.5%; Cost-plus 15.6% / 16.4% / 19.4%. Roughly 64% of revenue is recognized over time on a cost-to-cost percentage-of-completion basis. Services are delivered under master service agreements (MSAs), repair/maintenance contracts, and fixed/non-fixed-price construction. MSAs = 37% of 12-month backlog and 44% of total backlog — the programmatic, high-retention base — but it is demand-recurring, not contractually committed: the 10-K is explicit that customers “are not contractually committed to specific volumes” and “most of our contracts can be terminated on short notice.” So “recurring” here means high-renewal programmatic spend, not take-or-pay.

Verdict (§2): An Electric-dominated (81% of revenue), labor-intensive specialty-contracting business transforming from a pure utility T&D contractor into a converged utility + generation + data-center infrastructure platform. Revenue is programmatic and sticky but cancellable, and increasingly fixed-price. Headline GAAP margin is structurally thin and weighed down by the amortization cost of the acquisitions that built the platform.

3. Industry Dynamics

Market size and growth — among the best demand setups in industrials. The structural backdrop is a genuine, multi-decade capex supercycle after roughly two decades of flat US electricity demand:

• Utility capex: EEI reports US investor-owned utility capex hit a record $204.1B in 2024 — the 14th consecutive record year — with 2025 projected ~$208B and 2026 forecast to jump ~17% to ~$239B. Cumulative 2025–2029 capex is projected to exceed $1.1 trillion, with transmission build alone ~$178B over 2025–2028 and generation’s share rising toward ~30%. (Sources: Utility Dive / EEI; POWER Magazine; accessed 2026-06-12.)
• Data-center load: US data-center grid demand is forecast to rise from ~25 GW operating in 2024 toward ~130 GW by 2030 (S&P Global / 451 Research), roughly tripling; BNEF projects ~106 GW by 2035. AI compute is the accelerant.
• Quanta’s own TAM claim: ~$2.4 trillion, “twice as big” as a few years ago (Investor Day, 2026-03-31). Treat the absolute figure as directional marketing — PWR’s ~$28–35B revenue is ~1.2–1.5% of it — but the direction is corroborated by EEI and BNEF.

Demand drivers are unusually broad and mutually reinforcing: (1) load growth from AI/data centers, electrification (EVs, heat pumps), and manufacturing reshoring; (2) an aging grid and aging utility workforce driving replacement demand independent of load; (3) grid hardening/resiliency/wildfire mitigation; (4) renewable interconnection and storage; (5) a generation rebuild (new gas CCGT, nuclear life-extension, batteries) to feed large loads; and (6) grid congestion (>$10B/yr) as the political justification for building. Crucially, several of these are non-discretionary (reliability, replacement) and therefore less cyclical than a typical construction end-market.

Competitive intensity and barriers — candid and important. The 10-K is unusually frank and undercuts a naïve “wide moat” read: “there are relatively few barriers to entry into some of the industries in which we operate,” competitors include “large companies,” subcontractors “may develop the expertise necessary to compete with us,” and “a significant portion of our revenues is currently derived from unit price or fixed price agreements, and price is often an important factor… we could be underbid.” At the project level — especially small/commodity work and renewables EPC — this is a competitively bid industry with thin structural profitability (industry operating margins ~5–9%). The moat is not at the project level; it is at the scale/relationship/labor-aggregation level on the largest, most complex programs (§4).

Capital-cycle read (Marathon lens). This is mid-cycle with capital visibly flooding in — normally the Marathon yellow flag for future margin mean-reversion: every peer (MasTec, Primoris, EMCOR, MYR Group) is expanding, PE-backed entrants are arriving, and PWR itself is the largest acquirer. The mitigant is that the binding constraint is skilled labor, not capital — linemen and journeymen take years to train and cannot be conjured by a capital raise. This supply-side throttle delays the normal capital-cycle margin compression and structurally favors the largest labor aggregator. The open question is whether the labor scarcity protects margins (pricing power) or merely volume — so far consolidated GAAP margin has been roughly flat (~5.4–5.7%), with the improvement story carried by mix and adjusted EBITDA, not headline pricing.

Verdict (§3): Structurally GOOD industry — one of the best multi-year demand setups in industrials, with a decade-plus runway throttled by a labor shortage that favors scale. Two caveats keep it short of “great”: project-level work is competitively bid with low formal entry barriers (so industry-wide structural profitability is thin), and the renewables/pipeline pieces are policy-cyclical. The tailwind is real and durable; it does not by itself confer pricing power — that depends on the firm-specific moat.

4. Competitive Position

Moat type (Greenwald taxonomy): economies of scale fused with customer captivity — specifically scale economies in skilled-labor aggregation, training, supply chain, and geographic density, locked in by relationship/MSA captivity. Supporting elements: a cost advantage from self-perform scale and an intangible (the training pipeline, safety reputation, and 25-year head start). This is not a network-effects or consumer-brand business.

The specific, financially-tied advantages:

1. The largest fungible craft workforce in North America. ~69,500 employees, ~55,700 hourly, with a core of ~28,000 journeymen and foremen in the field daily (~36% under collective-bargaining agreements). In a labor-constrained industry, this is the scarce asset — and it is fungible across electric, generation, solar, storage, and data-center work, which is exactly what let PWR fold renewables into Electric and redeploy crews as policy cycles swing. Financial tie-in: it enables ~85% self-perform (90% on the SunZia project), capturing subcontractor margin and giving the schedule/cost certainty that wins negotiated (vs. bid) work.

2. A captive training pipeline. PWR owns Northwest Lineman College (10 campuses) and an Advanced Training Center; a new competency-based apprenticeship trains journeymen up to 30% faster, with enrollments +120% YoY. Unions route members through PWR’s programs. This is the closest thing to a true intangible barrier — a competitor cannot replicate 25 years of training infrastructure and union relationships quickly.

3. A vertical supply chain. PWR owns transformer, breaker, and prefab/modular manufacturing (expanding toward >5M sq ft of off-site fabrication, including doubling 765kV transformer capacity). With transformers carrying multi-year lead times, owning supply is “certainty” PWR sells to capacity-desperate utilities and hyperscalers; ~$2.5B of intracompany revenue is now eliminated in consolidation — work PWR increasingly does for itself rather than subcontracting.

4. Breadth and density → “total solutions.” PWR can deliver interconnect + substation + generation + data-center MEP on a single program, turning a “$50M relationship into a $5–7B opportunity over 5–7 years” (the NiSource-Indiana / SunZia model). Breadth lets it prime programs too large or complex for single-discipline competitors.

5. Bonding, balance sheet, and safety ($14.9B of performance bonds outstanding; investment-grade) screen out sub-scale bidders on the largest jobs.

6. Programmatic/negotiated contracting. ~75% of work is negotiated, not hard-bid; many relationships are strategic alliances with preferential bidding. Being “in the rooms where customers are planning their entire multiyear capital spend” is the captivity mechanism that removes the price shootout the 10-K admits exists at the commodity end.

Share-stability test (Greenwald): passes at the top end, mixed overall. PWR has held the #1 position in North American utility infrastructure services for years and is structurally larger than the next competitors; its ten largest customers are ~30% of revenue with no single customer ≥10% — moderate, stable concentration. But part of its share gain has been bought (Blattner, Cupertino, DSI), so “stable share” is partly a consolidation artifact, and at the small-project commodity end share is genuinely contestable.

ROIC test: the moat is real but moderate. Management targets and claims 12–15% ROIC; our strict GAAP computation is ~8% (below an ~8–9% WACC), with cash-ROIC (adding back after-tax amortization) at ~11%. The moat exists — it ties to the ability to win negotiated work (supporting the ~10% Electric segment margin), to adjusted EPS compounding ~25% over a decade, and to backlog growing faster than revenue — but the level of excess return is moderate (mid-teens ROIC at best, ~5.7% GAAP operating margin), not fortress-grade. This is a good business, not a spectacular one.

Direct competitor comparison:

PWR’s GAAP operating margin (~5.7%) is below EMCOR’s (~9.2%), so the moat does not show up as superior headline margin. It shows up as scale/breadth (2x the nearest pure peer; the only player able to self-perform a SunZia or a full data-center MEP-plus-interconnect program), consistency (two consecutive five-year plans beaten), and incremental returns. EMCOR’s higher margin reflects an asset-lighter building-services mix, not a wider moat in grid infrastructure.

Where the moat is thin — directly stated. (1) Renewables EPC is the most commoditized end, competitively bid with lower barriers; folding it into Electric partly masks this. (2) The rising fixed-price mix (49%→61% in two years) imports E&C performance risk the moat doesn’t neutralize — the disputed Canadian renewable-transmission claim ($900M+ of a $984M change-order/claim balance) is a live example. (3) Labor-cost pass-through is imperfect on some fixed-price/long-term MSAs — the labor shortage protects volume but is also a cost risk if PWR can’t reprice. (4) Subcontractor-to-competitor leakage and regulatory subcontracting mandates cap self-perform and dilute margin. (5) Utilities could in-source work. (6) The marquee data-center MEP capability is newly acquired (Cupertino <2 years; DSI brand-new) and unproven — competing with established general contractors.

Verdict (§4): A durable but MODERATE-width moat — narrower than the most bullish framing. PWR has a genuine, hard-to-replicate economies-of-scale-plus-captivity moat anchored in the largest fungible craft-labor force in North America, an unmatched training pipeline, ~85% self-perform, a vertical supply chain, and deep programmatic utility relationships — all tied to financial outcomes (mid-teens ROIC ambition, target-beating EPS compounding, record book-to-bill). But the moat is concentrated at the top end of the most complex programs and is thin at the commodity/renewables-EPC end, where the 10-K concedes “few barriers to entry.” It is being stretched by a rising fixed-price mix, and the fastest-growing vector is newly acquired. The durability question is whether scale plus labor scarcity keep competition rational, or whether the capital and labor flooding into the sector eventually compress the very returns the bull case extrapolates.

5. Growth History and Forward Opportunities

Historical growth — strong, but increasingly acquired. PWR has compounded revenue at ~14% and adjusted diluted EPS at ~25% over ten years, beating its last two five-year EPS targets ($4.91 actual vs. $3.98 target in the prior plan). FY2025 revenue grew +20.3% to $28.48B — but ~58% of that increase (~$2.8B) was acquired (Cupertino in Electric, plus the broader 2025 deal set in both segments), with organic growth ~8.5%. Over ten years PWR generated ~$7B of FCF while spending ~$10B on M&A; over the last four years, ~$5B FCF vs. ~$6B M&A. The compounding record is genuine, but it is the combined output of organic growth and a continuous, balance-sheet-and-equity-funded roll-up — it is not self-funding the way a high-ROIC compounder’s growth is. That is the central growth-quality caveat.

Forward drivers — the demand stack. Management maps six end markets: (1) technology/load center (data centers) — the single fastest-growing, but only ~10% of backlog today, with PWR sizing its addressable MEP scope at ~50–60% of a facility’s craft spend (~$13.5M/MW), roughly doubling if generation and HV interconnects are added; (2) the electric grid and gas utility backbone; (3) power generation and storage (~double-digit growth, gas CCGT + renewables/storage); (4) communications; (5) industrial; (6) pipeline (with gas-to-generation/LNG upside). Programmatic spend with anchor utilities is cited running to “2032, 2034,” with the AEP 765kV relationship and the NiSource integrated model as templates.

Visibility — backlog and book-to-bill. FY2025 total backlog $43.98B (+27%); firm RPO $23.76B (+42%); Q1 2026 record total backlog $48.5B. Implied book-to-bill is comfortably >1. The important caveat: ~46% of backlog is non-binding 12-month MSA estimate — nearly half of the headline “$48.5B” is a programmatic-spend estimate, not firm contracted RPO, and firm RPO (<1x forward revenue) is the harder number.

Verdict (§5): High-quality demand backdrop, medium-quality reported growth. The end markets are structurally advantaged and durable well beyond 2030, and the moat is real. But the recent revenue beat is majority-acquired, ~46% of backlog is non-binding MSA estimate, and renewables/storage growth is policy-contingent. The demand is more durable than the reported numbers are clean.

6. Financial Quality

The adjusted-earnings bridge is the crux. PWR’s 10-K MD&A is GAAP-only; the non-GAAP reconciliation lives in the earnings releases/calls. Reported results:

The GAAP→adjusted wedge is $3.95/share (37%), dominated by $498.8M of intangible amortization (~$3.30/share pre-tax; up 30% YoY on the Cupertino/DSI deals), plus rising acquisition/integration costs, a $31.2M contingent-consideration remeasurement (up from $7.1M), and $181.9M of non-cash SBC. The honest critique: ~$499M of annual amortization plus growing acquisition costs are the recurring cost of the M&A growth model, so “adjusted” EPS systematically overstates capital-charge-inclusive earning power and will never converge to GAAP. That said, the adjustments are mainstream for a roll-up — SBC dilution is low and there are no large restructuring/impairment add-backs polluting the number.

Margin structure: a volume compounder, not a margin-expansion story. Consolidated gross margin ~15.0%, GAAP operating margin ~5.7% (flat), net margin ~3.6%. Electric segment margin is flat at 10.3% (2024→2025) despite +21% revenue — the easy mix-driven gains were 2024, and 2025 was volume, not margin. Underground recovered to 7.3% (from a depressed 5.7%), the principal margin lever management points to for 2030. Critically, management itself tempers margin hopes: hiring ~6,000 employees organically per year “is pressure on those margins… we continue to have the same ratios,” and the 2030 target is only 10–11% adjusted EBITDA margin — roughly +30bps versus the 2026 guidance midpoint. Anyone underwriting margin expansion in the multiple is fighting management’s own guidance.

Returns on capital — the roll-up’s weak spot. Our reconciled FY2025 computations: ROE 12.7% (on a clean ~32%-of-assets equity base, so not buyback-inflated); NOPAT ~$1.21B; invested capital ~$14.7B; GAAP ROIC ~8% — at or below an ~8–9% WACC. Adding back after-tax amortization lifts cash-ROIC to ~11% — above WACC but not dramatically. The honest answer: PWR earns a modest positive spread on a cash basis and roughly breakeven on a strict GAAP basis. $7.32B goodwill + $2.91B intangibles (= 41% of total assets) drag the denominator; tangible common equity is negative (−$1.3B), making book-value metrics meaningless. Management’s own 12–15% ROIC target is a future promise — it concedes current returns need to improve.

Quality-of-earnings flags (none individually fatal, collectively worth weighting):

• Unbilled change orders and claims of $983.6M (+34% YoY) sit in contract assets as management estimates “probable of collection,” not approved/billed — the largest piece a disputed Canadian renewable-transmission project whose collection is built into 2026 FCF guidance. A growing slice of reported revenue depends on unconcluded negotiations.
• DSO rose to ~88 days from ~80, with AR +32% vs. revenue +20% and a −$854M working-capital drag from receivables — the classic POC-contractor warning combination (rising DSO + rising unbilled claims), partly explained by acquired-business AR and large-project mix.
• ~46% of backlog is non-binding MSA estimate (firm RPO $23.76B is the harder number).

Quality-of-earnings positives (genuine):

• PWR is net billed-ahead by $1.74B (contract liabilities $3.26B vs. contract assets $1.52B) — customers fund the work in advance, a structurally favorable working-capital posture and a major OCF tailwind.
• SBC is low (0.64% of revenue) and dilution minimal (145M→151M shares over three years, partly offset by buybacks).
• No large one-time items distort the run-rate; equity-method affiliates contribute (+$56M), not detract.

Balance sheet — a genuine strength. Total debt ~$6.0B, cash ~$0.44B, net debt ~$5.5B = ~1.9x adjusted EBITDA; $4.75B of fixed-rate senior notes laddered 2027–2041 (including legacy 2.90%/2.35% coupons), a benign maturity wall, ~$2.86B total liquidity, and ~11x adjusted interest coverage. The August 2025 $1.5B note issuance termed out the bridge financing for Dynamic Systems — disciplined funding. The $3.05B of FY2025 M&A was funded from OCF + notes + revolver/CP, not equity dilution.

Verdict (§6): Economics are GOOD-not-great, presented as great by adjustments. Adjusted EBITDA margin ~10.2% (flat) and adjusted EPS compounding ~20%, but GAAP ROIC ~8% is barely at/below WACC; cash-ROIC ~11% is the defensible figure. Economics improve only modestly with scale — this is a volume compounder, not a margin-expansion story — and the marquee EPS growth is majority-acquired. The balance sheet is a real strength; the returns profile is the real weakness.

7. Capital Allocation

The framework. Management’s stated priority stack: (1) organic growth (people, equipment, supply-chain/fabrication build-out); (2) acquisitions of “great family businesses” with accretive returns; (3) return excess capital via a growing dividend and opportunistic buybacks — explicitly measured against deal returns (“whatever drives a higher return is how we’ll move forward”). The deployment cadence is telling: over the last four years, ~$5B FCF generated and ~$6B spent on acquisitions; over ten years, ~$7B FCF and ~$10B of deals. M&A consumes ~100%+ of FCF — PWR is a net acquirer that outspends its own cash generation and funds the gap with debt and equity.

M&A track record. Goodwill rose from $2.1B (2020) to $7.32B (2025). The marquee deals: Blattner (2021, ~$2.7B) entered utility-scale renewables — strategically sound, but now the franchise most exposed to the OBBBA wind/solar credit phase-out (§8); Cupertino Electric (2024, ~$2.04B) bought the single hottest end-market (data-center electrical) ahead of the AI capex surge — the best-timed deal; Dynamic Systems (2025) extended mechanical/MEP into data-center/semi/healthcare. Discipline markers: PWR models no synergies, pays ~5–10x, and has owners roll into PWR equity and stay; ROIC expanded while deploying >$6B, and there are no goodwill impairments in the corpus despite the 2.5x goodwill build (acquired businesses are performing — rising earnout marks, +341% to $31.2M, signal acquisitions outperforming targets). Watch-items: M&A exceeds FCF every year (debt/equity-funded growth); the marquee 15–20% EPS algorithm is explicitly acquisition-dependent (management declined to give an organic EPS target); diluted shares crept up; and $7.3B goodwill (41% of assets) concentrates integration/impairment risk, with Blattner a specific renewables-impairment watch-item.

Capex is rising. From a historically asset-light ~2% of revenue, PWR is building a vertical supply chain (transformer/breaker factories, a $500–700M program, 765kV capacity doubling), which pressures near-term FCF conversion (hence only 55–60% conversion targeted by 2030) but deepens the moat.

Dividend and buyback are immaterial — by deliberate choice. The dividend is a token ~$0.41/share (~0.06% yield, ~$62M); buybacks were only $134.6M in 2025 ($0 in 2024) — anti-dilution mop-up that doesn’t even fully offset SBC + acquisition-equity dilution. Every dollar is prioritized to organic + M&A because management believes deployment returns (a 12–15% ROIC ambition) beat repurchasing stock at ~50x earnings. That is internally consistent — but it means per-share value depends entirely on EPS growth outrunning dilution, with no buyback support.

Incentives — well-aligned. CEO Earl “Duke” Austin Jr. total comp ~$15.7M (~70% equity, ~70% of equity performance-based PSUs); CFO Jayshree Desai ~$5.4M. PSUs (3-year cliff, 0–215%) vest on ROIC and cumulative adjusted EPS, with a relative-TSR-vs-S&P-Industrials modifier — textbook-correct metrics; the 2025 plan added cumulative adjusted EPS, strengthening alignment. The annual plan (60% adjusted EBITDA / 20% margin / 20% safety) penalized management for a safety miss in 2025, showing the plan isn’t a rubber stamp. The two yellow flags: plans paid out richly in the up-cycle (188.5% PSU / 126.8% annual), and say-on-pay was ~93% (a notch below the ~95%+ clean threshold — modest investor pushback). Insider ownership is low as a percentage (~0.6%) though large in dollar terms for Austin (~707,757 shares, ~$500M). The board is competent and sector-relevant (Waste Connections, Flowserve, ex-Ameren CEOs).

Insider activity — no buying, a notable sale. Across the full trailing Form 4 corpus (269 filings), there were zero open-market purchases (code P). The notable transaction: Austin’s discretionary, non-10b5-1 sale of ~155,992 shares on May 7, 2026 at ~$766–778 (~$120M) near the all-time high — a partial monetization, not an exit, but its discretionary nature and top-tick timing deserve a flag. Net insider direction is selling/grant-receipt, with no conviction buying at any price.

Verdict (§7): An above-average capital allocator running a disciplined-but-aggressive roll-up — tested at a cyclical peak. The ROIC expansion while deploying >$6B, the absence of impairments, and the negative-working-capital cash engine are real evidence of skill, and incentives are aligned to ROIC and per-share earnings rather than size. But the model is reinvestment/roll-up, not per-share return: M&A persistently exceeds FCF, the headline EPS algorithm is acquisition-dependent, buybacks/dividends are immaterial, and goodwill at 41% of assets concentrates risk. The thesis rests on continued accretive M&A and a sustained up-cycle — not on self-funded per-share compounding.

8. Changes and Headwinds — Last Two Years

The positives dominate the timeline. 2024: Cupertino Electric (~$2.04B) — the best-timed deal, anchoring the data-center segment now ~10% of backlog and the fastest-growing market. Q1 2025: segment reorganization (3→2), strategically coherent but a modest disclosure-quality negative. July 2025: Dynamic Systems plus seven other 2025 deals (~$3.05B), extending data-center MEP. Throughout: record results (FY2025 adjusted EBITDA $2.90B, adjusted EPS $10.75), the vertical supply-chain build-out, the March 2026 Investor Day 2030 framework, and a Q1 2026 beat-and-raise — FY2026 guidance lifted to revenue $34.7–35.2B and adjusted EPS $13.55–14.25, with record $48.5B backlog (Q1 2026 adjusted EPS $2.68 vs. $1.78). Leadership is stable (Austin CEO ~10 years; orderly bench development).

Two genuine headwinds to underwrite.

1. OBBBA wind/solar credit phase-out — the key structural overhang. The One Big Beautiful Bill Act (signed July 4, 2025) phases out the §45Y PTC and §48E ITC for wind and solar — terminated for projects placed in service after December 31, 2027 unless construction begins before July 4, 2026 (a 12-month safe-harbor window, then a cliff). Battery/storage, geothermal, hydro, and nuclear retain full credits through 2034. This is a direct headwind to PWR’s Blattner renewables franchise, creating a 2026 pull-forward (developers racing to safe-harbor) followed by a post-2027 air-pocket risk for new solar/wind starts. Management downplays it (“I’m not concerned with that business… there’s always going to be something”), citing proactive safe-harboring and a robust pipeline — but that is an asserted hypothesis, not proof, and the open question (unquantified by management) is what share of renewables backlog is safe-harbored versus exposed. The partial offset is real: the same load growth driving the cliff also drives gas generation, transmission (765kV), storage, and grid demand — all PWR strengths — and storage retains credits, so the net policy hit is cushioned by mix-shift toward credit-protected and non-credit-dependent work.

2. Wildfire tail-risk (Silverado matter). Since 2022, two PWR subsidiaries have received tenders of defense and evidence-preservation demands from Southern California Edison relating to the October 2020 Silverado Fire. PWR is not yet a named defendant — this is a contingent indemnity exposure — but utility-contractor wildfire liability is the sector’s catastrophic-loss vector (cf. PG&E) and warrants monitoring.

Verdict (§8): On balance, the last two years STRENGTHEN the thesis. Demand and execution dominate (well-timed data-center M&A, record raised guidance, $48.5B backlog, supply-chain verticalization, a credible 2030 framework). The renewable-policy cliff is the one durable structural overhang to track — real but partly mitigated — and the wildfire tender is a low-probability catastrophic-tail watch-item.

9. Risk Analysis

The dominant risk is unambiguously valuation: the business risks are mostly medium-likelihood/medium-impact and partly mitigated, but the price leaves no margin of safety, so even modest disappointment on growth, margin, or multiple translates into outsized downside.

10. Valuation Discussion (Embedded Expectations)

Where the multiple sits. At ~$707.74, PWR trades at ~51x forward adjusted P/E (FY2026 midpoint ~$13.90), ~30x forward EV/adjusted-EBITDA (FY2026 midpoint ~$3.57B), ~3.5x sales, and ~100x GAAP EPS — the 98th–99th percentile of its own ten-year history and a premium to every pure-play E&C peer (MTZ ~31x / 25.6x; EME ~25x / 19.3x; PRIM ~16.6x / 12.2x; ACM ~10.4x / 9x). The one richer comparable is FIX (Comfort Systems), itself near the top of its own valuation history at the peak-cycle. There is essentially no statistical margin of safety from the multiple.

What the price embeds. Management’s 2030 algorithm: 7–10% organic revenue CAGR (→ ~$44–49B revenue), 10–11% adjusted EBITDA margin, and 15–20% adjusted EPS CAGR off $10.75 → ~$21.60–26.75 by 2030 (a figure that includes assumed M&A; organic-only EPS growth is guided merely “double-digit”). The reverse logic, holding today’s price and asking what 2030 EPS × exit multiple clears a given return over ~4.5 years:

The embedded bet. At ~51x forward, the market underwrites that PWR executes the mid-to-high end of its own algorithm and retains a premium ~28–32x multiple at the end of it. Even hitting the plan midpoint ($24) with a de-rate to a still-premium 25x compounds at roughly zero-to-negative from here — because normal multiple maturation offsets ~15% earnings growth. To earn an equity-like return, the market needs both the high-end algorithm and a sustained rich multiple.

Illustrative scenarios (assumptions explicit; ~151M shares; not targets):

• Bear: organic decelerates to mid-single-digits, data-center capex rolls or renewables policy bites, margins stay ~10%, multiple compresses toward peers. Needs nothing exotic — a normal capex-cycle event plus a de-rate.
• Base: hits the algorithm midpoint (~15% EPS CAGR), margins ~10.5%, but the multiple de-rates to ~24–26x as a $50B-revenue company’s growth matures (normal). Management hits its plan and the stock still produces a flat-to-modestly-negative return.
• Bull: grid + data-center + gas “full stack” all accelerate (18–20%+ EPS CAGR), margins ~11%, multiple holds ~30x on sustained scarcity/visibility — a “double hold.”

The asymmetry. Unlike a cheap stock, the base case (management executes) still yields a flat-to-negative return because the entry multiple is so high; the bull case requires both high-end execution and multiple persistence to deliver only a mid-single-digit IRR; the bear case requires nothing unusual and implies ~−45–50%. The market is pricing the 2030 algorithm as a near-certain base case with the premium multiple assumed to persist. Margin of safety from valuation is negligible.

11. Variant Perception

Consensus. Near-universal “Buy” (~22 buy / 7 hold / 1 sell), with targets clustering ~$590–760 — several at or below the ~$708 price, i.e., the stock has run into/through the average target. Short interest is ~3.4% (low). This is a crowded, well-owned momentum long, not a contested short or a falling knife. The debate is not “is this a good company” (clearly yes) — it is “is the supercycle already in the price.”

Strongest bull case. Grid T&D is in a genuine multi-decade supercycle; data-center/large-load electrification is a structural demand shock at just ~10% of backlog and accelerating; PWR has a non-replicable moat (scarce craft labor, ~85% self-perform, vertical supply chain, deep programmatic customer integration) and a record $48.5B backlog with visibility to 2032–2034. Management has beaten its last two five-year EPS targets. In this view, 15–20%+ EPS CAGR is conservative, margins drift to 11%, and a premium multiple is earned by best-in-class durable compounding.

Strongest bear case. Peak-multiple × maturing-growth × M&A-funded compounding. At ~51x forward / 98–99th-percentile own-history, the price embeds the 2030 algorithm and a persistent premium multiple. Three vulnerabilities: growth is majority-acquired and the algorithm bakes in continued M&A (if deal flow slows or deal multiples rise, EPS misses); ~46% of backlog is non-binding MSA estimate and renewables is policy-contingent; data-center capex is a capital cycle, not a perpetuity. Mechanically, even hitting the plan midpoint yields a flat-to-negative return because the multiple must mature. ROIC is only ~8% GAAP / ~11% cash — below the 12–15% target, so the high returns are a future promise, not a present fact.

The 3–5 assumptions that matter most:

1. Is data-center/large-load demand secular (decade-scale) or cyclical (a capex bubble)?
2. Can PWR sustain 7–10% organic growth without the M&A crutch — and keep finding accretive deals to hit 15–20% total?
3. Does adjusted EBITDA margin actually expand to 10–11%, or stall at ~10%?
4. Does the premium ~50x multiple persist, or de-rate toward peers (~25x) as the company scales toward $50B revenue? (the single biggest swing factor)
5. Does ROIC actually climb from ~8–11% to the promised 12–15%?

Falsifying evidence. Falsifies the bull (toward de-rate): book-to-bill <1.0 for two consecutive quarters; firm RPO declining; a hyperscaler/utility capex guide-down; organic growth below ~6%; margin stalling at ~10%; ROIC not improving; a large deal at a stretched multiple; equity issuance for deals. Falsifies the bear (validates the premium): sustained 9–10% organic growth through a softer macro quarter; margin demonstrably climbing toward 11%; ROIC inflecting to 12%+; data-center backlog mix rising from 10% toward 20%+ with firm contracts; FCF conversion reaching 55–60%; continued beat-and-raise.

Framing. Momentum / quality-compounder-at-a-rich-price. The business quality and demand backdrop are real and durable; the question is whether ~51x forward already discounts the entire supercycle plus a persistently rich multiple — leaving the marginal buyer underwriting execution, multiple, and policy/M&A risk for a base-case return that is flat-to-negative. The supercycle is largely in the price; the margin of safety is thin.

12. Fact vs. Interpretation

13. Open Questions

1. What share of Blattner/renewables revenue and backlog is safe-harbored before July 4, 2026 versus exposed to the post-2027 cliff? Management asserts double-digit growth through 2030 but provides no exposure quantification.
2. What are the precise multiples paid for Blattner, Cupertino, and Dynamic Systems? PWR discloses only “5x / 9–10x framework” color; deal EBITDA multiples are not given.
3. Size and collection timing of the unresolved Canadian renewable-transmission receivable baked into 2026 FCF guidance — a concentrated, real collection risk.
4. How does management’s 12–15% ROIC definition reconcile to the ~8% GAAP figure? Almost certainly a cash-NOPAT-on-adjusted-capital measure; the gap matters for whether the roll-up creates value.
5. Was Austin’s May 2026 discretionary (non-10b5-1) ~$120M sale a one-off diversification or the start of a pattern? Currently isolated.
6. Does the labor constraint protect margins (pricing power) or only volume? Consolidated margin has been flat; the bull thesis needs the former.
7. Goodwill/intangible impairment sensitivity given $10.2B (41% of assets) of acquired intangibles, with Blattner the specific watch-item under OBBBA.

14. What Must Be True

For the bull case (the premium is earned):

• Data-center/large-load demand proves secular — the ~10%-of-backlog technology cohort scales toward 20%+ with firm (not MSA-estimate) contracts, and hyperscaler/utility capex stays on its multi-year path.
• PWR sustains 7–10% organic growth and keeps deploying capital accretively to reach 15–20% total EPS growth, with adjusted EBITDA margin climbing to 11% and ROIC inflecting to 12%+.
• The premium multiple persists (~28–32x) because scarcity, visibility, and best-in-class execution justify it even as revenue scales toward $50B.
• Falsification test: book-to-bill falls below 1.0 for two consecutive quarters, OR firm RPO declines, OR organic growth drops below ~6%, OR margin stalls at ~10% with ROIC failing to improve. Any one materially weakens the “earned premium” claim.

For the bear case (multiple compression dominates):

• Growth decelerates toward mid-single-digit organic as the data-center capex cycle matures and renewables hits the post-2027 cliff; the multiple normalizes toward peers (~20–25x) as a $50B-revenue company’s growth slows.
• The acquisition-dependent EPS algorithm falters — deal flow slows, deal multiples rise, or an integration/Blattner impairment surfaces — and ROIC fails to reach the 12–15% target, exposing the roll-up as value-neutral.
• Falsification test: PWR delivers 9–10% organic growth through a softer macro quarter, margin demonstrably climbs toward 11%, ROIC inflects to 12%+, the technology backlog mix rises with firm contracts, and the company sustains beat-and-raise — any of which would validate the premium and break the de-rating thesis.

The crux: both sides agree the business is good and the demand is real. The disagreement is entirely about price and the durability of a peak multiple. The base case — management executes its plan — still produces a flat-to-negative return from here, which is why the burden of proof rests on the bull to justify multiple persistence, not on the bear to explain a fundamental break.

15. Source Appendix

See the Source Appendix (Appendix B below) for the full list of primary and secondary sources, including: Quanta Services FY2021–FY2025 Forms 10-K; Q1 2026 Form 10-Q; DEF 14A (2026-04-10); the trailing Form 3/4/5 corpus; 8-K material-event filings; the Q4 2025 (2026-02-19), Q1 2026 (2026-04-30) earnings-call transcripts and the March 31, 2026 Analyst/Investor Day; EDGAR XBRL financial data; EEI/POWER Magazine utility-capex data; S&P Global and BNEF data-center-load forecasts; and OBBBA legal analyses (Simpson Thacher, Steptoe, Tax Foundation).

APPENDIX A — Standard Diligence Questionnaire — Quanta Services, Inc. (NYSE: PWR)

Supplemental to the research memo. Report date 2026-06-12. Labels: Fact / Interpretation / Assumption.

General

What thoughtful questions have other investors asked about this company? The most-asked institutional questions cluster on: (1) how much of growth is organic vs. acquired (Fact: ~58% of FY2025 growth was acquired; organic ~8.5%) and whether the 15–20% EPS algorithm survives without continuous M&A (management declined to give an organic EPS target, conceding it bakes in deployment); (2) margin trajectory — analysts repeatedly press on why scale isn’t lifting margins (management: hiring 6,000/yr “is pressure on margins… same ratios”); (3) data-center durability — is the ~10%-of-backlog technology cohort secular or a capex bubble; (4) renewables/OBBBA exposure — how much of Blattner backlog is safe-harbored vs. exposed to the post-2027 cliff (unquantified by management); (5) M&A multiples and discipline — the Investor Day “5x / 9–10x” framework Q&A; and (6) the valuation itself — whether ~51x forward can persist.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: a cyclical high in absolute terms — record revenue, adjusted EBITDA, adjusted EPS, and backlog — though management argues the demand cycle is secular (decade-plus utility/grid/data-center capex) rather than a near-term peak. The valuation is at a historical high (98–99th percentile own-history) regardless.

Driven by the external environment or internal actions? Both: the external utility/electrification/data-center capex supercycle is the dominant tailwind; internal actions (the roll-up, vertical supply-chain build-out, self-perform scale) amplify it.

How stable are revenues? Interpretation: more stable than typical construction because ~70% is regulated-utility-driven and ~44% of backlog is programmatic MSA spend — but it is demand-recurring, not contractually committed (most contracts terminable on short notice), and the rising fixed-price mix (61%) imports project lumpiness.

Outlook for products/services? Fact: FY2026 guidance raised to revenue $34.7–35.2B and adjusted EPS $13.55–14.25; 2030 targets of 7–10% organic revenue CAGR and 15–20% adjusted EPS CAGR.

How big will this market be? Fact: growing — US utility capex >$1.1T cumulative 2025–2029; data-center grid load roughly tripling toward ~130 GW by 2030; PWR’s claimed TAM ~$2.4T. Primarily North American (US + Canada); Latin American operations were wound down.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Interpretation: more capital is flooding in (Marathon caution), but the binding constraint is skilled labor, which favors the largest aggregator and delays margin mean-reversion. Competitive at the commodity/renewables-EPC end; less so on the largest, most complex programs.

How profitable is the business (ROIC, ROE)? Fact/Interpretation: ROE 12.7%; GAAP ROIC ~8% (at/below WACC); cash-ROIC ~11%. Returns are modest — a good business, not a spectacular one — with management targeting improvement to 12–15%.

How profitable is the industry — how many competitors, what barriers to entry? Fact: thin structural margins (~5–9% operating industry-wide); the 10-K concedes “relatively few barriers to entry into some of the industries.” Key peers: MasTec, EMCOR, Primoris, MYR Group, AECOM (adjacent).

Can the business be easily understood? Yes — a specialty-contracting services business; the complexity is in non-GAAP adjustments and the roll-up accounting.

Can it be undermined by foreign low-cost labor? No — work is on-site, domestic, union-heavy, skilled craft; not offshorable.

Do brands matter? Interpretation: not consumer brands, but reputation/safety record/bonding capacity function as a B2B credential that screens out sub-scale bidders.

What is the nature of competition? Negotiated/programmatic (~75%) on large work; competitively bid on smaller/commodity work.

Customers’ switching costs? Interpretation: high on programmatic alliances (embedded in multi-year capital plans, scarce qualified labor, safety/performance track record), low on one-off bid work.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Interpretation: the trained craft workforce and union/training relationships (the real moat) are not capitalized; conversely, $10.2B of goodwill+intangibles (41% of assets) is on the balance sheet at acquisition cost.

Off-balance-sheet liabilities? Fact: $14.9B of performance bonds outstanding (surety, not debt); operating leases; self-insurance reserves ($521M gross accrual); contingent consideration/earnouts ($31.2M FY25 remeasurement; $102.6M cash paid); the Silverado wildfire indemnity tender (contingent).

How conservative is the accounting? Interpretation: mixed. Conservative: net billed-ahead $1.74B (customer-funded working capital), low SBC, no impairment add-backs. Aggressive/judgment-laden: $984M unbilled change-orders/claims (+34%), percentage-of-completion estimates, heavy reliance on adjusted metrics excluding recurring amortization.

How CapEx-hungry is the business? Interpretation: historically asset-light (~2% of revenue), now rising with the vertical supply-chain build-out ($500–700M factory program), pressuring FCF conversion (55–60% targeted only by 2030).

Capital Allocation & Management

How much FCF, and how is it used? Fact: ~$1.7B FCF FY2025; priority stack = organic growth, then M&A, then a token dividend/opportunistic buyback. M&A consumes ~100%+ of FCF (net acquirer funding the gap with debt/equity).

Significant acquisitions recently? Fact: Blattner (2021, ~$2.7B, renewables), Cupertino Electric (2024, ~$2.04B, data-center electrical), Dynamic Systems (2025, mechanical/MEP) + ~$3.05B of 2025 deals total.

Buying back shares? Fact: minimal ($134.6M FY2025, $0 FY2024) — anti-dilution mop-up, not a return lever.

Issuing large amounts of new shares to insiders? Fact: SBC $182M (0.64% of revenue, low); acquisition equity issued (owners roll in); diluted shares crept 145M→151M over 5yr.

Compensation policy? Fact: CEO ~$15.7M, ~70% equity; PSUs on ROIC + cumulative adjusted EPS + relative-TSR modifier (well-aligned); say-on-pay ~93%; annual plan penalized a 2025 safety miss.

Motivations of management? Interpretation: aligned to ROIC and per-share earnings (not size); long-tenured CEO with ~$500M personal stake — but zero open-market buys and a discretionary ~$120M sale near the high.

Valuation & Market Data

ADR, MLP, or K-1 issuer? No — a standard US C-corp common stock (NYSE: PWR), Delaware-incorporated, files 10-K/10-Q.

Dividend policy? Fact: token ~$0.41/share (~0.06% yield), growing off a tiny base; not a yield story.

How profitable is the business? Interpretation: modestly — ~10.2% adjusted EBITDA margin, ~5.7% GAAP operating margin, ~3.6% net margin, ~8% GAAP / ~11% cash ROIC.

Is net income diverging from cash from operations? Fact: OCF ($2.23B) is ~2.1x net income — flattered by D&A + $499M amortization + a $824M contract-liability inflow; underlying receivable quality is softening (DSO 80→88d). Not a red flag in aggregate, but worth monitoring.

Risks & Downside

What factors would cause the stock to decline? Multiple de-rating (the dominant risk at 98–99th-percentile valuation), data-center/utility capex deceleration, the OBBBA renewables cliff, a fixed-price project blow-up or claim write-down, an M&A misstep/impairment, or a growth/margin miss.

Risk of a catastrophic loss? Interpretation: low-probability but real — utility-contractor wildfire liability (Silverado tender) is the sector’s catastrophic-loss vector (cf. PG&E). Mitigated by insurance and acquisition indemnities.

Chance of a total loss? Very low — investment-grade balance sheet (~1.9x net leverage), diversified end-markets, essential services. The realistic downside is multiple compression (a large drawdown), not insolvency.

Recent News & Events

Has the business environment changed recently? Fact: yes, favorably on demand (record $48.5B backlog, raised FY2026 guidance, accelerating data-center load) and unfavorably on policy (OBBBA wind/solar credit phase-out, July 2025).

Significant acquisitions? Fact: Dynamic Systems (July 2025) + ~$3.05B of 2025 deals; Cupertino Electric (2024).

Change in accounting policies? Fact: segment reorganization (3→2) effective Q1 2025, prior periods recast — reduced segment transparency, not an accounting-policy change per se.

Recent changes — new markets, facilities, management? Fact: expansion into data-center MEP (new market); vertical supply-chain factories (new facilities, 765kV capacity doubling); stable management with orderly bench development (Karl Studer elevated to President of Electric Operations).

APPENDIX B — Source Appendix — Quanta Services, Inc. (NYSE: PWR)

Report date 2026-06-12. Primary sources first. URLs accessed 2026-06-12 unless noted.

Primary — SEC Filings (EDGAR, CIK 0001050915)

1. Form 10-K, FY2025 (filed 2026-02-19; pwr-20251231.htm) — segments (Note 5), backlog/RPO, contract-type mix, customer mix, revenue recognition, intangibles, debt (Note 10), legal proceedings (Silverado), risk factors. Primary source for most business/financial facts.
2. Form 10-Q, Q1 2026 (filed 2026-04-30; pwr-20260331.htm) — Q1 results, updated debt, contract balances.
3. Forms 10-K, FY2021–FY2024 — multi-year revenue, margin, cash-flow, goodwill, and backlog trend; segment recast.
4. DEF 14A Proxy (filed 2026-04-10; d118470ddef14a.htm) — executive compensation, PSU/LTI performance metrics (ROIC + cumulative adjusted EPS + relative-TSR modifier), say-on-pay (~93%), insider ownership, board composition.
5. Form 3/4/5 corpus (269 Form 4 + 10 Form 4/A + 9 Form 3, trailing 5 years) — insider-transaction read: zero open-market purchases (code P); CEO Earl “Duke” Austin Jr. discretionary (non-10b5-1) sale of ~155,992 shares on 2026-05-07 at ~$766–778.
6. 8-K material-event filings (51 in corpus) — earnings releases (non-GAAP reconciliations), acquisition announcements, debt issuances (Aug-2025 $1.5B senior notes), guidance.

Primary — Earnings Calls & Investor Events

7. Q4 2025 Earnings Call (2026-02-19) — FY2025 adjusted EPS $10.75, adjusted EBITDA $2.90B; renewables/OBBBA/FEOC commentary; initial FY2026 framework.
8. Q1 2026 Earnings Call (2026-04-30) — record $48.5B backlog; raised FY2026 guidance (revenue $34.7–35.2B, adjusted EBITDA $3.49–3.65B, adjusted EPS $13.55–14.25); Q1 adjusted EPS $2.68.
9. Analyst/Investor Day (2026-03-31) — 2030 algorithm (7–10% organic revenue CAGR, 10–11% adjusted EBITDA margin, 15–20% adjusted EPS CAGR, 12–15% ROIC, $10–12B cumulative FCF); ~$2.4T TAM; capital-allocation framework; M&A-vs-FCF history; six-end-market growth map.
10. Bernstein Strategic Decisions Conference (2026-05-28) and prior conference presentations — supporting forward commentary.

Primary — Quantitative Data

11. SEC EDGAR XBRL — revenue, operating income, net income, diluted EPS/shares, OCF, capex, acquisition spend, goodwill, equity, intangible amortization (FY2020–FY2025). Authoritative for US-GAAP series.
12. Market data (accessed 2026-06-12) — price ~$707.74, market cap ~$106B, EV ~$108.6B, debt/cash, 52-week range, and peer multiples (MTZ, EME, FIX, ACM, PRIM). Reconciled to filings.
13. Valuation-percentile data (own ~10yr history, as of 2026-06-11) — composite ~98.8th percentile (P/E, P/B, P/S all 98–99th).

Secondary — Industry & Policy

14. Edison Electric Institute (EEI) via Utility Dive and POWER Magazine — US investor-owned utility capex record ~$204B (2024); >$1.1T cumulative 2025–2029; transmission ~$178B 2025–2028. Utility Dive; POWER Magazine.
15. S&P Global / 451 Research — US data-center grid demand ~75.8 GW (2026) → ~130 GW (2030). S&P Global.
16. BloombergNEF — US data-center power demand ~106 GW by 2035. Utility Dive/BNEF.
17. OBBBA (One Big Beautiful Bill Act, signed 2025-07-04) — phased elimination of §45Y PTC / §48E ITC for wind and solar (in-service after 2027 unless construction begins before 2026-07-04); storage/geothermal/hydro/nuclear retained through 2034. Simpson Thacher; Steptoe; Tax Foundation.

Secondary — Market / Consensus

18. Analyst consensus & short interest — consensus “Buy”; targets clustering ~$590–760; short interest ~3.4% of float. Benzinga; stockanalysis.com; WallStreetZen.
19. Peer data — MasTec (MTZ), EMCOR (EME), Comfort Systems (FIX), AECOM (ACM), Primoris (PRIM), MYR Group (MYRG) revenue/margin context via company filings and StockStory/Yahoo Finance.

Peer Cross-Read

20. Comfort Systems (FIX) — peak-multiple / data-center-cycle comparable.
21. Eaton (ETN), AMETEK (AME) — grid-electrification equipment peers; the equipment expression of the same secular thesis.

Analytical Frameworks

22. Greenwald & Kahn, Competition Demystified — moat-type taxonomy (economies of scale + customer captivity), share-stability and ROIC tests.
23. Marathon / Chancellor, Capital Returns — supply-side capital-cycle analysis (capital flooding in; labor as the binding constraint).