Argan, Inc. (NYSE: AGX) — A Turbine-Gated Windfall Priced as a Permanent Franchise

Independent Equity Research Report date: 2026-06-26 · Coverage: Initiation Price (2026-06-25 close): $753.07 · Shares (dil.): ~14.15M · Market cap: ~$10.6B · EV: ~$9.6B FY end: January 31 · Sector: Industrials · Engineering & Construction (Power Infrastructure)

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analysis that follows it is written to a position-free standard and carries no recommendation and no price target.

Verdict: HOLD / AVOID at $753 — a wonderful, fortress-funded business at a price that already pays you for a decade of flawless execution. Accumulate only on a deep pullback. Not a short. Conviction: medium. Directional valuation zone: fair value ~$430–560 (roughly 25–32x clean operating EPS / ~25–30x normalized EBITDA — a genuine premium for the quality, the net cash, and the real backlog); an attractive entry zone is ~$300–380 (~18–22x clean EPS), where the price stops underwriting permanence. At $753 the stock trades at ~59x EV/EBITDA, ~9x sales, and its richest multiple in company history (97.7th percentile of its own decade).

Argan is the best-positioned pure-play on the natural-gas EPC build-out that the data-center/AI load boom has unleashed, and it carries a balance sheet — ~$970M of cash and investments, essentially no debt — that almost no competitor can match. The bull case is real: heavy-duty gas turbines are sold out into ~2029–30, which rations gigawatt-scale fixed-price EPC work to a handful of bondable primes, and Argan’s backlog has doubled to ~$2.8B. But three things are simultaneously at a peak — gross margin (20.5%, flattered by the Trumbull early-completion close-out, an international rebound, and favorable mix), interest income (~16% of pretax profit, fading with rates), and reported cash flow (inflated by a ~$566M customer-prepayment float that reverses when backlog converts) — and they are stacked under the richest multiple the stock has ever carried. This is the textbook quality-at-a-fantasy-price setup: the moat is a turbine bottleneck (a cyclical scarcity, not a structural franchise), the business is fixed-price EPC (one bad gigawatt job erased a full year of profit in FY2020), and every insider sold into the all-time high — ~$80M+ in June 2026, the largest holder cutting ~52% of his stake, with zero open-market buys. The framing is a crowded momentum one-way street (beta 1.24, Momentum loading ~1.0, factor-neighbors are momentum ETFs), ~25x off its 2022 low — not a falling knife. I don’t short net-cash compounders mid-supercycle, but I won’t pay 59x EBITDA for a narrow-moat contractor at a cyclical peak either. The single fact that flips me bullish: a sustained $2B+ revenue run-rate with gross margin held near 20% and a clean, ex-float, ex-interest FCF that scales — proof the peak is a plateau. The single fact that flips me bearish (to “avoid entirely”): a fixed-price project write-down or a book-to-bill that falls below 1 for multiple quarters — either one cracks the margin, the float, the growth, and the multiple at once.

Tag: “Right business, right cycle, wrong price — the turbine shortage is the moat, and turbine shortages end.”

📈 Stock Price Action — Five-Year Event Map

Factual price history, not a recommendation. Price moves are Fact; attributed drivers are Interpretation. No price target, no support/resistance, no chart-pattern reading.

The arc. Argan is a parabolic five-year round-trip in a single direction. From a ~$30 low (Sep-26-2022) the stock has run to an all-time high of $790 (Jun-22-2026) — a ~25x move — and closed at $753.07 on Jun-25-2026, −4.7% off the high. The 52-week range alone is roughly $201 (Jul-2025) → $790. At $753 the stock trades ~59% above its 200-day EMA ($474), above the 50-day ($651) and the 21-day ($690). The move is an earnings ramp and a multiple re-rating stacked on top of each other, both driven by the data-center/AI gas-power narrative.

Cycle narrative. Through 2021–22 (1) Argan was a sleepy, range-bound contractor trading at single-digit earnings with no catalyst. The 2024 prints (2)–(3) — above all the +26% reaction to Q2-FY25 — are where the market re-cast Argan as the pure-play gas-EPC beneficiary of AI/data-center load growth, the moment turbine bottlenecks at GE Vernova, Siemens Energy and Mitsubishi made qualified gas-EPC primes scarce. Post-election power-demand optimism (4) doubled the stock into year-end 2024, but a −18.5% single-day air-pocket (5) showed the move was never one-directional beneath the surface. Each subsequent print (6)–(7) re-rated the multiple, not merely earnings. The defining event (8) was the +37.9% FY26 Q4 print (Mar-27-2026) — record results, backlog doubling to ~$2.9B, a 20.5% peak gross margin. The final leg to the $790 ATH (9) coincided with the Q1-FY27 beat and, tellingly, unanimous discretionary insider selling of ~$80M+ into the high.

1. Executive Summary

Argan, Inc. is a holding company whose crown jewel, Gemma Power Systems, is one of a small number of U.S. engineering-procurement-construction (EPC) primes capable of building gigawatt-scale natural-gas combined-cycle power plants on a fixed-price basis. Two smaller segments — The Roberts Company (TRC) industrial fabrication and SMC telecom/teledata — round out the group, but Power Services is ~80% of revenue and ~88% of gross profit and is the entire investment story. After two decades as a forgotten micro-cap, Argan has become the most direct listed pure-play on the gas-fired power build-out triggered by data-center/AI electricity demand, and the stock has risen ~25x off its 2022 low to a ~$10.6B market capitalization.

The operational results are genuinely excellent. Revenue grew from $573M (FY24) to $945M (FY26) and ~$1,042M on a trailing-twelve-month (TTM) basis; gross margin expanded from 14.1% to 20.5%; diluted EPS went from $2.39 to $9.74; ROIC reached ~28% and ROE ~39%; and the balance sheet is a fortress — ~$974M of cash and investments, essentially zero debt, funded in part by a ~$566M customer-prepayment float that exceeds total equity. Backlog has doubled to ~$2.8B (~79% natural gas), giving multi-year visibility. This is a well-run, disciplined, asset-light operator at the center of a real demand supercycle.

The problem is price and the quality of the peak. At $753 the stock trades at ~59x TTM EBITDA, ~9.2x sales, ~77x reported P/E (~92x on clean operating earnings), and ~23x book — its richest multiple ever (97.7th percentile of its own history) and roughly twice the richest pure-play E&C peer. Meanwhile the reported earnings flatter the durable run-rate on three fronts simultaneously: (1) the 20.5% gross margin is a cycle peak, lifted by the Trumbull early-completion close-out, an international rebound from prior U.K. losses, and favorable project mix; (2) ~16% of pretax profit is interest income on the cash pile that fades when rates fall; and (3) reported operating cash flow of ~$415M against $138M of net income is inflated by a customer-prepayment float that reverses as projects burn down — “clean” FCF is closer to $150–170M. The moat is narrow and substantially cyclical: it rests on a turbine-supply bottleneck and bonding capacity, not a structural franchise, and fixed-price EPC is a business where a single bad gigawatt project can — and in FY2020 did — erase a full year of profit. Capital allocation is disciplined but the ~$970M balance sheet sits lazy (no special dividend since 2020), executive comp contains no return-on-capital metric, and every insider sold into the all-time high with zero open-market buys.

The result is a high-quality, well-positioned business priced for permanence in a project-lumpy, narrow-moat, fixed-price industry with no margin of safety. The bull needs everything to break right — sustained $2B+ revenue at peak margins and a persistently rich multiple — merely to earn a low-double-digit return from here; the bear needs nothing exotic for a 50–70% drawdown.

2. Business Overview

Argan is a project-based holding company, not an operating company. Incorporated in 1961 and headquartered in Rockville, Maryland, it owns three construction/engineering subsidiaries that share a balance sheet and a head office but run independent project pipelines. The group reports three segments, and the concentration in the first is extreme.

Power Services (Gemma Power Systems). Gemma is the franchise. It is a turnkey EPC contractor — it engineers, procures, constructs and commissions large-scale power plants — predominantly natural-gas combined-cycle (CCGT) facilities, plus solar, solar-plus-storage, biomass and biofuel projects. Across its history Gemma has built ~18 GW of generating capacity. Atlantic Projects Company (APC), an Ireland/U.K. thermal-EPC and turbine-services business consolidated from around 2021, adds international peaker (simple-cycle / FlexGen) and thermal work and is the source of both the FY2020 disaster and the FY2025–26 international rebound. Power’s customers are independent power producers (IPPs), regulated utilities, and developers working alongside the turbine OEMs (GE Vernova, Siemens Energy, Mitsubishi Power). This segment drove essentially all of the consolidated margin expansion.

Industrial Services (The Roberts Company / TRC). Southeast-U.S. industrial construction, field services, and metal-component fabrication — piping systems, pressure vessels and storage tanks for industrial and energy customers. Revenue has been roughly flat at ~$167M for two years, but the forward catalyst is a ~$125M data-center thermal-expansion / energy-storage-tank fabrication contract (awarded Nov-2025), which is prompting a new ~$10–13M North Carolina fabrication facility — a genuine, if small, new adjacency that ties TRC to the same data-center theme as Power.

Telecom / Teledata (SMC). The smallest unit (~$20.6M): mid-Atlantic trenchless directional boring, fiber/cable installation, and structured cabling for commercial, data-center and federal/military-clearance sites.

How the business makes money — and the float. This is fixed-price, lump-sum EPC. Argan agrees a contract price, then earns a construction margin by delivering the project for less than that price. The prize and the peril are identical: under a fixed-price structure, cost overruns hit the contractor, not the customer. Revenue is recognized over time on a percentage-of-completion basis, which makes reported margin estimate-laden and sensitive to change orders, claims and cost re-estimates. Revenue is project-lumpy, not recurring — every plant is separately bid, and there is no contracted recurring revenue base.

The standout structural feature is the negative-working-capital, customer-funded model. Customers progress-bill and prepay ahead of costs incurred, generating large contract liabilities (billings-in-excess) — $565.8M at Apr-30-2026, larger than the company’s entire $474M of stockholders’ equity. This float funds operations and produces a ~$974M cash-and-investments balance with essentially no debt. Critically, bonding/surety capacity is the gate to winning large EPC work: clients require performance bonds, sureties underwrite project-by-project against balance-sheet strength, and management explicitly frames the fortress balance sheet as what expands bonding capacity and makes Argan a “bankable” prime. Net liquidity of ~$421M (cash net of float) is described by management as capable of supporting “several billion more” in backlog.

Verdict: A high-quality, asset-light, customer-funded EPC franchise concentrated almost entirely in one segment (Power) and increasingly in one end-market (gas-fired generation for load growth). The model’s elegance — negative working capital, fortress balance sheet, bonding-as-barrier — is real, but so is its lumpiness and its single-project tail risk.

3. Industry Dynamics

The central fact is a genuine, multi-year demand inflection. U.S. electricity demand was essentially flat for two decades; it is now inflecting upward as data-center/AI compute, manufacturing reshoring, and EV/electrification collide with an aging thermal fleet reaching end-of-life. The grid’s only dispatchable, financeable, near-term answer at gigawatt scale is natural-gas combined-cycle generation. Argan’s backlog reflects exactly this: ~79% natural gas / 13% renewable / 8% industrial, including four U.S. gas plants totaling >4.1 GW. The merit-order logic (cheapest plants dispatch first; gas typically sets the marginal clearing price; gas increasingly serves as the firming capacity behind intermittent renewables) underpins durable CCGT demand for years.

The supply bottleneck is the defining structural feature — and the source of Argan’s current pricing power. Heavy-duty gas turbines from the three OEMs (GE Vernova, Siemens Energy, Mitsubishi Power) are sold out into roughly 2029–2030. This gates the entire downstream: only developers who have secured turbine slots can build, and only a small set of qualified, bondable firms can EPC a gigawatt-scale combined-cycle plant on a fixed-price basis. The turbine shortage is, paradoxically, a moat-extender for the qualified EPC primes — it rations the available work to a handful of players rather than inviting a flood of new entrants, and it lets those primes hold price and select risk. Permitting, interconnection-queue congestion, air/water permits, and gas-supply access remain genuine project-timing frictions; management itself cites “air permits, access to gas, water permits, turbines, financing” as the gating milestones before an award converts.

Where Argan sits in the value chain. Downstream of the turbine OEM, as the EPC prime integrating the turbine, balance-of-plant equipment and civil works. Argan is a price-taker on the turbine (a procurement pass-through) and a margin-maker on execution. It captures none of the OEM’s economics and all of the construction execution risk.

Capital-cycle read (Marathon lens): the early-favorable phase. Returns are high and rising (ROIC 8.4% → 21% → 28.3%, FY24→26), but supply is not yet flooding in — it is constrained by three real frictions: turbine availability through ~2030, bonding/balance-sheet requirements that exclude smaller players, and scarce skilled construction labor and project-management talent. This is precisely the setup the capital cycle rewards: high returns that competitors cannot yet compete away. But it is the favorable phase, not a permanent state. Fixed-price EPC is historically a low-return, mean-reverting business; when turbine supply catches up post-2030 and competitors expand crews, returns normalize. The history of E&C is littered with cycles where elevated margins drew in capacity and then collapsed (Fluor’s serial fixed-price losses are the cautionary archetype).

Competitive intensity & the comparison set. The peer universe splits into three tiers. The true large-scale gas-EPC rivals — Bechtel, Kiewit, Black & Veatch, Sargent & Lundy — are all private, bigger, and deeper; they are the real competition for marquee jobs. The listed adjacents — Quanta (PWR), MasTec (MTZ), Primoris (PRIM) — are far larger but weighted to transmission/distribution, pipelines and renewables rather than gas-plant EPC primes; EMCOR (EME) and Comfort Systems (FIX) are mechanical/electrical specialty and facilities players with more recurring, less lumpy models. The cautionary mirror is Fluor (FLR), a diversified EPC with a troubled fixed-price-loss history. Argan competes in a genuinely small niche (gigawatt fixed-price CCGT primes) but does not have the scale of the privates or the diversification of the listed adjacents.

Verdict: Structurally attractive right now, cyclically — not a durably good industry. The next several years are a favorable, turbine-gated window with high returns rationed to a few primes. But EPC is a thin-margin, mean-reverting business; this is an early-favorable phase of the capital cycle, not a permanent structural advantage. Treat the attractiveness as cyclical and time-limited, not secular.

4. Competitive Position

Be skeptical: this is a thin-moat, execution-risk business currently wearing a cyclical halo. Fixed-price lump-sum EPC is the textbook bad-economics structure — a single mismanaged project can erase years of profit. Argan’s own FY2020 is Exhibit A: a $33.6M subcontract loss on the Tees Renewable Energy Plant (TeesREP), a 299 MW biomass project in northeast England where APC handled boiler mechanical installation, drove an FY2020 domestic operating loss and a consolidated operating margin of −21.3% (net loss −$40.7M, gross margin −2.9%). That is the downside of fixed-price contracting made concrete, and it is recent enough to anchor the bear case.

Greenwald taxonomy — the moat is narrow and is not one of the three genuine advantage types.

• No economies of scale + captivity. Argan ($945M revenue) is small versus the privately-held giants; there is no scale-cost lock-in. If anything, the scale advantage runs the other way — Bechtel and Kiewit are larger.
• No customer captivity / switching costs. Revenue is project-by-project with zero recurring component; each plant is re-bid. Receivables are concentrated (top-three customers = 29%/19%/11% of accounts receivable at Jan-31-2026), but no single customer exceeded 10% of revenue in FY24–26 — this is relationship-based repeat business, not contractual lock-in.
• No network effects.

What Argan does have is a reputational/intangible edge plus a balance-sheet-as-bonding moat:

1. Track record / execution reputation. Gemma’s roughly two-decade fixed-price record (“hasn’t had a lost job since we acquired them,” per management) and repeated early completions (Trumbull, the Midwest projects) matter in a business where one loss is catastrophic. But this is contestable — and the FY2020 TeesREP loss shows precisely how it can break.
2. Bonding capacity. The debt-free, ~$974M balance sheet underwrites surety capacity that smaller primes cannot match — a genuine barrier to qualifying for gigawatt-scale work.
3. Scarcity of qualified gas-EPC primes. The competitor set for “>1 GW fixed-price combined-cycle” is genuinely small, and the turbine bottleneck keeps it small.

Tie the moat to a financial outcome. The only moat claim that survives scrutiny is bonding capacity plus reputation gating who can bid. If that eroded — a major project loss damages the track record, or competitors build the balance sheet and crews to qualify — Argan’s pricing power and 22%+ Power gross margin would compress toward the high-teens (or worse) normalized rate. Management itself concedes blended margins run “high-teens to low-20s with meaningful variation” and that current early-stage jobs “have a lot of outstanding risks to account for.” A moat that depends on a turbine shortage is, by construction, a moat with an expiry date.

Direct competitor read. Against Bechtel/Kiewit/Black & Veatch (private), Argan is the smaller, more selective, more conservatively-financed player — it wins by discipline and bondability, not scale. Against PWR/MTZ/PRIM, it is a purer gas-EPC bet but lacks their diversification and recurring service revenue. Against EME/FIX, it has higher project-tail risk and lower revenue durability. Against FLR, it has a far cleaner balance sheet and risk record — Fluor is the value-destruction Argan is trying not to become.

Verdict: Narrow, reputation/bonding-based, partly cyclical advantage — not a wide or durable moat. Argan is a well-run, disciplined operator in a temporarily-scarce niche, not a structurally protected franchise. The advantage is real enough to earn above-cost-of-capital returns during the turbine-gated window, but it is contestable and one bad fixed-price project away from the FY2020 outcome.

5. Growth History and Forward Opportunities

Backlog is the single forward KPI for an EPC, and Argan’s has roughly doubled.

• Consolidated backlog $2.9B at Jan-31-2026 (Power $2.7B, Industrial $253.0M, Teledata $8.4M) — up from $1.4B a year earlier ($1.3B / $53.2M / $3.6M).
• $2.8B at Apr-30-2026 (Power $2.5B, Industrial $225M, Teledata $8M) — a slight sequential dip as Trumbull and the Midwest projects completed and rolled out of backlog.
• Mix ~79% natural gas / 13% renewable / 8% industrial; four U.S. gas plants = >4.1 GW.

Revenue trajectory (FY ends Jan-31): FY2020 $239M (loss year) → FY2021 $392M → FY2022 $509M → FY2023 $455M → FY2024 $573M → FY2025 $874M → FY2026 $945M → TTM ~$1,042M (through Q1-FY27). Q1-FY27 revenue rose ~50% YoY to $291M. Note the lumpiness: FY2023 declined from FY2022 — a reminder that this is project-cycle revenue, not a smooth secular ramp.

Organic vs. acquired. Growth is substantially organic U.S. gas EPC. APC (acquired ~2021) added Ireland/U.K. thermal work (ESB FlexGen peakers, Shannonbridge, Kilroot 2×330 MW, a 300 MW biofuel project for SSE Thermal, a 170 MW thermal) — meaningful but a modest fraction of revenue, and the source of the FY2020 loss. There has been no material M&A since; total goodwill is just $28M.

Named major projects (the backlog made concrete): Trumbull Energy Center (950 MW gas, Ohio — full notice-to-proceed Nov-2022; reached substantial completion in Q1-FY27, ahead of schedule — the close-out that flattered FY26 margins); Sandow Lakes / SLEC (1.2 GW combined-cycle gas, Lee County TX — NTP Apr-2025); CPV (~1.4 GW gas, TX); an 860 MW gas (TX); a 700 MW gas (U.S.); a Ward County TX combined-cycle (construction began Q4-FY26, ~CY2029 completion); the completed Midwest solar + battery trio; a Louisiana LNG subcontract (FNTP Jun-2024); and the Industrial segment’s $125M data-center tank-fabrication contract.

Forward cadence. Management guides to “a handful of new projects over the next 10–18 months,” a capacity of 10–12 simultaneous jobs (currently running ~8 power jobs: 6 thermal / 2 renewable), and frames $2B of annual revenue as achievable “down the road.” The data-center fabrication expansion is a genuine new adjacency that extends the theme into the Industrial segment.

The skeptic’s flag. Growth quality is high but cyclical-peak, not durable-compounder. It rests on a turbine-gated gas window; it is lumpy (backlog already dipped sequentially; management openly warns of “a gap between the completion of one job and the announcement of new jobs”); margins are at a cycle high (22%+ Power gross margin that management normalizes to “high-teens to low-20s”); and new awards depend on permitting/turbine/financing milestones outside Argan’s control.

Verdict: High-quality growth — but cyclical-peak high-quality, not durable. Real backlog, real projects, real margin expansion, fortress funding — yet a project-driven, turbine-supply-gated surge that will normalize when the cycle does. The backlog and 22% margins should be treated as a peak window, not a run-rate.

6. Financial Quality

The headline is real; the run-rate is not. Argan’s reported trajectory is genuinely spectacular — revenue $573M (FY24) → $874M (FY25) → $945M (FY26), TTM ~$1,042M; diluted EPS $2.39 → $6.15 → $9.74 (Q1-FY27 alone $3.24); gross margin 14.1% → 16.1% → 20.5%; operating margin 6.4% → 10.1% → 14.3%; ROIC ~28% and ROE ~39% in FY26. But three quality-of-earnings flags mean the FY26/TTM figures substantially overstate durable earnings power, and an investor paying a record multiple must normalize for all three.

(a) Margin expansion is mix + a favorable close-out + an international rebound — not a structural step-up. Management’s FY26 10-K MD&A attributes the gross-margin lift to “the changing mix of projects and contract types,” an international operation that “experienced a significant improvement in profitability,” and — critically — “disciplined execution on the Trumbull Energy Center [that] reduced project costs and enabled us to achieve substantial completion ahead of schedule.” That last item is the signature of a one-time favorable project close-out: under percentage-of-completion accounting, finishing a fixed-price job under budget releases accrued contingency to gross profit in the final periods. EPC gross margins are structurally high-single to low-double-digit; 20.5% is a peak, not a baseline. It coincided with the international segment swinging from multi-year U.K. losses (which had driven deferred-tax valuation allowances of +$2.1M in FY24 and +$1.4M in FY25) back to profit — a recovery, not a new normal. Q1-FY27 gross margin was 21.0%, so the elevated level persists for now, but it rests on the same favorable Power mix and is hostage to any single fixed-price project going the other way.

(b) ~16% of profit is interest on the cash hoard, not EPC profit. The “Other income, net” line was $25.8M in FY26 and $23.0M in FY25, which management states “primarily reflected income earned … on investments, cash and cash equivalents.” Against FY26 pretax income of $160.6M, that is ~16% of pretax income from T-bill interest, not operations (Q1-FY27: $8.4M of $53.8M pretax, ~15.6%; run-rate ~$33M/yr on the ~$974M pile). Taxed at the ~14% FY26 effective rate, the ~$26M leaves ~$22M after tax, or ~$1.55 of diluted EPS of pure interest income inside the $9.74. Clean operating EPS is therefore ~$8.0–8.2. A 200 bp decline in short rates removes roughly $0.90–1.00 of EPS with zero change in the underlying business — a direct headwind the market is unlikely to be discounting at a record multiple.

© The central catch — float-inflated operating cash flow that reverses. This is the most important QoE finding. Customer contract liabilities (billings-in-excess) rose $299.2M (FY25) → $514.0M (FY26) → $565.8M (Q1-FY27). Management explicitly labels the FY26 jump a “temporary increase in contract liabilities of $214.7M [that] represented a source of cash, primarily due to the net effect of the early phase of construction activities on certain projects” — i.e., customers prepaid ahead of work performed on newly-started jobs. That is why FY26 operating cash flow was $414.7M against net income of $137.8M, an OCF/NI ratio of 3.0x, with the working-capital swing contributing +$257.4M (of which “other operating assets/liabilities,” mostly the contract-liability inflow, was +$201.5M). Strip the float and clean OCF is roughly net income + D&A ($6.2M) + SBC ($7.9M) ± normal working capital ≈ $150–170M — not $400M+. The ~$974M cash balance is itself float-inflated: a large slice is customer money that flows back out as those projects burn down. The series proves the volatility — OCF/NI has been 7.4x (FY21), −0.9x (FY23, when OCF was actually negative −$30M), 3.6x (FY24), 2.0x (FY25), 3.0x (FY26). Argan is not a steady cash machine; it is a project ledger whose cash swings with the backlog phase.

Revenue recognition / concentration. Revenue is recognized over-time (percentage-of-completion) on fixed-price EPC contracts — inherently estimate-laden and exposed to change-order and cost-overrun risk. Customer concentration is extreme: in FY26 the top three Power customers were 23% / 16% / 11% = ~50% of consolidated revenue (FY25: 28/13/10; FY24: 19/16/15). Power is 80.1% of revenue. The loss or delay of one large gas-EPC award is a material event.

Returns are flattered by the model, not just the moat. ROIC ~28% and ROE ~39% (FY26) are inflated by (i) negative working capital / customer float funding operations, (ii) ~$974M of cash earning interest that lands in the return numerator, and (iii) a genuinely asset-light structure (goodwill $28M, intangibles $1.4M, PP&E ~$18M). Returns on the actual operating capital employed are high, but the denominator is tiny-to-negative, so the headline overstates durability — it rides peak margins plus float. SBC is modest ($7.9M FY26) and dilution is contained (diluted shares 13.55M → 13.91M → 14.15M, buybacks offsetting option/RSU creep).

Verdict: Earnings are real but low-quality at this point in the cycle. Peak gross margin (favorable mix + Trumbull close-out + international rebound), ~16% of profit from rate-sensitive interest, and float-inflated cash flow that will reverse. Use clean operating EPS ≈ $8 (plus fading interest) and FCF ex-float ≈ $150–170M — not the $9.74 EPS or the $400M+ headline OCF — for valuation. Economics do improve with scale on the operating side, but the reported numbers flatter the run-rate on three fronts at once.

7. Capital Allocation

The lazy ~$974M balance sheet is the defining capital-allocation question. At Apr-30-2026, cash & equivalents ($355.8M) plus investments ($617.7M) totaled $973.6M against equity of $473.5M, on a company earning ~$138M. Part of this is legitimate surety/bonding capacity — Argan must demonstrate balance-sheet strength to win and bond multi-hundred-million-dollar fixed-price contracts — but the pile is far above any reasonable bonding need and earns only T-bill yield. It is a genuine capital-allocation drag, and (as discussed above) a meaningful chunk is customer float, not truly free capital.

Special-dividend history. Argan has a track record of returning excess cash through specials — $1.00/share special in July 2020 and $1.00/share special in December 2020 (so FY21’s $3.00/share total = $1.00 regular + $2.00 of specials), and larger specials in prior cycles. The regular dividend has grown — $0.25/quarter for years, raised to $0.375 (FY25), then +33% to $0.50/quarter in September 2025 (~$2.00/yr, a ~0.27% yield at $753). FY26 dividends paid totaled $24.3M. Tellingly, no special has been declared since 2020 despite the cash pile tripling — management is hoarding rather than returning, the crux of the capital-allocation critique.

Buybacks — small and now chasing. The program began Nov-2021; the authorization was raised to $150M (Apr-10-2025) and then to $200M and extended to Jan-2030 (Apr-8-2026). Execution is tiny relative to the cash: FY26 repurchased just 64,699 shares for $9.9M at an average $152.66 (including a Dec-2025 tranche at $322.86). Treasury-stock cost rose from $114.4M (Jan-31) to $135.0M (Apr-30-2026), implying ~$20M more bought in Q1-FY27 — but at much higher prices. Historically Argan bought cheap; it is now buying dearer, in dollar amounts immaterial against a $10.5B market cap and $974M of cash.

M&A — disciplined bolt-ons only. Cash used for acquisitions was just $11.5M (FY26), $16.3M (FY25), $5.1M (FY24). APC (~2021) and the legacy TRC/Gemma businesses are the franchise; total goodwill is only $28M. There is no large, goodwill-funded, value-destroying M&A — a genuine and meaningful positive, and the clearest contrast with Fluor’s history.

Incentives & ownership. The DEF 14A (filed 2026-04-21) pays executives on Revenue, ROE, EPS and TSR — there is no return-on-invested-capital metric, and the chosen metrics are gameable via the lean-equity/float model and buybacks (ROE is mechanically inflated by negative working capital; EPS by repurchases). Insider ownership is just ~3% as a group (CEO Watson <0.5%; non-executive chairman; no founder/anchor block); the largest holders are BlackRock (7.36%) and Vanguard (6.64%) — index funds, not aligned owners.

Verdict: Above-average discipline, not elite. Positives: no destructive M&A, a growing and now-33%-higher dividend, opportunistic (if tiny) buybacks, an asset-light structure. Negatives: a lazy ~$974M balance sheet earning T-bills with no special since 2020, no return-on-capital metric in comp, a buyback too small to matter, and near-zero insider ownership. The cash hoard is the single biggest unresolved allocation question, and management has chosen to sit on it.

8. Changes and Headwinds — Last Two Years

Strategic and operational changes.

• Backlog roughly doubled from ~$1.4B (Jan-2025) to ~$2.9B (Jan-2026), almost entirely U.S. natural-gas EPC — the operational expression of the data-center/gas supercycle.
• Trumbull Energy Center reached substantial completion ahead of schedule (Q1-FY27), the single largest contributor to the FY26 margin peak via its favorable close-out.
• Major new awards: Sandow Lakes/SLEC (1.2 GW, NTP Apr-2025), Ward County TX combined-cycle (construction began Q4-FY26), the Louisiana LNG subcontract (FNTP Jun-2024), and the Industrial segment’s $125M data-center tank-fabrication contract (Nov-2025) with an associated new North Carolina facility.
• International rebound: APC swung from the prior U.K. losses back to profitability, contributing to the gross-margin step-up.

Capital-return and governance changes.

• Dividend raised +33% to $0.50/quarter (Sep-2025); buyback authorization raised to $200M and extended to 2030 (Apr-2026).
• Unanimous, discretionary insider selling into the all-time high in June 2026 (~$80M+, detailed below) — the most important recent governance signal.

Headwinds and watch-items.

• Margin normalization risk as the Trumbull close-out anniversaries and newer, earlier-stage jobs (carrying “outstanding risks”) dominate the mix.
• Backlog already dipped sequentially (Jan→Apr 2026), and management warns of award-timing gaps.
• Interest-income fade as rates decline.
• Float reversal as projects burn down.
• Litigation: the U.K. subsidiary sued EP NI Energy (Mar-2025) over a project on which it performed construction without being paid — a small but live reminder of fixed-price/counterparty risk.

Verdict: The last two years strengthened the operational thesis (backlog, margins, awards) but the changes that matter most for a buyer at this price — peak-margin close-outs, float-inflated cash, and unanimous insider distribution — collectively weaken the case for paying a record multiple. The business got better; the risk/reward got worse.

9. Risk Analysis

Risk synthesis. Solvency risk is negligible — net cash, no debt, asset-light. The real risk is to the valuation: a cluster of correlated peaks (margin, interest income, float, multiple) means a single crack — a project loss, an award gap, or a hyperscaler capex pause — could compress margins, reverse the float, decelerate growth and de-rate the multiple simultaneously. At ~59x EBITDA there is no cushion to absorb it.

10. Valuation Discussion

No price target; no recommendation. Embedded-expectations and scenario framing only.

Live multiple set, re-derived at $753 (ROIC’s EV snapshot is stale at the ~$668 Apr-30 quarter-end close):

The AZI composite valuation index sits at the 97.7th percentile of Argan’s own history; book and sales multiples are at all-time records.

Peer comparison — Argan is the most expensive name in E&C, by a wide margin. Live ROIC TTM enterprise values, cross-checked against comparable public E&C names:

Argan trades at ~2x the richest pure E&C peer (FIX) on both EBITDA and sales, and ~3–5x the cheaper peers (PRIM/EME/MTZ). The whole E&C set is already at 16–33x EV/EBITDA — roughly 2–3x normal mid-cycle multiples, a sector-wide re-rating — and Argan sits a full tier above that. The market’s justification: a gas-EPC pure-play (no fiber/pipeline/MEP dilution to the data-center-power theme), a net-cash fortress (only EME/FIX rival it; PWR/MTZ/PRIM carry leverage), peak optics (20.5% GM, 28% ROIC), and micro-float scarcity (~$10.5B cap on only ~14M shares, which amplifies momentum). But the object being priced is a narrow-moat, fixed-price EPC cyclical priced richer than franchise-grade peers.

Embedded expectations (reverse-DCF). To justify the ~$9.6B EV at a normalized through-cycle exit of ~12–14x EBITDA, the market needs steady-state EBITDA of ~$685–800M — roughly 4–5x the TTM $162.5M. At a through-cycle ~17% EBITDA margin that implies durable revenue of ~$4.0–4.7B, versus TTM ~$1.0B and management’s own “$2B down the road” framing. The alternative read — holding today’s ~59x — requires TTM EBITDA to be a floor that compounds ~20%+/yr for a decade with peak margins intact and the rich multiple held. Either path underwrites the bull case as the base case. What the market prices correctly: the demand is real and turbine-gated supply is genuinely scarce. What it prices incorrectly (the bear’s claim): permanence and peak-margin durability in a project-lumpy, fixed-price business with essentially no margin of safety.

Scenario analysis (explicit assumptions; outputs are scenario zones, not price targets):

The asymmetry is unfavorable. Most of the upside is already in the price: the bull scenario (a “double hold” — peak execution and a persistent peak multiple) delivers only a low-double-digit-to-mid-double-digit return, while the bear scenario — which requires nothing exotic, just normalization — implies a 60–70% drawdown, with a simultaneous float reversal pressuring the reported cash balance. The base case has the stock growing into much of its multiple over several years for roughly flat-to-modestly-negative return.

11. Variant Perception

Consensus belief. Argan is the best-positioned pure-play on a durable, multi-year data-center/AI gas-power supercycle: a fortress net-cash balance sheet, turbine-gated supply scarcity that protects EPC margins, and a backlog doubling to ~$2.8–2.9B that proves multi-year visibility. (Telling indicator of how far sentiment has run: even Lake Street’s bull-side note rates the stock Hold with a $600 price target (Jun-5-2026) — below the $753 price.)

Strongest bull case. The turbine-OEM bottleneck (GEV/Siemens Energy/MHI sold out to ~2029–30) gates large-scale gas EPC to a handful of qualified primes; Argan’s net-cash bonding capacity plus Gemma’s ~20-year fixed-price track record (“no lost job since acquisition”) make it one of the few that can bid gigawatt jobs. $2B+ revenue at high-teens/low-20s gross margin is reachable and durable; net cash funds growth with no dilution; the scarcity premium persists for years; and the micro-float amplifies the move.

Strongest bear case. Three correlated peaks, plus distribution and no cushion. (1) Peak margins — 20.5% GM is favorable mix + the Trumbull close-out + an international recovery, not a run-rate (management frames blended “high-teens to low-20s”). (2) Fixed-price EPC tail risk — the FY2020 TeesREP $33.6M loss is the exhibit; one bad gigawatt job erases a year of profit. (3) Concentration — top-3 customers ~50% of revenue. (4) Float reversal — contract liabilities ($566M) exceed equity; the reported $400M+ OCF and the $974M cash are float-inflated and reverse as backlog burns. (5) Rate-sensitive interest income (~16% of pretax) fades on cuts. (6) Unanimous insider distribution into the ATH with zero buys. The danger is that these are correlated: a single crack — a project loss, an award gap, or a hyperscaler capex pause — compresses the margin, reverses the float, decelerates growth, and de-rates the richest-ever multiple all at once. At ~59x EBITDA there is no margin of safety to absorb it.

The 3–5 assumptions that matter most. (1) Is 20.5% gross margin structural or peak? (2) Is the gas/data-center capex a decade-long plateau or a capital cycle that mean-reverts? (3) Can Argan 2–4x revenue to $2–4B at peak margins without a fixed-price blowup? (4) Does the ~$970M cash stay productive (bonding) or sit lazy (T-bills)? (5) Does a rich micro-cap multiple persist once momentum fades?

Falsification tests. The bull breaks if: a fixed-price project loss/write-down prints; book-to-bill falls below 1 and backlog rolls over for multiple quarters; or gross margin prints high-teens. The bear breaks if: a sustained $2B+ revenue run-rate is achieved with GM held ~20%; new multi-GW awards keep book-to-bill above 1; and clean ex-interest, ex-float FCF scales with revenue.

Where consensus may be offsides (factor read). Argan now trades as a crowded infrastructure-momentum vehicle — its factor-cosine neighbors are momentum and industrial ETFs (FFTY, PRN, PDP, AIRR), and momentum/grid names (STRL, PRIM, MYRG, MTZ, PWR, FIX, EME). Loadings are Momentum ~1.0, Market ~0.96, Industry:Infrastructure ~1.15–1.28; idiosyncratic vol is ~62% annualized (~two-thirds of variance is stock-specific). This is a parabolic momentum one-way street (beta 1.24, alpha 1.46, rs_12m +252.9, ~25x off the 2022 low, −4.7% off the ATH), not a falling knife — but momentum names crash at turns, and single down-days of −11% have already appeared (Sep-2025, Jun-2026). Consensus is offsides if it is extrapolating a turbine-gated cyclical peak into a secular-compounder multiple — pricing permanence and franchise-grade economics into a project-lumpy, narrow-moat, fixed-price contractor, precisely when insiders are selling into the high.

12. Fact vs. Interpretation

13. Open Questions

1. What is the true through-cycle gross margin for Power once the Trumbull-class close-outs anniversary and earlier-stage jobs dominate the mix — high-teens, or has the turbine-scarcity window structurally lifted it?
2. How large and how fast is the float reversal as the current cohort of early-stage projects burns down — and how visibly will it depress reported OCF and the cash balance?
3. Why no special dividend since 2020 despite the cash pile tripling — is there a bonding/regulatory constraint, an undisclosed M&A intent, or simply inertia?
4. What does the next 12–18 months of awards look like — does book-to-bill stay above 1, or does the sequential backlog dip (Jan→Apr 2026) become a trend?
5. Why is every insider selling — pure diversification at an obvious high, or a read on forward fundamentals the outside investor can’t see?
6. How exposed is the backlog to a single counterparty or a single hyperscaler’s capex plan, given ~50% top-3 customer concentration?
7. What is the company’s own capacity ceiling — can it actually staff and project-manage 10–12 simultaneous gigawatt jobs without execution slippage?

14. What Must Be True

For the bull case to be right (and the stock to grow into / exceed its multiple):

• The gas/data-center build-out is a decade-long plateau, not a 3–4 year capital cycle, and Argan converts it into a sustained $2B+ (toward $4B) revenue run-rate.
• Gross margin holds near 20% through that growth — i.e., the peak is a plateau, not a close-out artifact.
• Argan scales to 10–12 simultaneous gigawatt jobs without a single fixed-price blowup.
• Clean, ex-interest, ex-float FCF scales with revenue, and the rich multiple persists as momentum hands off to fundamentals.
• Falsification test: a fixed-price project write-down, OR a book-to-bill below 1 for multiple quarters, OR a gross margin that prints high-teens — any one falsifies the bull.

For the bear case to be right (and the multiple to de-rate):

• 20.5% gross margin mean-reverts to high-teens as favorable close-outs roll off.
• Backlog rolls over (the sequential dip becomes a trend) and/or a hyperscaler capex pause slows awards.
• The contract-liability float reverses, exposing how much of the cash and OCF was customer money.
• Interest income fades with rate cuts, and the market re-rates the richest-ever multiple toward the (already-elevated) peer cluster.
• Falsification test: a sustained $2B+ revenue run-rate with GM held ~20% and book-to-bill above 1, with clean FCF scaling — falsifies the bear.

The crux for both: whether the 20.5% gross margin and the doubled backlog are a plateau or a peak. Everything — the float, the interest income, the multiple — is correlated to that single question, which is why a crack in any one of them tends to crack all of them.

15. Source Appendix

See Appendix B below for the full source list. Primary sources: Argan FY2026 Form 10-K (filed 2026-03-26), Q1-FY27 Form 10-Q (filed 2026-06-04), DEF 14A (filed 2026-04-21), the June-2026 Form 4 cluster, and prior fiscal-year 10-Ks (FY2020–FY2025). Quantitative cross-checks from public financial data providers and a factor model. Peer context from comparable public E&C companies (PWR, MTZ, EME, FIX). Industry framing from published power-industry sources.

Prepared under the research framework. The institutional body carries no investment recommendation and no price target; the only position expressed in this document is the clearly-labeled “Claude’s Take” block at the top, which is Claude’s own subjective view and not the house view.

APPENDIX A — Standard Diligence Questionnaire

Supplemental to the research memo. Answers grounded in the research log; Fact / Interpretation / Assumption labels where they matter.

General

What thoughtful questions have other investors asked about this company? The recurring questions are: (1) Is the 20.5% gross margin sustainable or a cycle peak? (2) How much of the ~$974M cash and the $400M+ operating cash flow is real free capital versus customer-prepayment float that reverses? (3) Why hoard nearly $1B earning T-bill yield rather than return it (no special dividend since 2020)? (4) How durable is the data-center/gas demand — secular plateau or capital cycle? (5) Why are all insiders selling into the all-time high? (6) Does a record multiple (~59x EBITDA) on a fixed-price EPC contractor make sense given the FY2020 single-project blowup? These map directly to the memo’s QoE, capital-allocation, and variant-perception sections.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? A cyclical high (Interpretation, well-supported). Gross margin (20.5%), operating margin (14.3%), ROIC (~28%) and ROE (~39%) are all at or near record levels, lifted by the Trumbull early-completion close-out, an international rebound from prior U.K. losses, favorable project mix, and ~16% of pretax income from peak interest rates. Through-cycle margins are “high-teens to low-20s” by management’s own framing.

Driven by the external environment or internal actions? Both, but predominantly external: the data-center/AI-driven gas-power demand surge and the turbine-OEM supply bottleneck created the window; Argan’s disciplined bidding and fortress balance sheet let it capitalize. The margin level owes much to one-time project execution (Trumbull) and rates.

How stable are revenues? Project-lumpy and unstable. Revenue is fixed-price EPC recognized over time; FY2023 declined from FY2022, and backlog already dipped sequentially Jan→Apr 2026. There is no recurring revenue base. Backlog ($2.8B) provides 2–3 years of visibility but converts unevenly with award-timing gaps.

Outlook for products/services? Strong near-term: backlog ~$2.8B (~79% gas), >4.1 GW of gas plants, capacity for 10–12 simultaneous jobs, management framing $2B annual revenue “down the road,” plus a new data-center fabrication adjacency. The risk is post-2030 turbine-supply normalization and a capital cycle.

How big will this market be — growing, shrinking, domestic or international? The U.S. gas-EPC market is growing structurally for the next several years on load growth; Argan is ~90%+ domestic with a modest Irish/U.K. (APC) presence. The market is large relative to Argan’s ~$1B revenue, so capacity (labor, bonding, turbine slots), not demand, is the near-term constraint.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Less competitive near-term, structurally more competitive long-term. The turbine bottleneck currently rations work to a handful of bondable primes; when turbine supply catches up post-2030 and competitors expand crews, returns mean-revert.

How profitable is the business (ROIC, ROE)? Very — ROIC ~28%, ROE ~39% (FY26) — but flattered by negative working capital (customer float), a ~$974M interest-earning cash pile, and an asset-light structure (goodwill $28M, PP&E ~$18M). Returns on actual operating capital are high but the denominator is tiny-to-negative, overstating durability.

How profitable is the industry — competitors, barriers to entry? Fixed-price EPC is historically a thin-margin, mean-reverting industry. Barriers to qualifying for gigawatt work (bonding capacity, track record, turbine access) are real but the underlying economics are not structurally rich. A small number of large competitors, mostly private (Bechtel, Kiewit, Black & Veatch, Sargent & Lundy).

Can the business be easily understood? Yes — it builds power plants for a fixed price and earns a construction margin. The subtlety is the negative-working-capital float and percentage-of-completion accounting.

Can it be undermined by foreign low-cost labor? Largely no — gigawatt U.S. power-plant construction is site-based, permitted, unionized/skilled-labor-intensive, and bonding-gated; it is not offshorable.

Do brands matter? / Nature of competition / switching costs? “Brand” here is execution reputation and bondability, not consumer brand — it gates who gets invited to bid (Interpretation: a real but contestable intangible). Competition is project-by-project competitive bidding. Switching costs are essentially zero — every plant is separately bid; concentration (top-3 ~50% of revenue) is relationship-based repeat business, not contractual lock-in.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The franchise value of Gemma’s reputation and project-management bench is not capitalized (goodwill is only $28M). Conversely, much of the headline cash is not truly the company’s — ~$566M of contract liabilities is customer prepayment.

Off-balance-sheet liabilities? Performance/surety bonds and contractual completion obligations on fixed-price jobs are the main contingent exposures; operating leases are minimal (~$10M capital leases). Litigation: the U.K. subsidiary’s Mar-2025 suit against EP NI Energy.

How conservative is the accounting? Mixed. The balance sheet is conservative (net cash, no debt). The income statement relies on percentage-of-completion estimates (change orders, contingencies, close-out releases) that flatter peak periods — the Trumbull close-out is the clearest example. Revenue/margin recognition is estimate-laden, not aggressive per se, but cycle-sensitive.

How CapEx-hungry is the business? Very light — PP&E ~$18M, D&A ~$6M, capex a low-single-digit % of revenue. The new ~$10–13M North Carolina fabrication facility is the largest discrete capex item, and it is small.

Capital Allocation & Management

How much FCF does the business generate, and how is it used? Headline FY26 OCF was $414.7M, but clean FCF ex-float is ~$150–170M (the rest is reversing customer float). Uses: a growing regular dividend (~$24M/yr), tiny buybacks ($9.9M FY26), small bolt-on M&A — and a fast-growing idle cash pile. The capital-allocation critique is that ~$970M sits in T-bills with no special dividend since 2020.

Significant acquisitions recently? No — only small bolt-ons ($11.5M FY26, $16.3M FY25). APC (~2021) is the last meaningful deal. No value-destroying M&A is a genuine positive.

Buying back shares? Yes, but immaterially — 64,699 shares for $9.9M in FY26 (avg $152.66), with ~$20M more in Q1-FY27 at much higher prices. Authorization raised to $200M (Apr-2026) but execution is tiny versus the cash and market cap; they are now buying dear after historically buying cheap.

Issuing large amounts of new shares to insiders? No — SBC is modest ($7.9M FY26); diluted share count crept 13.55M→14.15M, partly offset by buybacks.

Compensation policy of directors/management? Pays on Revenue, ROE, EPS and TSR — no return-on-invested-capital metric (Fact, DEF 14A 2026-04-21). The chosen metrics are gameable via the lean-equity/float model and buybacks. A real governance weakness.

Motivations of management? Revealed preference is to sell: unanimous discretionary insider selling of ~$80M+ into the June-2026 ATH (largest holder cut ~52% of his stake; CEO and CFO both sold), zero open-market buys, ~3% group ownership. Interpretation: at minimum a strong “diversify at the high” signal; at most a read on forward fundamentals.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No — ordinary U.S. common stock, NYSE-listed, standard 1099 reporting.

Dividend policy? Regular dividend ~$2.00/yr (raised +33% to $0.50/quarter Sep-2025), ~0.27% yield; history of occasional special dividends (last in 2020). Payout is conservative (~18–20% of earnings).

How profitable is the business? Highly, at the cycle peak (see above), with the heavy caveats on durability and float/interest flattering.

Is net income diverging from cash from operations? Yes — OCF ran 3.0x net income in FY26 (and −0.9x in FY23). The divergence is the contract-liability float, which inflates OCF on the way up and will depress it on the way down. This is the central QoE point: reported cash flow is not a clean read on earnings quality.

Risks & Downside

What factors would cause the stock to decline? A fixed-price project write-down; gross-margin normalization to high-teens; a book-to-bill below 1 / backlog roll-over; a hyperscaler capex pause; interest-income fade; float reversal depressing reported cash; and — most powerfully — a de-rating of the richest-ever multiple. These are correlated, so one crack tends to cause several.

Risk of a catastrophic loss? To capital, low — net cash, no debt, asset-light, negligible bankruptcy risk. To the multiple, high — a 50–70% drawdown on normalization is entirely plausible from ~59x EBITDA (bear scenario).

Chance of a total loss? Negligible. The fortress balance sheet makes solvency a non-issue; the risk is valuation, not permanence of the business.

Recent News & Events

Has the business environment changed recently? Yes, favorably for fundamentals: backlog doubled to ~$2.9B, multiple gigawatt gas awards (SLEC, Ward County, Louisiana LNG), the $125M data-center fabrication contract, an international return to profit, a +33% dividend hike, and an expanded buyback authorization. But for a buyer at this price, the changes that matter — peak-margin close-outs, float-inflated cash, and unanimous insider distribution into the ATH — make the risk/reward worse, not better.

Significant acquisitions? No — bolt-ons only.

Change in accounting policies? None material identified; percentage-of-completion recognition unchanged.

Recent changes — new markets, facilities, management? New North Carolina fabrication facility (data-center adjacency); CEO/CFO transitions completed in recent years (Watson CEO, Baugher CFO); non-executive chairman. The June-2026 insider-selling cluster is the most notable recent governance event.

APPENDIX B — Source Appendix

Primary sources first. Accessed 2026-06-26 unless noted.

Primary — SEC filings (EDGAR, CIK 0000100591)

Primary — earnings call

Quantitative cross-checks (third-party aggregators — reconciled to filings)

Peer / industry context

Notes on reliability

• EV re-derivation: ROIC’s enterprise-value snapshot is struck at the Apr-30-2026 quarter-end (~$668); all current multiples in the memo re-derive EV at the live $753.07 close (2026-06-25).
• Aggregator caveat: ROIC/AZI/FactorsToday are third-party aggregated data; every material number is reconciled to the underlying SEC filing, which governs in any discrepancy.
• Management commentary (guidance, margin framing, backlog narrative) is treated as a hypothesis, validated against filings and financials, not as evidence.