Granite Construction Incorporated (NYSE: GVA) — Real Rock, Rented Cash Flow

Report date: 2026-08-09
Price reference: $128.32 at the 2026-08-07 close
Sector: Industrials · Construction & Engineering
Scope: Sections 1–15 below are evidence-led and position-free; they contain no recommendation or price target.

⚡ Claude’s Take

The author’s independent subjective opinion; general information only, not investment advice. The analytical body below carries no position.

Verdict: HOLD / avoid chasing at $128; accumulate only in roughly the $95–$110 zone, equivalent to about 12–14x an approximately $8 normalized forward-EPS sensitivity and a high-single-digit guide-EBITDA multiple after the convertible settlement is known.

Granite is a better company than it was three years ago. Construction gross margin rose from 10.9% in 2023 to 15.7% in 2025, committed and awarded projects reached $7.4 billion, and the company has assembled a real portfolio of permitted local quarries and asphalt plants. But the stock already prices the transformation as durable. At $128.32, the pre-settlement enterprise-value bridge is approximately $7.0 billion, roughly 14x TTM EBITDA and about 10x management’s adjusted 2026 EBITDA guide. No-growth earnings-power value is materially lower. The framing is quality-improving infrastructure exposure at a price, not deep value: the market is paying for Granite to become a ROAD-like vertically integrated compounder before it has disclosed deal-level returns or proved clean cash conversion.

The tape says the same thing more politely. GVA remains 38% higher year over year and above its 200-day EMA, but it is 21% below the June high and 9% lower over three months. Its factor identity is infrastructure/small-cap industrial rather than pure momentum, and 9.35% of float was short at July 15. That is a live disagreement, not capitulation. Conviction: medium. Bullish flip: acquired-capital ROIC above 12% with Materials cash margin above 30% and ex-claim Construction margin holding 15–16%. Bearish flip: ex-EAC Construction margin below 14% for two quarters or an unwind of billings/payables float that exposes structurally weak owner cash flow.

Tag: The rocks are real; the cash conversion still needs proof.

📈 Stock Price Action — Five-Year Event Map

GVA’s adjusted total return was 249% from August 2021 through August 7, 2026. The traded price fell from $38.67 to a $25.35 intraday trough in September 2022, reached $162.08 on June 26, 2026, and closed at $128.32—20.8% below that high and within a trailing-year range of $94.19–$162.08. Price history is from the AZI GVA CSV, accessed 2026-08-09.

The first six events mark the operating rerating: CAP rose from $4.1 billion in Q3 2022 to $7.2 billion by Q1 2026 as project selection, public funding, materials investment and acquisitions improved the earnings profile. Each move coincided with a filed earnings release (FY2021, Q3 2022, Q3 2023, Q3 2024, Q2 2025, and Q1 2026). Event 7 was a clean valuation reset—Investing.com reported the downgrade before a 12.5% one-day fall—while the Q2 2026 release recovered only part of that loss.

1. Executive Summary

Granite has executed a credible turnaround and a more coherent strategy. It exited much of its legacy mega-project risk, improved bid discipline, shifted toward vertically integrated home markets, and spent approximately $1.36 billion since 2023 buying aggregates, asphalt and regional construction assets. Revenue rose from $3.51 billion in 2023 to $4.42 billion in 2025; gross margin rose from 11.3% to 16.1%; operating margin increased from 2.3% to 6.4%; and reported free cash flow rose from $43 million to $331 million. Q2 2026 revenue grew 29%, and management raised full-year revenue guidance to $5.3–$5.5 billion with a 12.25–13.25% adjusted EBITDA margin. Those are real improvements, reconciled to the FY2025 10-K and Q2 2026 10-Q.

The competitive advantage is narrower than the headline. Granite owns 116 quarry properties, 86 in production, with approximately 2.081 billion tons of reserves and resources. Because aggregates are heavy and low-value relative to freight, permitted reserves close to demand can create a local delivered-cost and supply advantage. That is a real Greenwald-style local cost/scale barrier. Yet Materials was only 17% of FY2025 revenue and 19% of segment gross profit. The other 83% of revenue came from Construction: non-recurring projects, commonly rebid, with low inter-project switching costs. At FY2025, 91.5% of unearned revenue was fixed-unit-price or fixed-price. The best description is a narrow local materials moat embedded inside a largely unmoated, execution-driven contractor.

The financial-quality debate is not whether reported results improved; they did. It is how much should be capitalized. Q2 net favorable estimate-at-completion revisions were $14 million, or 5.9% of gross profit. H1 reported FCF was $85.7 million, but billings in excess rose $114.7 million and accounts payable supplied $169.9 million of cash. Removing only the billings increase produces a mechanical negative $29 million FCF sensitivity—not normalized FCF, but enough to show that contractor float is financing, not earnings. Headline CAP also needs qualification: of $7.419 billion at June 30, $5.056 billion was executed unearned revenue and $2.362 billion was softer “other awards,” and substantially all CAP can be canceled or modified.

Capital allocation is directionally sensible but unproven. Warren Paving, Papich, Cinderlite, Kenny Seng and earlier purchases add local reserve density and self-supply potential, but Granite discloses neither acquisition multiples nor deal-level ROIC. Debt carrying value rose from $655 million at 2023 year-end to $1.559 billion at Q2 2026. The company also exchanged/called its 2028 converts and issued $600 million of 6.375% senior notes. The resulting Q2 GAAP loss was mostly a conversion-accounting artifact, but the cash and higher coupon are economic. As of this report cutoff, capped-call proceeds and final convert settlement were still scheduled for August 11–12 and unknown.

At $128.32, actual shares imply $5.617 billion of equity value. A pre-settlement accounting bridge gives approximately $6.99–$7.08 billion of enterprise value, or 14.0–14.2x TTM EBITDA and 10.1–10.3x management’s guided adjusted EBITDA. No-growth EPV is approximately $3.3–$4.2 billion. Identifiable operating book assets are materially lower, while the replacement premium for permits and reserves cannot be measured from public disclosures. The market therefore assigns substantial value to profitable growth. A reverse DCF requires approximately a 6% normalized unlevered FCF margin and 7–8% revenue CAGR for a decade, or an equivalent margin/growth combination. That can happen, but it requires durable ex-claim margins, clean cash conversion, successful reauthorization, and acquisitions earning above their full cost.

Executive verdict: Granite is a materially improved operator with authentic local moat assets. The consolidated franchise is not yet wide, cash conversion is flattered by reversible float, and current valuation already assumes that acquisition-led vertical integration creates lasting returns.

The next two reported quarters matter more than another headline CAP record: they will expose settlement liquidity, winter cash conversion, ex-revision margins and the first fuller-year return evidence from recently acquired assets.

2. Business Overview

Granite is a U.S. infrastructure contractor and materials producer founded in 1922. The company operates two reportable segments. Construction builds and rehabilitates roads, bridges, rail, airports, ports, dams, reservoirs, aqueducts, tunnels, water infrastructure, solar/battery sites and other civil assets. Materials produces aggregates, recycled materials, asphalt concrete and liquid asphalt, both for Granite’s projects and third parties. Vertically integrated markets include Alaska, Arizona, California, Kentucky, Louisiana, Mississippi, Nevada, Oregon, Tennessee, Utah and Washington, with regional civil work extending into other states. The FY2025 filing is the primary source for scope, geography and segment data.

Construction revenue was 71% public and 29% private in 2025. Public customers include federal, state and local agencies; private customers include developers, utilities and industrial users. The value proposition to agencies is balance-sheet/bonding capacity, specialist labor, multiple delivery methods and the ability to execute large, complex jobs. It is not a proprietary product. For private customers, local materials, equipment and labor can reduce coordination and schedule risk. For materials customers, proximity and supply assurance matter most because hauling cost can overwhelm plant-gate price.

Revenue is project-based, not recurring in the software sense. CAP provides forward workload, but only the executed portion resembles a remaining performance obligation. At June 30, 2026, $5.056 billion of $7.419 billion CAP was executed unearned revenue; $2.362 billion was other awards such as probable later phases, options and task orders. Customers may cancel or modify substantially all of it. CAP is therefore both visibility and risk inventory.

Growth is becoming acquisition-assisted. In 2025, acquired businesses contributed $112 million of Construction revenue and $106 million of Materials revenue. In Q2 2026, acquired companies contributed $98 million and $60 million, respectively. The rest of Q2 growth was meaningful, but the correct model is no longer “organic turnaround only.” Granite is deliberately using M&A to form denser regional clusters around materials assets.

How Granite earns—and loses—money

Construction economics begin at bid selection. On fixed-price work Granite estimates the total cost, accepts the risk that actual labor, materials, schedule and subcontractor cost exceed that estimate, and recognizes profit over time. Fixed-unit-price work partly protects volume uncertainty but still exposes unit productivity and input cost. Negotiated and alternative-delivery work can improve risk allocation, yet the customer retains bargaining power. Gross margin therefore reflects four things at once: the quality of work won several years ago, current execution, risk retirement/claims, and the competitive environment when replacement work is bid.

Materials economics are more tangible. Granite extracts rock or sand/gravel, processes it, and sells aggregates directly or feeds asphalt plants. Cost per ton depends on stripping, blasting/crushing, energy, labor, haul distance, plant utilization and reserve quality. Price depends on local alternatives and freight. The same quarry may create value twice: third-party materials margin and a lower-risk internal supply chain for Granite’s paving work. Granite does not disclose internal transfers or per-ton KPIs, so investors cannot separate those two channels.

Seasonality matters. Weather restricts construction and quarry activity, especially in northern and wet markets, making the first quarter structurally weaker and working capital uneven. Project mobilization can consume receivables and contract assets before billing catches up; other projects receive advance milestone cash. H1 or quarterly cash flow should not be annualized mechanically.

The company also bears bonding and JV economics that a consolidated revenue view misses. Sureties provide bid, performance and payment bonds but rely on Granite’s balance sheet and indemnification. Consolidated JVs can make Granite appear to own cash that requires partner consent, while NCI partners retain part of CAP and earnings. The correct unit of analysis is Granite-attributable economics after partner interests and financing restrictions.

Public work is less discretionary than private development, but it is not recurring revenue. Each appropriation, letting and project phase must convert into an award, and each award must be executed profitably. A large CAP balance creates workload visibility, yet the value of that workload depends on margin and contract quality. Granite’s 2026 mix—68% executed unearned revenue and 32% other awards—is better than year-end 2025’s 59%/41% split, but remains far from a subscription backlog.

Materials makes the model more durable because reserves persist beyond one project and nearby customers repeatedly need rock and asphalt. However, 22.8 million annual tons against 2.081 billion tons of resources/reserves should not become a simple mine-life calculation: resources differ in economics, permit status, demand proximity and ownership/lease term. The reserve base is strategic optionality, not automatically monetizable value.

Customers, contract terms and operating leverage

The customer mix changes risk. Public agencies usually offer lower credit risk and long program visibility, but procurement rules can emphasize price and agencies retain termination rights. Private customers may negotiate faster and accept alternative delivery, yet payment can be slower and demand is more cyclical. Granite’s 71% public Construction revenue mix stabilizes demand without guaranteeing margin.

Contract type determines who absorbs inflation. Fixed-price work puts most estimating risk on Granite. Fixed-unit-price work adjusts for quantity but not necessarily unit cost/productivity. Alternative delivery lets the contractor collaborate earlier, reducing design uncertainty, but does not eliminate execution risk. The company reports contract mix across unearned revenue, not margins by type, so investors cannot quantify the risk premium Granite earns for each structure.

Operating leverage has two directions. When volume rises, equipment, plants, field supervision and overhead utilize better, expanding margin. When activity falls, fixed local capacity and idle crews reverse that benefit. Materials utilization can be particularly volatile because weather or quarry development reduces output faster than costs. The Q2 2026 Materials margin decline despite higher revenue demonstrates that acquisition volume alone does not ensure leverage.

Working capital is part of customer economics. Favorable billing terms allow Granite to fund projects with customer cash; unfavorable terms create contract assets and receivables. The quality of an award therefore includes margin, risk allocation and cash profile. CAP discloses only volume, not this economic bundle.

Verdict: Granite is understandable but not simple. The attractive materials assets and more stable public demand sit inside a contract-accounting business whose revenue is rebid, cancelable and execution-sensitive.

3. Industry Dynamics

The value chain divides into six layers: mineral rights and input suppliers; quarries; asphalt plants; civil contractors; sureties/insurers; and asset owners. Granite participates from extraction through delivery and construction. The profit pool is not uniform. Permitted aggregates near growing metros can earn scarcity economics. Asphalt also benefits from local density but carries liquid-asphalt input and utilization risk. General civil construction is fragmented, competitively tendered, labor intensive and vulnerable to fixed-price cost errors.

Freight economics explain why “local” matters. USGS describes aggregates as high-volume, low-value products whose transportation cost can exceed the plant-gate price; land-use conflict and environmental permitting constrain new supply near demand. Its 2023 producer census counted 1,322 crushed-stone companies across 3,531 quarries and 3,100 sand-and-gravel companies across 6,324 pits—national fragmentation coexisting with local scarcity (USGS aggregates data, accessed 2026-08-09). This is a better framework than treating U.S. aggregates as one market.

Demand is currently strong. The Census Bureau reported June 2026 public construction at a $544.1 billion seasonally adjusted annual rate and highway/street construction at $150.9 billion (June 2026 construction spending, released 2026-08-03). These are spending pools, not contractor TAM forecasts. Granite’s $3.655 billion FY2025 Construction revenue equals 2.4% of the highway figure, but the comparison is only a scale proxy because both numerator and denominator include different scopes.

Federal funding is supportive but reaches a decision point. FHWA states the IIJA provides approximately $350 billion for federal highway programs over FY2022–2026, with FY2026 federal-aid highway apportionments of $56.815 billion (FHWA IIJA funding). Current authorization expires September 30, 2026. Formula funds and state programs reduce near-term demand volatility, but successor legislation controls the next leg. A delay would more likely slow future lettings than halt work already obligated.

Regulation cuts both ways. Procurement rules, DBE requirements, Buy America provisions, OSHA/MSHA standards and termination rights add cost to Construction. Air, water, land-use, mining and reclamation permits make Materials more burdensome—but also make incumbent reserves harder to replicate. Regulation therefore weakens contractor flexibility while strengthening existing quarry barriers.

Competitive regime and supply-side cycle

Construction resembles a repeated prisoner’s dilemma. Contractors collectively benefit from rational pricing, but each has an incentive to keep crews and equipment utilized by accepting a thinner bid. Public bid transparency and fragmented local competition make discipline company-specific: Granite must refuse work even when competitors chase it. The 2023–2026 margin repair suggests better discipline, but record demand makes that behavior easier than it will be in a slowdown.

Materials is closer to a local oligopoly. A new quarry entrant must secure land, reserves, permits, community acceptance, capital and routes to customers. Freight can neutralize nominally cheaper plant-gate supply. Incumbents need not have national share; they need share inside the economic haul radius. Asphalt sits between rock and contracting: plants are local and the product time-sensitive, but a plant is more reproducible than a permitted reserve and margins depend on liquid-asphalt cost and utilization.

Industry capacity comprises people, equipment, bonding, reserves and plants. Labor and equipment can expand within a few years; quarry supply may take a decade or never receive approval. Strong IIJA demand therefore produces different responses: contracting capacity and M&A expand, while permitted rock remains constrained. Granite buys scarce assets rather than awaiting greenfield permits, but the purchase price transfers some scarcity rent to sellers.

Three capital-cycle indicators deserve priority. First, 2025 capex of $138 million was below $162 million D&A, but acquisition spending added far more capacity than organic capex reveals. Total asset growth and purchase-accounting step-ups are the better supply measure. Second, bid spreads and win rates would reveal discipline, but Granite does not publish them. Third, margin should be measured excluding individually material catch-ups. Until those disclosures improve, future returns remain an inference.

The $150.9 billion highway/street SAAR is a transparent reference, not SAM. It includes owner costs, subcontract scopes and markets outside Granite’s footprint. A proper Construction SAM would map state DOT, water, rail, federal and private programs only where Granite has crews, bonds and materials assets. A Materials SAM must be quarry-by-quarry. National aggregates statistics explain structure but cannot establish local share.

The capital-cycle read is mixed. Public spending, record CAP and strong margins are demand positives. Granite’s $1.36 billion of announced M&A since 2023, ROAD-like regional consolidation, and expansion of equipment/labor capacity are supply-side warnings: high returns attract capital. Construction capacity is easier to add than a permitted quarry, so future mean reversion should hit contracting before rock. The industry is not demonstrably at a peak, but it is clearly in expansion rather than scarce-capital recovery.

Verdict: A mixed industry. Permitted aggregates can be structurally attractive; asphalt is middling; civil construction is structurally mediocre. The winning model combines dense local materials with disciplined bid selection, not national size alone.

4. Competitive Position

Moat taxonomy

Local supply/cost advantage—real but partial. Granite reported 116 quarry properties across 31,797 acres: 86 producing, 15 in development and 15 in exploration. Reserves and resources totaled about 2.081 billion tons, 64% owned and 36% leased, against 22.8 million tons produced in 2025. Permits and proximity can lower delivered material cost and reduce supply risk. Those assets would be expensive and slow to reproduce.

Scale/bonding—qualification threshold, not broad pricing power. Approximately $3.9 billion of FY2025 CAP was bonded. Surety relationships, capitalization, workforce and technical experience reduce the eligible bidder pool for very large work. But once qualified, firms still compete on price, delivery and risk allocation. Scale admits Granite to the auction; it does not eliminate the auction.

Customer captivity—low across projects. Agencies can award the next project to a competitor. Live-project replacement is disruptive, but that is temporary contract lock-in, not durable switching cost. Materials buyers may be captive to nearby supply when alternatives require long hauls, but Granite does not disclose local shares or haul radii.

Intangibles—permits yes, brand weak. Purchase accounting valued Warren Paving permits at $20 million, corroborating the economic value of operating rights. Granite’s century-old reputation may help prequalification, but brand is not a moat in transparent procurement without proof of negotiated work or wins above lowest price.

Network effect—none. More project customers do not inherently improve the service for each customer. Denser assets improve utilization; that is scale, not a network effect.

Peer comparison

Vulcan Materials and Martin Marietta are purest expressions of the quarry moat, with aggregates dominating profit and pricing. Granite has the same mechanism in selected markets, but Materials remains a minority of economics; the whole company does not deserve a pure aggregates classification. Construction Partners is the closest model analogue: local roadbuilding plus aggregates/asphalt clusters. Sterling Infrastructure demonstrates the control case—excellent project selection and mix can raise margins without creating a protected franchise. PRIM, DY, FLR and private regional contractors compete across civil and specialty scopes; Granite’s breadth and bonding help, but its own filing acknowledges multiple competitors in every market.

The financial test is encouraging but short. Estimated filing-based ROIC rose from about 5.8% in 2023 to 13.3% in 2024 and 13.9% in 2025. That supports an advantage or better execution, but it covers only two strong years and mixes acquisition timing, JV/NCI and carrying-value effects. Share stability and local profitability—the strongest Greenwald tests—are not disclosed.

The missing disclosures matter: internal material self-supply, third-party tons, price per ton, cash GP per ton, local share, average haul radius, bid spread, win rate, and integrated versus non-integrated Construction margin. Without them, the mechanism is plausible but the magnitude is not auditable.

Durability and competitive response

A moat should produce a measurable counterfactual. If Granite’s local materials advantage disappeared, integrated-market bid margins should fall, purchased-material and haul cost should rise, schedule reliability should worsen, third-party pricing should weaken and asset utilization should decline. Granite publishes none of those comparisons. Consolidated margin and ROIC improved, but a turnaround in estimating and project selection can create the same outcome without a structural moat.

Market-share stability—the preferred Greenwald test—is unavailable. The relevant market ends at the economic haul radius, not the state or country boundary. Granite reports tons and quarry properties but not metro share. The report therefore grades the mechanism and the observable financial result, while explicitly withholding a wide-moat conclusion.

The reserve base is strongest where three conditions overlap: owned rather than short-term leased mineral rights; full extraction and processing permits; and proximity to durable demand. Granite discloses 64% ownership and says active quarries hold necessary permits, but it does not map demand or permitted tons by site. A remote reserve is not equivalent to one beside a constrained metro.

Bonding scale has a clearer effect. Large agencies need confidence that a contractor can mobilize and complete after a problem. Granite’s multi-surety program and $4.4 billion of bonded Q2 CAP reduce the bidder set. Yet the same capacity destroys value if it encourages larger fixed-price risks. Scale is an advantage only when paired with bid discipline.

Competitors can respond in several ways. ROAD can assemble denser Southeastern clusters; pure aggregates producers can sell material to Granite’s rivals; specialty firms can partner in JVs; and local contractors may underbid Granite’s overhead structure. Granite answers with breadth plus integration. Breadth diversifies demand but adds complexity. Integration is more defensible because a competitor cannot instantly reproduce a quarry.

Acquisition itself is not a moat. Rivals can bid for the same family-owned businesses, and abundant financing raises seller expectations. Granite creates value only if internal demand, regional density and operating capability make an asset worth more to Granite than to the next-highest bidder. Deal-level returns would test that proposition; management does not disclose them.

Verdict: Weak/narrow competitive advantage. Granite owns genuine local moat assets, but the consolidated company remains predominantly an execution business. Upgrade only after materials pricing and acquired-capital returns hold through weak volume; downgrade if Construction margins revert or Materials cannot price over volume.

5. Growth History and Forward Opportunities

The 2021–2022 decline reflected portfolio repair and a weak exit from legacy risk. Growth resumed in 2023 and accelerated in 2024–2026. The quality improved because Construction gross margin rose with revenue, but three growth engines must be separated.

First, public infrastructure is the demand engine. Roughly 70% of 2025 Construction revenue was agency funded, and 82% of Q2 2026 CAP was public. Bridges, highways, water and federal tactical infrastructure support volume. Q2 CAP included $624 million of U.S. Customs and Border Protection projects expected largely in 2026–2027.

Second, portfolio repair is a finite margin engine. Granite has been more selective, reduced legacy mega-project exposure and entered work with better risk terms. Construction gross margin rose from 10.9% in 2023 to 15.7% in 2025. But favorable claim/EAC revisions contributed, so not every basis point is repeatable operating leverage.

Third, M&A is now structural. Warren, Papich, Cinderlite and Kenny Seng broaden reserves and home-market density. This can create local scale, but it also makes EBITDA growth purchasable. Management said on the Q2 call that another $200–$400 million of acquisitions could occur in H2 2026 and described $300–$800 million as a possible annual range in later years. That is management guidance, not committed fact, and it makes return-on-acquired-capital the central KPI.

The forward opportunities are water infrastructure, federal/tactical work, rail/transit, mission-critical site development and further materials clusters. The largest opportunity is not an abstract TAM; it is increasing Materials profit share and using owned rock to earn better Construction economics in the same local market.

Growth quality by engine

Public infrastructure growth is visible but policy-supported. It is higher quality than speculative private construction because projects address long-lived needs and formula funding smooths budgets. It is lower quality than recurring contracted revenue because authorization expires, agencies can delay lettings, and awards carry execution risk. The right leading series is executed unearned revenue, not headline CAP.

Margin-repair growth is high quality while it comes from avoiding bad work. It also has a ceiling. Once legacy projects roll off and standards normalize, further improvement requires productivity, mix or structural pricing. Net favorable estimate changes obscure measurement, making ex-revision margin the test of whether repair became a new base.

Materials/M&A growth offers the largest strategic gain and allocation risk. A purchased quarry can produce decades of cash and strengthen internal bids; it can also be overpaid at a cycle high. Warren’s mineral-reserve step-up shows Granite paid explicitly for scarcity. The unanswered question is whether Granite captured synergy or prepaid it.

Adjacency growth is least proven. Federal tactical infrastructure and mission-critical site development can diversify highways and command attractive margins, but they attract sophisticated competitors and can create project/customer concentration. The bull case should not assign a premium before delivery evidence exists.

Verdict: Growth is currently high quality in revenue and reported margin terms, but increasingly acquisition-assisted. It creates value only if ex-claim margins persist and the full acquired capital base earns above the cost of capital.

6. Financial Quality

Five-year operating record

The evidence supports a genuine inflection. Revenue compounded about 12% from 2023 to 2025 while gross profit rose 34% annually and operating profit more than tripled. Both segments improved in 2025: Construction gross margin reached 15.7%, and Materials reached 17.8%. Filing-based directional ROIC rose from about 5.8% in 2023 to 13.9% in 2025. That estimate uses 75% of operating income over average year-end equity plus debt less cash and securities; it is not company-reported and does not fully normalize JVs, leases, acquisition timing or purchase accounting.

The Q2 2026 picture is more mixed than the 29% revenue headline. H1 revenue rose 29.7% to $2.368 billion, but gross margin fell to 14.7% from 15.5%. Construction H1 margin was nearly flat at 15.2%; Materials margin fell to 12.1% from 16.0%. Q2 Materials revenue rose 32%, yet gross profit declined 12%. Management attributed the gap to roughly $10 million of severe-weather cost, $5 million of quarry-development cost and purchase-accounting depreciation/amortization. The first two may normalize; purchase-accounting charges follow the acquisition strategy and should not be wished away.

Percentage-of-completion earnings

Granite recognizes Construction revenue over time using cost incurred relative to estimated total cost. That is standard and economically sensible, but it requires continual estimates of price, productivity, schedule, claims and cost to complete. PwC identifies multi-year fixed-price revenue/cost estimates as a critical audit matter in the 2025 10-K.

In 2025, eight individually material upward revisions added $64.9 million of project gross profit; three downward revisions subtracted $33.5 million and reduced Granite-attributable net income by $25.6 million, or $0.48 per diluted share. In H1 2026, upward revisions added $30.9 million and one downward revision removed $14.2 million. Q2’s net favorable $14.0 million equaled 5.9% of reported quarterly gross profit.

These revisions are neither fake nor automatically non-recurring. Claims settle, risks retire and productivity changes. But cumulative catch-ups move lifetime economics into the current quarter, so the right analytical margin is a range: reported margin and margin excluding individually material net favorable revisions. Granite does not provide a full sensitivity because management says one is not reasonably meaningful. That increases, not decreases, the importance of the disclosed project bridge.

Cash flow and contractor float

H1 2026 CFO was $141.5 million and capex $55.9 million, producing $85.7 million of reported FCF. The cash bridge shows why this is not owner earnings. Billings in excess increased $114.7 million to $435.3 million—81% of H1 CFO. Accounts payable supplied another $169.9 million, while receivables consumed $268.4 million. A mechanical sensitivity that removes only the billings increase turns FCF to negative $29.0 million.

That sensitivity is intentionally incomplete. It does not normalize payables, receivables, contract assets, acquisitions, seasonality or project mix. It is not a forecast. Its purpose is to identify the economic category: advance billing is customer financing that reverses as work is performed. Capitalizing management’s 11%-of-revenue OCF target without separating that float would overstate earnings power.

Cash earnings, normalization and returns

The five-year pattern shows why neither extreme is right. Reported FCF was negative in 2021–2022, $43 million in 2023 and about $320–$331 million in 2024–2025. The improvement is too large to dismiss as working capital: gross profit and operating income also increased. Yet cash conversion is more volatile than operating progress because claims, retention, advance billing and payables move cash across periods.

A useful normalized framework starts with NOPAT, adds D&A, subtracts maintenance capex, and applies a through-cycle working-capital requirement. It should not begin with one year’s CFO. In 2025 D&A exceeded capex, partly because acquired assets lifted depreciation. That does not prove capex can stay below depreciation indefinitely; strategic materials development and fleet replacement remain economic reinvestment.

Reported TTM FCF near $472 million also rolls in late-2025 collections and 2026 advance billings. Using its roughly 8.4% headline equity yield assumes those financing flows persist. The valuation scenarios instead use 5–8% revenue cash margins after ordinary working capital. This is less precise but more economically honest.

The Q2 $360 million convertible charge is non-operating and should not drive P/E or operating quality. Conversely, acquisition/integration expense, SBC, recurring disposal gains and the new senior-note interest are economic even when adjusted away. Normalization should remove fair-value noise and discrete legacy litigation while retaining the recurring cost of the strategy.

Project revisions require symmetric treatment. One cannot subtract downward revisions as “one-time” while retaining favorable claims. The conservative method observes GAAP margin and a second margin excluding net individually material favorable revisions, without pretending all estimate changes disappear.

Directional ROIC is the strongest positive evidence. Revenue can be bought, but a move from mid-single-digit to low-teens returns suggests better economics. Still, 2025 deals closed throughout the year, so the full acquired denominator was not employed for a complete period. 2026–2027 ROIC will be more informative and must include goodwill, mineral-reserve step-ups and intangibles. Excluding them may describe operating return but not shareholder value creation.

Accounting conservatism and historical context

Granite’s current filings report no material weakness, and the five-year corpus shows meaningful balance-sheet repair. Still, the business model deserves a permanently higher accounting burden than a simple manufacturer. Estimated costs to complete determine revenue and profit; claims and change orders require probability judgments; joint ventures complicate ownership; and reserve/permit purchase accounting adds valuation assumptions.

“Other costs, net” also recur more often than a one-time label suggests: $101 million in 2021, $24 million in 2022, $50 million in 2023, $40 million in 2024 and $41 million in 2025. The composition changes—litigation, SEC resolution, acquisition/integration and former-officer defense—but the aggregate shows that a clean adjusted base is rarely costless. Normalization should remove truly discrete items while retaining a reasonable ongoing restructuring/integration burden.

Property disposal gains were $66 million in 2021, $13 million in 2022, $28 million in 2023, $9 million in 2024 and $20 million in 2025. They are real cash but not core construction earnings and should not be capitalized at a recurring operating multiple. Similarly, debt-extinguishment and conversion charges are non-operating, yet repeated capital-structure resets signal real financing decisions.

Balance-sheet analysis should separate liquidity from control. Consolidated JV cash may not be freely distributed, unconsolidated JV cash is absent, and surety providers rely on Granite’s credit. Restricted or jointly controlled cash cannot provide the same downside protection as parent unrestricted cash.

The best quarterly checklist is therefore unusually concrete: gross margin excluding disclosed net favorable EAC revisions; contract liabilities and AP contribution to CFO; executed versus other-award CAP; deal-level acquired contribution and carrying value; parent-available cash; and actual shares after settlement. Reported EPS alone is the least useful summary during 2026.

Balance sheet, dilution and liquidity

At June 30, cash was $877.1 million and marketable securities $54.4 million. However, $154.5 million sat in consolidated construction JVs; Granite’s economic share was $95.9 million and distributions require joint control. Current plus long-term debt carrying value was $1.559 billion, and a separate $630.5 million conversion derivative represented the remaining 2028 notes. Granite equity fell to $752.7 million after the H1 GAAP loss and a $178.8 million equity charge on convertible repurchase.

Actual shares were 43.764 million. The H1 GAAP loss makes weighted diluted shares antidilutive and therefore economically too low for normalized-profit valuation; 2025’s 53.132 million diluted count, conversely, reflected if-converted accounting without cleanly incorporating capped-call economics. The only honest approach before settlement is actual basic shares plus a transparent convert/RSU sensitivity.

SBC rose from $10.5 million in 2023 to $39.2 million in 2025. It is not yet software-like dilution, but the acceleration should be monitored. Open-market buybacks were modest relative to the acquisition program and paused in H1 2026.

Verdict: Economics demonstrably improved with scale and better bidding, but financial quality is medium, not high. POC catch-ups, reversible contractor float, acquired growth and unsettled capital structure make headline EBITDA and FCF less durable than they appear.

7. Capital Allocation

Granite’s strategy is now capital allocation. From 2023 through April 2026 it announced approximately $1.36 billion of acquisitions:

The rationale is coherent. Warren’s purchase allocation included $275.3 million of mineral reserves, $142.5 million of goodwill and $46.8 million of intangibles; the deal adds real scarce assets rather than a vague revenue synergy. But neither EBITDA multiples nor deal-level ROIC are disclosed. Warren contributed $129.7 million of 2025 revenue and $21.1 million of gross profit; Kenny Seng contributed $29.8 million and $3.5 million in Q2 2026. Those contributions show scale, not value creation.

The timing deserves skepticism. Granite bought aggressively during record public demand, financed 2025 deals with a $600 million term loan, $75 million delayed draw, revolver borrowings and cash, and plans further acquisitions. Marathon’s capital-cycle warning applies: asset growth and M&A at a favorable industry point can transfer economics to sellers. The right scorecard is post-deal ROIC including goodwill, reserve step-ups, integration cost and incremental interest—not adjusted EBITDA growth.

Evaluating the roll-up as an owner

Each deal should clear three hurdles. First, the target must own a defensible local position—permitted reserves, asphalt density, customers or crews that cannot be assembled cheaply. Second, cost synergies must be observable: lower purchased-material/haul expense, better plant utilization, reduced duplicate overhead or higher third-party throughput. Revenue “cross-sell” is less reliable because public jobs are rebid. Third, the purchase price must leave those synergies with Granite shareholders rather than the seller.

Public disclosure currently clears only the first hurdle directionally. Warren’s reserve and permit allocations show real assets; Papich, Cinderlite and Kenny Seng build clusters. Contribution data confirms revenue and gross profit. But no purchase EBITDA, synergy realization, local margin or deal ROIC is available. Goodwill and mineral step-ups are bookkeeping records of price paid, not evidence of returns.

Leverage changes the hurdle. The 6.375% notes create a visible pretax cost, while integration management time and acquisition expense add less visible costs. A deal that looks accretive before interest can dilute owner value after the full capital stack. With public demand strong, the board should require returns above a normalized, not current-peak, earnings base.

The best positive signal would be restraint. If attractive permitted assets are unavailable at the right price, Granite should allow leverage to fall and return capital. Management’s suggestion that annual M&A could range from $300 million to $800 million risks turning a strategic option into a growth commitment. A fixed acquisition quota would be the wrong response to a cyclical supply of targets.

The falsification timeline is short enough to monitor. By 2027, full-year Warren/Papich/Cinderlite/Kenny Seng earnings will be visible; purchase-accounting noise will be better understood; and the acquired capital will have sat in the denominator longer. Consolidated ROIC, Materials cash margin and net leverage should improve together. EBITDA growth without those three outcomes would validate the capital-cycle bear case.

Convertible reset

Granite exchanged $100 million principal of 3.75% 2028 converts in March 2026 for $289.7 million cash, with no shares issued, and received $56.7 million from a partial capped-call unwind. It then called the remaining $273.7 million and issued $600 million of 6.375% senior notes due 2034, whose annual coupon is $38.25 million. The Q2 income statement recorded a $359.7 million non-operating convertible loss, mostly fair-value/bifurcation accounting rather than operating deterioration.

The accounting charge should be normalized; the economics should not. Granite exchanged a low coupon and dilution exposure for substantial cash use and a higher fixed coupon. The August 4, 2026 8-K scheduled capped-call funding for August 11 and note settlement for August 12, after this report’s cutoff. Final cash proceeds, settlement cash and any shares were unknown, so no post-settlement leverage conclusion is stated.

Dividends, repurchases and incentives

The annual dividend has remained $0.52 per share, approximately $23 million of cash. The $300 million repurchase authorization had $157.6 million remaining at Q2, but H1 purchases were only employee tax withholding. Acquisitions and conversion settlement clearly outrank returns of capital.

The 2025 annual incentive was 80% adjusted EBITDA and 20% adjusted OCF as a percentage of revenue, modified for safety and individual performance. The OCF payout was 175% because actual margin reached 11.0%. That metric can reward billings/payables float. Annual measures also exclude acquired businesses and preapproved acquisition/integration, SBC and other costs in the acquisition year. Long-term compensation is better: 50% relative TSR, 25% three-year adjusted RONA and 25% time-based RSUs. Three-year RONA was 12.7% versus 11.8% target.

Insider activity is distribution-heavy but nuanced. Across the five-year Form 4 corpus, open-market buys totaled only 2,400 shares/approximately $0.26 million versus sales of 184,627 shares/approximately $18.2 million. Major CEO sales were disclosed as 10b5-1. Two officers sold approximately $2.0 million combined near $141 in June 2026 without a 10b5-1 box; that identifies transaction form, not motive. Small director/officer buys in 2025–2026 are supportive but not decisive.

Verdict: Strategic direction is credible; price discipline and returns remain unproven. Compensation includes a useful RONA guardrail but rewards adjusted EBITDA and working-capital-sensitive OCF while excluding acquisition costs. Capital allocation merits a provisional, not positive, grade.

8. Changes and Headwinds — Last Two Years

The most important change is strategic: Granite moved from legacy-project repair toward a vertically integrated regional roll-up. D&B, Warren, Papich, Cinderlite and Kenny Seng expanded materials capacity and home-market density. This is economically more attractive than pursuing undifferentiated mega-project volume, but it changes the risk from project selection alone to acquisition price, integration and leverage.

The second change is operating leverage. Construction gross margin rose from 10.9% in 2023 to 14.4% in 2024 and 15.7% in 2025. CAP rose to $7.4 billion by Q2 2026. The improvement survived strong revenue growth in Construction during Q2, but favorable EAC/claim revisions remain a meaningful contributor.

The third is a broader growth mix. Federal tactical infrastructure contributed $624 million of CAP; water, rail/transit and mission-critical/private site development are increasingly relevant. These adjacencies diversify DOT exposure but introduce their own customer, schedule and project-risk profiles.

The fourth is capital-structure simplification with economic cost. The 2028 converts created volatile GAAP charges and potential dilution. The exchange/call should remove that complexity after settlement, but the 6.375% notes raise recurring cash interest and settlement consumes liquidity.

Headwinds are visible in Materials. Q2 GAAP margin fell 800 basis points despite 32% revenue growth, driven by management-attributed weather, quarry development and purchase accounting. This is a useful reminder that the materials “moat” does not immunize utilization, development cost or deal accounting.

Policy is the next external inflection. IIJA highway authority expires September 30, 2026. Existing appropriations do not vanish, but a delayed or smaller successor can slow lettings. Granite also disclosed a rare approximately $300 million public-highway project cancellation in Q1 2026, proving that CAP can disappear even in a strong market.

Verdict: The last two years strengthened the operating thesis and weakened the balance-sheet/capital-cycle margin for error. The company is better; the proof burden has shifted from turnaround to durable returns.

9. Risk Analysis

Catastrophic loss is plausible at the project/JV level through a severe cost overrun, legal claim, safety failure or environmental liability. A total equity loss is low probability because Granite has diversified work, hard assets and substantial liquidity, but it is not impossible under a combined scenario of settlement cash drain, project losses, funding contraction and covenant/liquidity stress. Historical 10-year factor data includes an 84.7% maximum drawdown, which is not a forecast but demonstrates realized tail risk.

The most dangerous risk correlation is procyclical: strong public demand lifts CAP and encourages M&A; the balance sheet grows; labor and equipment tighten; fixed-price margins then weaken as demand slows. That sequence can make reported backlog and acquired EBITDA look safest just before returns mean-revert.

Verdict: Risk is medium-high. Diversification and hard assets reduce existential risk, but POC estimates, working-capital float, M&A and public funding all load on the same infrastructure cycle.

10. Valuation Discussion — Embedded Expectations

Capitalization at the information cutoff

The 2026-08-07 close of $128.32 multiplied by 43.764 million actual shares gives $5.617 billion of equity value. At June 30, debt carrying value was $1.559 billion, the embedded convert derivative was $630.5 million, NCI was $52.6 million, and available cash/securities were approximately $777–$873 million depending on treatment of JV cash. The resulting pre-settlement accounting bridge is approximately $6.99–$7.08 billion EV.

That bridge is not a clean post-settlement EV. June debt contains a conversion discount; the derivative is marked to conversion economics; capped-call proceeds and final settlement were pending. A mechanical post-settlement sensitivity remains broadly around $6.8–$7.2 billion, but no precise settled value is knowable by the August 9 cutoff.

Peer context

Third-party screening using August 7 prices and latest reported balance-sheet data places ROAD near 19x TTM EBITDA, STRL near 24x, PRIM near 18x on depressed earnings, DY near 18x and EME near 17x. Pure aggregates peers VMC/MLM screen around 18–19x. These are orientation, not filing-reconciled apples-to-apples comps: peer share/balance dates and EBITDA definitions differ. The premium median is also earned by better mix, higher margins or stronger balance sheets. Granite’s discount is therefore not self-evidently mispricing.

Earnings power and asset value

A sustainable EBIT range of approximately $440–$500 million recognizes acquisition annualization and operating improvement while haircutting management add-backs, claim benefits and peak demand. At a 25% cash tax rate, NOPAT is $330–$375 million. Capitalized at 9–10%, no-growth EPV is approximately $3.3–$4.2 billion of enterprise value.

Asset reproduction is less precise. June net operating working capital excluding cash/securities is only roughly $0.1 billion, net PP&E is about $1.3 billion, and JV/affiliate investments are approximately $0.23 billion. Permitted quarries, mineral reserves and organizational capability have replacement value above book, but Granite does not disclose the local economics needed to quantify that premium. The defensible conclusion is qualitative: EPV appears above identifiable operating book assets, consistent with a narrow franchise; market EV stands materially above EPV, so profitable growth supplies a large part of value.

The EPV diagnostic deliberately sits below management’s guide. Guide-midpoint EBITDA of roughly $689 million reflects acquired annualization and management adjustments in strong demand. Sustainable EBIT of $440–$500 million credits structural improvement while refusing to capitalize every add-back or favorable estimate. If the guide proves durable on a GAAP and clean-cash basis, EPV rises; if it requires continual acquisitions and advance billing, EPV overstates standalone power.

Reading the multiples

The approximate 10x guided adjusted EBITDA multiple is the bull’s cleanest statistic and the easiest to misuse. The numerator is unsettled until the capped-call/convert closes, while the denominator excludes costs and includes acquired contribution. The 14x TTM EBITDA figure is less optically cheap but uses a more conservative earnings base. Neither says much without a durability view.

ROAD’s higher screening multiple shows what investors pay for an integrated road/materials roll-up; it does not prove Granite deserves convergence. ROAD differs in regional density, asphalt mix and growth. EME, STRL and DY earn premiums for specialty/private/service exposure or balance-sheet quality. VMC and MLM earn scarcity multiples because rock dominates economics. Granite’s justified multiple should be a blend: mostly contractor until Materials becomes a larger, demonstrably price-over-volume share of profit.

Peer figures are third-party screening data, not a fully filing-reconciled comp sheet. Prices were cross-checked at August 7, but share, debt and EBITDA dates differ by company. Their purpose is to reject false analogies and locate a broad range—not to produce a target value.

Valuation sensitivities

Margin dominates value. A 100-basis-point change on $5.4 billion of revenue is $54 million of EBITDA before tax and reinvestment, comparable in present value to several years of modest revenue growth. Ex-claim Construction margin and Materials cash margin matter more than a small change in CAP growth.

Acquisition reinvestment is next. The reverse DCF assumes cash margins after necessary reinvestment. If Granite must spend $300–$800 million annually on M&A to sustain high growth, that cash outflow belongs in owner economics. Purchased EBITDA cannot be treated as organic free growth.

Capital structure is the third sensitivity. Settlement cash may reduce liquidity; capped-call proceeds offset part of conversion value; limited shares may issue. Because the flows occur after cutoff, the report uses a pre-settlement range. Any update should replace the full bridge rather than mechanically changing one debt line.

Reverse DCF

Using $5.4 billion starting revenue, a 9.5% WACC, 2.5% terminal growth and a ten-year explicit period, approximately $7.0 billion of EV requires one of three broad combinations:

• about an 8% normalized unlevered FCF margin with only 3–4% revenue CAGR;
• about a 6% margin with 7–8% CAGR; or
• about a 5% margin with roughly 10% CAGR.

The model assumes the cash margin already includes maintenance/reinvestment; separately adding acquisitions would make the hurdle higher. The market is therefore underwriting persistence, not merely the 2026 guide: mid-teens Construction gross margin, Materials cash margin around or above 30%, supportive funding, value-creating M&A and cash conversion after contractor float.

These are operating scenarios, not target values. The bear assumes reauthorization delays, ex-claim margin mean reversion, weak Materials utilization and float reversal. The base assumes high-single-digit organic growth, acquisition annualization and some margin persistence. The bull assumes >10% organic growth through 2027, >30% Materials cash margin and ROIC above 12–14% on the full acquired-capital base.

Valuation verdict: The stock screens below premium contractors on TTM EBITDA and far below pure aggregates on sales, but no-growth EPV supports only part of current EV. The market prices a durable margin regime and successful reinvestment. The central question is not whether 2026 earnings are strong; it is whether Granite can earn franchise returns on purchased assets after cash-flow float reverses.

11. Variant Perception

Consensus belief

The consensus story is that Granite has completed a portfolio turnaround, entered a multi-year public infrastructure upcycle, and is using acquisitions to become a vertically integrated materials-and-construction compounder. Record CAP, raised guidance and margin expansion support that view. The stock’s 249% adjusted five-year return shows that this is no longer a hidden turnaround.

Strongest bull case

The bull argues the market still underestimates local scarcity. Granite’s 2.081 billion tons of reserves/resources, permits and asphalt network can replicate ROAD’s cluster economics across multiple states. Public funding provides demand duration; home-market density improves hauling, utilization and bid competitiveness; Materials profit share rises; and the legacy-project drag disappears. If Construction margins hold 15–16%, Materials cash margin exceeds 30%, and acquired-capital ROIC rises into the low-to-mid teens, current earnings are the start of a compounder rather than a peak.

Strongest bear case

The bear argues the market mistakes a favorable cycle, claim settlements and customer financing for a franchise. Construction remains rebid and fixed-price; 31.8% of CAP is softer than an executed obligation; Q2 favorable EAC revisions were 5.9% of gross profit; and H1 cash flow depended on billings and payables. Granite is buying assets during record public spending without disclosing acquisition multiples. Reauthorization, labor and bidding pressure can expose lower normalized margins just as higher interest and integration costs become fixed.

Factor-positioning read

At August 7, price was 1.8% below the 50-day EMA but 3.5% above the 200-day. Raw returns were -9.1% over three months, +2.2% over six months and +38.1% over twelve months. The FactorsToday All-Factors model—observation dated 2026-07-31, retrieved August 9—loaded on Market +0.999, Infrastructure +0.592, Small Size +0.464 and Industrials +0.332; Momentum was only +0.040 and Quality was zeroed. R² was 40.1%, with 26.6% specific volatility. This is infrastructure/small-cap industrial exposure with a large idiosyncratic component, not a pure crowded-momentum factor bet (FactorsToday GVA loadings).

Short interest was 4.052 million shares, or 9.35% of float, on July 15 (MarketBeat). The positioning corroborates an active valuation/funding debate but says nothing about which side is right.

Load-bearing assumptions

1. Construction margin durability. Does reported 15–16% survive with fewer favorable revisions?
2. Materials unit economics. Can price/cash GP per ton rise through weak volumes?
3. Acquired-capital returns. Do $1.36 billion of purchases earn above cost including goodwill and reserve step-ups?
4. Cash conversion. Is FCF positive after contract-liability and AP timing normalize?
5. Funding duration. Does successor legislation preserve real highway formula spending?

Verdict: The differentiated view is not “infrastructure demand is weak”; it is that demand can remain strong while shareholder returns disappoint if purchased growth and contractor float are overcapitalized.

12. Fact vs. Interpretation

13. Open Questions

1. What percentage of Construction materials demand is internally supplied in each home market?
2. What are local quarry/asphalt shares, average haul radii, third-party tons, prices and cash gross profit per ton?
3. What margin gap exists between integrated and non-integrated Construction markets?
4. What are CAP win rates, bid spreads and expected margins by executed versus other-awards categories?
5. How much 2025–2026 Construction margin came from claims and favorable EAC changes versus bid quality/productivity?
6. What are full-cost post-synergy ROIC results for LRC/MSG, D&B, Warren, Papich, Cinderlite and Kenny Seng?
7. What acquisition multiples and hurdle rates does the board approve?
8. What is normalized OCF after removing retention/claim collections and contract-liability/AP timing?
9. What were final August 11–12 capped-call proceeds, convert cash settlement and share delivery?
10. How will annual incentives prevent advance billings and acquisition exclusions from overstating performance?
11. How much of the 2.081 billion tons is economically permitted for near-term extraction versus long-dated resource?
12. What successor highway authorization does Granite assume beyond September 2026?

14. What Must Be True

Bull case

• Local Materials moat proves measurable: Materials cash gross profit grows faster than revenue over two years, cash margin recovers above 30%, and price rises through flat/down same-store tons. Falsification: cash margin remains below the high-20s or pricing requires volume growth.
• Vertical integration improves Construction: integrated markets sustain at least a 200-basis-point margin advantage without higher loss provisions. Falsification: no disclosed/observable gap, or it disappears after acquisition accounting normalizes.
• Acquisitions create value: consolidated ROIC including goodwill/mineral step-ups rises above 12% by 2027 while leverage declines. Falsification: EBITDA rises but ROIC stagnates or goodwill/reserves impair.
• Funding persists: successor authorization avoids a real formula-funding cut and executed unearned revenue continues to grow. Falsification: letting delays reduce executed work and book-to-burn falls below 1x.

Bear case

• Margin is cycle/claims: ex-favorable revisions, Construction gross margin falls below 14% for two consecutive quarters. Falsification: 15–16% holds through slower volume with few catch-ups.
• CAP overstates visibility: other awards grow faster than executed work, options lapse or cancellations accelerate. Falsification: executed unearned revenue grows and converts on schedule.
• Materials lacks pricing power: same-store price and cash GP per ton fall in weak volume. Falsification: cash GP per ton rises despite lower tons.
• Negative capital cycle arrives: labor/equipment/M&A costs rise, bid spreads narrow and post-deal ROIC declines. Falsification: Granite walks away from expensive deals and returns rise as capex/depreciation normalizes.
• Cash is rented: billings/AP reverse and multi-year FCF settles below 4% of revenue. Falsification: FCF remains above 6% after a flat contract-liability year.

15. Source Appendix

The detailed source inventory, public-source inventory, document dates, access dates, URL checks and evidence limitations are in Appendix B of the combined report. Primary analytical anchors are the FY2025 10-K, Q2 2026 10-Q, Q2 2026 earnings exhibit, 2026 proxy, and August 4 convert-settlement 8-K. Management calls are treated as hypotheses and reconciled to filings.

APPENDIX A — Standard Diligence Questionnaire

Company: Granite Construction Incorporated (NYSE: GVA)

As of: 2026-08-09. Facts come primarily from Granite’s FY2025 10-K, Q2 2026 10-Q, Q2 earnings exhibit, and 2026 proxy.

General

What thoughtful questions have other investors asked?

Recent calls focus on five questions: how much CAP is executed versus optional; whether 15–16% Construction gross margin is sustainable after claims; why Materials margin fell despite acquisitions; how much 2027 growth is organic; and how management evaluates acquisition returns. The most useful unanswered versions are more specific: disclose same-store materials price/tons/cash GP per ton, integrated versus non-integrated market margin, bid spreads, deal-level ROIC, and cash flow excluding advance billing.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low?

Interpretation: Earnings are above mid-cycle, though not necessarily at the peak. Public demand and CAP are records, margins recovered sharply, and M&A is expansionary. Construction’s capital cycle is mid-to-late expansion; permitted aggregates remain supply constrained.

Are results driven by external conditions or internal actions?

Both. IIJA/state programs and private site demand support volume. Portfolio repair, bid discipline and acquisitions lifted mix/margin. Q2 net favorable EAC revisions and contractor float also improved reported earnings/cash. The thesis requires internal gains to persist after external growth slows.

How stable are revenues?

Public funding makes demand more stable than private construction, and $5.056 billion of Q2 CAP was executed unearned revenue. But projects are non-recurring, cancelable/modifiable and execution sensitive; $2.362 billion of CAP was softer other awards. Revenue stability is medium, not subscription-like.

What is the outlook for products/services?

Highways, bridges, water, transit and federal tactical infrastructure have strong near-term pipelines. Management raised 2026 revenue guidance to $5.3–$5.5 billion and expects >10% 2027 organic growth; those are management hypotheses. Reauthorization after September 2026 and conversion of CAP into profitable work are the external/internal tests.

How large is the market?

June 2026 public construction spending was a $544.1 billion SAAR and highway/street $150.9 billion (Census). These are reference pools, not Granite SAM. Materials markets are local haul-radius markets; public disclosure does not permit a defensible local SOM calculation.

Business Quality & Competitive Moat

Is the industry becoming more or less competitive?

Interpretation: Demand is strong, but contracting supply discipline is deteriorating as labor, equipment and acquisition capital enter. Quarry supply remains constrained by permits. Competition is therefore worsening faster in contracting than in aggregates.

How profitable is Granite?

FY2025 gross/operating margin was 16.1%/6.4%; directional analyst ROIC was approximately 13.9%, versus 5.8% in 2023. Q2 2026 Construction margin was 16.5%, but Materials fell to 16.1%. Returns improved but have only two strong years and do not yet prove through-cycle advantage.

How profitable is the industry; what are barriers?

Civil contracting is fragmented and generally earns modest margins; bonding, expertise and labor limit mega-project bidders but do not create pricing power. Permitted aggregates near demand can earn superior returns because freight and land-use regulation deter entry. Asphalt is between the two.

Can the business be easily understood?

The operating activities are understandable. Accounting is not simple: POC/EAC estimates, claims, JVs, NCI, CAP categories and convert derivatives require careful reconciliation.

Can foreign low-cost labor undermine it?

Not directly. Construction and hauling are local. Foreign inputs can affect steel, equipment and liquid asphalt; tariffs/Buy America rules affect sourcing. Domestic skilled-labor shortages are the larger risk.

Do brands matter?

Reputation matters for prequalification and surety confidence, but public procurement limits brand pricing. Permits, reserves, proximity and execution history matter more.

What is the nature of competition?

Local contractors compete on bid price and execution; national/specialty firms compete on complex delivery and bonding. Materials competition is local and delivered-cost based. Granite has multiple competitors in every market.

What are switching costs?

Low between projects. During a live project replacement is disruptive and bonded performance creates lock-in, but the next award can go elsewhere. Materials customers face moderate proximity captivity if alternative quarries are far away.

What is the moat verdict?

Narrow local materials moat embedded in a largely unmoated contractor. No network effect. Scale is a qualification advantage; permits/reserves create the defensible cost/supply mechanism.

Financial Condition & Balance Sheet

Are assets understated?

Permitted reserves and local operating networks may have reproduction value above book. But acquisition accounting already recognized large mineral-reserve and goodwill step-ups. Without site-level economics, no hidden-value amount is defensible.

Are there off-balance-sheet liabilities?

Surety indemnities, JV obligations, environmental/reclamation exposure, claims and cancelable awards matter. Unconsolidated JVs also separate cash/obligations from Granite’s balance sheet. Review guarantees and commitments in the 10-K notes.

How conservative is accounting?

GAAP POC is appropriate but estimate intensive. Q2 net favorable EAC revisions equaled 5.9% of gross profit. Management cannot provide meaningful sensitivity. Current controls are effective, but analytical conservatism requires ex-revision margin and cash-flow reconciliation.

How capex hungry is it?

Organic capex is moderate—$138 million in 2025 versus $162 million D&A—but this understates reinvestment because M&A is the main asset-growth channel. Management guides $140–$160 million 2026 capex plus continued acquisitions.

What is liquidity and leverage?

At Q2: $877.1 million cash, $54.4 million securities, $1.559 billion debt carrying value and $630.5 million convert derivative. $154.5 million cash was in consolidated JVs. Final convert/capped-call settlement was pending after cutoff. Use a pre-settlement range, not a precise net-debt figure.

Capital Allocation & Management

How much FCF is generated and how is it used?

Reported FCF was $331 million in 2025 and $86 million in H1 2026. H1 billings in excess rose $115 million and AP supplied $170 million, so reported cash is not normalized owner earnings. Cash use prioritizes acquisitions, capex and convert settlement; dividend cash is ~ $23 million annually and buybacks are secondary.

What is management’s philosophy?

Build vertically integrated home-market clusters, maintain bid discipline and compound through materials acquisitions. The logic is good; the lack of disclosed deal multiples/ROIC prevents proof of discipline.

Significant acquisitions?

Approximately $1.36 billion since 2023, led by Warren Paving ($540 million), LRC/MSG (~$278 million), Papich ($170 million), Kenny Seng ($164 million), D&B and Cinderlite. No purchase EBITDA multiple or post-deal ROIC is disclosed.

Is Granite buying back shares?

The authorization remains, but no program repurchases occurred in H1 2026; small purchases were employee tax withholding. Acquisitions and debt/capped-call actions dominate.

Is it issuing large amounts to insiders?

SBC rose to $39.2 million in 2025 from $10.5 million in 2023. This is material but not extreme relative to revenue. Settlement/RSU dilution must be tracked on actual shares because GAAP diluted counts are distorted by the loss/convert.

Compensation policy and motivations?

Annual pay is 80% adjusted EBITDA and 20% OCF/revenue, with safety/individual modifiers. LTI is 50% relative TSR, 25% adjusted RONA and 25% RSUs. RONA is constructive; OCF can reward advance billing and annual metrics exclude acquisitions/integration. CEO 2025 compensation was $9.861 million. Form 4 history is distribution-heavy, though major CEO sales were 10b5-1 and three modest open-market buyers appeared.

Is management aligned?

Mixed. TSR/RONA and ownership guidelines help. Acquisition exclusions, working-capital-sensitive OCF, RSUs counting toward ownership and limited open-market buying weaken the alignment signal.

Valuation & Market Data

ADR, MLP or K-1?

No. GVA is NYSE-listed U.S. common stock; it is not an ADR, MLP or K-1 issuer.

Dividend policy?

$0.13 quarterly/$0.52 annually, unchanged, costing approximately $23 million per year.

How profitable is the business?

2025 operating income was $282 million and directional ROIC about 13.9%. TTM GAAP earnings are distorted by the Q2 convertible charge. Management’s adjusted EBITDA guide is useful but excludes real strategy costs.

Is net income diverging from CFO?

Yes. H1 GAAP net loss primarily reflects convertible fair-value charges, while CFO is positive and working-capital aided. Both require normalization: remove conversion accounting from earnings, but also remove temporary billings/AP support from cash.

What is priced in?

At $128.32, pre-settlement EV is approximately $6.99–$7.08 billion, 14.0–14.2x TTM EBITDA and about 10.1–10.3x guided adjusted EBITDA. A reverse DCF requires roughly 6% normalized FCF margin and 7–8% revenue CAGR for a decade, or an equivalent tradeoff. That embeds durable improvement and productive acquisitions.

Risks & Downside

What could cause a decline?

Construction margin below 14%; unfavorable EAC revisions; contract-liability/AP reversal; CAP cancellations; weak reauthorization; Materials price/margin disappointment; acquisition impairment/low ROIC; settlement cash/dilution surprise; labor inflation; safety/environmental events; or a higher cost of capital.

Risk of catastrophic loss?

Medium at an individual project/JV and low at the company level. Fixed-price overruns, environmental liabilities, litigation or a major safety event can create large losses. Diversification, hard assets and liquidity reduce but do not eliminate the risk.

Chance of total loss?

Low but not zero. It requires several correlated failures—project losses, funding contraction, acquisition impairment, settlement/liquidity strain and surety/covenant pressure. Historical price drawdowns show equity can lose most of its value without the company failing.

Recent News & Events

Has the environment changed?

Yes. Q2 revenue/CAP set records and guidance rose, while Materials margin weakened. IIJA authorization approaches expiration. The July downgrade focused attention on valuation/funding. Convert settlement is changing capital structure after the report cutoff.

Significant acquisitions?

Kenny Seng closed April 2026 for $164.1 million; Warren/Papich/Cinderlite closed in 2025. Management indicated more M&A is possible. Treat all forward amounts as management plans, not facts.

Accounting policy changes?

No thesis-changing revenue-policy change. The convert bifurcation/remeasurement caused extraordinary 2026 non-operating volatility. POC remains the critical accounting judgment.

New markets, facilities or management?

The company is expanding Southeast/Western home markets, federal tactical infrastructure and mission-critical site development. Kyle Larkin remains CEO. The strategic change is acquisition-led vertical integration rather than a leadership reset.

Diligence Conclusion

The decision hinges on evidence Granite does not yet publish. Request same-store materials unit economics, local share/haul radii, integrated-market margin, ex-EAC Construction margin, deal-level ROIC and normalized cash conversion. Until then, the safest conclusion is that Granite’s operating turnaround is proven, its consolidated moat is narrow, and its reinvestment economics remain an open question.

APPENDIX B — Source Appendix

Company: Granite Construction Incorporated (NYSE: GVA)
Research date / accessed: 2026-08-09 unless otherwise stated
Source policy: Public sources are listed below. Dynamic webpages are identified. Management releases and commentary support what management reported or asserted, not independent validation of forward claims.

Public sources

Granite Construction and SEC filings

1. Granite Construction Incorporated, Annual Report on Form 10-K for the year ended December 31, 2025. Granite Construction Incorporated; filed 2026-02-13; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000004/gva-20251231.htm
   Used for the business and segment descriptions; FY2025 revenue and gross profit; public/private mix; CAP definition, composition and cancellation risk; contract-type mix; quarry properties, reserves/resources, acreage, ownership/lease mix and production; bonding; regulation; acquisitions and purchase accounting; revisions in estimates; and risk disclosures.

2. Granite Construction Incorporated, Quarterly Report on Form 10-Q for the quarter ended March 31, 2026. Granite Construction Incorporated; filed 2026-04-30; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000018/gva-20260331.htm
   Used for Q1 2026 reported results, CAP additions and cancellation, acquisition updates, and current balance-sheet/cash-flow evidence.

3. Granite Construction Incorporated, Quarterly Report on Form 10-Q for the quarter ended June 30, 2026. Granite Construction Incorporated; filed 2026-07-30; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000029/gva-20260630.htm
   Used for Q2 2026 segment results, acquired-company contributions, CAP, material-weather/quarry-development effects, purchase accounting, debt and cash-flow evidence. This filing is the preferred source where it overlaps with the earnings release.

4. Granite Reports Second Quarter 2026 Results. Granite Construction Incorporated; published 2026-07-30; accessed 2026-08-09; issuer earnings release / Form 8-K exhibit. https://investor.graniteconstruction.com/news-and-events/press-releases/granite-reports-second-quarter-2026-results
   Used for reported Q2 2026 revenue, segment tables, adjusted measures and reconciliations, CAP and tactical-infrastructure disclosure, acquisition contribution, and FY2026 guidance. Forward-looking statements and causal explanations are management claims.

5. Granite Construction Incorporated, Current Report on Form 8-K reporting Q2 2026 results. Granite Construction Incorporated; filed 2026-07-30; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000027/gva-20260730.htm
   Used as the filing record for the Q2 earnings release furnished under Item 2.02.

6. SEC XBRL Company Facts: Granite Construction Incorporated (CIK 0000861459). U.S. Securities and Exchange Commission; continuously updated, retrieved 2026-08-09; regulatory structured-data API. https://data.sec.gov/api/xbrl/companyfacts/CIK0000861459.json
   Used to resolve the legal registrant name/CIK and to cross-check reported financial series. Period context and form type must be checked before using any datapoint.

7. Board of Directors / Corporate Governance. Granite Construction Incorporated; no publication date shown, current page accessed 2026-08-09; issuer governance webpage. https://investor.graniteconstruction.com/corporate-governance/board-of-directors
   Used only to identify Kyle T. Larkin as President and Chief Executive Officer. The URL redirects to Granite’s corporate-governance landing page but resolved successfully when tested on 2026-08-09.

Historical financials, governance, financing and insider filings

1. Annual Report on Form 10-K for the year ended December 31, 2021. Granite Construction Incorporated; filed 2022-02-28; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000143774922004482/gva20211231_10k.htm
   Used only for the five-year historical financial, segment, cash-flow, capital-allocation and balance-sheet series. Older than 18 months and not evidence of current conditions.

2. Annual Report on Form 10-K for the year ended December 31, 2022. Granite Construction Incorporated; filed 2023-02-21; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000143774923004014/gva20221231_10k.htm
   Used only for the five-year historical financial, segment, cash-flow, capital-allocation and balance-sheet series.

3. Annual Report on Form 10-K for the year ended December 31, 2023. Granite Construction Incorporated; filed 2024-02-23; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145924000006/gva-20231231.htm
   Used only for the five-year historical financial, segment, cash-flow, capital-allocation and balance-sheet series.

4. Annual Report on Form 10-K for the year ended December 31, 2024. Granite Construction Incorporated; filed 2025-02-14; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145925000007/gva-20241231.htm
   Used for historical comparisons and 2024 capital allocation/acquisition evidence.

5. Definitive Proxy Statement for the 2026 Annual Meeting. Granite Construction Incorporated; filed 2026-04-23; accessed 2026-08-09; SEC proxy filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000012/gva-20260423.htm
   Used for executive pay, annual- and long-term-incentive design, adjusted EBITDA/OCF/RONA/relative-TSR goals and outcomes, ownership guidelines, CEO compensation and governance.

6. Current Report on Form 8-K — convertible-note exchange agreements. Granite Construction Incorporated; filed 2026-02-18; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000006/gva-20260218.htm
   Used for the $100 million principal exchange of the 3.75% 2028 convertible notes. Cash paid, accounting effects and capped-call proceeds are reconciled to the Q2 2026 10-Q.

7. Current Report on Form 8-K — 6.375% Senior Notes due 2034. Granite Construction Incorporated; filed 2026-06-02; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000143774926019166/gva20260601_8k.htm
   Used for the $600 million senior-notes issuance and intended use of proceeds. Annual cash coupon is an analyst calculation from principal times 6.375%.

8. Current Report on Form 8-K — capped-call unwind and pending 2028-note conversion settlement. Granite Construction Incorporated; event/filing date 2026-08-04; accessed 2026-08-09; SEC regulatory filing. https://www.sec.gov/Archives/edgar/data/861459/000086145926000032/gva-20260804.htm
   States that capped-call cash settlement was expected on/about 2026-08-11 and note conversions on 2026-08-12, after the report cutoff. Settlement amounts were not yet disclosed. Current enterprise value cannot treat the transactions as completed or use a June debt carrying value without qualification.

9. Statement of Changes in Beneficial Ownership — Alan Krusi. U.S. SEC / Granite reporting person; filed 2025-05-12; accessed 2026-08-09; Form 4. https://www.sec.gov/Archives/edgar/data/861459/000112760225013979/xslF345X05/form4.xml

10. Statement of Changes in Beneficial Ownership — John Timothy Romer. U.S. SEC / Granite reporting person; filed 2026-06-15; accessed 2026-08-09; Form 4. https://www.sec.gov/Archives/edgar/data/861459/000208694726000006/xslF345X06/wk-form4_1781556069.xml

11. Statement of Changes in Beneficial Ownership — Carlos Hernandez. U.S. SEC / Granite reporting person; filed 2026-08-06; accessed 2026-08-09; Form 4. https://www.sec.gov/Archives/edgar/data/861459/000117938626000006/xslF345X06/wk-form4_1786048435.xml

12. Statement of Changes in Beneficial Ownership — Kyle T. Larkin. U.S. SEC / Granite reporting person; filed 2026-03-30; accessed 2026-08-09; Form 4. https://www.sec.gov/Archives/edgar/data/861459/000171995026000008/xslF345X06/wk-form4_1774909854.xml
    Used for a reported sale explicitly disclosed as pursuant to a Rule 10b5-1 trading plan.

13. Statements of Changes in Beneficial Ownership — Michael Tatusko and Bradley Williams. U.S. SEC / Granite reporting persons; filed 2026-06-08; accessed 2026-08-09; Forms 4. https://www.sec.gov/Archives/edgar/data/861459/000179846226000006/xslF345X06/wk-form4_1780951441.xml ; https://www.sec.gov/Archives/edgar/data/861459/000193495626000006/xslF345X06/wk-form4_1780951221.xml
    Used for the two June 2026 open-market officer sales. “Discretionary” means the forms do not mark the transactions as pursuant to a Rule 10b5-1 plan; it does not establish motive.

Government and industry sources

8. Monthly Construction Spending, June 2026. U.S. Census Bureau; released 2026-08-03; accessed 2026-08-09; official statistical release / dynamic current-release landing page. https://www.census.gov/construction/c30/current/index.html
   Used for June 2026 seasonally adjusted annual rates for total, public, and highway-and-street construction. The page is dynamic and will roll to later months; the report must state the June 2026 reference period and 2026-08-03 release date wherever cited.

9. Infrastructure Investment and Jobs Act — Funding. Federal Highway Administration, U.S. Department of Transportation; page updated 2026-06-18 as displayed at research date; accessed 2026-08-09; official program webpage. https://highways.dot.gov/iija/funding
   Used for FHWA’s statement that IIJA provides approximately $350 billion for federal highway programs over FY2022–FY2026. A direct automated request returned HTTP 403 on 2026-08-09 because the FHWA host blocks this client; the URL is an official FHWA endpoint and should not be described as independently reachable by automation.

10. Apportionment of Fiscal Year 2026 Highway Infrastructure Program Funds, Notice N 4510.905. Federal Highway Administration, U.S. Department of Transportation; dated 2025-10-01; accessed 2026-08-09; official funding notice. https://highways.dot.gov/laws-regulations/directives/notices/n-4510905
    Used for FY2026 federal-aid highway apportionments. Cite the notice/table directly for the $56.815 billion figure. The host returned HTTP 403 to the automated test on 2026-08-09.

11. U.S. Transportation Secretary Kicks Off Surface Transportation Reauthorization. Federal Highway Administration, U.S. Department of Transportation; published 2025-07-17; accessed 2026-08-09; official agency news release. https://highways.dot.gov/newsroom/us-transportation-secretary-sean-p-duffy-kicks-off-surface-transportation-reauthorization
    Used for the reauthorization process and the current authorization’s 2026-09-30 expiry. The host returned HTTP 403 to the automated test on 2026-08-09.

12. Infrastructure Investment and Jobs Act — FHWA Implementation. Federal Highway Administration, U.S. Department of Transportation; dynamic implementation page, accessed 2026-08-09; official program webpage. https://highways.dot.gov/infrastructure-investment-and-jobs-act
    Used for Build America, Buy America / federal-aid implementation context. Cite the specific rule or implementation page, rather than Granite management, for regulatory requirements. The host returned HTTP 403 to the automated test on 2026-08-09.

13. USGS Aggregates Time Series Data by State, Type, and End Use. U.S. Geological Survey, National Minerals Information Center; published/updated 2025-04-03; accessed 2026-08-09; official dataset description and download page. https://www.usgs.gov/media/files/usgs-aggregates-time-series-data-state-type-and-end-use
    Used for the economic characterization of aggregates as high-volume, low-unit-value, local/regional products and for industry structure/history. Historical producer-count evidence must be labeled with its underlying census year rather than presented as current 2026 capacity.

14. Construction Sand and Gravel Statistics and Information. U.S. Geological Survey, National Minerals Information Center; no single publication date shown, accessed 2026-08-09; official commodity statistics webpage. https://www.usgs.gov/centers/national-minerals-information-center/construction-sand-and-gravel-statistics-and
    Used for official construction-sand-and-gravel industry context and links to annual commodity data.

15. Mineral Commodity Summaries 2026. U.S. Geological Survey; published February 2026; accessed 2026-08-09; official annual government report (PDF). https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf
    Used for 2025 U.S. mineral-commodity production/value context for crushed stone and construction sand and gravel. It supports national industry context, not a Granite-specific addressable market or local market share.

Price, factor, positioning, transcript and event sources

16. GVA Historical Price CSV. AZI Trading; observations through 2026-08-07, downloaded 2026-08-09; market-data CSV. https://azitrading.com/controls/download-data.php?t=GVA
    Used for daily adjusted and unadjusted OHLC prices, volume and 21-/50-/200-day exponential moving averages. Event-table prices use the unadjusted/traded-price fields; multi-period total returns use adjusted closes. The last available trading observation before the Sunday 2026-08-09 report date is Friday 2026-08-07.

16A. Peer Historical Price CSVs. AZI Trading; final observations dated 2026-08-07, downloaded/cross-checked 2026-08-09; market-data CSVs. ROAD: https://azitrading.com/controls/download-data.php?t=ROAD ; STRL: https://azitrading.com/controls/download-data.php?t=STRL ; PRIM: https://azitrading.com/controls/download-data.php?t=PRIM ; DY: https://azitrading.com/controls/download-data.php?t=DY ; EME: https://azitrading.com/controls/download-data.php?t=EME ; VMC: https://azitrading.com/controls/download-data.php?t=VMC ; MLM: https://azitrading.com/controls/download-data.php?t=MLM
Used only to verify the peer-table prices. All seven prices match the 2026-08-07 unadjusted closes. Peer shares, debt, cash, NCI, revenue, EBITDA and FCF denominators were not independently filing-reconciled and remain third-party screening data unless exact issuer filing/date support is added.

17. FactorsToday Methodology. FactorsToday; current methodology page accessed 2026-08-09; third-party model documentation. https://www.factorstoday.com/about
    Used to describe the ElasticNet factor model, annualization and reported risk statistics. These are model outputs, not audited company facts.

18. GVA Stock Loadings API. FactorsToday; Base / Base+Sector / Base+Sector+Industry observations dated 2026-08-07 and All Factors observation dated 2026-07-31; retrieved 2026-08-09; third-party quantitative-model API. https://www.factorstoday.com/api/stock-loadings/GVA
    Used for model-specific factor betas and R-squared. The All Factors date is one week older than the other current models and must be disclosed when its loadings are quoted.

19. GVA Leaderboard API. FactorsToday; data page dated 2026-08-09; retrieved 2026-08-09; third-party performance/risk API. https://www.factorstoday.com/api/leaderboard/GVA
    Used for annualized return, volatility, Sharpe, Sortino and maximum drawdown by horizon. Three- and six-month returns are annualized by the provider; use the raw AZI series for ordinary unannualized short-horizon moves.

20. GVA Stock Information API. FactorsToday; latest market observation 2026-08-07, retrieved 2026-08-09; third-party market/factor API. https://www.factorstoday.com/api/stock-info/GVA
    Used to cross-check the 2026-08-07 close, 6-/12-month relative strength, beta and alpha fields.

21. GVA Stock-Specific Volatility API. FactorsToday; 252-trading-day window through the current model observation, retrieved 2026-08-09; third-party quantitative-model API. https://www.factorstoday.com/api/stock-specific-vol/GVA
    Used for 26.57% annualized specific volatility and model fit statistics.

22. GVA Related Stocks API. FactorsToday; retrieved 2026-08-09; third-party similarity-model API. https://www.factorstoday.com/api/related-stocks/GVA
    Used only for factor-similar names; similarity is a model output, not a fundamental peer determination.

23. Historic Factor Returns API. FactorsToday; retrieved 2026-08-09; third-party factor-return API. https://www.factorstoday.com/api/factor-returns/historic
    Used for the cited near-term factor-return/regime observations. Each horizon and z-score should retain the provider’s date/context.

24. Granite Reports Fourth Quarter and Fiscal Year 2021 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2022-02-25; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000143774922004386/ex_306493.htm
    Used for the February 2022 price-event attribution. This source is older than 18 months and is used solely as historical event context.

25. Granite Reports Third Quarter 2022 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2022-10-27; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000143774922024861/ex_408416.htm
    Used for the October-November 2022 price-event attribution; historical context only.

26. Granite Reports Third Quarter 2023 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2023-10-31; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000086145923000023/gva-20230930xexx991.htm
    Used for the October-November 2023 price-event attribution; historical context only.

27. Granite Reports Third Quarter 2024 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2024-10-31; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000086145924000043/gva-20240930xexx991.htm
    Used for the October-November 2024 price-event attribution; older than 18 months at the research date and retained only for the mandated five-year price history.

28. Granite Reports Second Quarter 2025 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2025-08-07; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000086145925000039/gva-20250630xexx991.htm
    Used for the August 2025 price-event attribution and then-current guidance.

29. Granite Reports First Quarter 2026 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2026-04-30; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000086145926000016/gva-20260331xexx991.htm
    Used for the April-May 2026 price-event attribution and reported Q1 results/guidance.

30. Granite Reports Second Quarter 2026 Results (Exhibit 99.1). Granite Construction Incorporated; filed/furnished 2026-07-30; accessed 2026-08-09; SEC-filed issuer earnings release. https://www.sec.gov/Archives/edgar/data/861459/000086145926000027/gva-20260630xexx991.htm
    Used for the July-August 2026 price-event attribution and reported Q2 results/guidance. This is the SEC-hosted version of source 4.

31. Events and Presentations Archive. Granite Construction Incorporated; dynamic issuer page accessed 2026-08-09; issuer IR event archive. https://investor.graniteconstruction.com/news-and-events/events-and-presentations
    Used to locate company earnings materials. Earnings-call statements used in the research were retrieved through ROIC.ai direct transcript pulls for call dates 2026-02-12, 2026-04-30 and 2026-07-30 and must be identified as management commentary.

32. Granite Construction downgraded by Goldman Sachs. Investing.com / InvestingPro syndication; published 2026-07-09; accessed 2026-08-09; secondary financial-news report. https://ca.investing.com/news/stock-market-news/goldman-sachs-downgrades-granite-construction-stock-rating-to-sell-93CH-4726420
    Used only as secondary evidence that a Goldman Sachs downgrade occurred before the 2026-07-09 selloff and for the reported rationale. The page returned HTTP 403 to the automated URL test. The sell-side note itself is not public primary evidence; do not reproduce its price target or elevate its rationale to fact.

33. Granite Construction Short Interest Overview. MarketBeat; data as of 2026-07-15, accessed 2026-08-09; third-party exchange-data aggregation. https://www.marketbeat.com/stocks/NYSE/GVA/short-interest/
    Reports 4,052,256 shares short and 9.35% of public float. Its displayed days-to-cover data are internally difficult to reconcile with the simultaneously displayed volume, so omit days to cover unless the denominator and calculation window are separately established.

34. ROIC.ai company data and direct earnings-call transcripts for NYSE:GVA. ROIC.ai; transcript dates 2026-02-12, 2026-04-30 and 2026-07-30; accessed 2026-08-09; third-party financial/transcript database. https://www.roic.ai/
    Used for verbatim management-call context and aggregator cross-checks. The transcript list endpoint ignored the identifier and was discarded. ROIC valuation/price fields are period-end, not live, and ROIC’s calculated ROIC series remains third-party until reconciled to filings. The public homepage returned HTTP 403 during URL validation.

Evidence controls and known limitations

• Granite’s CAP is not synonymous with executed backlog. At 2025 year-end, executed unearned revenue was $4.123 billion (59.2% of CAP), while $2.846 billion (40.8%) was other awards that included probable but not yet executed/funded construction phases, unexercised options and unissued task orders. Substantially all CAP may be modified or canceled. Any report wording that calls the full $7.4 billion Q2 2026 CAP “contracted backlog” should be cut or qualified.
• Granite does not disclose local aggregates share, average haul radius, internal self-supply, third-party tons, unit prices, cash gross profit per ton, bid spreads, win rates, or integrated-versus-non-integrated margins. Claims that quantify a materials moat, local share, price leadership, or a vertical-integration margin uplift must be labeled assumption/open question or removed.
• The 2023 USGS establishment counts are historical industry-structure evidence. They should not be labeled a 2026 current count without a newer census.
• Granite’s Q2 2026 release supports management’s attribution of weaker Materials profitability to severe Southeast weather and quarry-development costs, while the 10-Q also identifies higher purchase-accounting charges. A report should include all disclosed drivers and not present management’s causal explanation as independently proven.
• All current valuation and momentum calculations must be keyed to the final row of the AZI five-year price CSV and dated 2026-08-07 (the last trading day before the 2026-08-09 report date), not to a stale aggregator market-cap or enterprise-value field. A single precise current EV is not supportable from the June balance sheet alone: the 2028-note conversion and capped-call unwind remained pending at the cutoff, the capped-call proceeds/settlement cash were undisclosed, and June debt carrying value was distorted by bifurcation of the $630.5 million conversion derivative. Use an explicitly dated pre-settlement/pro forma range or disclose the limitation.
• The peer multiple table’s prices are independently AZI-verified, but its market caps, enterprise values and TTM denominators use ROIC.ai plus unspecified “latest available” share feeds/filings. Without an exact per-ticker filing/share/balance date trail, present it as a third-party screening comparison, not a fully filing-reconciled comp table.
• Reproduction value and EPV must be compared on the same enterprise/equity basis. A cash-inclusive estimate of approximately $1.04 billion for “net current operating assets” would include approximately $931.5 million of cash/marketable securities; true net operating working capital from the June balance sheet is only approximately $0.13 billion before classification refinements. The published report therefore does not use the invalid cash-inclusive reproduction-value range and limits the asset-value conclusion to a qualitative comparison.