Granite Writes a 2029 Check the Market Won't Cash
Record backlog, record EBITDA, a fresh 10–12% growth target — and a stock 24% off its June high. The delta is a pivot into AI-adjacent site work just as that tape rolls over.
GVA · Earnings Call · 2026-10-29
A record quarter the tape refuses to believe
On the page, this was the best quarter Granite Construction has ever produced. Revenue rose 23% to $1.8 billion, gross profit 29% to $336 million, and adjusted EBITDA climbed $67 million to $283 million. The Construction segment — the engine — grew revenue 27% and expanded gross margin 210 basis points in the quarter, 80 basis points year-to-date, comfortably ahead of the roughly 50 basis points a year management had been promising. Then Kyle Larkin did something companies rarely do mid-drawdown: he put a longer, harder number on the table. “we are targeting an organic revenue CAGR of 10% to 12% through 2029” — Kyle Larkin, President and Chief Executive Officer · 2026-10-29, alongside a 14%–16% adjusted EBITDA margin, operating cash flow of ~13% of revenue, free cash flow of ~10%, and a ~2.5x net leverage ceiling. The tape is unimpressed. GVA has compounded 256% over its full history but is down 19.9% in the last 90 trading days alone, a single unbroken downtrend, sitting 24.1% below its June 26 peak of $160.39. That is the tension of this report: operating momentum at an all-time high, price action at a three-month low. And the valuation hasn't actually reset much — at 1.4x revenue the stock is still priced at the richest level in a decade, with price-to-gross-profit of 8.9x against a 2019 peak of 9.4x. The market is not doubting the quarter. It is doubting the 2029 bridge.Mission-critical: the fourth leg, and a brand-new word — space
For a decade Granite has been a public-works contractor that occasionally did private work. The new top theme in this quarter's transcript is unmistakable: critical infrastructure sits at the very top of the company's keyword ranking, well above the highway and materials language that usually dominates. Federal contract backlog stands at $1.3 billion, of which $569 million is tactical infrastructure. Mission-critical — data centers, battery storage, and now space — is the newest leg. “At the end of the third quarter, mission-critical infrastructure CAP totaled approximately $331 million. Less than 5% of 2026 revenue is expected to come from these projects, so we see opportunity for growth.” — Kyle Larkin, President and Chief Executive Officer · 2026-10-29 That is a trajectory, not an event. Three months ago Larkin told investors “last year, at this time, we had CAP of around $65 million within the data center space. Today, it's around $225 million” — Kyle Larkin, President and Chief Executive Officer · 2026-07-30 — and he already wanted it at 10% or better of revenue. What is genuinely new this quarter is the fourth bucket. “what's new in the last quarter is space. And so we've added just under about $100 million of CAP within space within that mission-critical infrastructure” — Kyle Larkin, President and Chief Executive Officer · 2026-10-29. Project sizes, he noted, have moved from a historical $20–30 million to roughly $100 million today. Here is the contrast worth flagging. Granite is scaling into civil site work for AI infrastructure at precisely the moment the public market has turned on that complex: AI data center is among the worst-performing thematic clusters over the last 90 days, with dozens of decliners against a handful of advancers. The bull case is that Granite sells grading, paving, utilities and mass excavation — civil scope that gets paid regardless of which model wins. The bear case is that the "at least 10% of Construction segment revenue" ambition is being underwritten into a demand curve the market has just stopped believing in.Collaborative contracting goes federal — the actual new thing
The mechanism that supposedly de-risks all of this is collaborative contracting, and the honest read is that it is not new. Michael Dudas was probing the best-value/bid-build split back in February, and Larkin answered at length then about how the model lets Granite be chosen on qualifications rather than low bid. What is new is where it is being applied:That is the single most consequential sentence in the call, because it changes the character of the federal market business. Granite's federal growth to date has leaned on fast-burn, politically-exposed tactical work. Now it can pursue Progressive Design Build and CMAR awards on airfield recapitalization in Alaska, naval modernization in Washington State, and missile-defense infrastructure in Guam — qualification-selected, higher-quality backlog. And the company has spent a decade building exactly that capability, including the Pacific Northwest waterborne aggregate position that underpins it. One caveat management disclosed quietly: the $569 million of tactical infrastructure that was expected to finish in Q2 2027 now finishes in Q3 2028 on right-of-way availability. Profitability unchanged, revenue pushed right.The 2026 National Defense Authorization Act introduces collaborative contracting procurement to federal work, including opportunities for progressive design build, construction manager at risk and other flexible procurement approaches for military construction.
The bill: energy surcharges, leverage, and a quiet Materials problem
Record CAP of $7.5 billion is a genuine headline — but read the increment. Backlog grew just $37 million sequentially in what management itself called its busiest construction quarter. Net additions are decelerating off a very high base, even as gross wins outpace burn. The Materials segment is where the optics and the economics diverge. Revenue rose $21 million to $292 million while gross margin fell 420 basis points and cash gross margin fell 100 basis points — and the reason is mechanical, not operational.That is higher energy costs diluting a percentage, not a dollar. Management expects cash gross margin back up roughly 3% when energy normalizes. But note what has faded: the vertical integration and pull-through story that dominated 2025 calls — Warren Paving, the Kentucky quarry, Memphis stone — has slipped from center-stage to "resilient." The keyword history is blunt about it: the Southeast platform and liquid asphalt themes have both cooled in consecutive quarters. Yet the 2029 margin bridge needs roughly 150 basis points a year out of Materials. And the balance sheet has quietly become the constraint. Effective net cash is now negative $1.0 billion, down $758 million in a single year against a 2017 peak of positive $210 million. The convertible was retired — $567 million of net cash and 662,000 shares — which removes dilution but adds debt. And the roll-up continues: “a total spend up to about $750 million if all 3 deals get across the finish line” — Kyle Larkin, President and Chief Executive Officer · 2026-10-29. That is the real wager. The 2029 plan needs free cash flow of roughly 10% of revenue to simultaneously fund acquisitions, buy back stock, and walk leverage back to 2.5x. GAAP earnings are a poor guide to any of it — the last three filed quarters swung from a $59 million profit to a $278 million loss on charges unrelated to operations. The story here is not the record quarter. It is that Granite is re-underwriting itself as a federal, rail and mission-critical civil contractor — and asking shareholders to keep paying 1.4x revenue for that identity — at the exact moment the market has decided to discount anything that touches AI infrastructure. Management spent the call explaining why the pivot is de-risked by procurement reform. The tape spent the last 90 days disagreeing.Higher energy costs were largely recovered through surcharges, which increased revenue and offset higher costs in absolute dollars. This resulted in a lower incremental margin even as underlying unit economics remained solid.