Open in interactive viewer → charts, metric popovers & call review

Cardinal Infrastructure Grew 65% and Lost Its Margin: A $13 Million Estimate Revision That Ended Guidance

Nine weeks after calling margin pressure 'transitional, not structural,' Cardinal withdrew full-year guidance, lost 13 points of EBITDA margin to three Carolina projects, and announced its CFO's exit.
CDNL · Earnings Call · 2026-11-12

The quarter that broke the story

Cardinal Infrastructure Group is a $229 million market cap Southeast site-work and paving roll-up that came public recently, and on November 12 it delivered the kind of print that resets a narrative. “Revenue grew 65% this quarter to $202 million, but our third-quarter adjusted EBITDA margin totaled 4.2%, well below our expectations.” — Jeremy Spivey, Chairman and Chief Executive Officer · 2026-11-12 Adjusted EBITDA fell to $9 million from $21 million; gross margin collapsed to 5.4% from 15.9%; the quarter produced an $11 million net loss against $10 million of net income a year earlier. The stock has already voted. From a peak of $30.73 on October 2, shares are down 81.6% — a decline so violent it dwarfs the earnings miss itself, implying the market is repricing the durability of the whole roll-up model, not just one bad quarter. The anatomy matters because it is unusually concentrated. Management attributed roughly 6.5 points of the 13.3-point margin decline to project estimate revisions — specifically a $13.2 million upward revision to cost estimates — with a further 3.6 points from residential market pricing and about 3.5 points from higher fuel and transportation costs. And the revisions were not diffuse: “10 projects account for 61% of the estimate revision, and simplifying that even further, 3 projects accounted for 50% of the estimate revision.” — Jeremy Spivey, Chairman and Chief Executive Officer · 2026-11-12 All three of the worst offenders were residential, and all sit in the Carolinas. Then the other shoe:

Given the magnitude of the project estimate revisions recorded in this quarter and the work underway to strengthen our project-level forecasting, we are withdrawing our previously issued full-year 2026 revenue and adjusted EBITDA margin guidance.

Mike Rowe, Chief Financial Officer · 2026-11-12
The CFO, Mike Rowe, is stepping down. The proposed Hughes Brothers acquisition — a keyword that ranks among the most prominent in this quarter's company vocabulary — has been postponed. A $50 million buyback was authorized against a $229 million market cap, which reads less like capital-allocation confidence and more like a signal sent from a company that believes its own equity is mispriced.

What management stopped saying

The most revealing artifact here is linguistic. In the second quarter, the margin story was exogenous. Rowe framed four headwinds — one-time labor, deployment shifts, Georgia weather, new-public-company SG&A — and delivered the line that now looks ruinous:

it is transitional. it is not structural. it is very much transitional.

Mike Rowe, Chief Financial Officer · 2026-08-11
He closed the same answer by saying the company would not repeat the misses, and Jeremy Spivey told the same analyst two questions later that on customer demands for pricing concessions, “we obviously have a red line” — Jeremy Spivey, Chairman and Chief Executive Officer · 2026-08-11 — if a project did not clear the margin bar, Cardinal would walk away. Now compare vocabularies. The old keyword was onetime cost — a word that implicitly promised the problem lived outside the building. In the third quarter that term has vanished entirely, replaced at the very top of Cardinal's keyword list by estimate revision, project review, and project selection. That is an admission of an internal process failure, and to management's credit they say so plainly: the CEO calls the results "not acceptable," installs the ALGC operating chief over the whole platform, and hires an independent third-party adviser to audit estimating and project-review practice. The word that never got said out loud, but is implied by the entire remediation plan, is controls. There is a second silence worth flagging. Eight weeks ago Cardinal had spent real airtime on a big data-center win — a marquee, mission-critical end market that global markets are still arguing about. In the current call, data center does not appear in the top-30 keyword set at all; it surfaces only when an analyst asks about it in Q&A, and the answer is a Phase 1 change order representing a 50% increase in the original contract with Phase 2 slipping to late Q1 or early Q2 of 2027. That is a theme quietly demoted from growth narrative to footnote.

Which headwinds are genuinely shared

Honesty requires separating the company-specific from the cyclical. The fuel and transportation piece is real and it is not Cardinal's alone — global keyword traffic shows high fuel prices as a persistent cross-sector theme, and this reporting season alone produced airlines, container lines and REITs all flagging elevated fuel cost as a live margin drag. Cardinal gets little credit here, but it also deserves little blame. Residential is more interesting, because it is where the tape disagrees with the anecdote. Cardinal says the softness is a broad Southeast phenomenon, and there is corroboration from a home-improvement distributor reporting residential trade plumbing revenue down 2% on new residential construction. Yet the market is bidding housing-adjacent demand higher over the last 30 days, not lower. The reconciliation is that Cardinal is levered to land development volume and to builder margin compression being pushed down the supply chain, not to repair-and-remodel spend. When Rowe says “when interest rates hit in Q3, we saw expedited softness” — Mike Rowe, Chief Financial Officer · 2026-11-12, he is pointing at the same rising interest rate environments that global keyword work has flagged for several quarters — a policy variable Cardinal cannot estimate its way out of. Where Cardinal stands alone is execution. The global tape has a keyword for exactly this: strong project execution, a top-90-day decliner across a basket of contractors and engineers, down roughly 38% as a group. That is the cross-section Cardinal just joined, and it is why the miss landed as a model question rather than a weather question. Note also that the global data-center complex is simultaneously de-rating — AI data center showed a wide majority of constituents negative over 30 days — so even the one growth vector Cardinal could point to is fighting a tape headwind.

Cash, credibility, and the one clean catalyst

The balance sheet is the reason this is a credibility story rather than a solvency story. Cardinal exited with $335 million of cash and roughly $136 million of net cash, plus undrawn capacity. But the cash conversion is thinning: year-to-date operating cash flow of $30 million against $42 million of capital expenditures implies negative free cash flow, consistent with free cash flow of approximately -$12 million in the latest reported period, a sharp reversal from slightly positive a year ago. Funding a $50 million repurchase out of that position is a statement of intent, not a surplus of capital. The one genuinely constructive, specific fact is buried in the Q&A: “most of these projects on the revision are over 75% complete.” — Jeremy Spivey, Chairman and Chief Executive Officer · 2026-11-12 The bad work finishes this fiscal year or early in the next one. That gives fourth-quarter adjusted EBITDA a mechanical tailwind and makes management's claim of sequential improvement the most falsifiable thing they said all morning — a low bar, as the analyst on the call noted, but a real one. So the setup for the next two quarters is a clean test. Either the project review regime produces visible, boring, on-plan margins — in which case an 81% drawdown on a $229 million market cap with net cash and a finishing backlog is an extraordinary option — or the estimate revisions were the first disclosure of a system that cannot estimate, and the buyback will have bought stock into a second leg down. Cardinal has one quarter to prove which.