GPGI: When the Operating System Meets a $20 Million Hole
Husky's delayed cost actions cost GPGI its full-year EBITDA guide — and quietly moved the leverage target by half a turn, not that the chairman mentioned it
GPGI · Earnings Call · 2026-11-05
The operating system was the story. This quarter it became the excuse.
For five straight quarters, GPGI sold investors a management system, not a set of businesses. In the first-quarter 2026 keyword set the company's own language was dominated by the Resolute Operating System, high performance culture and "seed planting" — the pitch being that Dave Cote's playbook, imported from Honeywell and Vertiv, would compound margins at anything GPGI touched. By the second-quarter call, the phrase of the day was deployment of ROS, cost absorption and ROS-led efficiency gains.
Now look at the words that actually spiked in this third-quarter print: cost actions at the very top, then growth algorithm, margin headwinds, delayed realization, direct labor and the S&OP process. Meanwhile cultural transformation shows up as one of the quarter's biggest keyword decliners — the inspirational half of the narrative quietly collapsing into the operational half. That is the whole call in a word shift.
The admission is unusually blunt for a company that once promised "the market may not see it today." Husky's CEO: “we have been slow in executing the cost actions identified early in the year.” — Robert Domodossola, Executive, likely CFO or similar financial leadership role · 2026-11-05 And Cote, in Q&A, on the direct-labor piece: “There was kind of a hope that sales might turn out to be better than they were. So we carried direct labor a little longer than we should have.” — David Cote, Executive, likely CEO · 2026-11-05
These delays, net of EBITDA contribution from incremental sales represents approximately $20 million of EBITDA headwind. We are focused on addressing this to position the company well for 2027.
Twenty million dollars is roughly 4% of the guided full-year EBITDA midpoint. It is also, notably, a self-inflicted number: not lost volume, not price, but an execution miss management had already flagged the actions for.
Two businesses, two very different tapes
Strip out the noise and GPGI is two assets moving in opposite directions. CompoSecure delivered record adjusted net sales of $140.7 million, up about 16%, with adjusted EBITDA of $61.2 million, up 23%, and margins at 43.5% — up 230 basis points year-over-year and 220 sequentially, after roughly a 30% cut in manufacturing rates over the past year. Husky, ten months into ownership, delivered $368.9 million of adjusted net sales, up 0.5%, and $88.8 million of EBITDA, down 0.9%, at a 24.1% margin that was 30 basis points lighter year-over-year.
The genuinely new company-specific theme is on the CompoSecure side. debit card landed at rank four this quarter, alongside the sibling tag debit, and Graham Robinson was explicit about why a metal debit card matters: “Metal debit cards for premium clients allow financial institutions to promote a differentiated and distinguished customer experience, which in turn should help retain those clients' deposit base.” — Graham Robinson, Executive, likely COO or similar operational leadership role · 2026-11-05 It is a small revenue line today but a real wedge — CompoSecure has 200-plus active programs and still sits at ultra-low penetration of the card market.
Husky's problem is macro residue. The resin price shock that followed the Iran war is the same Middle East conflict that rattled packaging and shipping names across the tape this year, and it froze capital purchases at bottlers overnight. The read-across from packaging is instructive rather than damning: Ball, reporting the same week, talked up strong volume growth in cans. End demand for beverage containers is not broken. What broke was the timing of machinery orders — which is exactly why the recovery case rests on orders (+7% year-over-year, +21% in October) and a backlog management expects to finish the year up double digits rather than on any current-quarter number.
The arithmetic of a cut that isn't quite a cut
GPGI now guides full-year 2026 to $2.034–2.064 billion of pro forma adjusted net sales — up about 1%, at the higher end of the prior range — but $520–550 million of adjusted EBITDA, down about 2%, margins of 25–27% (down 80 basis points), and free cash flow of $250–300 million. Tom Knott framed the drivers plainly: “we are lowering our outlook for adjusted EBITDA, margins and free cash flow due to margin headwinds at Husky, primarily due to delayed realization of identified cost actions, product mix, higher freight costs and continued investments necessary to position the business for growth in 2027.” — Thomas Knott, Executive, likely CFO or similar financial leadership role · 2026-11-05
The fourth-quarter guide — $630–660 million of sales (+1%) and $178–208 million of EBITDA (+8% at the midpoint) — implies that roughly a third of the year's EBITDA lands in one quarter, and that the quarter is up year-over-year while the full year is down. That only works because of historical seasonality, which the company has leaned on for at least three calls now: back in May, the then-CFO noted that “seasonally, second half represents roughly 60% of our revenue base.” — Kevin Moriarty, Executive · 2026-05-07 Fine — but seasonality is also the reason a fourth-quarter miss would be maximally painful, and GPGI has now told investors twice that it is "adjusting" a guide whose second half was supposed to be the easy half. Last quarter the same management team insisted “the guide does not assume things materially improve.” — Graham Robinson, Executive at CompoSecure (a GPGI business) · 2026-08-06 It didn't, and the guide still came down.
Cote's framing has consequently migrated to 2027: “while 2026 has been a foundational year for GPGI, we expect to see a return to our long-term growth algorithm next year.” — David Cote, Executive, likely CEO · 2026-11-05 The word growth algorithm is doing a lot of work here — mid-to-high single-digit organic growth, 100-plus basis points of annual margin expansion, double-digit EBITDA growth — and it is now a forward promise rather than a demonstrated track record.
The half-turn nobody asked about
Here is the quiet item. In March, GPGI's then-CFO was asked about leverage and answered: “We would expect total leverage to be below 3x.” — Thomas Knott, Executive (likely CFO or similar senior role) · 2026-03-12 On this call, Cote described a plan to cut net debt by more than $200 million through year-end and to end the year at approximately 3.5x leverage, with the long-term target range reached next year. Nobody on the call pressed the gap. Half a turn of leverage is not a rounding error when the equity has already fallen 47.6% from its January peak of $26.00 — and it makes the "return to the growth algorithm" story contingent on Husky's 2027 recovery actually arriving.
What the market is being asked to underwrite is a company trading at roughly $3.95 billion of market cap, levered about 3.5x against guided EBITDA, with one business compounding in the high teens and another that just told you it under-executed. The reported fundamental series is no help: post spin-off, GPGI Holdings is accounted for under the equity method, and the reported parent shows essentially nil revenue — which is why reported total revenue falls to zero in the most recent quarters and why price-to-revenue reads a nonsensical 40.9x. Both are artifacts of presentation. Investors have to rebuild the consolidated economics from the pro forma disclosures.
Which may be exactly why Cote closed with his signature line: “My family personally owns nearly 18% of GPGI today, so there is significant skin in the game. As you've heard me say many times, and I still think it's true, buy now while supplies last.” — David Cote, Executive, likely CEO · 2026-11-05 A year ago he said the same thing about a business he was still learning. The difference now is that the promise has a deadline attached — 2027 — and the operational slack that was supposed to be the easy part has already cost him $20 million and 80 basis points of margin.