Celanese's $800 Million Answer to a $9.3 Billion Debt Problem
A fourth straight quarter of the same free-cash-flow promise, while the growth story quietly narrows to electronics and data centers
CE · Earnings Call · 2026-11-03
The Story That Isn't Changing
Celanese reported third-quarter 2026 into a stock still down roughly 71% from its 2022 peak of $173.82 — a name that has spent three years being de-rated, not discovered. And the very first question on the call, from Baird's Ghansham Panjabi, was not about demand or pricing or the tow market. It was about whether the $700 million to $800 million of free cash flow underpinning next year is real. That tells you where this company's center of gravity now sits: the cash machine, not the chemicals. CFO Chuck Kyrish answered with the same line he has used for several quarters. They are “demonstrating our capability to generate 7 to 800 million of free cash flow in 2 very different environments the last couple of years” — Chuck Kyrish, Chief Financial Officer · 2026-11-03. Back in August he had already conceded the deeper point: “I would consider the $700 million to $800 million of free cash flow as our sort of baseline sustainable level” — Chuck Kyrish, Chief Financial Officer · 2026-08-05. So this is not new information — it is a re-underwriting of the same covenant-in-spirit, now stretched to 2027. The balance sheet explains the obsession. Effective net cash sits around negative $9.3 billion, and interest coverage has collapsed to roughly 1.5x from north of 60x in late 2020. The equity is now priced as an option on deleveraging: the stock trades at about 0.5x revenue, down more than 80% from its 2021 multiple.Deleveraging Is the Goal; Divestitures Are the Lever
The $1 billion divestiture target has been management's other refrain, and this quarter finally produced a concrete installment: the NutraNova joint venture, where Mitsui is taking a larger share and Celanese expects a little over $150 million in proceeds. Scott Richardson framed it as the template — “it is likely to get to close the gap on the remaining 350 or so, $340 million of additional divestiture proceeds” — Scott A. Richardson, President and Chief Executive Officer · 2026-11-03. This is the same promise made in August (“We are committed to the $1 billion target of divestitures by the end of 2027” — Scott Richardson, President and Chief Executive Officer · 2026-08-05) and again in the prior quarter. The progress is real but the cadence is slow, and the market has heard it before. Management's own math on the deleveraging glide path: finish 2026 near $10 billion of net debt, 2027 near $9 billion, crossing 5x leverage this year and targeting 4x next — as Kyrish laid out in the August call (“We're going to finish this year around $10 billion. Next year, we feel really good about finishing at $9 billion” — Chuck Kyrish, Chief Financial Officer · 2026-08-05).The Only Engine Actually Growing
Strip out the balance sheet and one part of Celanese is genuinely outperforming: Engineered Materials, where electronics and medical are doing the heavy lifting. Richardson called it a generational spike in demand tied to the AI data-center buildout — “we are seeing kind of a generational spike in demand coming from kind of the global data center build out around AI... that is gonna be a multi-year kind of tailwind” — Scott A. Richardson, President and Chief Executive Officer · 2026-11-03. Even with auto volumes down (largely China builds), EM EBITDA is growing roughly 20% year-over-year ex-Micromax, and about 15% including it. Here is the honest caveat: this is not a fresh discovery. On the February call, the same electronics and data center build-out was already the identified bright spot (“electronics is what I would say the bright spot right now... a global build out from AI, as well as as data centers” — Scott A. Richardson · 2026-02-18). What is new for 2027 is the specificity — three EM projects commercializing in the first half, two in Asia, one in Europe, plus a robotics relationship. That is a real pipeline, but it is a small revenue base against a company this size. And there is a divergence worth flagging: while Celanese talks up AI data-center content, the market tape has been voting against that exact cluster. Over the last 30 days the global AI data center theme is down more than 20% with 55 decliners against 10 advancers. The narrative and the tape are pointed in opposite directions.What Management Stopped Saying
The Acetyl Chain is where the story leaks. Richardson was candid that margins were “coming off through the quarter and finishing at lower levels at the end of quarter” — Scott A. Richardson, President and Chief Executive Officer · 2026-11-03, and the Q4 guide embeds a roughly $25 million sequential decline from paints and coatings seasonality, plus an unexpected Chinese supplier outage and low Rhine River water levels. The Ibn Sina venture — long a volatility wildcard — is now expected to be roughly flat year-over-year rather than the $10 million drag previously flagged, a mild positive. What is striking is what has quietly faded from the script. The reliability of supply theme, which management hammered for quarters as its Western Hemisphere weapon, has receded — arguably because it worked, converting into 2027 contracts rather than talking points. And Frankfurt, for so long the swing-asset centerpiece, is now barely mentioned beyond a note that the German plant is the largest remaining gas exposure. When a company stops advertising the flexibility of its network, it usually means the flex is already spent.That refusal came in response to Fermium's Frank Mitsch, who asked plainly whether 2026 was a year that Celanese over-earned — noting 2026 results sit roughly where an average of 2019 and 2020 did. Management would not say the word. Instead, 2026 is being reframed as a base: “we do feel like we can make 2026 a base that we can build off of going into 2027” — Scott A. Richardson, President and Chief Executive Officer · 2026-11-03, supported by $30-40 million of carryover cost savings, about $40 million of interest savings, and a $150 million gross productivity program to offset inflation. The follow-through will hinge on a tow market that is merely stabilizing and an Acetyl Chain whose western-margin recovery keeps being pushed a quarter out.And I said this to our Acetyl Chain team. I believe this is the business that never sleeps... I do not think we are gonna put labels on 2026.