JELD-WEN Holding, Inc. JELD

Key data

The business

JELD-WEN makes doors and windows: the molded interior door in most American tract homes, the exterior door on the front of the house, and the windows around it. It sells them through the big-box retailers, through homebuilders, and through building-products distributors. North America is 65% of revenue ($528.5M last quarter), Europe the rest ($289.3M). Demand rides two waves it does not control: new housing starts and repair-and-remodel spending, both of which fall when mortgage rates are high and homes stop changing hands. This is a commodity product sold on price and availability. The moat is thin, mostly scale in distribution and shelf space, which is exactly why a downturn hurts this much: there is no premium margin to cushion the fall.

The numbers

Revenue peaked at $4.54B in 2022 and has fallen every year since, to $3.21B in 2025, down 29%, about 11% a year. That is the whole demand story. But look at the last two quarters instead of the last three years: Q2 2026 revenue of $817.8M was down just 0.7% from a year earlier, and gross margin recovered to 16.8% from a 12.8% trough in Q1. The operating loss narrowed to $5.1M from $13.9M a year ago. The operating bottom looks close.

The Q3 2025 loss carried a large asset write-down; H1 2025 also absorbed a $137.7M goodwill impairment. Strip those and the underlying business was losing far less than the headline suggests, which is the point of the recovery table below.

The operating leverage is the entire case. On a mostly fixed cost base, every dollar of lost revenue drops hard to the loss line on the way down, and reverses just as hard on the way up. Watch it working in reverse already:

What the market is not pricing: that the operating decline has stopped, and this is now a balance-sheet problem, not a business problem. Consensus, from only two or three analysts, sees revenue roughly flat through 2027 and losses shrinking but not gone. The single print that settles it is Q3 2026 revenue holding, paired with a refinancing that pushes out the debt.

Management

The tell here is what insiders did not do. With the stock near $2, down two-thirds in a year, not one insider bought a share. There were three sales in the last twelve months totaling $37,562, trivial amounts, plan status not disclosed. Buybacks fell sharply after 2022 (2023: zero, 2024: $24.3M, 2025: zero) and the prior authorization has not been renewed. Pay and guidance are secondary when the question is solvency.

How it fails or surprises you

The debt wall. Net debt is about $1.3 billion against $198M of equity and negative tangible book, and cash fell to $57.2M from $136.1M in six months. A refinancing on punishing terms, or a covenant trip, hands the company to the lenders and wipes the equity to zero. First signal: any refinancing announcement and the rate on it before mid-November.

Housing turns (right tail). If mortgage rates ease and repair-and-remodel demand returns, revenue reflates toward $3.5B-plus on a cost base already cut to the bone, and the decremental margins above run in reverse into fat incremental margins. The market pays nothing for this today. First signal: year-over-year revenue turning positive in Q3 or Q4 2026.

The recovery is seasonal, not real. The fact my read explains least: cash still drained about $79M in the first half even as margins recovered, so the "bottom" has not yet stopped the bleeding. If H2 free cash flow stays negative, the operating recovery is a mirage and the clock on the debt keeps running. First signal: H2 2026 cash flow.

Closing thoughts

This is close to binary, and a named print resolves it. The left tail is a permanent zero: a restructuring that leaves current shareholders with nothing. The right tail is several times your money, as a housing recovery reflates earnings and a refinanced balance sheet stops being the story. Because the downside is total, survivability governs everything, and the fat question is not how good the business gets but whether it reaches the next cycle solvent. The three-days-ago watch on this name still stands open: no refinancing has been announced and Q3 has not printed, but Q2's $817.8M at least argues the revenue stabilization was not a one-quarter head-fake.

The bet is still that Americans start buying doors and windows again, and that JELD-WEN survives its mountain of debt long enough to be there when they do. What breaks it is the debt coming due before the demand comes back. The one pair of numbers that tells you first: a refinancing rate on the wire by mid-November, and Q3 2026 revenue above roughly $780M. Miss both and the equity is a lottery ticket, not a position.