Cavco Industries, Inc. CVCO

Key data

The business

Cavco is one of the country's largest makers of factory-built homes. It builds HUD-code manufactured houses and modular homes in plants across the U.S., trucks them to a site, and sells them through company-owned stores and a wide network of independent dealers. A Cavco home sells for substantially less than a median site-built house, and that price gap is the entire reason the company exists. A second, smaller segment lends to its own buyers and sells them insurance, which smooths the cyclicality of the building side and throws off steady spread income. The moat is unglamorous: scale in a business where a plant's local density and a retailer relationship are hard to copy, plus a balance sheet carrying almost no debt while much of the industry runs levered. What Cavco sells is not a house so much as the cheapest legal path to owning one, and demand for that path rises exactly when the rest of housing gets unaffordable.

The numbers

Revenue is climbing to a company record while profit slides.

Revenue of $610.0M in the June quarter was up 9.5% on the year and a record, yet net income fell 18% and EPS dropped to $5.43 from $6.42, landing under the two-analyst estimate of $5.69. The driver is gross margin: 22.1% last quarter, the lowest of the five, down from 23.3% a year earlier, with operating margin off about 220 basis points. Volume is filling the plants, and each home earns less doing it. That squeeze, not the record top line, is the number that matters, and it has not eased since it first surfaced in this filing.

Over four years revenue compounded 8.4% a year, but diluted EPS only 3.0%, and FY2026's $23.98 still sits below the FY2023 peak of $26.95. Most of the lift that reached shareholders came from buybacks: $581M repurchased over five years shrank the share count about 16%, from roughly 9.3M to 7.8M. Return on equity holds near 16.5% and the company carries roughly $203M of net cash. This is a steady, well-run compounder, not a fast one, and the two analysts covering it model EPS falling in FY2027 to $22.55 before recovering to $27.87 in FY2028.

Backlog is the builder's leading tell, and it just jumped 49% on the year to $298M, up $103M in a single quarter. That is orders the plants have not yet shipped, and it argues the record revenue is early, not late. The market is not paying for it. At 25 times a consensus that has EPS declining, the price implies the margin squeeze is structural rather than a passing mix of promotions and input costs. The view worth holding is that backlog this size converts into volume that restores operating leverage by FY2028, and the print that settles it is gross margin over the next two quarters. Stabilize near 22% while revenue compounds and the down-EPS consensus is wrong.

Management

The record here is mixed and worth reading closely. In February, insider William C. Boor put $495,000 of his own money into the stock on the open market, the kind of purchase that carries weight at a company this size. Against that, insider Allison Aden sold about $1.6M across two June sales, and the filings do not state whether those were pre-scheduled, so plan status is not disclosed and the signal is muted. Capital allocation is consistent and aggressive: $159.9M of buybacks in FY2026, higher every year, funded entirely from cash flow with no debt raised. The one caution is price. Management is retiring stock at 3.8 times book, near the richest multiple in twelve years.

How it fails or surprises you

Backlog converts to a volume surge (right tail). Orders on the books rose 49% on the year to $298M, up $103M in one quarter. If the plants ship that at even flat margins, revenue accelerates past the down-EPS consensus and FY2028's $27.87 comes into view early. The market prices none of this today. The tell is sequential backlog and revenue over the next two quarters.

Record revenue, falling profit is the crack. Gross margin fell to 22.1% last quarter, the lowest of the five, while sales hit a record and net income dropped 18%. If that is structural, from mix shifting to cheaper homes and heavier incentives rather than a passing cost bump, volume growth stops reaching the bottom line and the stock is a value trap at 25 times. The print is gross margin next quarter.

A rate stumble empties the backlog. Manufactured-home demand runs on chattel lending, which reprices fast when rates rise. The 49% backlog build assumes affordability holds. If mortgage and chattel rates climb, cancellations rise and the leading indicator reverses before revenue shows it. At 3.8 times book, the 92nd percentile of its own history, there is no valuation cushion. Watch cancellations and sequential orders.

Closing thoughts

Gross margin over the next two quarters is what decides it. Hold near 22% while the $298M backlog ships and the down-EPS consensus is simply wrong, with FY2028 earnings pulling the stock up. Keep sliding and the record revenue is a mirage, the multiple compresses from its twelve-year high, and you have paid a peak price for a business whose profitability is quietly eroding. The fatter near-term tail is the downside: you are paying 25 times for a year analysts already expect to shrink, with no margin of safety if demand or margin disappoints. What is at risk if margin breaks is a de-rating from 3.8 times book toward the 2.3 times low end of its range, a long way down. What the upside is worth, if backlog converts, is the FY2028 number arriving a year early.

The bet is still that Americans priced out of regular houses keep buying the cheaper factory-built homes Cavco makes, the pile of unfilled orders keeps growing, and Cavco shrinks its share count by buying back stock. What breaks it is a backlog that stops growing or a gross margin that keeps falling. The one pair of numbers that tells you first is next quarter's backlog against next quarter's gross margin. Rising orders at a stable margin confirm the bet. A record top line at a 21% margin refutes it.