Everus Buys a Factory, Not a Contractor: Inside the $295M Epsilon Deal

Everus closed a deal on September 1 to buy Epsilon Industries for $295 million in cash, announced barely a month earlier, on July 31, and funded with money it had on hand plus borrowings under its credit lines. That is quick and clean for a company whose total market value is around $5.8 billion and whose 2026 revenue guidance runs to about $4.5–4.7 billion. The deal is roughly 5% of Everus, and roughly 5% of its annual revenue. So the price tag is not really the story. What is worth a moment is what Everus bought, and what the language around the deal says about how it thinks.

Epsilon is not, despite the press release's careful wording, exactly a construction contractor. It is a designer and manufacturer of complex, modular mechanical and electrical building infrastructure systems — pipe racks, prefabricated electrical and mechanical guts for buildings — that it builds off-site in factories in the U.S. and Canada and ships to data centers, advanced-manufacturing plants, and hospitals. Think of it as the difference between building a building's nervous system on the job site versus assembling it in a plant and bolting it in. Everus, for its part, is a specialty electrical-and-mechanical contractor and a transmission-and-distribution contractor, and it is deeply exposed to the thing that is currently minting money for a certain kind of contractor: the data-center buildout. Its backlog was $4.55 billion as of June 30, up 53% from a year earlier, with data centers the largest piece.

This is Everus's second purchase in a quarter. It closed on SE&M Constructors in the second quarter, and SE&M knocked a good chunk of its reported growth — Everus's revenue rose 34% year over year in Q2, and management flagged that organic growth, excluding SE&M, was 30%. Everus is not a passive holder of these businesses; it is assembling manufacturing and contracting capacity as fast as the data-center boom will let it, and it is doing it on a balance sheet that is basically flush.

Here are the arithmetic and the carefully chosen words. Epsilon expects roughly $250 million of revenue in calendar 2026 and an EBITDA margin management describes as "low double digits" — call it about $25–30 million of EBITDA. Against the $295 million price, that is a multiple around 10 times EBITDA. Everus's own stock trades for roughly 16 times trailing EBITDA. Buying a business at 10 times that the market values your own at 16 times is, mechanically, a multiple upgrader — you are swapping expensive paper-free cash for cheaper earnings. And the company says the deal will be "financially cash accretive."

That phrase is doing honored, careful work, and a retail investor should read it as written. "Cash accretive" is not the same as earnings-per-share accretive. Pay for this thing partly with borrowed money and the EBITDA you get has to cover the interest first; at $295 million of debt at, say, 6%, that is about $18 million of annual interest against $25–30 million of EBITDA. Cash flows positive, sure. Whether it lifts reported EPS after interest and the amortization of the intangibles you invariably record in an acquisition is a separate and less certain question, which is presumably why Everus also declined to reconcile Epsilon's EBITDA margin to GAAP. The structure is honest; the language is chosen precisely.

The funding is the part that makes the whole thing mundane and rational. Everus ended the second quarter with $157.4 million of cash, $277.5 million of gross debt, net leverage of just 0.3 times against a stated comfort range of 1.5–2.0 times, and $222.8 million of available revolver. Whatever blend of cash and credit it used, a $295 million all-cash deal does not stress that balance sheet. A construction company with strong cash flow and near-zero leverage buying a same-margin manufacturer at a discount to its own multiple is not a gamble; it is a roll-up running on schedule.

The thing for someone who owns or watches Everus to decide is not, then, whether this one shop was worth $295 million. It is whether the strategy that the deal is part of keeps working. The risks are the ones baked into any fast roll-up: SE&M is still being integrated, Epsilon now has to be folded in too, and a backlog that leans so heavily on data centers means a single slowdown in construction spending in that niche would hit the largest part of the pipeline at once. And the stock, at roughly 16 times trailing EBITDA and above 30 times forward earnings after being up around a third year to date, has already paid you for a lot of the enthusiasm. Everus is buying cheap earnings with a clean balance sheet, which is comforting; it is paying you in multiples for the chance to keep doing it at a scale large enough to matter, which is worth watching.