Everus Construction Group ECG

Current Stats

Post-Spinoff Performance

Computed from split-adjusted closing prices. Returns are total return — (end price + dividends received − start price) ÷ start price — with no reinvestment assumed.

Post-Spin Analysis — September 16, 2026

Written against the September 15 close of $109.22. Primary sources: Everus’s Form 10-K for 2025 (filed 2026-02-25) and for 2024, its Form 10-Q for the quarter ended June 30, 2026 (filed 2026-08-05), the second-quarter 2026 results release, the Form 10 information statement (10-12B/A, 2024-10-09), the 2026 proxy statement, the September 2026 8-Ks for the Epsilon Industries closing and the credit agreement amendment, and MDU Resources’ Form 10-Ks for 2008 through 2023, which report this business as the Construction services segment. Peer figures are computed from each peer’s own 10-Q and 10-K. Price, volume and cohort figures are from this site’s own daily archive.

The one-line version

Everus is a well-run contractor at the top of a data-center construction cycle, and almost all of the margin expansion it has reported since the spin is the re-estimation of work it had already done.

The growth is real and it is extraordinary. Revenue rose 31.5% in 2025 and 30.0% organically in the second quarter of 2026. Backlog is $4.55 billion, up 53% in a year and 2.3 times what it was at the end of 2023. Management has raised its full-year guidance at four of its last five quarterly reports, and finished 2025 7% above the top of its final range.

The margin is the problem, and the company discloses it. In the second quarter, favorable changes in estimates on work performed in earlier periods added $75.7 million to gross profit — 41% of the quarter’s total, and $1.09 of its $1.64 of earnings per share. Strip those catch-ups out and operating margin has not expanded at all: it has fallen from 5.1% in 2023 to 2.9% in the second quarter. In the same quarter MYR Group’s estimate changes were worth 0.9 points of gross margin and Primoris’s were negative.

The longer record says this business has two speeds. As MDU’s construction-services segment, revenue fell 37% between 2008 and 2010 and operating income fell 59%; in 2015 revenue fell 17% and operating income 47%. Reported operating margin has never been above 8.2% in any full year since 2008. It was 9.1% last quarter.

The price has already fallen 35% from its May peak — mostly with the sector — and still asks a lot. At $109.22 the shares require today’s margin to hold permanently and revenue to compound at about 11% a year for five years. From the 2008 peak to 2025, through a full cycle, this business compounded at 6.6%. Knife River, spun from the same parent seventeen months earlier, rose 150% in its first two years and has since given back almost half its peak value.

Rating: HOLD · Target $97 · Grade: B+ (3.18)

What the business actually is

Two segments that run at very different margins and are growing at very different rates.

Segment revenue is before $23M of eliminations; segment margins are before $43M of corporate cost.

The mix has turned over in five years. Commercial work — which is where data centers sit — was 26.7% of contract revenue in 2021 and 61.8% in the first half of 2026. Utility work fell from 30.5% to 17.1%, and industrial from 22.1% to 9.9%. Everus does not disclose a data-center revenue figure, but its 10-Q says it “currently generate[s] a significant portion of our revenues from data center and other similar high tech and advanced technology contracts”, and commercial revenue rose $877 million in 2025 on “higher data center, hospitality and commercial submarket activity.”

What it sells is labor. About 9,400 employees at the end of 2025, 85% of them union, working on some 44,000 projects for 4,000 customers. 52% of 2025 contract revenue was fixed-price, 43% cost-reimbursable and 5% unit-price. T&D is the steadier half: 56% of its revenue comes from master service agreements with utilities.

The customer base is more concentrated than those project counts suggest. The top ten customers were 43% of 2025 revenue. The largest single customer — in E&M — was 16.5% of total revenue in 2025 and 18.3% in the first half of 2026, or 22.4% of the segment.

Management Team

This is the team that ran the business inside MDU, plus two outside hires in the seats a public company needs.

Thiede’s record is checkable, and it is the eighteen-year table below. He took over in 2013, ran the business through the 2015 downturn — operating income down 47% — and has grown revenue from about $1.0 billion to $3.7 billion since, with backlog up roughly tenfold to June 2026. That is a record of growth, not of margin: operating margin in his thirteen full years ranged from 4.7% to 8.2% and averaged under 7%. The hiring of Sznewajs in 2025 and two acquisitions in the following eighteen months say the strategy has changed from organic growth to organic growth plus buying.

The board

Eight directors, seven independent, elected annually. Dale Rosenthal, a former CFO of Clark Construction, chairs. Betty Wynn — chief financial officer of MYR Group from 2015 to 2023, the closest listed competitor — chairs the Compensation Committee. David Sparby, a former Xcel Energy group president, chairs Audit. Four of the seven independent directors came from MDU’s board.

The governance is clean on every conventional screen: one share class, no poison pill, no classified board, majority voting and proxy access. Stockholders cannot call special meetings or act by written consent, and the company has not opted out of Delaware’s anti-takeover statute.

How management is paid — and the bar that was set

The annual bonus is 80% EBITDA and 20% safety; the long-term plan is half EBITDA and half relative total shareholder return. There is no return-on-capital or per-share earnings measure in either.

Two facts about that design matter.

The 2025 bonus target was set inside a guidance range the market punished. In February 2025 Everus guided to $210–225 million of 2025 EBITDA — below 2024’s $232 million — and the shares fell 18.2% and 11.5% on consecutive days. A securities class action followed in April and was voluntarily dismissed in May. The bonus plan’s EBITDA target for that year was $224.1 million, inside the guided range. Actual EBITDA was $319.8 million, 142.7% of plan, and the bonus paid at its 200% maximum for every named officer. The guidance was conservative, the bar was set to the guidance, and the bar was cleared by 43%.

And EBITDA rewards buying EBITDA. With a dedicated acquisitions executive and two debt-funded deals in five months, a plan that measures EBITDA and not return on capital pays for acquired earnings at any price. The two prices paid so far have been sensible (see Capital allocation); the structure does not require that they stay so.

Insiders have bought, not sold. Directors Della Rocca and Ryan and the CFO bought in February 2025 after the guidance collapse, at $38–48; director Hernandez bought at $92 in December 2025; director Sparby bought 1,000 shares at $144.95 on August 17, 2026. The only open-market sale since the spin was the chief legal officer’s 3,300 shares at $136.69 in August 2026. Directors and officers together own under 1%.

The financial record — eighteen years, one segment

Everus has only filed two annual reports of its own, and its Form 10 carries 2021 onwards. That is exactly the window this site’s methodology warns about — a spinco’s own filings begin where the parent chose to begin them, and in this case they begin in the middle of the strongest construction market the business has seen. MDU reported this business as its Construction services segment for decades, so the longer record exists.

The join is exact. For 2021, 2022 and 2023, which MDU and Everus both report, revenue, gross profit, SG&A, operating income, net income and backlog are identical to the dollar on the two bases. MDU’s segment figures already carried the corporate allocations Everus later disclosed. The only difference is where MDU booked some 2023 interest, which does not touch operating income. So the series below is one line, not a splice.

2008–2020 from MDU Resources’ 10-Ks; 2021 onwards from Everus. 2011 is omitted rather than estimated. Backlog before about 2019 excluded time-and-materials work, so the early figures understate on today’s definition. 2019 and 2020 include two small acquisitions (Redmond, Washington and PerLectric).

Three things in that table matter.

First, the downturns are severe. From 2008 to 2010 revenue fell 37.2% and operating income 59.1% — 1.6 points of profit lost for every point of revenue, on a revenue decline far deeper than EMCOR’s U.S. electrical construction segment (−32%, ratio 1.2) or Quanta’s (−12%, ratio 1.3) over the same downturn. In 2015 revenue fell 17.2% and operating income 47.3% after large Western projects completed. This is a business whose profits fall much faster than its revenue, which is what a fixed-cost, fixed-price labor contractor does.

Second, the margin range is narrow and the current figure is outside it. Across seventeen full years, operating margin ran between 4.2% and 8.2%, and never exceeded 7.1% in any year from 2015 to 2025. The first half of 2026 was 8.4% and the second quarter 9.1%.

Third, the long growth rate depends on where you start. From 2012 to 2025 revenue compounded at 11.2% a year. But 2012 was still below 2008. Measured from the 2008 peak — one full cycle — it compounded at 6.6%. That distinction comes back in the valuation.

What the separation costs, and it is already in the numbers. Everus’s own record carries both standalone costs this site’s methodology checks for. Corporate cost, separated from the segments in every year, went from $17.4 million in 2023 to $43.3 million in 2025 and is running at about $55 million a year in 2026 — almost exactly the $25.5 million of recurring dis-synergies the Form 10 forecast. And interest expense is already standalone: the $300 million term loan has been in place since October 31, 2024, and interest was $21.5 million in 2025. Unlike most recent spinoffs, there is no hidden carve-out flattery left to remove from the 2025 and 2026 figures. What there is instead is the next section.

The first half of 2026:

$ millions except per share. Southeast Electrical & Mechanical (SE&M), bought April 1, 2026, contributed $33.4 million of second-quarter revenue.

Operating cash flow of $196.8 million flatters the half too. $98.9 million of it is an increase in contract liabilities — customers paying in advance of work. That is a sign of bargaining position in a tight market, and it reverses when the work is done.

The margin that comes from re-estimating finished work

This section changed the conclusion of this analysis, and every figure in it is from Everus’s own filings.

Everus recognizes revenue on its contracts as costs are incurred. When it revises its estimate of a contract’s final profit, the revision to the work already performed is booked immediately, in the current period, as a “cumulative catch-up.” A contractor that estimates conservatively and then executes well will book positive catch-ups routinely, and they are real profit. Everus has booked net positive catch-ups in every year disclosed. What has changed is their size.

Catch-ups and their EPS effect are as stated in the 10-K and 10-Q revenue notes (“changes in estimates associated with performance obligations that were satisfied or partially satisfied in prior periods positively net impacted operating revenues, and in turn gross profit, by $75.7 million … and diluted earnings per share by $1.09”). The margin excluding them is my arithmetic. Q1 and Q2 do not sum to H1 because the prior-period reference point differs.

Read the last two rows together. Reported operating margin rose by 2.4 points between 2023 and the second quarter of 2026. Catch-ups rose by 4.5 points of revenue over the same period. The margin on newly performed work — the price Everus is winning jobs at, net of its own higher standalone costs — has gone down, not up. All of the reported improvement, and more, is prior work turning out better than its estimate.

There is a benign reading, and it is substantial. The company attributes the favorable revisions to “labor efficiencies and favorable impacts from projected cost changes, including material costs, project risk mitigation and change orders.” In a tight market for skilled electrical labor, a contractor that bids conservatively and has the crews to execute will keep beating its own estimates, and change orders on fast-moving data-center jobs are genuine additional scope. The shift toward cost-reimbursable work — E&M’s fixed-price share fell from 63.9% to 52.5% between 2024 and 2025 — should, if anything, reduce catch-up volatility. And the contract-loss provision is tiny: $2.1 million at June 30.

The uncomfortable reading is about what is repeatable. Catch-ups are a function of conditions — scarce labor, customers paying for speed, change orders approved without argument — and those are the conditions of a boom. At the 2021–2023 rate of about 1.8% of revenue, the first half of 2026 would have produced an operating margin of about 5.1%, not 8.4% — in the lower half of the eighteen-year range. The reported margin is outside that range only because of the catch-ups.

Against its peers in the same quarter, Everus is an outlier by a wide margin:

From each company’s 10-Q for the period ended June 30, 2026. The disclosures are not defined identically, so read the magnitudes, not the decimals.

Primoris is the reminder that the same line turns negative: its Energy segment took $87.6 million of negative revisions in the first half.

Two further facts belong beside this, without implication. Everus’s annual bonus is paid on EBITDA, which includes catch-ups: the 2025 plan target was $224.1 million, actual EBITDA was $319.8 million, and the plan paid at its 200% maximum. And Everus changed auditors for 2026, from Deloitte to KPMG, with the 8-K reporting no disagreements. Estimates are audited, and nothing in the filings suggests they are wrong. The point is narrower: the single largest source of Everus’s earnings growth is also the most judgment-dependent line in a contractor’s income statement, and 2026 is the first year a new auditor examines it.

Management’s own guidance appears to assume the catch-ups fade. The August guidance of $410–425 million of 2026 EBITDA, less the $217.5 million earned in the first half, implies $200 million in the second half — up 15% on a year earlier, after first-half growth of 49%, at a margin of about 8.6% against 9.6% in the first half. Everus has beaten every guidance range it has issued, so this may be conservatism. It is also what normalization would look like.

The backlog, and who it depends on

$4.55 billion at June 30, 2026, against $2.98 billion a year earlier. The company reported “project bookings of more than $2 billion” in the second quarter alone. $3.64 billion of it is expected to convert within twelve months — about 85% of trailing revenue already in hand.

All of the growth is in E&M, and E&M’s growth is data centers. T&D backlog is flat to down, which matters because T&D is the higher-margin, MSA-based, lower-volatility half of the company — the part a utility investor would pay for. The business is becoming more concentrated in its most cyclical end market while it grows.

The customer concentration moves with it. The largest customer has been 16–19% of revenue in three of the last four disclosed periods, and the top ten are 43%. Everus does not name the customer. A backlog this size built this quickly on a handful of hyperscale-driven relationships is simultaneously the bull case’s best evidence and its single point of failure.

Price and volume: a quiet debut, then a round trip on the sector

MDU distributed Everus shares with regular-way trading beginning November 1, 2024.

The when-issued prints are in the archive but are not the day-1 price: they traded a few percent of normal volume, before the distribution.

The forced selling was modest and brief. Day-1 volume was 3.9 times the name’s own baseline, the first week averaged 2.7 times, and volume was back to normal within nine sessions. For comparison, Versant’s day 1 was 19.2 times and nVent’s 47.9 times. The likeliest reason is the index: S&P added Everus to the SmallCap 600 effective the morning of its first regular-way session, so index demand met the utility holders’ selling on the same day.

The 56% drawdown that followed was not forced selling, and it was partly self-inflicted. It came in three steps:

It ended at the April 4, 2025 tariff low, when the S&P 500 total return index was down 10.9% over the same span. The recovery came from the business: +17.3% on 6.2 times normal volume on May 14, 2025, when first-quarter revenue came in 32% higher, and +24.8% on February 25, 2026, on the 2025 results and the first 2026 guidance.

The guidance that caused the worst week in the stock’s history turned out to be wrong by 47% — 2025 EBITDA was $319.8 million against a $217.5 million midpoint.

The interval record (price return; Everus pays no dividend, so it is also the total return):

The 2026 decline is mostly the sector. From the May 6 high to September 15 Everus fell 35.4%. Every one of eight electrical, mechanical and utility contractors checked fell too — median −23.1% — with MasTec down 48.6% and MYR Group down 40.3%, the latter despite a record quarter. Everus’s largest down days line up with sector-wide events, not company ones: the early-July unwind of AI infrastructure stocks, the late-July selloff after Alphabet’s capital-spending increase and Vertiv’s miss, the August 18 rise in the 30-year Treasury yield to a two-decade high, and the September 14 slump in AI-exposed stocks. Its own news in that window was good: a second-quarter beat and a guidance raise on August 4, and a director buying stock at $144.95 on August 17.

About 12 points of the decline are Everus-specific, and no single event explains them. The candidates are the ones this analysis dwells on — the most data-center-concentrated small cap in the group, two debt-funded acquisitions in five months — and one register change: Wasatch, a 5.9% holder at March 31, had cut its stake to 2.3% by June 30.

Liquidity is not a constraint. About $55 million trades a day, volume is 0.75 times its baseline, and the float is 99.6%.

Capital allocation

What MDU took, and what it left. Unlike many parents, MDU did not starve this business of investment before separating it: gross capital expenditure ran at 1.35 to 1.91 times depreciation in every year from 2021 to 2024. It did take the cash. Net transfers to MDU totaled $464 million over 2021–2024 against $301 million of free cash flow, and at separation Everus borrowed a $300 million term loan and used it to repay $230 million owed to MDU’s holding company and pay MDU a $60 million dividend, keeping $42 million. For a business earning $232 million of EBITDA, that is a light load — net leverage of about 1x at the spin, well inside the company’s own 1.5–2.0x target.

What Everus has done with its own cash:

This is a coherent posture and a well-priced one. Everus is buying private contractors for cash at 8–12 times EBITDA while its own shares trade at about 15 times, and both targets widen the footprint — the Southeast, and prefabrication, which is the capability that most distinguishes the highest-margin peers. Epsilon is the strategically interesting one: off-site modular construction is how Comfort Systems and EMCOR convert labor scarcity into margin, rather than merely into catch-ups.

The caution is timing. Two cash acquisitions totalling about $460 million, and a term loan that went from $277.5 million to $477.5 million, at what the rest of this analysis argues is a cyclical high in the target market. Pro forma net leverage is about 1.0 times EBITDA at the top of the cycle; the 2008–2010 record says EBITDA can halve.

Valuation

The capital structure, pro forma for Epsilon. 51.06 million shares at $109.22 is a market capitalization of $5,577 million. Debt is the $477.5 million term loan; cash after paying $95 million of Epsilon’s price from the balance sheet is roughly $78 million, before third-quarter cash generation. Net debt about $400 million; enterprise value about $5,980 million. Operating leases of $90 million are excluded throughout.

Discounted cash flow

A forward DCF on a contractor at a cyclical high mostly measures the assumption about the margin, so I have built it to make that assumption visible.

Assumptions, all stated:

• Starting point: 2026 revenue of $4.85 billion (guidance midpoint $4.6 billion plus a full year of Epsilon) at a 9.2% EBITDA margin
• Revenue: grows at a five-year rate (base 10%), then fades in a straight line to 3% by 2035. The company’s own long-term target is 5–7% organic plus acquisitions; the backlog supports well above that for 2027
• Margin: glides from 9.2% to a terminal EBITDA margin by 2030 — base 8.0%, the average of 2021–2025. That average already includes the boom years’ catch-ups, so it credits the data-center mix, prefabrication and scale with holding on to part of today’s improvement. The 2016–2020 average was 7.5%
• Depreciation and amortization 0.9% of revenue; net capital expenditure 1.85% of revenue (the company’s 2.0–2.3% gross framework, less asset sales)
• Working capital 8% of each dollar of revenue growth — 2025’s revenue grew $897 million and absorbed $80 million
• Tax 26%; discount rate 9.5%, for a union labor contractor with a cyclical earnings record; terminal growth 3%

Base case: $82 a share, 25% below the price.

Sensitivity on the load-bearing assumptions — terminal EBITDA margin down the side, five-year revenue growth across the top:

About $16 a share for each percentage point of terminal margin, and about $5 for each point of five-year growth. A point on the discount rate is worth about $15 ($100 at 8.5%, $70 at 10.5%).

Read it backwards and the question becomes plain. The market price sits between today’s 9.2% margin held forever with roughly 11–12% growth, and an 8% margin with 15% growth. The first requires a margin this business has never sustained for a full year; the second requires growth faster than its thirteen-year record from a trough. Neither is impossible — the backlog alone makes 2027 look like a 15–20% growth year — but both require the current cycle to be the new normal rather than a cycle.

Earnings-based cross-check

2027 earnings on three margin assumptions, on revenue of $5.34 billion (the 2026 base plus 10%), with D&A of 0.9% of revenue, about $24 million of net interest on the enlarged term loan, a 26% tax rate and 51.3 million diluted shares:

At 20 times normalized earnings — below the peer group’s trailing median of about 28 times, because those peers’ trailing earnings are also cyclical highs while this figure is normalized — the shares are worth $102.

Price target

$97 is 11% below the current price. The third method is effectively the market’s own verdict on the whole peer group; the first two are what the business has actually delivered through a cycle. The target is a blend of the two because the honest answer is that nobody knows when this cycle ends — only that this business has shown twice what its earnings do when one does.

What this most resembles

Five cuts against this site’s universe of spinoffs, each testing one characteristic. Figures are day-1 close to the latest close.

1. The sibling — unfavorable, and the closest comparison available. MDU spun off Knife River, its construction-materials business, on June 1, 2023. Same parent, same register, the same infrastructure-spending narrative.

Knife River more than doubled in its first two years and has since given back nearly half its peak value, and now trails the S&P 500 by 7.5 points a year since spin. Everus is on the same path about seventeen months behind, and is already a third of the way down it.

The two separations were not identical, and the differences favor Everus. MDU kept about 10% of Knife River and disposed of it later in a debt-for-equity exchange, and Knife River left carrying $425 million of 7.75% notes plus a term loan and revolver, used to repay $825 million owed to MDU’s holding company. Everus was distributed 100% with a $300 million term loan. Knife River is a materials company with a heavier balance sheet; the resemblance is the path, not the business.

2. Spinoffs that doubled in their first two years — unfavorable. Across the universe, 35 names returned 100% or more in their first two years. From year two to year five their median further return was +11.8%, against +16.6% for all 215 names with both figures, and 9 of the 35 (26%) lost more than 30% from their year-two level. An early double carries no momentum in this data; if anything it mean-reverts.

3. Industrial contracting and infrastructure separations — mixed, and the failures are instructive.

The winners here sell equipment; the failures sold projects. nVent, Arcosa and Carrier make products. Babcock & Wilcox is the one to study, and it is an operating failure rather than a structural one, so it belongs in the comparison. Its losses came through the same line this analysis is about: “changes in the estimated revenues and costs to complete” fixed-price renewable energy contracts in Europe — $141.1 million in 2016 and $233.0 million in 2018, per its 2018 Form 10-K. It is the extreme version of the risk here: estimates on long projects can move in either direction, and they move together when conditions change. Amentum is a merger with a different structure and is shown for completeness, not weight.

4. First-quarter return — favorable, and already paid. Everus returned +34.3% in its first quarter, the 87th percentile of 343 names. Spinoffs with a positive first quarter have a median total return of +76.0% against +21.2% for the rest, and beat the S&P 500 on an annualized basis 27% of the time against 22%. Everus has already delivered more than the median of that better group.

5. Float — favorable. At 99.6% float, Everus is in the full-float bucket: median total return +99.4%, 30% beating the index on an annualized basis, against 12% for names with 50–90% float. There is no retained stake and no overhang.

The two cuts that describe the business and its path are unfavorable; the two that describe the distribution are favorable and have mostly played out. Everus ranks 10th of 330 names on annualized excess return, but on a window of 1.9 years, where annualizing magnifies everything.

In depth: competitive dynamics

Electrical and mechanical contracting is fragmented and local, and every company in it says so. Everus’s 10-K describes competition as “generally local and regional” and “based primarily on price, reputation for quality, safety and reliability”, and names Comfort Systems, EMCOR, IES, MasTec, MYR Group, Primoris, Quanta and Sterling alongside private firms Rosendin, M.C. Dean, Pike and ArchKey. On EC&M’s 2025 ranking of the top 50 electrical contractors, Everus is fifth with $2.85 billion of 2024 revenue — 4.8% of the list’s total, and a far smaller share of the whole market. Quanta alone is 22.7% of the list. ENR ranked Everus twelfth among all specialty contractors.

In T&D the structure is different at the top. Quanta’s $28.5 billion is several times anyone else’s; MasTec, Primoris, MYR Group and the private Pike form a second tier, and Everus’s $849 million of T&D is regional — significant in Missouri, California, Montana and Oregon. Large transmission builds behave like an oligopoly; distribution work and small projects do not.

Where Everus meets each competitor:

The union question cuts both ways. Everus’s workforce is 85% union, as is MYR Group’s craft labor; Comfort Systems and IES are essentially non-union and report segment margins of 15–19% against Everus’s 9–11%. That is a correlation, not a cause the companies disclose — mix and prefabrication differ too. Everus argues its union relationships let it “scale both up and down”, which is true and is precisely the flexibility the 2008–2010 record shows being used. 21% of its collective bargaining agreements expire within a year, in the tightest electrical labor market in decades.

Where the moat is: local relationships, a skilled union labor pool that can be deployed quickly, surety capacity, and a safety record that clears prequalification for large owners. Where it is weak: none of those confer pricing power in a downturn, and the company discloses no price-versus-volume split to test whether it has any in the upturn.

In depth: valuation against peers

Computed from the filings, not taken from a screen. Trailing twelve months to June 30, 2026; EBITDA is GAAP operating income plus depreciation and amortization for every company, so Everus’s figure here excludes the joint-venture income its own EBITDA includes. Debt includes finance leases and excludes operating leases. Prices are September 15, 2026 closes.

Each company’s 10-Q for the period ended June 30, 2026 and its 2025 10-K. IES’s fiscal year ends September; its trailing figures use nine-month periods, its share count is adjusted for its August 2026 2-for-1 split, and its P/E is flattered by $56 million of unrealized investment gains. Primoris’s trailing earnings are depressed by Energy-segment losses. Backlog definitions differ by company and are not strictly comparable.

On the face of it Everus is cheap against its growth. It trades level with the peer median on EBITDA and at a quarter below it on earnings, while growing faster than all but Comfort Systems and carrying a higher operating margin than the median.

The catch-up table is why that is not the right reading. Everus’s operating margin is 7.8% trailing — and about 3–4% excluding catch-ups that are running at roughly seven times MYR Group’s rate. On underlying margin Everus is not a 7.8% business next to MYR’s 5.6%; it is a 3–4% business next to MYR’s roughly 5%. Restated that way, its 15.5× EBITDA and 21.9× earnings are a premium to its closest analogue, not a discount.

Against the data-center leaders it is priced for what it is. Comfort Systems and IES trade at 22–27× EBITDA because they earn 12–17% operating margins, own prefabrication capacity, and do it without union labor constraints. Everus at 15.5× is not a mispriced version of them. Epsilon is the first step toward that model, and whether it works is the most important strategic question for the next three years.

In depth: is Everus an acquisition target?

Possible from November 1, 2026, strategically logical, and priced well above anything this industry has paid.

The clock runs out in six weeks. The separation was tax-free under §355, backed by an IRS private letter ruling. The tax matters agreement bars Everus, without MDU’s consent or a new ruling or opinion, from merging, selling 30% or more of its core business, or letting anyone acquire a 50% interest “until the first day after the two-year anniversary of the Distribution Date” — through November 1, 2026. Everus’s indemnity to MDU for any tax triggered by an acquisition has no cap. The same clause is why Everus has bought back no stock, and it binds Everus as an acquirer too — which is why both of its acquisitions were paid in cash.

The strategic logic is real. Specialty contracting is consolidating quickly, and the buyers are Everus’s own peers:

Multiples computed from the EBITDA each acquirer disclosed in its announcement 8-K; ranges reflect disclosed ranges or margins. Quanta’s Dynamic Systems, Primoris’s PayneCrest and MYR Group’s Valley Electric did not disclose EBITDA.

The price is the obstacle. Every precedent above is a private company, bought at 7–13 times EBITDA. Everus trades at about 15 times trailing EBITDA including Epsilon before any premium. A conventional 25–30% control premium would imply roughly 19–20 times — half again the richest multiple any of these buyers has paid for the same kind of business. A strategic buyer would be paying for the cycle at its high, in public, with its own shareholders watching.

Who could: Quanta ($92 billion) and EMCOR ($32 billion) could absorb a $6–7 billion purchase and both run union-heavy workforces that fit Everus’s 85%. MasTec and Primoris have the appetite but carry 1.8–1.9 times leverage already. MYR Group, the closest strategic fit, is smaller than Everus. Comfort Systems and IES are merit-shop and would inherit a very different labor model.

What a buyer inherits: a change of control is a default under the credit agreement, so the $477.5 million term loan would be refinanced; change-in-control severance at three times salary and target bonus for the CEO and two times for other officers; multiemployer pension participation across 85% of the workforce; and one customer at 18% of revenue.

What this means for the score: Acquisition Potential is held at 3.0. The industry, the size and the clock all point toward a deal; the price, and management’s evident preference to be the buyer, point away from one.

Tail scenarios

Upside tails

The cycle is longer than it looks (2σ). Data-center and grid spending hold through 2028, the backlog converts at today’s 9%+ margin, and 2027 revenue reaches $5.6 billion. At 14× EBITDA of $520 million less net debt, about $135 a share (+24%). This is the bull case, and it needs nothing new — only that the current quarter repeats for two more years.

Epsilon changes the business (2σ). If off-site modular construction lifts E&M’s underlying margin by two points permanently — the gap between Everus and EMCOR — the terminal margin in the DCF moves from 8% to 10%, worth about $30 a share. This is the only upside tail that is about the business rather than the cycle.

A takeout after November 1 (2–3σ). A strategic buyer paying a 25% premium — about 19 times trailing EBITDA, far above any precedent — implies about $137 a share (+25%). It is in the upside tails rather than the base case because no buyer in this industry has yet paid that multiple.

Downside tails

Catch-ups revert to their 2021–2023 rate, and nothing else changes (2σ). Removing 3.3 points of revenue from a 9.2% EBITDA margin on $4.85 billion of revenue leaves roughly $285 million of EBITDA. At the peer median of 15.4×, less net debt, that is about $78 a share (−29%) — without any fall in revenue at all.

The largest customer pauses (2σ). One customer was 18.3% of first-half revenue and 22% of E&M. Losing it — or a hyperscaler deferring a campus — takes roughly $850 million of revenue at an incremental margin well above the average. At 15× on about $340 million of EBITDA, about $92 a share (−16%), before any multiple compression.

A 2008–2010 replay (3σ). Revenue −37%, operating income −59%, and the multiple falls to 12× trough EBITDA: about $40 a share (−63%). This is not a forecast; it is what the business did the last time its main end market turned, and it started that downturn with no acquisition debt.

A large fixed-price project loss (2–3σ). 52% of revenue is fixed-price and surety bonds outstanding are $827 million. A single loss of $50–75 million — the size of one quarter’s favorable catch-ups, reversed — would be worth $1.00–1.10 of EPS and, more importantly, would reset how the market reads every catch-up since.

What would change this view

Better:

• Underlying operating margin — reported margin less catch-ups — rising above 5% for two consecutive quarters. That is the single test that would show the improvement is in the price of new work, not the re-estimation of old work
• T&D backlog growing again, which would rebalance the mix toward the steadier segment
• Epsilon disclosed as a margin contributor with prefabrication used across E&M

Worse:

• Catch-ups staying above 5% of revenue while underlying margin keeps falling
• The largest customer rising above 20% of total revenue
• Another cash acquisition taking net leverage above 1.5× before the cycle has been tested

Dated:

• November 1, 2026 — the tax matters agreement’s restrictions on a sale of the company, and on buybacks, expire
• Third-quarter results, expected early November 2026 (not yet scheduled; the 2025 report was November 4) — the first guidance including Epsilon, and the third quarter of catch-up disclosure at the new scale. The most important scheduled event for this name
• The 2026 Form 10-K, February 2027 — the first year audited by KPMG, and the first full-year catch-up figure it signs off
• Collective bargaining renewals — 21% of agreements within a year of December 31, 2025

Thesis-breaking:

• A net unfavorable period of estimate changes. Everus has not reported one in any period it has disclosed; the first would mean the tailwind has become a headwind
• Backlog falling for two consecutive quarters with no offsetting growth in T&D

Investment Scorecard

What moved and what deliberately did not.

This is Everus’s first investment grade on this site. It entered the universe as a backfill row in 2026 and has never been covered in a monthly report.

The catch-up finding is charged once, to Financial Profile. It would be easy to charge it three times — to Financial Profile for earnings quality, to Management for the bonus structure, and to Competitive Position for what it says about pricing power. It is described under management because the bonus fact belongs there, but it is not scored there: the bonus plan is ordinary for a contractor and the guidance record argues against aggressive estimating. And the downturn record is charged to Competitive Position on its own evidence — eighteen years of filings — not because of the catch-ups.

Grade and rating point in different directions, and should. B+ says this is a good business: well-capitalized, growing, competently run, cleanly separated. The rating says the price assumes the best eighteen months in its history are the base case. A B+ contractor at 22 times earnings that are two-thirds re-estimation is not a better proposition than the same contractor at 15 times earnings that are not.

Rating: HOLD. Target $97. The target is 11% below the price. It is a HOLD rather than anything stronger because the backlog makes the next four to six quarters unusually visible, the balance sheet can absorb a disappointment, and the sector has already taken a third off the price. Do not add above $97. Accumulate below $80, where the DCF no longer needs the margin to stay above 8%, and where a reversion of the catch-ups to their pre-boom rate is already in the price.

Nothing on this page is investment advice. Figures are drawn from the primary filings linked at the top and from this site’s own price archive; where a figure could not be primary-sourced, or is my own estimate, that is stated in the text.