ECG — Everus Construction Group — Initiation of Coverage

Business Overview

Everus Construction Group is a $3.75B specialty electrical and mechanical contractor that was spun off from MDU Resources on October 31, 2024. Fifteen operating companies, nineteen local brands, two segments: Electrical & Mechanical (E&M, 78% of revenue) and Transmission & Distribution (T&D, 22%).

The business was a boring $2.7B contractor doing steady mid-single-digit growth for three years. Then data centers happened. Commercial revenue — overwhelmingly data center subcontracting — went from $1.2B to $2.1B in a single year (+73%). That one submarket accounted for $877M of the $890M E&M revenue increase. Strip out data center and E&M was approximately flat. The T&D segment grew 1%.

How it makes money: ECG deploys skilled electrical and mechanical tradespeople (85% IBEW union) to wire and plumb large construction projects. They work as specialty subcontractors to general contractors or directly for end-use customers. Revenue is recognized over time using cost-to-cost percentage-of-completion. The cost structure is ≈88% cost of sales (dominated by labor), ≈5% SG&A, leaving ≈7% operating margin. This is an asset-light, labor-intensive business — capex is 1.8% of revenue, rising to 2.2-2.4% as they invest in prefabrication facilities.

Contract mix: 52% fixed-price, 43% cost-reimbursable, 5% unit-price. The shift toward cost-reimbursable (from 34% in FY2024) is significant — it partially mitigates tariff risk on materials, as cost-plus contracts pass through price increases. But the majority of work is still fixed-price, where material cost inflation eats margin directly.

Scale context: Ranked 12th on ENR Top 600 Specialty Contractors, 5th on EC&M Top 50 Electrical. Mid-tier by revenue. Smaller than FIX ($9B), EME ($16B), PWR ($26B), MTZ ($13B). Comparable to IESC ($3.4B) and MYRG ($3.5B).

Financials

Four-Year Trajectory

The financial story has three distinct chapters. FY2022-2024 was steady-state: $2.7-2.9B revenue, gradual margin expansion from project mix improvement, healthy FCF conversion north of 90%. FY2025 was the inflection — revenue jumped 31.5% on the data center wave, margins expanded another 40bp, net income grew 41%. FY2026 guidance tells a third story: revenue +10% to $4.1-4.2B, but EBITDA flat to +5% at $320-335M, implying 30-90bp of margin compression. The deceleration is real and acknowledged.

The Margin Question

This is the single most important financial debate on ECG. There are three different margin numbers floating around:

8.5% EBITDA margin — FY2025 reported. This is what trailing multiples capitalize.

7.6-8.2% — FY2026 guidance range. Management's own expectation, incorporating "normal" execution assumptions.

7.0-7.5% — What CFO Marcy explicitly called the normalized run-rate on the Q2 2025 call: "If you adjust for the strong execution that has benefited our results this year, our margins have been relatively consistent in the low to mid-7% range." He doubled down in Q3: "We believe our fourth quarter projected margin is a good starting point for our 2026 outlook."

8.4% — What Q4 actually delivered. Q4 2025 EBITDA was $84.8M on $1.01B revenue = 8.4% margin (8-K, Feb 24, 2026). This is the number the CFO said to use as a "starting point for 2026." At 8.4%, the starting point is essentially the FY2025 trailing average (8.5%), not the 7.0-7.5% "normalized" range Marcy quoted mid-year. Either Marcy was being deliberately conservative (consistent with a management team that beat guidance by 65% average across four quarters), or Q4 benefited from the same execution upside he said to strip out. The answer matters: 8.4% starting point with some compression yields the high end of guidance (8.2%); 7.0% starting point would imply margin below the guide range itself.

The gap between market-implied margin (≈9.3%, reverse-engineered from the forward P/E) and guidance (7.6-8.2%) is 110-170bp — narrower than the 180-230bp gap versus the "normalized" 7.0-7.5%. On $4.15B of revenue, the gap is $45-70M of EBITDA. This is still the most important debate on the stock, but the Q4 actual result tilts the evidence toward the higher end of the guidance range.

Catch-Up Earnings

Cumulative catch-up estimate revisions contributed $1.64/share to FY2025 EPS of $3.95 — that's 41.5%. These are revisions to prior-period project profitability estimates that flow through the current period under percentage-of-completion accounting. The trend is accelerating: $0.67/share (FY2023), $1.21 (FY2024), $1.64 (FY2025). For large projects individually exceeding $1M impact: $12.9M (FY2023), $28.6M (FY2024), $60.1M (FY2025) — nearly 5x growth in two years.

Two interpretations. Bull: management is systematically conservative on initial estimates, which means real earnings power is higher than first recognized — the catch-ups are proof of better-than-expected execution. Bear: profitability is being recognized faster than earned, and the trend could reverse when large projects hit unexpected costs. The truth is likely somewhere in between, but the magnitude matters — at $1.64/share, the run-rate EPS ex catch-ups is ≈$2.31. At $122 per share, that's a 52.9x P/E on run-rate earnings.

Cash Conversion Deterioration

FCF declined from $152M (FY2023) to $129M (FY2024) to $100M (FY2025) while net income grew from $137M to $143M to $202M. FCF conversion (FCF/NI) fell from 111% to 90% to 50% — a steady three-year decline, not a sudden collapse. For context, FY2022 operating cash flow was negative (-$25.5M), showing working capital swings can be violent in this business. The current deterioration is driven by growth: contract assets grew $89M (work performed but not yet billed) and receivables grew $180M. Capex is ramping from $48M to $67M and guided to $90-100M.

The EBITDA-to-FCF conversion is 31% (vs 55.5% in FY2024). On $6.2B market cap, FCF yield is 1.6%. If revenue growth moderates to the guided 10%, working capital absorption should moderate proportionally — FY2024's 55.5% conversion at 6% revenue growth supports this. But the market is paying a growth multiple for a declining FCF profile, which works only if growth moderates enough for cash to catch up without the multiple compressing.

Balance Sheet

Net debt $114M (0.4x EBITDA). $375M total liquidity. Covenant headroom to 3.0x net leverage implies theoretical M&A capacity of $500-700M in incremental debt. No dividends, no buybacks — all capital reserved for M&A and organic growth. The balance sheet is a weapon that hasn't been deployed.

Competitive Position

The industry is fragmented — thousands of contractors, no one with more than low-single-digit market share. ECG's moat is narrow and time-limited:

Labor scarcity (moderate moat). 85% IBEW-unionized workforce provides scalable labor access during a 439,000-worker industry shortage. Data center electrical work pays 30%+ premiums. It takes years to build a trained workforce at scale. But labor isn't proprietary — workers move between contractors, and every major peer (FIX, EME, IESC) has comparable union relationships. The moat exists because demand is growing faster than workforce supply, not because ECG has unique labor access.

Customer relationships (moderate moat). 28 years of operating history, 20-30 year utility MSAs in T&D, design-assist involvement that embeds ECG early in project lifecycles. Top 10 customers now represent 43% of revenue (up from 33%), suggesting deepening relationships. But contracts are awarded competitively, and hyperscalers work with multiple contractors simultaneously.

Execution track record (moderate moat). Ranked 5th in EC&M Top 50. Complex data center work favors incumbents with proven on-time, on-budget delivery. But this is table stakes among top-10 contractors.

What's commoditized. The core service. The 10-K lists price FIRST among competitive factors, ahead of reputation for quality and safety. This is not a differentiated product business — it's skilled labor deployment where the moat is consistency at scale.

The Margin Gap

The most important competitive question: ECG's EBITDA margin (8.5%) trails peers. But which peers? FIX at 16% is a misleading benchmark — FIX operates 3M+ sq ft of modular manufacturing facilities, increasingly a manufacturer-contractor hybrid with structurally higher margins, plus decades of mature SG&A absorption as a public company. The fairer comparison is EME's Electrical Construction segment at 11.3% EBITDA margin — same trade, closer business model. Against EME Electrical, the gap narrows to 280bp.

Possible explanations for the remaining gap: (a) less prefab/modular capability at ECG's scale; (b) younger data center practice still climbing the learning curve; (c) ≈$42.6M in corporate overhead (vs $29.1M allocated pre-spin) still being absorbed; (d) EME has been expanding into new geographies with "lower profitability where reduced labor productivity" — suggesting ECG may face similar margin pressure as it expands.

Whether this 280bp gap to EME Electrical converges determines whether ECG is a value trap or a margin expansion story. The Q4 2025 actual margin of 8.4% is already closer to EME than the mid-year normalized quotes suggested.

Peer Valuation

ECG trades at a discount to FIX (34.5x vs 61.4x) — deserved given the margin gap. Premium to EME (32.3x) and IESC (30.0x) — reflecting higher DC concentration and recent momentum. The group as a whole is expensive.

Management & Governance

CEO Jeff Thiede — 28-year tenure, built the business from within MDU Resources. Operator DNA, knows every operating company. On calls he's direct and operationally detailed — talks about moving crews between geographies, project selection discipline, customer relationships. Not a promotional CEO.

CFO Max Marcy — First standalone CFO. Notably candid about margin normalization, explicitly guiding the market away from trailing margins. This is unusual — most CFOs let the market over-earn on consensus. Marcy is either genuinely conservative or trying to set a low bar.

VP Corp Dev Tim Snevis — Hired mid-2025, 20 years of middle-market construction M&A banking. The signal: geographic acquisition strategy is being professionalized. The constraint: Tax Matters Agreement restricts material transactions until October 2026.

Board: Seven directors identified. No DEF 14A filed yet (proxy due ~April 2026). Board quality, independence, and compensation structure cannot be assessed.

Insider ownership: Minimal open-market buying. One director (Helena Hernandez) purchased 250 shares at $92.18 in December 2025 — $23K, negligible. No executive purchases. All other transactions are RSU grants. This is not the bullish insider signal you'd want from a management team that built the business over three decades.

Capital allocation track record: Conservative to a fault. Net leverage 0.4x vs 1.5-2.0x target. No acquisitions executed despite active pipeline. No shareholder returns. The VP Corp Dev hire and "funnel is broader and deeper" language (Q3 2025) suggest something should land in 2026-2027, but execution is unproven.

Factor Profile

Factor regression (250 trading days, SPY + MTUM + XLI):

              β         Var Contribution
MTUM        +1.93        45.3%    ← DOMINANT
SPY         -0.72       -12.4%    (artifact of MTUM collinearity)
XLI         +0.26         4.3%    (not significant)
Idio                     62.8%

α = 90.8% annual    σ_idio = 47.4%    R² = 37.2%

63% idiosyncratic — below the 75% target. The dominant factor is momentum (MTUM at 45% of variance). This isn't ECG-specific — the entire DC construction peer group loads heavily on momentum (FIX 63%, IESC 57%, EME 57% momentum variance contribution).

Style decomposition reveals two additional significant exposures:

• Small-cap tilt (SMB β = +1.22, t=3.96): ECG behaves like a small cap despite $6.2B market cap. Low analyst coverage (7), recent IPO, retail-heavy holder base.
• Growth tilt (HML β = -1.22, t=-3.70): Trades as a growth stock. When value rotates, ECG underperforms.

Market beta (SPY) and sector beta (XLI) are NOT statistically significant after controlling for momentum and style. ECG is more correlated with QQQ (0.557) than XLI (0.506) — it's trading as a tech proxy, not an industrial.

Versus peer basket (equal-weight FIX + EME + IESC): 59% idiosyncratic, 0.79 beta, ≈0% alpha. ECG's outperformance over the market comes from the group's momentum loading, not ECG-specific alpha orthogonal to peers.

Volatility: 59.9% annualized (vs SPY 19.4%). Positive skew (+1.23), extreme kurtosis (8.42). Max drawdown -20.3%. Up/down capture ratio vs XLI: 1.64x (favorable asymmetry — captures 1.92x upside, only 1.17x downside).

Caveat: 15 months of trading history. ECG began trading November 1, 2024. This regression captures the entire post-spin price discovery period, initial momentum rally, and investor base formation. The factor loadings — particularly MTUM at 45% and the SMB/growth tilts — may be artifacts of this discovery phase rather than persistent factor exposures. As institutional ownership deepens and analyst coverage expands beyond 7, these loadings should moderate. Treat this factor profile as preliminary; re-run in 6-12 months with a full cycle of data.

Bottom line on factor profile: This is currently trading as a momentum + small-cap growth stock dressed as an industrial contractor. Two bets compound: (1) the DC construction momentum factor (≈45% of variance), available cheaply via any peer or MTUM; and (2) ECG-specific drivers (≈63% idio) — backlog execution, M&A, margin trajectory. To own ECG specifically versus peers, you need an ECG-specific thesis. But given the short trading history, don't over-anchor on the 63% idio figure.

Forward Expectations Gap

Current Pricing (Feb 26, 2026)

What Current Price Requires

The forward P/E of 26.6x implies ≈$4.59 EPS. Management guides net income of $205-215M, or $4.02-4.22 EPS. The market is pricing an 11% beat over guidance midpoint.

For current price to be fair at 25x (reasonable for a 10% grower), ECG needs $4.89 EPS — requiring 9.5% EBITDA margin on $4.2B revenue. That's ABOVE FY2025's record 8.5%, which the CFO called non-recurring.

The Disconnects

Margin: the big one. Market-implied EBITDA margin (≈9.3%) vs management guidance (7.6-8.2%) vs CFO's mid-year normalized quote (7.0-7.5%). But Q4 actual came in at 8.4% — and the CFO said Q4 is the "starting point for 2026." If the starting point is 8.4% and guidance midpoint is ≈7.9%, you get modest compression, not a reset. The market is still pricing above guidance, but the gap is narrower than the mid-year "normalized" quotes implied. The resolution depends on whether Q4's 8.4% reflected the same execution upside management warns is non-recurring, or whether it represents the new steady state.

Serial sandbagging. ECG beat consensus EPS by 67%, 75%, 80%, and 41% in four consecutive quarters — averaging 65%. This is not stale consensus; this is a management team systematically setting a low bar and clearing it by a wide margin. The pattern applies to guidance too: Q2 2025 EBITDA guidance of $240-255M became $319.8M actual (+25-33% beat). If the same conservatism is embedded in 2026 guidance ($320-335M EBITDA, $4.02-4.22 EPS), even a modest 15-20% beat yields EPS of $4.62-5.06, putting the forward P/E at 24-26x — expensive for a 10% grower but in-line for a company with this secular tailwind. Analysts are scrambling to catch up — DA Davidson just bumped from $102 to $130 post-earnings. But even post-revision, the stock trades above 6 of 7 targets.

Revenue: aligned. Guidance ($4.1-4.2B) was 5-8% above pre-earnings consensus. Street will revise to match. No disconnect here.

FCF: unresolved. Market pays a growth multiple but FCF is declining. Cash conversion can only improve if (a) revenue growth slows (reducing working capital absorption) or (b) billing cycles accelerate. Neither is guided. The implicit bet is that growth moderates just enough for cash to catch up, without the growth multiple compressing.

Options: mixed. Open interest is heavily call-biased (P/C 0.13). But on Feb 26, 1,000 $100 puts traded against 16 OI — a 62.5x ratio, indicating a large new protective position at $100 (18% below current). Someone is hedging.

Key Risks

1. Data center concentration. Management ADDED a new risk factor in the FY2025 10-K — did not exist in FY2024 — explicitly warning: "uncertainties surrounding the longevity of this type of work." Commercial (DC-heavy) revenue went from $1.2B to $2.1B in one year, and the growth driver is formally acknowledged as potentially transitory. Without DC growth, E&M revenue would have been approximately flat.

2. Customer concentration. Single customer: <10% → 17% ($638M) in one year. E&M level: one customer 21.1%, another 10.4%. Neither customer is named across four quarters of earnings calls. The opacity is deliberate — prevents investors from sizing single-customer risk.

3. Margin compression. FY2026 guidance implies 30-90bp margin decline from trailing. But the picture is less clear than it appeared mid-year. The CFO quoted 7.0-7.5% normalized in Q2-Q3, then Q4 came in at 8.4% and he called it "a good starting point for 2026." The guidance range of 7.6-8.2% sits between these signals. Drivers of potential compression: loss of execution upside (one-time project closeouts, catch-up earnings timing), insurance cost increases ("not going down anytime in the future" — CFO), and standalone public company costs still being absorbed. But if Q4's 8.4% is the true starting point, the compression risk is 20-80bp, not the 100-150bp the mid-year quotes implied.

4. Tariff exposure on fixed-price contracts. 52% of revenue is fixed-price, where ECG "generally cannot adjust prices." Materials at risk: copper, aluminum, steel, electrical components, plastics. Partially mitigating: E&M cost-reimbursable revenue doubled from $670M to $1,347M (33% → 46% of E&M), and these contracts pass through material cost inflation. The shift is rapid and structural — hyperscaler customers increasingly prefer cost-plus for schedule certainty. But the fixed-price book ($1.95B) remains exposed.

5. Backlog lumpiness. Book-to-bill was below 1.0x in Q2 and Q3 as revenue ran at $1B/quarter and sequential backlog declined. But Q4 reversed: backlog recovered from $2.95B to $3.23B (+$283M sequential, +16.1% YoY), implying Q4 book-to-bill of ≈1.28x (8-K, Feb 24, 2026). Year-end backlog is a record. The mid-year declines look more like timing lumpiness during peak execution quarters than demand deterioration — particularly given Eaton's 11-year electrical equipment backlog commentary. But order visibility beyond the backlog remains limited, and revenue beyond the guidance horizon depends on continued bookings.

6. Tax Matters Agreement. Restricts material M&A, buybacks >$100M, and equity issuances until ~October 2026 (uncapped indemnity to MDU). Capital allocation flexibility is constrained during the prime window for geographic expansion.

7. Catch-up earnings sustainability. $1.64/share (41.5% of EPS) came from prior-period estimate revisions. Trend is accelerating but could reverse. Magnitude makes reported EPS a poor proxy for steady-state earnings power.

8. Multiemployer pension costs. $309M in annual contributions (8.2% of revenue), up 28% YoY. 21% of CBAs expire within one year. IBEW Local 683 contributions nearly doubled, reflecting workforce ramp in DC markets.

9. Floating rate debt. The $285M term loan is SOFR + applicable rate (5.67% weighted average as of Dec 31, 2025). Interest expense was $21.5M in FY2025 (8.1% of operating income). A 100bp rate move on the term loan balance = ≈$2.85M pre-tax impact. Not existential, but unhedged rate exposure exists.

10. Key person / succession. CEO Thiede is a 28-year veteran who IS the institutional knowledge. No succession plan is discussed in filings. The management team is thin for a $6.2B market cap company. CFO Marcy and VP Corp Dev Snevis are both relatively new hires. If Thiede departed, the operational continuity risk is real.

11. Disputed project. A customer is withholding $31.3M on a large time-and-materials project since October 2023. ECG has filed a mechanics' lien. Revenue was recognized but cash has not been collected. Outcome uncertain. Additionally, $69.9M in change orders are included in revenue but not yet formally executed by customers — up 24% YoY. Both items create collection risk.

12. JV opacity. Equity method JV income was $16.3M in FY2025 (8.1% of net income), up from $5.0M in FY2023 — 3.3x growth in two years. The JV counterparties, project types, and margins are undisclosed. This is a material and growing earnings contributor with zero transparency. If JV profitability reverses, the impact on net income is non-trivial.

What to Watch

Near-term (next 2 quarters):

• Q1 2026 earnings (May 12): Does margin reset to ≈7% as guided, or does execution upside repeat? Street estimate $0.79 EPS.
• Book-to-bill recovery: Does backlog stabilize or continue declining sequentially?
• TSA expiration (March 2026): Any cost surprises from standing up remaining functions?
• First M&A announcement: Snevis pipeline, geographic expansion target. Quality of deal, price paid, accretion math.

Medium-term (2-4 quarters):

• Tax Matters Agreement restriction lifts ~October 2026: Opens full M&A/capital allocation flexibility.
• DEF 14A proxy filing (~April 2026): First full disclosure of compensation, board, governance.
• ASU 2024-03 (effective FY2027): Will force disaggregation of cost of sales — reveals labor vs materials split for the first time.
• Customer concentration trajectory: Does 17% single-customer exposure grow or diversify?

Structural:

• Hyperscaler capex trajectory: META ($115-135B), AMZN (≈$200B), GOOGL ($75B+) — any capex guide-down is a direct headwind to ECG's growth.
• DC construction capacity expansion: ETN says 11 years of backlog, but new contractors are entering. As workforce expands, ECG's labor scarcity moat erodes.
• Margin convergence with EME Electrical (11.3%): If ECG's margins expand toward the fairer peer benchmark (from 8.5% toward 11%), the stock rerate opportunity is meaningful. If the 280bp gap is structural, current valuation is full.
• Interest rate environment: Affects both ECG's borrowing costs and hyperscaler project economics.

Source Register