IES Holdings

Phase A — Understand the business

Lens 1 · Company Overview

IES Holdings designs and installs integrated electrical and technology systems and provides infrastructure products/services across four self-managed segments; the Sugar Land, TX corporate office (plus a Greenwich, CT executive office — Gendell/Tontine's seat) handles only capital allocation, M&A, and segment-leadership selection. The four segments and their FY2025 revenue share:

• Communications — $1,140.6M, 33.8% of revenue (+46.9% YoY). Nationwide network-infrastructure for data centers (co-location + managed hosting), e-commerce distribution centers, high-tech manufacturing; serves "Fortune 100 and 500" tech/social/e-commerce brands. 41 offices, HQ Tempe AZ. Originally established 1984. This is the data-center pure-play segment.
• Residential — $1,304.4M, 38.7% (−6.1% YoY). Single-family + multi-family electrical, plus HVAC & plumbing in some markets; 99 locations, Sun-Belt/Western/Mid-Atlantic/Midwest/Northeast. Majority of single-family revenue from Texas and Florida. Still the largest segment by revenue but shrinking and margin-pressured.
• Infrastructure Solutions — $498.7M, 14.8% (+42.0% YoY). Electro-mechanical: custom generator enclosures (used in data centers), metal-enclosed bus duct, power-distribution equipment, structural steel; plus industrial apparatus repair (AC/DC motors, gen sets). 15 locations, HQ Massillon OH. Highest-margin segment and a second data-center lever.
• Commercial & Industrial — $427.7M, 12.7% (+16.3% YoY). Electrical/mechanical design-build & maintenance for data centers, industrial facilities, offices, healthcare, T&D, wind/solar. 17 locations, HQ Houston TX.

Contract structure / payment terms: A "significant portion" of revenue is under fixed-price contracts — bid-risk and commodity exposure (copper, aluminum, steel) live here; percentage-of-completion accounting. Q2-FY26 disaggregation shows the mix: of $974.3M total, $783.4M fixed-price vs $190.9M time-and-material. Communications is the most T&M-weighted ($148.0M of $367.7M) — a quality tell, since T&M de-risks the data-center growth. Backlog converts to revenue over ~1–2 years (mgmt: ~$1.4B of the FY25-end backlog lands in FY26, ~$1.0B in FY27).

Customers: Diversified — no single customer >10% of consolidated revenue in FY2025; one Residential customer was 12.0% in FY2024/FY2023. Concentration risk is segment-level, not consolidated.

Lens 2 · Supply Chain

Map: commodity & component suppliers → IES segments (fabrication/installation) → OEM / EPC / end-customer.

• Upstream inputs: electrical fixtures & system components, copper, aluminum, raw steel, certain plastics; plus long-lead electrical switchgear and power generators (lead times "several months or more"). Generally multi-sourced domestically at competitive prices, so no single-supplier chokepoint on raw material — but switchgear/generator lead-time is the real bottleneck and it is industry-wide (the same shortage throttling all data-center build-out). IES partially hedges via early commodity buys and escalation/escape clauses, but "such protections are not included in every contract."
• The company: four segments add fabrication (generator enclosures, bus duct, structural steel — now hugely expanded by Gulf Island's 450,000-sq-ft Houma, LA campus ) and field installation labor.
• Downstream / route to end-customer: Custom-engineered products are "principally sold in partnership with an OEM or to an EPC firm on behalf of the end-user". Communications sells direct to hyperscaler/co-lo/enterprise data-center owners as a preferred provider. The named end-buyers are the hyperscalers and co-location operators driving data-center capex (IES doesn't name them in filings beyond "Fortune 100/500"; customer identities are confidential).
• Chokepoints / single-source dependencies: (1) skilled craft labor — repeatedly flagged as the growth constraint ("the pace of growth in this business may also be slowed by the availability of labor"); (2) two primary surety-bond providers — bonding access is "at the sole discretion of our surety providers," a genuine single-point dependency for winning large contracts; (3) long-lead switchgear/gen sets as above.

This lens is grounded in the 10-K's own supply/sourcing disclosures (no commercial-layer supply-chain.md exists for the datacenters topic — it is "missing" per the Step-0 briefing).

Lens 3 · Competitive Advantages (moats)

For a fragmented, low-barrier-to-entry contracting industry, IES's moats are real but bounded:

• Balance-sheet / bonding moat (the strongest one). IES competes mostly against "small, privately owned contractors who generally have limited access to capital." Its debt-free balance sheet and surety relationships let it bond and bid large, mission-critical projects competitors cannot. In electrical contracting, access to capital + bonding IS the moat.
• Preferred-provider / switching costs in data centers. Mission-critical facilities "significantly rely upon our past performance record, technical expertise and specialized knowledge"; a "significant portion" of Communications volume is "long-term, repeat customers" using IES as a preferred provider. Re-qualifying a structured-cabling/power vendor on a live data-center program is costly — modest but durable switching cost.
• Scale + capability breadth. 174 locations and the ability to self-perform across electrical, mechanical, fabrication, and structural steel (post-Gulf Island) let IES offer a one-stop scope national homebuilders and data-center owners value.
• Process / execution edge (margin proof). Consolidated gross margin climbed 18.7% → 24.2% → 25.5% (FY23→24→25) and Communications gross margin 19.7% → 23.2% YoY, attributed to "a more disciplined bidding process" and "successful project execution". Margin expansion of this magnitude in a commodity trade is itself evidence of an operating-discipline moat.

Bargaining power: Over suppliers — moderate (multi-sourced commodities, but exposed on long-lead switchgear). Over customers — improving in data centers (preferred-provider, record backlog, labor scarcity gives IES pricing power), weak in residential (homebuilders hold the whip; "few barriers to entry" and price competition). Net: the moat is widening exactly where growth is (data centers) and thinnest where the drag is (residential). No commercial-layer positioning.md exists (missing per Step-0), so this lens is filing-grounded.

Lens 4 · Segments

FY2025 segment P&L, all ``:

The story in one line: the two data-center segments (Communications + Infrastructure Solutions) now do $1,639M, ~49% of revenue and carry the highest/rising margins, while the largest single segment (Residential, 38.7%) is in revenue decline. The mix is shifting toward the high-margin, secular-growth half — the bull's whole case.

Geography: Single-family revenue concentrated in Texas & Florida; multi-family in Texas + Midwest/Mid-Atlantic/Southeast. Otherwise diversified national footprint; no foreign-revenue concentration disclosed.

Latest-quarter inflection (Q2-FY26, quarter ended 2026-03-31):

• Communications op income $61.2M @ 16.6% margin (from $39.6M @ 14.5%) — still climbing.
• Residential op income collapsed to $6.4M @ 2.2% margin (from $22.7M @ 7.1%) — the housing drag is now severe and is the single ugliest line in the model.
• Infrastructure Solutions $41.9M @ 21.8% (rev +64% to $192.4M, Gulf-Island-aided).
• C&I $21.4M @ 16.9% (margin expansion).

---

Phase B — Measure performance

Lens 5 · Earnings Result — latest print (Q2 FY2026, quarter ended 2026-03-31)

Headline:

• Revenue $974.3M, +16.8% YoY ($833.96M PY). Driven by Communications (+34.7% to $367.7M), Infrastructure Solutions (+63.6% to $192.4M, Gulf Island contributed $37.5M), C&I (+0.9%); Residential −9.5% to $287.6M.
• Operating income $112.3M, 11.5% margin (from $92.7M, 11.1%).
• Net income attributable $109.9M, +55.6% ($70.7M PY); diluted EPS $5.44, +55.4% ($3.50 PY).
• Gross margin 26.2% (from 25.0%) — up in Communications & C&I, down in Residential & Infrastructure Solutions.
• H1-FY26: revenue $1,845.2M (+16.5%), diluted EPS $9.95 (from $6.22), operating cash flow $131.0M (from $62.1M).

vs consensus / market reaction: The print was a clear beat — the stock +17.8% after the Q2 report. Net income "+56%" and "backlog surges" were the headlines.

Guidance/outlook tone: No numeric EPS guide (IES doesn't give one), but the qualitative tone strengthened: FY25 10-K said data-center demand "remains particularly strong" and backlog is at "record levels," while flagging Residential weakness into FY26; FY26 capex guide raised hard to $110–130M (from $67.3M actual FY25) — a capacity build-out for Infrastructure Solutions + Communications.

Balance-sheet flags (Q2-FY26):

• Cash $48.7M (down from $127.2M FY25-end) — deployed into Gulf Island + the $40M Edmonson NCI buyout.
• First debt in years: $35M drawn on the $300M revolver (vs $0 at FY25-end) to fund Gulf Island. $253.7M availability; in compliance; max leverage covenant 3.0x with enormous headroom. Effectively still net-cash-ish — $35M debt against $48.7M cash + a $63.9M Jett/CB&I equity-method stake.
• Total equity $1,072.7M (from $884.0M) — retained earnings compounding fast.
• Receivables/WC building with revenue (normal for a growing contractor).

Anything unusual vs its own history: (1) taking on debt for the first time; (2) buyback essentially switched off — only 4,112 shares repurchased in H1-FY26 @ $418.31 vs 139,362 @ $178.40 in H1-FY25 — management is hoarding cash for M&A at >$400/share rather than buying back at >$700; (3) a $2.8M remeasurement gain on the previously-held Gulf Island toehold (Gendell-style: build a trading-security stake, then take the whole company).

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research-layer shelf (transcripts=0 per Step-0). From web + filings, the management narrative has shifted decisively toward data centers/infrastructure and away from residential over the last ~4 quarters:

• Recurring phrases: "record backlog," "data-center demand remains particularly strong," "scalability of the business," "disciplined bidding".
• What they've started saying: nuclear/oil-&-gas optionality via Gulf Island ("attractive opportunities in oil & gas as well as the potential to support U.S. nuclear buildout over time").
• What they've de-emphasized: residential growth — now framed defensively ("we expect a reduction in multi-family revenue for fiscal 2026").
• Tone: confident on demand, consistently cautious on labor availability as the throttle. Provenance caveat: sentiment here is inferred from filings + press, not a transcript read — label /, not a verbatim call analysis.

Lens 7 · Comps

Peer set = IES's own proxy peer group (Comfort Systems, MYR Group, Sterling Infrastructure, Primoris, Installed Building Products) + Quanta (the bellwether). Multiples are ``, dated; where not sourced, "n/a." Never fabricate.

Read: The entire data-center-infrastructure cohort is richly valued (FIX/PWR ~53x, STRL ~79x) on the same secular thesis. IESC at ~38.7x trailing is the cheapest large name in that cohort — and it grew EPS ~50%+, has zero net debt, and the highest incremental-margin segment. A bull frames IESC as the relative-value way to own the theme. A bear notes (a) trailing P/E understates richness because earnings are near a cyclical/data-center peak, (b) IESC deserves some discount for the ~54% Tontine control overhang, thin float, and residential cyclicality, and (c) every multiple here is hostage to the same single macro variable (hyperscaler capex). 5-yr average ROE not separately sourced — but FY25 ROE ≈ $306.0M / ~$748M avg equity ≈ ~41%, an exceptional return-on-equity for a contractor.

Lens 8 · Stock-Price Catalysts (what moves it >5%)

IES is one of the great small-cap compounders of the cycle. Five-year cumulative total return (index to $100 at 2020-09-30):

~19x in five years — beating even its high-flying peer group (~11x) and crushing the Russell 2000 (~1.7x). And it kept going: from $359.63 (2025-11-17) to an all-time-high $749.83 (2026-06-12) and ~$712 (2026-06-19) — roughly another double in seven months.

What the tape reacts to (mostly ``):

• Earnings beats + backlog prints — +17.8% on the Q2-FY26 beat. This is the dominant catalyst.
• Data-center/AI-capex sentiment — IESC now trades as an AI-infrastructure derivative; hyperscaler capex headlines move it.
• M&A — the Gulf Island deal (2026-01-16) was received as strategically expanding data-center + nuclear capacity.
• Sentiment/liquidity air-pockets — "a stock that can shed nearly 7% in a single session without a major negative news catalyst"; thin float (only ~46% non-Tontine) amplifies moves both ways.
• Insider sells — Gendell sold ~$9.34M of stock — periodically flagged, though trivial vs his ~54% stake.

(Phase B note — operating variant: Lens 5/7 run as-is; no clinical/private overlay.)

---

Phase C — Judge people & books

Lens 9 · Management

• Jeffrey L. Gendell, 66 — Executive Chairman (since 2025-07-01; was CEO 2020–2025, Chairman since 2016). Founder/managing member of Tontine Associates, IES's ~54% controlling shareholder; ex-Odyssey Partners, started at Smith Barney 40+ yrs ago. This is the most important fact about the company. Gendell runs IES as a long-horizon capital-allocation vehicle — he still "focuses on strategic issues...organic growth, acquisitions and capital allocation". The IES record on his watch is extraordinary (19x/5yr, gross margin 18.7%→25.5%, EPS $4.54→$15.02). The flip side: a controlling shareholder whose interests can diverge, and who can sell/trigger change-of-control provisions across the credit/surety/severance agreements.
• Matthew J. Simmes, 50 — President & CEO (since 2025-07-01). 31-year IES insider (ran IES Communications 2017–2021, COO 2021–2023, President & COO 2023–2025) — the orderly, planned succession of an operator who built the data-center segment. Continuity, not a pivot.
• Tracy A. McLauchlin, 56 — SVP/CFO since 2015 (CPA, ex-PwC/Dynegy). Long tenure spanning the entire turnaround.
• Track record (quantified): EPS $4.54 (FY23) → $9.89 (FY24) → $15.02 (FY25) diluted; operating margin 6.7%→11.4%; near-41% ROE. Capital allocation = bolt-on M&A (Greiner FY24 $67M; multiple FY25 deals $52.4M; Gulf Island FY26 ~$152M EV) + the $40M Edmonson NCI buy-in + a Tontine-style investment portfolio ($104.6M marketable securities + a $63.9M Jett/CB&I equity-method stake financing the McDermott storage-business carve-out). Buybacks are opportunistic — heavy at <$180 (FY25), switched off at >$400 (FY26). This is disciplined, value-aware allocation, not empire-building.
• Skin in the game: ~54% via Tontine — extreme alignment (and extreme control).
• Red flags: (1) the controlling-shareholder overhang itself; (2) Gendell's $9.34M insider sale; (3) the equity-investment portfolio (Jett/CB&I, marketable securities) injects non-core, harder-to-value risk onto an otherwise clean contractor balance sheet — a "hedge-fund-manager-runs-an-industrial" quirk that bulls should underwrite consciously. Founder/owner-operator archetype, decisively — with all the upside and key-man/control risk that implies.

Lens 10 · Forensic Red Flags

Acting forensically across the three statements:

• Revenue recognition: percentage-of-completion on fixed-price contracts — inherently estimate-laden ("uncertainties inherent in the use of percentage-of-completion accounting, which could result in the reduction or elimination of previously recorded revenues and profits"). The #1 accounting-risk surface. No restatement; auditor attestation on ICFR is present; no error-correction checkbox marked.
• Cash vs earnings: clean — FY25 operating cash flow $286.1M vs net income $311.8M (0.92x); H1-FY26 OCF $131.0M vs net income $202.1M (lower conversion as WC builds with growth — normal for an accelerating contractor, watch it). No earnings-without-cash divergence.
• Receivables/inventory vs revenue: trade receivables $552.2M and inventory $111.5M at FY25-end grew roughly with revenue; "costs & estimated earnings in excess of billings" (unbilled, $69.2M) < "billings in excess of costs" ($176.8M) — i.e. the company is billing ahead of cost, a positive working-capital sign, not a red flag.
• Leases / related parties / contingencies: operating leases on B/S (ROU $88.4M). Related-party = the Tontine relationship (disclosed, Note 3). Legal: Item 3 / Note 19 — "ordinary course"; "None of these proceedings, separately or in the aggregate, are expected to have a material adverse effect". Self-insurance reserves $12.1M; surety cost-to-complete on bonded projects ~$199.8M (disclosed, normal for a contractor).
• SBC / non-GAAP: SBC is modest (Employee+Director PSUs ~$3.5M/qtr) and IES reports on a GAAP basis (no aggressive non-GAAP bridge) — a quality marker. Diluted share count is stable-to-down (~20.2M) — no dilution games. Goodwill $107.8M→$129.2M (Gulf Island) is small vs $1.0B+ equity — not a goodwill-stuffed serial acquirer.
• Tax: effective rate ~24.6% (FY25 provision $96.8M / pretax $393.9M) — clean; small NOLs ($3.8M federal, Sec. 382-limited).

Regulatory findings:

• SEC Litigation Releases: none naming IES Holdings since 2021-06-24 (EDGAR EFTS, LR forms).
• SEC AAERs: none since 2021-06-24 (EDGAR EFTS, AAER forms).
• Non-SEC (FTC/DOJ/FDA/CFPB/etc.): web search "IES Holdings" (FTC OR DOJ OR... ) enforcement surfaced no material enforcement actions or consent decrees.
• 10-K Item 3 (Legal Proceedings): company's own disclosure — ordinary-course only, no material matters.
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-24. The standing accounting risk is intrinsic (POC estimates + fixed-price bids), not enforcement-driven.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (FY2026E–FY2028E)

Bottom-up from FY2025 actuals + H1-FY26 run-rate + guidance. Outputs ``, every input labeled. No forecast.ts create (watchlist/breadth mode — not committing a tracked forecast).

FY2026E (FYE 2026-09-30): H1-FY26 diluted EPS is already $9.95. H2 historically ≈ H1 with a Q4 seasonal lift; assume H2 modestly > H1 on Gulf Island ramping and continued data-center strength, partly offset by deepening Residential weakness.

• Base FY26E EPS ≈ $20.5. Consensus corroborates: one tracker shows next-FY EPS $19.69 and another $19.01 TTM already — so ~$20 is reasonable, not heroic.
• Bull FY26E ≈ $22+ (Gulf Island accretes faster, data-center backlog converts, Residential stabilizes).
• Bear FY26E ≈ $18 (Residential op income keeps bleeding toward breakeven, Gulf Island integration loss persists — it ran a −$1.2M op loss in its stub quarter ).

FY2027E: drivers — Communications + Infrastructure Solutions continuing to grow on record ~$3.9B backlog (vs $2.37B a year earlier ); Residential a flat-to-down anchor; operating leverage on a scaling corporate cost; share count ~flat (buyback off, minimal dilution).

• Base FY27E EPS ≈ $24.
• Bull ≈ $27 / Bear ≈ $20 (data-center capex digestion + housing recession).

FY2028E:

• Base FY28E EPS ≈ $27–28. Bull ≈ $32 / Bear ≈ $21 (cyclical rollover).

Valuation implication: at ~$712, the stock trades ~35x base-FY26E and ~30x base-FY27E EPS — not cheap on forward earnings if data-center capex normalizes, even though it screens "cheapest in cohort" on trailing P/E. Multiple independent DCFs cited fair value $398–$579 — all below the current price. Brier forecast to log when conviction warrants (not now): IESC FY26 diluted EPS >= $20.0 — p≈0.65, resolves 2026-09-30.

Lens 12 · Bull vs Bear

Bull case. IES is the relative-value, debt-free way to own data-center electrification. Two data-center segments (~49% of revenue) are growing 40–65% with rising, sector-leading margins; record ~$3.9B backlog gives multi-year revenue visibility; Gulf Island adds fabrication capacity and nuclear/oil-&-gas optionality; an owner-operator (Tontine, ~54%) with a 19x-in-5-years record and disciplined, value-aware capital allocation (buying back at <$180, hoarding cash at >$700); ~41% ROE; zero net leverage; and a trailing P/E (~38.7x) below every large peer (FIX/PWR ~53x, STRL ~79x) despite faster growth. Earnings surprises have repeatedly been upward (+56% net income last print). If hyperscaler capex persists through 2027, $24–27 of EPS at even a 30x multiple implies meaningful upside.

Bear case (2–3 ways it permanently/severely impairs).

1. Single-variable thesis. The valuation now requires sustained hyperscaler/AI data-center capex. A capex pause or digestion (the entire cohort's shared risk) would hit growth and the multiple simultaneously — a double de-rate from ~38x trailing. Data-center concentration "leans heavily" and is the named #1 risk.
2. Residential is a real, present drag — not theoretical. The largest segment by revenue (38.7%) saw Q2-FY26 op income collapse to $6.4M @ 2.2% margin (from 7.1%) on housing affordability/rates. If housing worsens into a recession, ~$1.3B of revenue keeps de-margining and partly offsets data-center gains.
3. Cyclical-peak earnings + fixed-price risk. Contractor margins this far above the 5-yr base (gross 25.5% vs 18.7% two years ago) tend to mean-revert; a few bad fixed-price/POC contracts could force downward revisions of "previously recorded revenues and profits". Plus control overhang, thin float (~7% single-session drops on no news), and Gulf Island integration risk (stub-quarter op loss).

Pre-mortem (18 months out, thesis broke): AI data-center capex visibly decelerated in late-2026/2027; IESC's Communications backlog growth flattened; Residential slid toward breakeven in a housing downturn; a couple of large fixed-price data-center jobs took POC margin write-downs; the ~38x trailing multiple compressed to ~20x as "peak-cycle contractor" reasserted itself — and the stock round-tripped a large chunk of its 2026 doubling. Gendell's insider sales, in hindsight, looked like a tell.

Are multiples too high? On trailing P/E, IESC is the cheapest in its cohort — defensible. On forward, mid-cycle, controlling-shareholder-discounted earnings, ~30–35x for a cyclical contractor with a declining largest segment is rich and prices in continued perfection. The DCF community ($398–579 fair value) agrees.

Contrarian view (what the market refuses to see): Bulls treat IESC purely as an "AI data-center pick-and-shovel," but ~half the business is housing/commercial that is cyclical and currently deteriorating, and the equity-investment portfolio (Jett/CB&I, marketable securities) is an un-modeled, hedge-fund-flavored wildcard. The market is paying an AI multiple for a company that is also a levered bet on US housing and on one investor's capital-allocation calls.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• What structurally breaks the model: a hyperscaler capex air-pocket. Revenue is concentrated in fixed-price/POC contracting where a demand pause cancels backlog ("customers often have no obligation...to assign or release work...many contracts may be terminated on short notice" ). The "record backlog" is not a guarantee — it's signed work that can be deferred.
• Where revenue is concentrated / what if it shifts: consolidated has no >10% customer, but the growth and the multiple are concentrated in a handful of unnamed hyperscaler/co-lo data-center buyers. If two of them trim 2027 capex, Communications growth halves.
• Why the moat may be weaker than bulls think: it's still electrical contracting — "relatively low barrier for entry," price competition, and the moat (capital/bonding) protects against small private contractors, not against Quanta/Comfort Systems/Sterling chasing the same data-center dollars. Margin gains from "disciplined bidding" can reverse when the cycle turns and everyone bids harder.
• Most dangerous competitor bulls underestimate: the larger, better-capitalized public peers (Quanta $40B+, Comfort Systems, Sterling) targeting the same data-center scope with more scale — plus the hyperscalers' own self-perform/EPC arms.
• Worst capital-allocation/incentive concerns: the controlling shareholder running a side investment portfolio through a public contractor (Jett/CB&I equity-method stake, $104.6M marketable securities) — capital that could fund buybacks/core M&A is instead in Gendell's investment ideas; plus the change-of-control triggers across material agreements if Tontine sells, and the insider selling.
• Assumptions that must hold for today's price: data-center capex stays strong through ≥2027; Residential doesn't deteriorate further into a housing recession; Gulf Island integrates accretively; fixed-price contracts don't produce material POC write-downs; the ~38x trailing multiple doesn't compress as growth normalizes.
• If growth disappoints 20–30%: FY27E EPS slips from ~$24 toward ~$17–19, and a cyclical-contractor multiple (~18–22x) reasserts → a fair value in the ~$350–420 zone, i.e. ~40–50% downside from ~$712. That asymmetry — limited modeled upside, large multiple-compression downside — is the short case in one line.
• Single permanent-impairment scenario (and plausibility): a sustained AI/data-center capex retrenchment coinciding with a US housing recession would hit both halves of IES at once; permanent impairment is unlikely (debt-free, real business, real backlog), but a 40%+ de-rating is entirely plausible — moderately likely on a 2–3 yr view, given the valuation.

Lens 14 · Management Questions (ordered by information value)

1. What share of Communications + Infrastructure Solutions revenue is tied to your top 3 data-center customers, and how concentrated is the ~$3.9B backlog among hyperscalers?
2. How much of the record backlog is enforceable vs "agreements without an enforceable obligation," and what cancellation/deferral have you seen historically in a capex pause?
3. What is the normalized, mid-cycle gross margin for Communications once "disciplined bidding" tailwinds and the current supply/demand tightness fade?
4. Gulf Island ran an operating loss in its first stub quarter — what's the integration timeline to your target Infrastructure Solutions margin, and what synergies are underwritten?
5. With the buyback essentially paused above $400, what is your capital-allocation priority stack today (M&A vs buyback vs the investment portfolio), and at what price would buybacks resume?
6. Please walk through the Jett/CB&I and marketable-securities positions — what role does a non-core investment portfolio play inside a public contractor, and who decides those allocations?
7. How exposed is the fixed-price book to copper/aluminum/steel and long-lead switchgear, and what % of contracts carry escalation/escape protection?
8. What is the realistic FY26/FY27 trajectory for Residential op margin (now ~2%), and at what housing scenario does it go breakeven or negative?
9. Labor is repeatedly cited as the growth throttle — what is your craft-labor attrition/hiring run-rate, and how much backlog could you convert if labor were unconstrained?
10. With FY26 capex guided to $110–130M (≈2x FY25), what capacity/return are you building, and what payback do you underwrite?
11. The change-of-control provisions tied to a Tontine sale touch your credit, surety, and severance agreements — how do you mitigate that overhang for minority holders?
12. What is the succession/key-man plan around Mr. Gendell's role as Executive Chairman and Tontine principal?
13. How should we think about your bonding capacity ceiling with two primary sureties as project sizes scale with data-center work?
14. Where are you in the nuclear/oil-&-gas opportunity you flagged with Gulf Island — pipeline, timeline, and required investment?
15. Under a 20–30% data-center demand shortfall, what levers (cost structure, the "low and variable" Residential model) protect margins and cash flow?

---