Atkore

Phase A — Understand the business

Company Overview

Atkore makes the unglamorous metal and plastic plumbing of the electrical grid: conduit (steel & PVC), electrical cable & flexible conduit, installation accessories, metal framing (Unistrut), mechanical tube, perimeter security and cable management. It runs two reportable segments:

• Electrical — FY25 net sales $1,998.2M (70% of total); products sold through the electrical wholesale distribution channel to contractors.
• Safety & Infrastructure (S&I) — FY25 net sales $852.2M (30%); metal framing, mechanical pipe, perimeter security, cable management, sold to contractors, OEMs and end-users.

The company claims #1 or #2 US share by net sales in a significant number of its products. Brand house consolidated under one master brand "Atkore" in 2020, over sub-brands Allied Tube & Conduit, AFC Cable, Unistrut, Power-Strut, Heritage Plastics, Calpipe, FRE Composites.

Customers — concentrated in distribution. Top 10 customers ≈ 40% of net sales; Sonepar USA alone > 10% in each of FY23–25. Named distributors: CED, Graybar, Rexel, Sonepar, Wesco, plus super-regionals (US Electrical Services, Crescent Electric) and buying groups (AD/IESD, STAFDA). ~83% of sales are distribution-based; ~11.5% OEM (solar racking, conveyor). 88% of FY25 sales were in the US.

Contract structure — the single most important business fact: Atkore sells on a spot basis, not under long-term contracts, and "historically has not engaged in material hedging strategies for raw material purchases". So in a rising-commodity world it can push price; in a falling-commodity world customers claw price back. There is no take-or-pay, no recurring revenue, no contracted backlog cushion. Price is the entire P&L swing factor — and that is exactly what the last three years prove.

Supply Chain

Map: commodity producers → Atkore (38 plants, 8.6M sq ft, 8 countries, HQ Harvey IL) → electrical/industrial distributors (12,000+ US branches) → electrical & mechanical contractors → non-resi construction / MR&R / data centers / solar / infrastructure.

Named upstream suppliers:

• Steel: Cleveland-Cliffs, Steel Dynamics (SDI), Nucor.
• Copper: AmRod, SDI LaFarga, Nexans.
• PVC resin: Westlake, Formosa, Oxy Vinyls.
• HDPE resin: LyondellBasell, Baystar.
• Also exposed to zinc (galvanizing), electricity, natural gas, and diesel/freight (delivers most product by truck).

Chokepoints / dependencies: Atkore is a price-taker on inputs and a spot price-setter on outputs — the spread is the business. No single-source supplier risk is flagged as critical (multiple suppliers per commodity), but the distribution channel is the chokepoint downstream: with Sonepar >10% and top-10 = 40%, the wholesale distributors hold real bargaining power, and they actively destock/restock with the commodity cycle (visible in the FY24–25 receivables and inventory drawdowns). A structurally relevant positive shift: management cites declining Mexican steel-conduit imports (down into the "mid-to-high tens" percent) as a tailwind to domestic conduit volume and price. (No supply-chain.md in the research layer — kb/energy commercial layer is empty for this name; mapped from filings.)

Competitive Advantages (moats)

Honest read: this is a scale-and-distribution moat, not a pricing-power moat. Durable advantages:

• Scale + national footprint + "must-stock" status — for many of 12,000+ distributor branches Atkore's lines are must-stock staples; breadth of portfolio + short order cycle times + ReliaRoutes hub-and-spoke trucking = a real service/logistics edge in a low-value-density, freight-sensitive product.
• Process/technology edge in galvanizing — Allied Tube's in-line Flo-Coat galvanizing and Cellular Core co-extrusion are genuine low-cost manufacturing advantages.
• #1/#2 share in many categories → purchasing economies and shelf dominance.

But: patents and trademarks are explicitly "not material" to the competitive position. There are no switching costs (purchase-order relationships, not contracts; customer-relationship intangibles amortized over 6–14 yrs on attrition assumptions). The products are commoditized — competition is "product offering, innovation, quality, service and price". Bargaining power runs against Atkore on both ends: input commodities are set by global markets; the downstream channel is consolidated (Sonepar/Graybar/Rexel/Wesco are giants). Conclusion: a wide-ish cost/distribution moat around a commodity product — defensible on volume and share, undefendable on price. The 2021–25 margin round-trip (below) is the proof.

Competitors:

• Electrical: Zekelman Industries, Mitsubishi, Nucor, Southwire, Dura-Line, Prysmian.
• S&I: Zekelman, Eaton, ABB, Hubbell, nVent, Haydon.

Segments

All ``. The story is entirely an Electrical-segment margin round-trip; volume barely moved.

Electrical segment — Adj EBITDA & margin by fiscal year:

FY25 Electrical net sales −15.1% YoY, driven by avg selling prices −11.9%, volume only +0.7%. This is a price unwind, not a demand collapse.

S&I segment held up: FY25 net sales $852.2M (+0.5%), Adj EBITDA $109.2M (12.8%) — actually up 21% YoY because input-cost declines outpaced price declines.

Geography (FY25): US $2,501M, International $349M. Notable in the latest quarter: Asia-Pacific revenue jumped to $39.5M (Q2-FY26) from $12.7M (PY) — early traction in international cable management / Global Megaprojects.

Latest quarter (fiscal Q2 FY26, ended 2026-03-27), the inflection:

• Electrical net sales $532.5M (+8.1% YoY); Adj EBITDA $74.4M, margin 14.0% (vs 18.5% PY) — still compressing as input costs out-ran price, but volume +8.1%.
• S&I net sales $199.1M (−4.9%); Adj EBITDA $17.3M, margin collapsed to 8.7% (from 17.2% PY) — the soft spot this quarter.
• Product-line detail (Q2-FY26): Metal conduit & fittings $129.0M (+15%), Electrical cable $125.1M (flat), Plastic pipe & conduit $149.5M (−7%) — PVC still the drag; Other Electrical $128.9M (+35%, the data-center/international lines).

---

Phase B — Measure performance

Earnings Result (latest print: fiscal Q2 FY26, ended 2026-03-27)

The headline GAAP number is ugly and the underlying number is the opposite — the gap is the thesis.

• Net sales $731.4M, +4.2% YoY — the first YoY sales increase since Q4 FY2022, composed of volume +$32.3M and price +$10.2M (first positive price in ~13 quarters) plus FX +$8.2M, less divestitures −$12.6M.
• GAAP net loss $(124.1)M, diluted EPS $(3.65). But that swallows three one-timers: a $136.5M PVC antitrust litigation settlement, an $11.6M HDPE goodwill impairment, and a $25.7M loss on assets held for sale.
• Adjusted diluted EPS $1.23, beat consensus by ~16%; stock +3% on the print.
• Total segment Adj EBITDA $91.7M (Electrical $74.4M + S&I $17.3M); consolidated margins pressured by input-cost lag.
• Balance-sheet flags: cash $442.3M (down from $506.7M at FYE — partly working capital + the litigation overhang); inventory drawn to $401.1M from $484.8M (destocking); debt flat at $760.6M; equity $1,280.9M. H1-FY26 FCF was negative $(53.5)M vs +$97.3M PY — seasonally weak half plus litigation/working-capital timing; capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. H1 only $26.2M (down from $63.6M).
• Guidance MAINTAINED for FY26: Adj EBITDA $340–360M, Adj diluted EPS $5.05–5.55, net sales $2.9–2.95B, mid-single-digit volume growth.
• Vs the company's own history: trough-ish. FY23 Adj EBITDA (consolidated segment) was ~$1.1B; FY26 guide is ~$350M — a ~68% peak-to-trough EBITDA compression that is almost entirely PVC price reverting to mid-cycle.

Earnings Calls (sentiment trend)

No transcripts on the research shelf (transcripts/ empty); sourced ``. Tone arc over the cycle:

• FY24 → early FY25: defensive — "normalization of pricing," managing the down-cycle, heavy buybacks framed as conviction.
• Q1-FY26 (Feb 2026): stabilizing — maintained full-year outlook, leaning on volume and the cost program.
• Q2-FY26 (May 2026): the most constructive in two years — management explicitly flags the first YoY sales growth since Q4-FY22, "double-digit growth in product lines tied to data-center electrification," solar mechanical-tube demand, and an improving conduit market as Mexican steel-conduit imports decline. Recurring new theme: portfolio focus — "since February 2025, Atkore has divested five businesses" to concentrate on domestically-manufactured electrical infrastructure for high-growth markets. What they stopped saying: the "Global Megaprojects water-pipe" growth narrative — HDPE/water is being exited, not promoted.

Comps

ATKR is the cheap, low-quality (commodity-pricer) name in the electrical group; the premium names are differentiated-product compounders. Multiples with date, or `n/a`. ATKR's own implied multiples are from sourced price + guidance.

ATKR EV/EBITDA derivation: EVEnterprise valueWhat it would cost to buy the whole company: its market value plus its debt, minus the cash you would get with it. Often very different from market cap. ≈ $2.5B equity + $760.6M debt − $442.3M cash ≈ $2.82B; FY26 guided Adj EBITDA mid ≈ $350M → EV/EBITDA ≈ 8.0x on guided EBITDA, or ~5–6x on the ~$450M+ "normalized/mid-cycle" EBITDA bulls underwrite. P/E ≈ $77.5 / $5.30 ≈ 14.6x FY26E adj EPS. The discount to nVent (28x) / Hubbell (~20x EV/EBITDA) is deserved — ATKR has no pricing moat and a legal cloud — but a ~2–3 turn EV/EBITDA discount and a half-the-multiple P/E for a #1/#2 share, FCF-rich franchise is wide. 5-yr avg ROE: n/a (note FY25 ROE was negative on the impairment; FY23 net income $689.9M on ~$1.5B equity ≈ 45%+ at the peak — illustrating how cycle-distorted ROE is for this name).

Stock-Price Catalysts (what moves ATKR >5%)

Pattern over the cycle, + for the events:

1. PVC conduit price prints / guidance — the dominant driver. The 2021–22 spike took the stock to ~$180+; the price unwind drove it down to the ~$60s–$70s. Every quarter the market re-reads the price line.
2. Q2-FY26 earnings beat (May 2026): +3% premarket on the first positive YoY sales + price.
3. The PVC antitrust headlines (2024→2026): the original suits (Q4-FY24), the DOJ grand-jury subpoena (Feb 2025) and two securities class actions (Q2-FY25) were a major leg of the de-rating; the $136.5M (April 2026) + $50M (June 2026) settlements are the recent re-rating catalysts.
4. CEO retirement announcement (Aug 2025): governance overhang.
5. AI data-center / electrification demand narrative (2026): a new positive driver as ATKR re-pitches itself as an electrification beneficiary.

The tell: this stock reacts to the commodity/price cycle and to legal headlines far more than to volume or secular demand — it is priced as a cyclical, and the antitrust saga added an idiosyncratic discount on top.

---

Phase C — Judge people & books

Management

• CEO transition in progress (the key governance fact). President & CEO William E. Waltz, Jr. notified the board on Aug 4, 2025 of his intent to retire after 12 years at Atkore / 7 as CEO; he stays until a successor is named, and as of the latest filings/search no successor has been announced. A leaderless transition during a down-cycle and an active DOJ probe is a real uncertainty — but also a classic set-up for a fresh-eyes portfolio reset (which is arguably already happening via the divestitures).
• Track record: Waltz's tenure spanned the 2021–22 super-cycle (Adj EBITDA ~$1B, EPS $17+) and the subsequent normalization. He executed the IPO-era de-leveraging and the brand consolidation, and returned enormous capital (below). Quality of the underlying execution is hard to separate from the PVC windfall.
• Capital-allocation history — aggressive, and increasingly questionable in hindsight: buybacks of $491.0M (FY23) + $381.0M (FY24) + $100.0M (FY25) = ~$972M in three years, retiring shares from ~40M+ toward 33.7M. They bought back heavily at peak-cycle prices (FY23–24, stock often >$130–160) and then paused entirely — zero repurchases in Q4-FY25 and zero in H1-FY26 — with $328.1M still authorized under the May-2024 $500M plan. Buying high and stopping at the lows is the textbook bad-timing pattern; charitably, conserving cash through the antitrust settlements is prudent. Dividend initiated Nov 2023, now $0.33/quarter ($1.32/yr), maintained through the trough.
• The HDPE/water-pipe misadventure: management invested capex into HDPE and "water pipe to support Global Megaprojects," then took a $214.4M FY25 impairment ($194.5M HDPE long-lived + $18.9M Mechanical goodwill) and put HDPE up for sale. A growth bet that was written off inside ~2–3 years — a genuine capital-allocation black mark.
• Founder vs professional manager: professional-manager / PE-legacy company (Atkore was a CD&R-controlled spin from Tyco; CD&R has long since exited). No founder dynamic.
• Red flags: the securities class actions (below) explicitly allege management made misleading disclosures; the DOJ subpoena names the company's pricing practices. Comp/related-party items appear ordinary in the filings.

Forensic Red Flags

Accounting quality is, ironically, decent — the problems are real (legal, impairments), not cosmetic. All `` unless noted.

• Revenue recognition: simple, point-in-time on shipment, single performance obligation, no long-term contracts, immaterial returns/variable consideration. Low manipulation surface.
• Impairments — large and concentrated, but disclosed and non-cash: $214.4M FY25 + an additional $11.6M HDPE goodwill in Q2-FY26 (HDPE goodwill now fully impaired) + $25.7M loss on HDPE assets held for sale. The 10-KForm 10-KA company’s audited annual report to the US regulator. The most complete thing it publishes. explicitly warns the Conduit & Fittings and EMEA reporting units' fair value exceeds carrying value only modestly — "less than significant changes in valuation assumptions could result in" further impairment. Watch for more impairments if the cycle stays low.
• Inventory / LIFO: ~79–81% on LIFO; a Q2-FY26 LIFO liquidation added ~$2.6M to COGS; excess-&-obsolete reserve grew to $29.6M from $23.2M. FIFO inventory would be only ~$31.8M higher — modest LIFO reserve, no red flag.
• Cash flow vs earnings: historically clean — FY25 operating cash flow $402.8M on a GAAP net loss, bridged transparently by the $214.4M non-cash impairment. No suspicious accrual build. Receivables ($447.0M) and inventory ($484.8M FYE) both fell with revenue — the opposite of a channel-stuffing tell.
• SBC: $23.6M FY25 (~0.8% of sales) — modest; not flattering non-GAAP egregiously.
• Leverage: clean. Total debt $760.6M (Senior Notes $400M @ 4.25% due 2031; Term Loan ~$370M @ SOFR+2.00%/5.3% due 2032), cash $442.3M → net debt ~$318M, ~0.9x guided FY26 EBITDA. ABL undrawn ($325M available). Split-rated (some agencies IG, some non-IG). Low rate sensitivity ($3.8M per 100bps).

Regulatory findings (required sub-section).

• SEC EDGAR EFTS: No SEC Litigation Releases and no AAERs naming Atkore in 2021-06-30→2026-06-30.
• DOJ — the serious one: on Feb 13, 2025 Atkore received a DOJ Antitrust Division grand-jury subpoena (N.D. California) for documents on PVC pipe/conduit pricing; the DOJ intervened in the civil case (Oct 2025) and obtained a discovery stay extended through July 1, 2026 — the criminal investigation is ongoing. This is the largest unquantified risk in the file.
• Civil antitrust — In re: PVC Pipe Antitrust Litigation (N.D. Ill. 24-cv-07639): putative class actions filed Q4-FY24 alleging PVC manufacturers shared confidential pricing via the OPIS "PVC & Pipe Weekly" report (and direct communications), violating Sherman Act §1, covering ~2021–present. Now substantially settled (no admission of liability): DPP class $72.5M + NCSP class $64M = $136.5M (April 28, 2026, recognized Q2-FY26), plus the End-User class $50M (signed June 3, 2026; preliminary court approval June 10, 2026). Total civil settlements ≈ $186.5M. A parallel suit in British Columbia (Sept 2025) is unquantified.
• Securities class actions: two putative §10(b)/§20(a) suits (Westchester Putnam Local 60 v. Atkore; Coles v. Atkore, both N.D. Ill., filed Q2-FY25) against the company and current/former officers re: disclosures.
• Environmental: legacy IEPA stormwater/zinc matters at Harvey IL and groundwater remediation at Wayne MI — long-running, management deems immaterial.
• Net: No SEC accounting enforcement. The accounting is honest. The exposure is antitrust — civil now ring-fenced at ~$186.5M, criminal DOJ still live and unquantifiable.

---

Phase D — Project & stress-test

Forward Projection (FY26 / FY27 / FY28 adjusted diluted EPS)

Built bottom-up from FY26 company guidance + the cycle dynamics. Output ; FY26 base anchored to guidance.

Anchor — FY26 (guided): Adj EBITDA $340–360M; Adj diluted EPS $5.05–5.55 (mid ~$5.30); net sales $2.9–2.95B; mid-single-digit volume growth. Share count ~33.8M, buybacks paused.

• Base case — mid-cycle stabilization. FY26 ≈ $5.30 (guidance mid). FY27: volume +mid-single-digit, price flat-to-slightly-positive (Mexican-import relief + DC/electrification demand absorbing capacity), modest margin recovery as input-cost lag washes through; some buyback resumption once DOJ tail clears. → FY27 ≈ $5.75. FY28 ≈ $6.25.
• Bull case — re-acceleration. Electrical margin recovers toward 22–25% (still below the 30%+ bubble), data-center/grid drives high-single-digit volume, aggressive buyback at depressed prices. FY27 ~$6.75, FY28 ~$8.00+.
• Bear case — extended trough / price re-collapse + DOJ fine. Non-resi construction rolls over with rates, PVC price gives back the FY26 gain, S&I stays sub-10% margin, a material DOJ criminal penalty lands. FY27 ~$4.25, FY28 ~$3.75.

Brier forecast (logged in the dossier, not via our model — watchlist mode skips create): ATKR FY26 (FYE 2026-09-30) adjusted diluted EPS ≥ $5.05, p ≈ 0.78 — i.e. likely lands in or above the guided range given the Q2 beat and maintained guide, with downside only on a sharp H2 non-resi air-pocket.

Bull vs Bear

Bull case (narrative). You are buying a #1/#2-share, cash-gushing electrical-infrastructure manufacturer at ~14x trough-ish earnings and ~8x guided / ~5–6x mid-cycle EBITDA, while the market is still anchored on the PVC bubble's unwind and a now-mostly-resolved lawsuit. Three things inflect together in 2026: (1) price turned positive YoY for the first time in 13 quarters and volume is growing mid-single-digit on data-center electrification, solar, grid, and declining Mexican conduit imports; (2) the civil antitrust is capped at ~$186.5M — a known, paid number, removing the open-ended overhang; (3) management is refocusing the portfolio (5 divestitures since Feb 2025, HDPE exit) onto high-margin domestic electrical, and has $328M of buyback authorization plus a clean ~0.9x-levered balance sheet to repurchase a 33.8M-share float at depressed prices once the DOJ tail clears. Normalized EPS power is plausibly $6–8 within two years; a re-rate to even 16–18x on $6+ is a $100+ stock vs ~$77 today. FCF yield is double-digit even at the trough.

Bear case (narrative). This is a commodity manufacturer with no pricing moat whose entire 2021–24 earnings base was a one-time PVC-conduit supply shock that is never coming back — the "normalized" earnings bulls underwrite may simply be the new, lower reality, and FY26's $5.30 could prove the ceiling, not the floor. Three permanent-impairment risks: (1) DOJ criminal antitrust — a guilty plea / large fine / debarment-type consequence is unquantifiable and unsettled, and the very thing that made the margins (allegedly coordinated PVC pricing via OPIS) is now illegal-until-proven-otherwise, structurally capping future Electrical margins; (2) non-residential construction is cyclical and rate-sensitive — Dodge starts are above historical average (1,209M sq ft FY25), so the construction cycle is a headwind-from-a-high, not a tailwind-from-a-low; (3) the HDPE write-off shows management will chase growth and destroy capital, and a leaderless CEO transition raises the odds of another value-destructive pivot or an ill-timed acquisition with the dry powder. At ~14x, the market is not pricing a re-collapse — a 20–30% EPS disappointment to ~$3.75 on a de-rate to 11x is a sub-$45 stock.

Pre-mortem (18 months out, thesis broke): non-resi softened as rates stayed high, PVC price round-tripped its FY26 gain, the DOJ extracted a nine-figure criminal fine and imposed a monitor, a new CEO "rebased" guidance lower and bought something expensive — and the multiple compressed as earnings fell. The stock is at $45.

Are multiples too high? No — they are low, deliberately. The risk is not multiple compression from a high base; it's that the earnings base itself is overstated by residual bubble margins and that the DOJ tail is a left-skew the multiple can't see.

Contrarian view (what the market refuses to see): The market is treating the antitrust saga as a permanent scarlet letter, but the civil cases just got capped at a knowable $186.5M — and if the DOJ matter resolves as a fine rather than a structural remedy, the overhang lifts on a name that is simultaneously a levered call on US electrification and data-center buildout trading at half the multiple of nVent. The bubble-margin reversion is done (FY25→FY26 is the trough); the next surprise is more likely to be volume-driven margin recovery than further price collapse.

Devil's Advocate (short-seller)

Dismantling the bull case.

• The moat is a cartel, not a competitive advantage. The bull's "pricing power" in 2021–22 is exactly what the DOJ is criminally investigating and what Atkore just paid $186.5M to settle civilly. Strip out any coordinated-pricing benefit and the true normalized Electrical margin may be the low-teens we see now, not the 20s bulls model. The peak was the anomaly; the trough may be the truth.
• Revenue concentration: Sonepar >10%, top-10 = 40%. These consolidated distributors have every incentive to claw price back in a soft market — and no contracts stop them. A single large-distributor destock swings a quarter.
• Most dangerous competitor bulls underestimate: imports + Zekelman/Southwire on conduit. The current "Mexican imports falling" tailwind is policy/trade-driven and reversible — if trade dynamics shift, low-cost conduit floods back and price re-collapses. Bulls are extrapolating a transient.
• Worst capital allocation: bought ~$872M of stock at FY23–24 peak prices, then stopped at the lows, and wrote off $214M+ on the HDPE/water-pipe growth bet. This is a management team that has demonstrably bought high, sold (impaired) low, and is now leaderless with $328M of authorization — a setup for another mistake.
• Assumptions that must hold for $77: (1) FY26 $5.30 EPS is a floor not a ceiling; (2) DOJ ends in a manageable fine, not a structural remedy; (3) non-resi doesn't roll over from above-average starts; (4) data-center demand is large enough to matter to a $2.9B-revenue conduit maker (it's a mix helper, not a needle-mover at scale).
• −20–30% growth scenario: EPS to ~$3.75, de-rate to ~11x → ~$41, ~45% downside.
• Single scenario that permanently impairs: a DOJ criminal resolution that forces a structural change to how conduit is priced/sold in the US (the OPIS-report mechanism), permanently resetting Electrical margins to the low-teens — turning today's "trough" into the new normal.

Management Questions (ordered by information value)

1. On the DOJ criminal antitrust investigation: what is the realistic range of outcomes (fine, monitor, structural remedy), and has the company taken any reserve beyond the civil settlements?
2. What do you believe the true mid-cycle Electrical Adjusted EBITDA margin is, now that the 2021–22 pricing environment is gone — low-teens, high-teens, or low-20s — and why?
3. How much of the 2021–24 Electrical margin do you attribute to industry pricing conduct now under DOJ scrutiny vs. genuine cost/scale advantage?
4. With ~$328M of buyback authorized, a 33.8M-share float, and the stock near multi-year lows, why are repurchases paused — and what specifically must resolve before they resume?
5. On the CEO succession: timeline, internal vs external, and what mandate (steward the franchise vs. transform the portfolio) is the board hiring against?
6. Was the ~$872M of FY23–24 buybacks at peak prices, followed by the pause at the lows, the right capital-allocation call in hindsight — and what changes in the framework?
7. What is the normalized FCF conversion and the FY26/FY27 free-cash-flow guide, given the H1-FY26 negative FCF and the litigation cash outflows still to come?
8. How large and how durable is the data-center / electrification demand for your specific products — quantify it as a share of Electrical revenue and its growth rate.
9. The "Mexican steel-conduit imports declining" tailwind — how much is policy/trade-driven and therefore reversible, vs. structural?
10. After the HDPE/water-pipe write-off, what is the go/no-go discipline for the next growth investment, and is M&A on or off the table during the CEO transition?
11. Which reporting units are closest to further impairment (Conduit & Fittings and EMEA were flagged as modest cushions), and what cycle assumptions underpin those carrying values?
12. S&I margin collapsed to 8.7% in Q2-FY26 — is that cyclical (input-cost lag) or structural, and what is the recovery path?
13. How should investors think about price vs. volume in FY27 — is the company now structurally a volume growth story rather than a price story?
14. What is the long-term capital-return framework (buyback vs. dividend vs. M&A) and the target leverage range through a full cycle?
15. With ~88% US revenue, how real is the international / Global Megaprojects opportunity (Asia-Pacific tripled this quarter) — is it a needle-mover or a rounding error?

---