ATI Inc. (NYSE: ATI) — A Real Aero-Alloy Franchise Wearing a Supercycle’s Richest Multiple

An independent equity research note Report date: 2026-06-19 · Price reference: $201.34 (close 2026-06-18) · Market cap ~$26.9B · EV ~$28.3B

⚡ Claude’s Take

This block is the author’s own subjective opinion and general information only — not investment advice. The analysis that follows is deliberately position-free and carries no price target; this fenced block is the single exception.

Call: HOLD / AVOID-at-this-price / NOT-a-short. Accumulate-on-weakness only in the ~$110–140 zone (~18–22x mid-cycle EBITDA, ~25–30x normalized EPS). Do not chase near the $200 all-time high. Low-to-medium conviction.

ATI is a genuinely better business than it was five years ago — and the market knows it, in full. The repair is real: the company exited commodity stainless, annuitized away a multi-billion-dollar pension overhang (now just $200M), de-levered to ~1.6x, and concentrated the portfolio on qualification-gated aerospace, defense, and nuclear alloys where it is sole-source on 5 of the 7 most advanced jet-engine nickel grades. Operating margin tripled from 3.8% (2021) to 14.0% (2025); ROIC went from 3.9% to 13.7%; the order backlog is at an all-time-high $4.1B. This is a real franchise core (HPMC engine alloys + the near-monopoly zirconium/hafnium exotics) wrapped in a still-capital-intensive, still-cyclical metals body. The problem is not the business; it is the price. At $201 the stock has compounded +145% in twelve months and +75% year-to-date, sits ~1% below its all-time high, and trades at ~66x trailing and ~46x forward (FY26 guide) earnings, ~33x trailing and ~27x forward EBITDA — the 98th percentile of its own ten-year valuation history, the richest in a cohort (CRS, HWM, GE, RTX, FTAI) where even the best franchises are judged full-price.

The framing — grounded in the factor tape (beta 1.47, positive alpha +0.41, Aerospace & Defense loading ~0.88, y1 return +145% at a 3.35 Sharpe) — is crowded momentum with genuine idiosyncratic alpha, not a falling knife and not a bargain. The alpha is real now; what it is not is durable through a cycle. This is structurally a high-beta specialty-metals cyclical (lifetime max drawdown −94.7%) earning at a cyclical peak and priced as if the peak is permanent. The mid-teens ROIC at the top of the best aero demand environment in a generation is the tell: capital intensity and the commodity AA&S tail cap the through-cycle economics well below the multiple’s implied expectations. Two governance smells reinforce caution: a late-2025 executive “Enterprise Value Acceleration” plan whose $195 max stock-price hurdle was blown through within ~6 months (windfall, not stretch), and zero open-market insider purchases in 5.5 years — nobody bought the $30–45 bottom, and insiders are net sellers into the run. Bull-flip: the $5B-revenue / 20%±margin / ~$1.0B±EBITDA run-rate proves durable through a Boeing/Airbus build-rate wobble, with naval-nuclear + missiles structurally lifting the next trough’s floor — i.e., the cycle stops being a cycle. Bear-flip: any commercial-aero air-pocket (build-rate cut, destock, recession) collapses operating leverage the way 2020–21 did, and a 14%-ROIC cyclical does not hold a 27x EBITDA multiple. Tag: “Right franchise, wrong tape — buy the metal, not the momentum.”

📈 Stock Price Action — Five-Year Event Map

ATI has traveled a roughly 14x round-trip-and-beyond over five years: from a COVID intraday low of ~$4.95 (Mar-2020) and a trailing-5yr low near ~$13.85 (Nov-2021), through a steady 2022–2024 grind, into a near-vertical 2025–2026 aerospace-supercycle re-rating that carried it to an all-time intraday high of ~$203.60 (12-Jun-2026). It now trades at $201.34 (18-Jun-2026), ~1.1% off that high, inside a 52-week range of ~$70.42–$203.60 — i.e., at the very top of its own history.

1. 2021 — recovery stall. After the 2020 COVID collapse, ATI traded sideways near the mid-teens through 2021 as commercial-aerospace build rates were only beginning to recover; the trailing-5yr low (~$13.85) was set in Nov-2021.
2. 2022 — cyclical inflection. Specialty-metals and aerospace demand inflected off a depressed base, roughly doubling the stock to ~$30 by year-end as volumes and pricing recovered.
3. 2023–2024 — steady ramp. The multi-year commercial-aero and defense ramp drove a methodical climb to the mid-$50s, with the stock range-bound ~$52–67 through late-2024 before the next leg.
4. Apr–Jun 2025 — re-rating begins. Following Q1-2025 results (30-Apr) and accelerating jet-engine/missile materials demand, ATI broke out ~+59% to the mid-$80s — the first sharp leg of the supercycle re-rating.
5. Sep 2025–Jan 2026 — beat-and-raise momentum. A Q3-2025 beat-and-raise (28-Oct) and a building record-backlog narrative (plus nuclear/missile thematic interest) carried the stock from ~$81 to ~$120.
6. Jan–Feb 2026 — the big month. The Q4-2025 print (3-Feb-2026) was a strong beat-and-raise that spotlighted a record ~$4.1B backlog, driving a single-month ~+41% surge to ~$164.
7. Mar–Jun 2026 — to all-time highs. After a brief March pullback, the Q1-2026 beat-and-raise (30-Apr) and the 11-Jun BWXT naval-nuclear material supply agreement pushed ATI to an all-time intraday high of ~$203.60 (12-Jun), settling at $201.34.

Price moves are FACT (5-year daily price data); attributed drivers are INTERPRETATION cross-referenced to earnings dates and the news feed. No price target, no recommendation.

1. Executive Summary

ATI Inc. (formerly Allegheny Technologies; Dallas, TX; CIK 0001018963) is a vertically integrated specialty-materials producer — titanium and titanium alloys, nickel- and cobalt-based superalloys, and reactor-grade zirconium/hafnium exotics — sold primarily into commercial aerospace, defense, and specialty energy/nuclear. It reports two segments: High Performance Materials & Components (HPMC) — melt-through-machined-component engine and structural alloys, ~53% of revenue, ~92% aerospace & defense, 23.6% segment EBITDA margin — and Advanced Alloys & Solutions (AA&S) — flat-rolled, zirconium, nickel and titanium products, ~47% of revenue, 16.3% margin, a higher-cyclicality mix that still carries automotive, conventional-energy and electronics tails.

The investment story is a genuine business repair colliding with an extreme valuation. The repair (real, value-creating): revenue grew from a $2.80B COVID trough (2021) to $4.59B (2025); operating margin rose 3.8% → 14.0%; EBITDA margin 8.9% → 17.7%; ROIC 3.9% → 13.7%. Management exited commodity stainless, executed an October-2023 Athene pension annuitization that cut the legacy DB overhang to ~$200M, de-levered net debt to ~1.6x EBITDA at falling coupons (7.25% → 5.875%), and rotated the book toward qualification-gated, sole-source aero/defense/nuclear alloys. Backlog is at an all-time-high $4.1B (~3/4 HPMC); the company renewed a naval-nuclear agreement at ~$1B over five years (~2x prior) and a Cameco nuclear-fuel agreement at ~$250M, and missile-related revenue more than doubled year-on-year. FY2026 guidance (raised at Q1) is adjusted EBITDA $1.010–1.060B (+20%), adjusted EPS $4.20–4.48, adjusted FCF $465–525M.

The valuation (the problem): at $201.34 the stock trades at ~66x trailing and ~46x forward EPS, ~33x trailing and ~27x forward EBITDA, ~16x book — the 98th percentile of its own 10-year history, the richest of an aero cohort (CRS, HWM, GE, RTX, FTAI) in which every member is already judged full-price. The factor tape reads as crowded aerospace-supercycle momentum with genuine positive alpha (+0.41), up +145% in a year — not a falling knife, but a one-way street up priced for the peak to be permanent.

The core tension: ATI is a real-but-narrow, cyclical franchise — intangible qualification incumbency plus switching costs and niche melt/forge scale, concentrated in HPMC engine alloys and the Zr/Hf near-monopoly, diluted by a commodity AA&S tail — earning a mid-teens ROIC at a clear cyclical peak, while the market prices it as a secular compounder. The moat dampens but does not eliminate cyclicality (the 2020–21 trough printed ~10% gross margin and a $1.57B loss). The business deserves a premium; this price embeds near-perfection.

This memo takes no position and sets no price target

2. Business Overview

What ATI does. ATI converts commodity metal inputs into highly engineered, performance-critical alloy products and components for applications where failure is not an option — jet-engine rotating parts, airframe structures, naval-nuclear reactor components, missile bodies and propulsion, and power-generation turbines. The company is integrated across the value chain: melting (vacuum induction melting/VIM and remelting via VAR/ESR), forging (it is a global leader in isothermal and hot-die forging), rolling, finishing, testing, machining, and additive/powder manufacturing. This melt-to-component integration is deeper than a pure mill-product peer (Carpenter Technology) and is a key driver of HPMC’s premium margins.

Segments and revenue mix (FY2025, from the 10-K MD&A “Results by Business Segment”):

HPMC end markets (FY2025): commercial jet engines $1,649.6M (68% of segment, +21% YoY), commercial airframes $312.1M (13%, −16%), defense $277.7M (11%, +24%) — A&D ~92% of the segment. HPMC product mix: nickel-based & specialty alloys 43%, precision forgings/castings/components 40%, titanium & Ti alloys 17%. This is the franchise engine: jet-engine rotating-part alloys, where ATI is sole-source on 5 of 7 of the most advanced nickel grades.

AA&S end markets (FY2025): A&D $872.1M (41%; airframes 22%, defense 13%, jet engines 6%), conventional energy $322.3M (15%), automotive $238.1M (11%), specialty energy $181.5M (8%), electronics $184.8M (9%), medical $84.2M (4%), construction/mining $117.7M (5%). Product mix: nickel-based & specialty alloys 50%, zirconium & related alloys 19% (~$400M — the distinctive reactor-grade Zr/Hf exotics franchise), titanium 19%, precision-rolled-strip 12%. AA&S is the more diversified, more cyclical, lower-margin half — but it houses the near-monopoly zirconium/hafnium business and is rapidly mix-shifting toward aero-like defense/nuclear margins.

How it makes money / recurring vs. non-recurring. ATI sells materials and components under long-term agreements (LTAs) with most major aerospace OEMs, plus index/surcharge-priced commercial sales. Revenue is not subscription-recurring, but it is unusually visible: LTAs with minimum quantities and frozen order windows, multi-year qualification lock-in, and a $4.1B backlog (~1 year of sales) give multi-quarter forward sightlines rare for a metals producer. A growing share is aftermarket/spares (engine refurbishment), which is more annuity-like and accelerates as the in-service fleet ages and flies. Customers are a concentrated oligopsony: GE Aerospace, Rolls-Royce, Pratt & Whitney (RTX), Safran (engines); Boeing, Airbus, Embraer (airframes); BWX Technologies and the naval-nuclear program; Cameco (nuclear fuel). International sales are ~42% of total.

Verdict. A vertically integrated specialty-alloy producer that has deliberately concentrated into the highest-value, qualification-gated aero/defense/nuclear niches, with a genuine franchise core (HPMC engine alloys, Zr/Hf exotics) and a still-meaningful commodity tail (AA&S flat-rolled/auto/conventional-energy). High revenue visibility for a metals company, but not recurring revenue in the SaaS sense.

3. Industry Dynamics

Structure: a consolidating, qualification-gated oligopoly. The relevant market is not “specialty metals” broadly but the premium-melt tier of flight-critical and nuclear alloys — a fewer-than-ten-player Western oligopoly. ATI’s named competitors: in HPMC, Berkshire Hathaway’s Precision Castparts (the largest, vertically integrated melter/caster/forger — Special Metals, TIMET), Howmet Aerospace (Ti and Ti-based), Carpenter Technology (legacy Ni superalloys and specialty steel), and Aubert & Duval (forgings); in AA&S nickel/superalloys, Haynes International and VDM Metals (Acerinox). The set passes Greenwald’s “count the leaders on one hand” test for the premium tier.

Consolidation is recent and structural. In under twelve months the public pure-play universe shrank: Haynes International was acquired by Acerinox/North American Stainless (closed Nov-2024, ~$970M EV) and Universal Stainless & Alloy Products by Aperam (closed Jan-2025, ~$537M EV). That leaves Carpenter and ATI as the only scaled public pure-plays. Simultaneously, Russia’s VSMPO-AVISMA — the world’s largest titanium producer — was effectively removed from the Western supply chain (ATI terminated its Uniti JV with VSMPO-AVISMA), tightening Western titanium structurally.

Demand: a multi-year, partly uncorrelated stack. (i) Narrowbody engine ramp — CFM LEAP and Pratt GTF, with >30,000 jet engines on firm order; specialty-materials demand leads aircraft deliveries by ~6–12 months. (ii) Aftermarket/spares — an early-innings supercycle as an aging in-service fleet flies longer (new-engine shortages and GTF powder-metal inspections keep old metal flying), driving recurring rotating-part demand; ATI’s content on next-gen LEAP/GTF is ~2x its legacy-engine content. (iii) Defense — +14% in FY2025 and accelerating: naval nuclear (Virginia/Columbia submarines), military engines, titanium armor, and a missile business up >100% YoY (PAC-3, THAAD, Tomahawk). (iv) Specialty energy / nuclear renaissance — ATI is the named pioneer/Western leader in reactor-grade Zr/Hf cladding; SMRs, life-extensions, and data-center power demand are tailwinds. (v) Land-based gas turbines — the same nickel-superalloy technology, bid by data-center power needs.

Marathon capital-cycle read — favorable supply side. High and rising returns are not yet attracting competing capacity, because the supply response is throttled by qualification, not steelmaking: no Western greenfield premium-melt mill has been built in decades; additions are brownfield and qualified furnace-by-furnace over years (Carpenter’s Athens VIM furnace is ~$400M for one furnace, qualifying to ~2030). Consolidation reduced the player count; there is no IPO wave and no bubble-priced empire M&A. This is the configuration under which elevated returns persist longer than a normal industrial cycle — but it is not permanent. The watch item is cohort-wide capex: ATI itself is adding nickel remelt (online Q4-2026) and VIM melt (2027) and qualifying new titanium PQ capacity; Carpenter and others are expanding. Cumulative qualified-capacity additions by ~2028–2030 are the variable that eventually mean-reverts returns.

Cyclicality is structural and acknowledged. ATI’s filings flatly state demand is cyclical; commercial aerospace “has historically been cyclical,” and AA&S’s oil & gas, auto, and construction/mining exposures are “highly cyclical.” The 2020–21 aero collapse (revenue −25%+, gross margin to ~10%, a net loss) is the live memory of the downside. Raw materials (scrap, nickel, titanium sponge, zirconium sand, ferrochromium, molybdenum, cobalt, niobium, hafnium) are commodity and multi-sourced; ATI passes cost through surcharges/indices and selective hedges — there is no input cost advantage.

Verdict: structurally GOOD — for the premium-melt / qualified A&D + nuclear tier. A consolidating, qualification-gated oligopoly with disciplined brownfield-only supply against a multi-year, partly uncorrelated secular demand stack. Two firm qualifiers: (1) the verdict applies to HPMC and the Zr/Hf exotics, not to AA&S’s commodity flat-rolled/auto tail (fragmented, weak pricing); (2) the whole sector remains demand-cyclical — “good industry” means good through-cycle structure and a currently favorable supply side, not low cyclicality. The capital cycle is in a favorable up-leg; what ends it is a wave of qualified-capacity adds or a demand air-pocket — neither imminent, both worth monitoring.

4. Competitive Position

Name the moat (Greenwald). ATI’s advantage is intangibles (AMS/customer-qualification incumbency) + switching costs, reinforced by niche economies of scale + captivity in capital-intensive vacuum melt (VIM/VAR) and isothermal/hot-die forging, plus a near-monopoly intangible position in reactor-grade zirconium/hafnium. There is no supply/cost advantage (commodity, multi-sourced inputs, surcharge pass-through). The moat is real but concentrated in HPMC engine alloys and the Zr/Hf exotics; the AA&S commodity tail has little to none.

The qualification mechanism — the heart of switching costs. A&D contracts “involve meeting highly exacting performance requirements and product specifications,” and ATI touts a “fully qualified asset base.” A specific nickel or titanium grade, melted via a specific VIM-VAR route on specific assets, is qualified onto a named engine or airframe program and certified by the OEM and aviation regulators (FAA/EASA via AMS specs). ATI is qualified on 6 of the 7 most advanced jet-engine nickel alloys and is sole-source on 5 of 7. To dual-source, an OEM must fund a multi-year requalification, accept flight-safety and liability risk, and disrupt a program — which it will not do absent a quality or capacity failure. This converts qualified positions into quasi-locked, multi-year demand on platforms that fly for decades, and is the source of ATI’s pricing power, its all-time-high $4.1B backlog, and ~2-year lead times on premium-quality titanium. On the Q1-2026 call, management was explicit that in a constrained, sold-out market it is repricing LTAs upward even into an OE ramp (step-ups, escalators, resets) — the opposite of the normal OE price-down — and picking up share where other suppliers cannot deliver. That is a moat surfacing as pricing power.

The moat shows up in the financials at the top of the cycle — but not at the bottom. HPMC segment EBITDA margin 23.6% (FY25, up from 20.3%/20.5%); AA&S 16.3% (up from 15.4%/13.5%); company ROIC 13.7% (25) / 11.7% (24) / 14.3% (22). But gross margin collapsed to 9.8% (2020) and 11.9% (2021), operating margin to ~3%, ROIC to 3.9%, with a 2020 net loss. A wide-moat franchise (e.g., Howmet) stayed profitable through COVID; ATI did not. The Greenwald financial test therefore reads: a genuine but narrow-to-medium, cyclical advantage — returns are moat-consistent (mid-teens ROIC, 20%+ HPMC margins) near the peak, but the moat dampens, it does not eliminate, cyclicality.

The pressure-test — why ROIC is only ~14%, not 30%+. Three reasons. (1) Capital intensity: this is an asset-heavy melt/forge business — net PP&E $1.94B on $4.59B sales, managed working capital ~32.5% of annualized sales (inventory $1.40B), because long lead times trap cash in WIP. High invested capital is the ROIC denominator drag. (2) Commodity mix dilution: roughly half the company (much of AA&S — flat-rolled, PRS, auto, conventional energy) is genuine commodity specialty metals with no captivity, dragging the blended return below HPMC-only economics. (3) Cyclical under-earning: the multi-year average ROIC is well below the 13–14% peak because of the 2020–21 trough. ATI is a capital-intensive, cyclical specialty-metals producer with real pockets of captivity — not a clean 25%+ ROIC franchise.

Direct peer comparison. Precision Castparts (Berkshire) is the largest, most integrated player — both customer and competitor, private, run for cash. Howmet (HWM) sits at a different node (investment castings, forgings, fasteners) and is the higher-quality franchise (profitable through COVID, dominant gas-turbine-blade share, ~21% ROIC). Carpenter (CRS) is the closest public pure comp — smaller, melt-through-mill-product, less downstream-integrated than ATI, and more commodity-exposed at the trough (it printed ~0% gross margin in FY21). Moat-quality ranking: HWM > ATI ≈ CRS > the now-captive Haynes/Universal. ATI’s edge over CRS is its deeper downstream integration (melt-to-machined component), the Zr/Hf exotics near-monopoly, and a broader, more diversified A&D book; ATI’s relative weakness vs. HWM is its more cyclical mill-product/AA&S exposure and more cyclical history.

Verdict: a DURABLE advantage — narrow-to-medium width and cyclical. Not a wide moat, not a crowded commodity. Intangibles (qualification incumbency, sole-source on 5 of 7 advanced engine nickel alloys, Zr/Hf near-monopoly) + switching costs + niche melt/forge scale are genuine, surface in sticky qualified positions, a record backlog, 20%+ HPMC margins and mid-teens peak ROIC, and would clearly cause economics to deteriorate if removed. But the level of returns is heavily operating-leverage/cycle-driven, capital intensity caps ROIC at ~14% even at the peak, and roughly half the company carries little captivity. A real franchise core wrapped in a capital-intensive, cyclical metals body — currently earning at/near a cyclical high.

5. Growth History and Forward Opportunities

Historical growth. Revenue: $2.982B (2020) → $2.800B (2021, COVID trough) → $3.836B (2022) → $4.174B (2023) → $4.362B (2024) → $4.587B (2025), a ~13% CAGR off the trough but only +5.2% in FY2025 as the company deliberately deemphasized lower-value medical/electronics/industrial volume under its 80/20 program. The more important growth story is in quality, not tonnage: operating income rose from $106M (2021) to $642M (2025), EBITDA from $250M to $810M, as mix shifted to aero/defense/nuclear and operating leverage compounded. FY2025 incremental operating margins were ~38%; management now guides ~40% consolidated incremental margins.

Q1-2026 acceleration. Revenue $1.15B (69% A&D); adjusted EBITDA $232M at a 20.1% margin (+300bps YoY); adjusted FCF $75M (vs. a $143M use a year earlier — a $218M swing). Jet engine +12%, defense +9%, specialty energy +22%; airframe −9% (timing/destock, expected to ramp in H2). Backlog grew 10% sequentially to a record $4.1B. Management raised FY2026 guidance to adjusted EBITDA $1.010–1.060B (+20%), adjusted EPS $4.20–4.48, adjusted FCF $465–525M, and framed “$5B of revenue at 20%+ margins in clear sight.”

Forward opportunities — organic, contracted, capacity-gated. (1) Jet engine (largest market, ~41% of sales): mid-teens growth guided on OE build rates + aftermarket shop visits, with ATI sole-source on 5 of 7 advanced nickel alloys and ~2x content on next-gen LEAP/GTF vs. legacy. (2) Defense mid-teens, with a renewed naval-nuclear agreement at ~$1B over 5 years (~2x prior, aero-like margins) and missiles up >100% YoY off a small base, plus the new BWXT naval-nuclear supply agreement through FY2030. (3) Specialty energy / nuclear mid-teens, anchored by the Cameco ~$250M Zr extension and land-based gas-turbine demand from data centers. (4) Airframe recovery weighted to H2-2026 and 2027 as widebody build rates ramp and titanium SQ demand returns. (5) Capacity unlock: a nickel debottlenecking that already lifted primary-melt weekly output >15% YoY, with nickel remelt online Q4-2026, VIM melt in 2027, and new titanium PQ in qualification — each pre-sold against the backlog.

Quality of growth. This is high-quality growth where it counts: contracted, qualification-protected, margin-accretive, and concentrated in the franchise core. The caveats are that (a) it is being delivered into the best aero demand environment in a generation — i.e., cyclically flattered; (b) capacity additions into a backlog peak carry the Marathon late-cycle risk; and © headline revenue growth is modest because management is (correctly) trading volume for margin.

Verdict: HIGH-QUALITY growth — but cyclically amplified and contracted-not-secular. The mix-up, margin-accretive, capacity-gated profile is exactly what you want; the question is durability through a demand down-leg, not the quality of the up-leg.

6. Financial Quality

Revenue, margins, operating leverage. Operating margin tripled from 3.8% (2021) to 14.0% (2025); EBITDA margin rose from 8.9% to 17.7%; Q1-2026 ran at 20.1%. The trajectory is genuine operating leverage on a fixed melt/forge asset base plus mix-shift and LTA repricing — incremental margins ~38–40%. HPMC (23.6%) is both higher-margin and faster-improving than AA&S (16.3%); the gap is structural (downstream integration, near-pure aero).

(Source: company filings. Note 2023 net income was flattered by a $128M tax benefit; 2024 EPS dipped on share count/mix.)

Cash flow and capital intensity. FY2025 operating cash flow $614M, capex $281M, FCF $334M — a meaningful step up as working capital normalized. The business is capital- and working-capital-hungry: inventory $1.40B (much of it WIP held through long lead times), managed working capital ~32.5% of annualized sales, net PP&E $1.94B. This is the structural reason a real moat still only earns mid-teens ROIC. Capex is rising (~$280–300M gross guided, partly customer-funded) to add qualified nickel/titanium melt. FY2026 adjusted FCF guide $465–525M (~$495M midpoint) implies a ~1.8% FCF yield at $201 — thin for the multiple, and flattered by the cyclical earnings peak.

Balance sheet — materially de-risked. Total debt $1.749B, cash $417M, net debt $1.333B (~1.6x EBITDA); current ratio 2.7x. The signature win is the October-2023 Athene group-annuity pension buyout, which cut the legacy defined-benefit liability to ~$200M from historically multi-billion-dollar levels — structurally removing the overhang that depressed equity value and cash for a decade. New notes were issued at falling coupons (7.25% in 2023 → 5.875% in 2026). Goodwill is modest ($225M); the equity base is small (book ~$1.8B), which makes ROE (143%) and P/B (~16x) meaningless as standalone metrics — use ROIC and EV-based multiples instead.

Quality of earnings. Earnings quality is reasonable: cash conversion is solid (OCF > net income in 2025), the 2020 loss was a one-time impairment/pension event, and 2023’s tax benefit is a clean normalize-out. The honest caveat is cyclical position, not accounting: every margin and return figure above is at or near a cyclical peak, and the company reports “adjusted” EBITDA/EPS/FCF that strip restructuring and other items — reconcile to GAAP (GAAP diluted EPS $2.85 in FY25 vs. adjusted figures management guides on).

Verdict: economics DO improve with scale — within the cycle. Operating leverage is real and the balance sheet is genuinely repaired. But the mid-teens peak ROIC, heavy working-capital absorption, and rising capex mean this is a good, not great, return profile — and one observed at the top of the cycle. The structural floor is higher than last cycle (better mix, de-risked balance sheet), but the cyclicality is intact.

7. Capital Allocation

The structural repair (genuinely excellent). Management’s 2020–2024 program is textbook good capital allocation. It exited standard/commodity stainless and grain-oriented electrical steel, idled or sold commodity flat-rolled assets, and reorganized into HPMC + AA&S — withdrawing capital from low-return businesses (the favorable side of the Marathon cycle). It executed the Athene pension annuitization, removing the legacy DB overhang. It de-levered net debt from ~$1.84B (2021) to $1.33B at falling coupons. And it let ROIC re-rate from 3.9% to 13.7% as capital exited the bad businesses. These moves earned the re-rating.

The capital-return record (procyclical — the blemish). ATI pays no dividend (suspended years ago, not reinstated); buybacks are the sole return vehicle. The cadence is the textbook value-destroying pattern of buying high, not low: repurchases were minimal ($96M) in 2022–23 when the stock was $30–45 (cheapest, lowest book multiple), and large and accelerating ($286M in 2024, $504M in 2025, $75M in Q1-2026, with a fresh $500M authorization in February 2026 — $545M remaining) now that the stock is $201 at a ~16x-book, 98th-percentile valuation. ATI is buying the most stock at the highest price. FY2025 buybacks ($504M) actually exceeded FCF ($334M), funded partly off the balance sheet. Do not credit the buyback as accretive at $200.

Growth capex into a backlog peak. Capex is ramping (~$280–300M gross) to add nickel/titanium melt into a record backlog. This is largely contracted/LTA-backed (naval nuclear, jet engine), which mitigates the risk — but adding melt capacity at the top of an aero supercycle is precisely when Marathon warns that returns later mean-revert as supply catches demand. It reads as disciplined brownfield expansion for now; it is the variable that turns the cycle.

Compensation and incentives. The annual cash plan is sensibly returns/cash-anchored: EBITDA 60% / FCF 30% / strategic-individual 10% — no volume/tonnage/empire metric. The long-term plan is PSUs 70% (100% relative-TSR vs. peers, 3-year cliff, negative-TSR cap) + RSUs 30% — defensible and market-aligned, but notably with no ROIC/operating-returns gate to ensure capital discipline underlies the TSR. The blemish is a late-2025 “Enterprise Value Acceleration” (EVA) one-time grant: 4-year absolute-stock-price-hurdle awards earning target at $135 and max (300%) at $195 (20-day average) — hurdles set when the stock was ~$110–130 and blown through within ~6 months as the stock hit $201. A “stretch” retention plan that became a near-automatic windfall is a governance yellow flag and a poorly calibrated use of equity.

Insider behavior (no conviction tell). A scan of all ~221 Form 4 filings from 2021 through mid-2026 finds zero open-market purchases (code P) — not a single insider bought a share with personal cash, including through the $30–45 trough. All activity is routine grants/vesting/tax-withholding/sales, with net selling into the 2026 run-up. Aggregate insider ownership is ~1%. There is no bullish insider signal supporting the current valuation; if anything the tape is mild distribution.

Verdict: GOOD on structure, POOR on buyback timing. The de-leveraging, pension annuitization, and commodity-exit/portfolio focus are genuinely value-creating and earned the re-rating. But the capital-return cadence is procyclical (minimal buybacks at the bottom, peak buybacks at a 98th-percentile valuation), the EVA plan is a windfall, and insiders show no conviction. An above-average operator-allocator; a below-average returner of cash at this price.

8. Changes and Headwinds — Last Two Years

Strategic / portfolio. Completion of the commodity-stainless exit and the 80/20 portfolio rationalization (deemphasizing medical/electronics/industrial to concentrate on aero/defense/specialty-energy); the FY2024 precision-rolled-strip divestiture (PRS fell from 19% to 12% of AA&S product mix). The result is a higher-margin, A&D-led book (~65–70% A&D).

Balance sheet / pension. The October-2023 Athene pension annuitization (a portion of US DB obligations transferred; liability now ~$200M; related litigation had claims recommended for dismissal in Aug-2025). Note refinancings at falling coupons ($425M 7.250% 2030 notes in 2023; $450M 5.875% 2033 notes in 2026). Credit-facility amendments/extensions (2022, 2025).

Leadership / governance. A complete C-suite/Chair transition: Kimberly Fields became President & CEO July-1-2024 (from President & COO) and Board Chair effective May-2026; predecessor Robert Wetherbee retired from the board at the 2026 annual meeting. Rob (James) Foster became CFO January-1-2026, succeeding Donald Newman. The late-2025 EVA compensation plan (see above).

Commercial / demand. Record $4.1B backlog (Q1-2026); naval-nuclear renewal ~$1B/5 years (~2x prior); Cameco ~$250M Zr extension; BWXT naval-nuclear supply agreement through FY2030 (June-2026); missile revenue >100% YoY; GE/RR/Pratt/Safran and Boeing/Airbus LTA repricing upward. The February-2026 $500M buyback-authorization increase.

Headwinds to monitor. (1) Commercial-airframe timing — airframe revenue fell ~9–16% on destock/build-rate phasing, with the ramp pushed to H2-2026/2027. (2) Boeing/Airbus execution risk — the entire demand stack rests on OEM build rates that have repeatedly disappointed. (3) Middle East / fuel / retirements — management says no impact to date but is monitoring fuel-driven retirements (mitigated by ~2x content on next-gen engines). (4) Tariffs — currently passed through; refund policy unresolved. (5) China export controls on zirconium/hafnium/critical materials (supply risk, but also a reshoring tailwind). (6) Energy costs — rising power prices (data-center demand) managed via pass-throughs and gas hedges.

Verdict: net STRENGTHEN the business, but the thesis risk has shifted to valuation and cycle-timing. The last two years materially improved the franchise (portfolio, pension, balance sheet, backlog, defense/nuclear legs). What has not improved is the entry point — the same period re-rated the stock to its richest-ever multiple, and the live headwinds (airframe timing, OEM execution, the EVA windfall, absent insider conviction) all argue for patience rather than chase.

9. Risk Analysis

Risk of permanent capital loss / total loss. A permanent impairment is plausible from the entry price (#1 + #2 combined: a de-rating into a demand down-leg could halve the stock without the business being broken — exactly what 2020 showed). A total loss is very unlikely: net debt is only ~1.6x EBITDA, the pension is de-risked, the franchise core is real, and the asset base has takeout value (the sector has been consolidating). This is a valuation/cyclicality risk, not a solvency risk.

10. Valuation Discussion (Embedded Expectations)

Where the multiple sits. At $201.34: market cap ~$26.9B, EV ~$28.3B (net debt $1.33B + minority $0.11B). Against FY2025 results — EV/EBITDA ~35x, EV/sales ~6.2x, P/E ~66x trailing, P/B ~16x. Against FY2026 guidance (EBITDA ~$1.035B, adj EPS ~$4.34) — ~27x EBITDA, ~46x earnings. FCF yield ~1.8%. Own-history valuation percentiles read composite 98.1th percentile, with P/E 94th, P/B and P/S effectively 100th — the richest valuation in ATI’s traded history.

Peer comparison (the crux).

PCC is private (Berkshire); Haynes and Universal Stainless were acquired 2024–25. The honest read: ATI screens marginally cheaper than CRS/HWM on TTM EV/EBITDA (~33–35x vs. ~42x), but sits below both on every quality metric (lower margins, lower ROIC, more commodity AA&S mix) — and carries the highest own-history valuation percentile of the three. The lower headline EV/EBITDA is a quality discount, not a margin of safety. ATI is the lower-quality, more cyclical member of the trio trading at its own richest-ever multiple, inside a cohort where even HWM and GE are judged full-price.

Embedded-expectations analysis — what the price requires. At ~27x forward EBITDA and ~46x forward earnings on a business earning a mid-teens ROIC at a cyclical peak, the market is underwriting that (a) the FY2026 ~$1.0B+ EBITDA / 20%+ margin run-rate is a new structural baseline, not a peak; (b) the supercycle (engine ramp + aftermarket + defense + nuclear) extends multiple years with little cyclical interruption; © the naval-nuclear/missile/Zr legs structurally raise the through-cycle margin floor; and (d) capacity additions are absorbed without mean-reverting returns. Several of those are plausible; collectively they price near-perfection. The market is pricing ATI as a secular compounder; the financial fingerprint (14% ROIC, capital intensity, a commodity tail, a −94.7% lifetime drawdown) is that of a high-quality cyclical at a top.

Scenario analysis (illustrative ranges, NOT price targets).

• Bear (~$90–120): a commercial-aero air-pocket or destock pulls EBITDA back toward ~$700–800M and the multiple de-rates toward a still-premium ~15–18x — the way cyclicals re-rate when the cycle turns. The 2020–21 trough (gross margin ~10%) is the reminder the downside is real.
• Base (~$140–180): the supercycle holds, EBITDA reaches ~$1.0–1.2B, but the multiple normalizes from 98th-percentile toward a high-teens/low-20s EV/EBITDA as growth becomes “expected” — i.e., earnings grow into a still-full multiple, modest downside-to-flat from $201.
• Bull (~$230–280+): the ~$1.0B EBITDA proves a baseline, the defense/nuclear legs lift the floor, $5B revenue at 20%+ margins arrives ahead of plan, and the market keeps paying a secular-compounder multiple through the cycle. This is the scenario the current price largely already embeds.

Verdict. A real franchise at an unreal multiple. The business merits a premium to a commodity metals producer; this price embeds a premium to itself at its best, forever. The asymmetry from $201 is unattractive — the bull case is largely priced, the bear case is a genuine de-rating, and the catalyst path (continued beats) is the consensus everyone already owns.

11. Variant Perception

Consensus belief. ATI is a prime beneficiary of a multi-year aerospace/defense/nuclear supercycle — sole-source on the hardest engine alloys, record backlog, repricing power, expanding margins, a de-risked balance sheet, and a credible “$5B at 20%+ margins” path. The tape agrees: momentum is positive, factor loadings read aerospace-defense/momentum, and the stock is +145% in a year with positive alpha — i.e., the market has concluded this is a quality re-rating, not a mere beta ride. Consensus is now ON the trade.

Strongest bull case. The structural argument is genuine: qualification incumbency + sole-source positions + a consolidating, supply-disciplined oligopoly + a multi-year, partly uncorrelated demand stack (engine OE and aftermarket and defense and nuclear) mean the next trough is higher and shallower than 2020–21. The de-risked balance sheet and 80/20 mix mean ATI keeps far more of each revenue dollar. If naval-nuclear ($1B), missiles (+100%), and Zr/Hf nuclear structurally raise the floor, the business is less cyclical than its history — and a less-cyclical 20%-margin franchise can defend a premium multiple.

Strongest bear case. This is still a capital-intensive, high-beta specialty-metals cyclical earning a mid-teens ROIC at the top of the best demand environment in a generation, priced at the 98th percentile of its own history. The moat is real but narrow and did not prevent a ~10% gross margin and a $1.57B loss in 2020. The entire thesis rests on Boeing/Airbus build rates that have serially disappointed; airframe revenue is already declining. Buybacks are being executed at the peak; insiders bought zero shares at the bottom and are selling into strength; the EVA comp plan paid out a windfall in months. When the cycle turns, a 27x-EBITDA multiple on a 14%-ROIC cyclical does not hold.

The 3–5 assumptions that matter most. (1) Is the FY2026 ~$1.0B+ EBITDA / 20%+ margin a baseline or a peak? (2) Do the defense/nuclear legs structurally raise the through-cycle floor, or are they a one-time reset? (3) Will commercial-aero build rates hold through 2027? (4) Do cohort-wide capacity adds mean-revert returns by ~2028–30? (5) Will the market keep paying a secular multiple, or re-rate ATI back toward a cyclical one?

What would falsify each side. Bull falsified if margins/ROIC roll over with the airframe cycle and the multiple compresses toward history (the cyclical-peak read wins). Bear falsified if ATI sustains ~$1.0B+ EBITDA and 20%+ margins through a Boeing/Airbus build-rate wobble, proving the floor has structurally risen (the secular read wins). The factor read frames the risk precisely: the alpha is real now; the open question is whether it survives the cycle, and the price assumes it does.

12. Fact vs. Interpretation Table

13. Open Questions

1. Peak or baseline? How much of FY2025/FY2026 record margin (HPMC 23.6%, ~20% consolidated) is durable moat economics vs. cyclical operating leverage on near-sold-out demand + LTA repricing?
2. Through-cycle ROIC? What is the HPMC-only franchise’s normalized ROIC vs. the blended ~14% — i.e., how much moat is masked by the AA&S commodity tail and capital intensity?
3. Does the floor rise? Do the naval-nuclear (~$1B), missiles (+100%), and Zr/Hf nuclear legs structurally lift the next trough’s margin floor above the 2020–21 lows, or are they a one-time reset?
4. Capacity discipline? Is rising capex (Ni remelt 2026, VIM 2027, Ti PQ) disciplined brownfield expansion of qualified capacity (moat-deepening) or the start of a cohort-wide over-build that mean-reverts returns by ~2028–30?
5. Airframe timing. Does the H2-2026/2027 airframe ramp materialize on contracted orders, or does Boeing/Airbus phasing keep slipping it?
6. Capital return. Will management keep repurchasing aggressively at a 98th-percentile valuation, and would it have the discipline to stop and accumulate cash if the stock stays expensive?

14. What Must Be True

Bull case — what must be true (and its falsification test). ATI must prove that the ~$1.0B+ EBITDA / 20%+ margin / mid-teens-ROIC run-rate is a structural baseline, not a cyclical peak — i.e., that the franchise core (sole-source engine alloys, Zr/Hf, naval-nuclear, missiles) plus the de-risked balance sheet and 80/20 mix have permanently raised the floor, so the business compounds through a Boeing/Airbus build-rate wobble rather than collapsing into it. Falsification: if, in the next commercial-aero down-leg or destock, consolidated margins fall below ~15% and ROIC drops toward high-single-digits the way they did in 2020–21, the “secular compounder” thesis is dead and the 98th-percentile multiple was a mistake.

Bear case — what must be true (and its falsification test). The bear must be right that this is a high-beta, capital-intensive cyclical at a top, where a 27x-EBITDA / 46x-earnings multiple on a 14%-ROIC business cannot survive either a demand wobble or simple multiple normalization — and that the procyclical buyback, absent insider conviction, and EVA windfall confirm a late-cycle setup. Falsification: if ATI sustains ~$1.0B+ EBITDA and 20%+ margins through an OEM build-rate stumble while the defense/nuclear legs visibly raise the floor — and the multiple holds because the cash flow proves durable — then ATI is genuinely less cyclical than its history and the bear’s “cyclical peak” call is wrong.

15. Source Appendix

See the separate Source Appendix (below) for the full list of primary and secondary sources. Principal sources include: ATI Inc. FY2025 Form 10-K (filed 2026-02-20, ati-20251228.htm); FY2021–FY2024 10-Ks and FY2021–Q1-2026 10-Qs (SEC EDGAR, CIK 0001018963); 2026 DEF 14A (filed 2026-03-24); the 8-K corpus 2021–2026; the Form 4 corpus 2021–2026; the Q1-2026 earnings-call transcript (2026-04-30); aggregated fundamentals and ratios; own-history valuation percentiles and 5-year price history; a quantitative factor model; and the public filings of peers Carpenter Technology (CRS), Howmet (HWM), GE Aerospace, RTX, and FTAI used for cohort/comparison.

APPENDIX A — Standard Diligence Questionnaire

ATI Inc. (NYSE: ATI) — as of 2026-06-19

Supplemental to the research memo. Answers given with Fact / Interpretation / Assumption labels where material.

General

What thoughtful questions have other investors asked about this company? From the Q1-2026 call, the sell-side pressed on: (1) whether pricing/margin can keep rising into an OE build ramp (it is — LTA step-ups/escalators/resets in a constrained market); (2) the source of the EBITDA-guide raise (defense + jet engine contracts, bias to the high end); (3) the durability of the second-half airframe ramp (contract-driven, orders placed 12–18 months out, frozen windows); (4) capacity additions (nickel remelt Q4-2026, VIM 2027, titanium PQ in qualification); (5) incremental margins (~40% consolidated); (6) Middle East / fuel / retirement risk (no impact to date; ~2x content on next-gen engines is a mitigant); (7) the missile ramp (small but +100%+ YoY). The unasked-on-the-call but central investor question is valuation: can a mid-teens-ROIC cyclical hold a 98th-percentile multiple?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? High (Interpretation, high-confidence). Operating margin (14.0%), EBITDA margin (17.7% → 20%+ Q1-2026), and ROIC (13.7%) are at/near record levels, achieved in the strongest aerospace/defense demand environment in a generation. The 2020–21 trough (gross margin ~10%, a $1.57B loss) is the cyclical low for comparison.

Driven by external environment or internal actions? Both. External: the aero supercycle (engine ramp + aftermarket + defense rearmament + nuclear). Internal: genuine self-help (commodity-stainless exit, 80/20 mix-up, pension de-risking, de-leveraging) that structurally raised margins independent of the cycle. The internal actions are durable; the external tailwind is cyclical.

How stable are revenues? Visible but cyclical. A $4.1B backlog (~1 year), LTAs with minimum quantities and frozen windows, and qualification lock-in give unusual forward visibility for a metals company — but demand is structurally cyclical (the 10-K says so explicitly), and 2020–21 saw revenue fall >25%.

Outlook for products/services; how big is the market? Growing for years: >30,000 jet engines on firm order; an aging in-service fleet driving aftermarket; defense rearmament; a nuclear renaissance (SMRs, life-extensions, naval nuclear). ATI is deliberately shrinking low-value end markets (medical/electronics/industrial, −low-to-mid-single-digits) to concentrate on A&D + specialty energy. Domestic and international (~42% of sales ex-US).

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Less, at the premium tier. Consolidation removed two public pure-plays (Haynes → Acerinox Nov-2024; Universal Stainless → Aperam Jan-2025); Russian VSMPO-AVISMA titanium is out of Western supply. Carpenter and ATI are the only scaled public pure-plays left. No greenfield Western melt capacity in decades.

How profitable is the business (ROIC, ROE)? ROIC 13.7% (FY25, peak-cycle), up from 3.9% (2021). ROE (143%) and P/B (~16x) are meaningless due to a small equity base (book ~$1.8B after the 2020 loss); use ROIC and EV multiples. Mid-teens ROIC is good, not great, for a business with this revenue visibility — capped by capital intensity and the commodity AA&S tail.

How profitable is the industry — competitors, barriers? Fewer-than-ten qualified Western premium-melt players; barriers are high in the premium tier (multi-year per-part/alloy/program qualification, AMS certification, flight-safety risk, capital-intensive VIM/VAR melt + isothermal forging). Barriers are low in AA&S commodity flat-rolled/auto.

Can the business be easily understood? Yes, with effort — a specialty-alloy melt/forge producer; the nuance is which parts are franchise (HPMC engine alloys, Zr/Hf) vs. commodity (AA&S tail), and where in the cycle the numbers sit.

Can it be undermined by foreign low-cost labor? No for flight-critical/nuclear (qualification + regulatory barriers + national-security reshoring); yes in principle for commodity AA&S — but ATI is exiting that.

Do brands matter? Nature of competition? Switching costs? Not consumer brands — but the qualified-supplier reputation functions as one. Competition is on qualification, quality, capacity, and reliability, not price (price is rising). Switching costs are very high in HPMC/exotics (multi-year requalification, flight-safety/liability risk), near-zero in commodity AA&S.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The qualified-position/qualification incumbency intangible (the real moat) is not capitalized. Goodwill is modest ($225M). The Zr/Hf near-monopoly franchise is not separately valued.

Off-balance-sheet liabilities? Materially reduced. The October-2023 Athene pension annuitization cut the legacy DB liability to ~$200M (Fact). Operating leases and surcharge/contract commitments exist but are not unusual. Pension-annuitization litigation had claims recommended for dismissal (Aug-2025).

How conservative is the accounting? Reasonable. The 2020 loss was a one-time impairment/pension event; 2023 net income was flattered by a $128M tax benefit (normalize out). Management guides on “adjusted” EBITDA/EPS/FCF — reconcile to GAAP (FY25 GAAP diluted EPS $2.85).

How CapEx-hungry is the business? Quite — net PP&E $1.94B on $4.59B sales; capex ~$280–300M (rising), partly customer-funded; managed working capital ~32.5% of sales with $1.40B inventory (long-lead WIP). Capital intensity is the structural cap on ROIC.

Capital Allocation & Management

How much FCF, and how is it used? FY2025 FCF $334M; FY2026 guide $465–525M. Use: 100% to buybacks (no dividend) + de-leveraging + growth capex. Buybacks are procyclical — $96M at the $30–45 lows (2022–23) vs. $504M + a fresh $500M authorization at the $200 peak (Interpretation: buying high).

Significant acquisitions recently? No — the story is divestiture/exit (commodity stainless, PRS) and organic capacity, not M&A. Clean.

Buying back shares? Issuing to insiders? Buying back aggressively (at peak valuation). Equity issuance to insiders is routine grant/vesting; share count is roughly flat-to-down. The late-2025 EVA absolute-price comp plan is a governance flag (max hurdle cleared in ~6 months = windfall).

Compensation / incentives / motivations. Annual plan EBITDA 60% / FCF 30% / strategic 10% (returns-anchored, no empire metric — good). LTI: PSUs 70% (100% relative-TSR, negative-TSR cap) + RSUs 30% — no ROIC gate (yellow flag). CEO Kim Fields FY2025 comp $13.5M. Insider ownership ~1%; zero open-market insider purchases 2021–2026.

Valuation & Market Data

ADR, MLP, or K-1 issuer? No — ordinary US C-corp common stock (NYSE), 1099 reporting.

Dividend policy? No dividend (suspended years ago, not reinstated). Buybacks only.

How profitable is the business? See above — mid-teens ROIC at a cyclical peak.

Is net income diverging from cash from operations? No material divergence in FY2025 (OCF $614M > net income $404M — healthy conversion). FY2023 OCF was depressed by working-capital build and the tax benefit (a divergence year). Generally cash-backed.

Risks & Downside

What factors would cause the stock to decline? A commercial-aero build-rate cut/destock/recession (operating-leverage collapse); valuation de-rating from the 98th percentile; a cyclical earnings peak proving to be the top; cohort-wide capacity adds mean-reverting returns; buyback impairment at peak prices.

Risk of catastrophic / total loss? Permanent capital loss from $201 is plausible (a de-rating into a demand down-leg could halve the stock without breaking the business — see 2020). Total loss is very unlikely: net debt ~1.6x EBITDA, de-risked pension, a real franchise core, and consolidation/takeout value support the floor. This is a valuation/cyclicality risk, not a solvency risk.

Recent News & Events

Has the business environment changed recently? Yes, favorably and recently — record $4.1B backlog (Q1-2026), naval-nuclear renewal ~$1B/5yr (~2x prior), Cameco ~$250M Zr extension, BWXT naval-nuclear supply agreement through FY2030 (Jun-2026), missiles +100% YoY, raised FY2026 guidance — all of which drove the +75% YTD / +145% 1-year run.

Significant acquisitions? None (organic + divestitures).

Change in accounting policies? None material (fiscal-year/bylaw housekeeping in 2022/2025).

Recent changes — markets, facilities, management? New capacity (nickel remelt Q4-2026, VIM 2027, titanium PQ); a full C-suite/Chair transition (CEO Kim Fields since Jul-2024, Chair May-2026; CFO Rob Foster since Jan-2026); 80/20 portfolio rationalization toward A&D/specialty energy.

APPENDIX B — Source Appendix

ATI Inc. (NYSE: ATI) — Research as of 2026-06-19

Primary sources first. This appendix lists the underlying sources. Third-party aggregated data is reconciled to primary filings.

1. SEC filings (primary) — EDGAR, CIK 0001018963

• FY2025 Form 10-K, filed 2026-02-20 (ati-20251228.htm) — Item 1 Business (segments, end markets, principal competitors, raw materials, R&D, backlog/cyclicality, energy), Item 1A Risk Factors, Item 7 MD&A “Results by Business Segment” (segment & by-market revenue tables). Primary source for segment/end-market mix, competitors, moat language, raw materials, cyclicality.
• FY2021–FY2024 Form 10-Ks (ati-20211231, ati-20221231, ati-20231231, ati-20241229) — multi-year trend, portfolio reshaping, pension annuitization disclosure.
• Form 10-Qs, Q2-2021 through Q1-2026 (ati-20260329.htm latest) — quarterly trend.
• 2026 DEF 14A, filed 2026-03-24 (FY2025 compensation) — incentive metrics (APP EBITDA 60%/FCF 30%/strategic 10%; LTI PSU 70% relative-TSR + RSU 30%), the EVA absolute-price plan, CEO Kim Fields comp ($13.5M), CEO/CFO/Chair transition, insider ownership (~1%).
• 8-K corpus, 2021–2026 — debt issuances ($425M 7.250% 2030 notes Aug-2023; $450M 5.875% 2033 notes Jun-2026; credit-facility amendments 2022/2025), Oct-2023 Athene pension annuitization (7.01 disclosure Nov-2023), leadership transitions (Fields CEO 2024, Foster CFO 2025/26, Chair 2025/26), Feb-2026 +$500M buyback authorization, BWXT naval-nuclear supply agreement (Jun-2026), quarterly earnings 8-Ks.
• Form 4 corpus, 2021–2026 (~221 filings) — insider-transaction read: zero open-market purchases (code P); all activity routine grants/vesting/withholding/sales; net selling into the 2026 run-up.

2. Earnings call transcript (primary management commentary; treated as hypothesis)

• Q1-2026 earnings call, 2026-04-30 — revenue $1.15B, adj EBITDA $232M (20.1% margin), adj FCF $75M, backlog $4.1B ATH, raised FY2026 guidance (adj EBITDA $1.010–1.060B, adj EPS $4.20–4.48, adj FCF $465–525M), naval-nuclear ~$1B renewal, Cameco $250M, missiles +100%, sole-source 5 of 7 advanced jet-engine nickel alloys, capacity plans, 80/20 portfolio actions, ~40% incremental margins. Speakers: Kimberly Fields (CEO), James “Rob” Foster (CFO).

3. Quantitative data services (third-party; reconciled to filings)

• Aggregated fundamentals databases — income statement, balance sheet, cash flow (FY2020–FY2025); profitability ratios (ROIC, ROE, margins); enterprise value and valuation multiples; per-share data. Used for the financial spine and EV.
• Own-history valuation percentiles — composite 98.1, P/E 94.3, P/B 99.98, P/S 99.98 (as of 2026-06-18, price $201.34). Own-history context only, not cross-sectional.
• 5-year daily price history — price history for the Five-Year Event Map: ATH intraday $203.60 (12-Jun-2026), 52-wk range ~$70.42–$203.60, year-end close path.
• Financial news wires — sparse (7 rows); material item: BWXT naval-nuclear supply agreement (2026-06-11, Benzinga, ids 401485/401546).

4. Factor / price-action model (third-party statistical estimates)

• Quantitative factor model — stock-loadings (beta 1.47, market ~1.25–1.42, Aerospace & Defense industry ~0.88, alpha +0.41), leaderboard (y1 +145% Sharpe 3.35, m3 +306% ann, y5 +53.6%/yr, lifetime max DD −94.7%), stock-info (rs_ytd +75.4%, rs_12m +137.7%), related stocks (MOG-A, NUKZ, momentum/midcap-growth ETFs, NVT).

5. Peer / cohort comparison (figures from each issuer’s public filings)

• Carpenter Technology (CRS) public filings ( — closest pure comp; industry consolidation, ~42x EV/EBITDA, 17.2% ROIC, 96.8 pctile).
• Howmet (HWM) public filings ( — higher-quality franchise; ~24.8% op margin, ~21% ROIC, ~42x EV/EBITDA, 95 pctile).
• GE Aerospace, RTX, and FTAI public filings (aero-supercycle demand framing: engine ramp + aftermarket annuity + defense + gas-turbine power).

6. Industry framework references

• Published aerospace and metals-industry primers were used for value-chain framework context only (not current data).

7. Analytical frameworks

• Greenwald & Kahn “Competition Demystified” (moat taxonomy, barriers-to-entry/ROIC/share-stability tests) and Marathon/Chancellor “Capital Returns” (supply-side capital-cycle analysis). Applied throughout the relevant section, the relevant section, the relevant section

Note: this is independent, fresh coverage; the author holds no position-related disclosure and this note is position-agnostic.