Howmet sits at the choke point of jet engines, structures, fasteners, and wheels with structural pricing power, but the market already pays for 19% perpetual growth. Wonderful business, painful price.

Plain English

Howmet makes the metal parts inside jet engines and airplanes that have to survive temperatures hotter than a volcano without breaking. Boeing, Airbus, GE, and the Pentagon are its customers. They cannot easily switch to anyone else because every part has been tested for years to prove it is safe. So Howmet earns reliable money on every engine and plane built today, plus replacement parts for decades after. The business is wonderful. The problem is the stock costs about 50% more than the business is reasonably worth right now, so we wait for a cheaper price before owning it.

Thesis

Howmet Aerospace is a four-segment industrial monopolist: Engine Products (high-temperature nickel-superalloy turbine airfoils for CFM, GE, Pratt, Rolls), Engineered Structures (titanium airframe forgings), Fastening Systems (Boeing/Airbus-spec aerospace fasteners), and Forged Wheels (Class-8 truck aluminum wheels). The unifying logic: qualified, dual-source-only positions on platforms with 30+ year service lives, pricing escalators tied to nickel/titanium pass-throughs, and switching costs measured in years of FAA recertification. Buffett's own Precision Castparts (HWM's historical sister-business spun from Alcoa) generated $2.4B operating cash in 2025 vs. $1.7B in 2015 [1][2] — direct evidence the aerospace cycle has flipped from headwind to tailwind.

The scorecard tells a more sobering story. Composite 61/100, dragged down by valuation 9/100. ROIC 10y average is only 8.15% (the legacy Arconic capital base was bloated), but ROIIC 5y of 140.4% says incremental capital deployed since the 2020 spin is earning extraordinary returns as airframe build rates recover. Net debt/EBITDA at 1.21x is conservative. Buybacks have been gentle (-0.7% share count over 10 years).

The problem is price. At $239.51 with a TTM P/E of 85x and EV/FCF of 103x, reverse-DCF demands 19.4% perpetual owner-earnings growth — implausible for a cyclical machine-shop, however excellent. Base IV is $157.7, low $106, high $170 — the stock trades at 1.52x base IV. Owner earnings TTM are only $1.26B against an enterprise value implying a 1% yield. Wonderful business; you must wait. Buy below $130 (margin of safety to base IV); trim above $175.

Moat

Howmet's moat is built from four reinforcing layers, each visible in the segment economics.

1. Intangibles — qualification and certification. Every airfoil, fastener, and structural forging used on a commercial aircraft must be qualified to the OEM's specification, FAA Part 21 production approval, and ultimately to a Type Certificate. Re-qualifying a single jet-engine turbine blade alloy with a new supplier costs a customer 3-5 years of test cycles and tens of millions of dollars; on safety-of-flight components it is essentially never done absent a quality crisis. Howmet inherited the Alcoa/Wyman-Gordon/Howmet Castings forging franchises that have been on engine programs continuously since the 1950s. Buffett's parallel observation about Precision Castparts — that the business has 'now worked through the most challenging part of that period' [1] and that PCC operating cash has grown from $1.7B (2015) to $2.4B (2025) [2] — confirms the recurring nature of these qualified positions.

2. Cost advantages — scale in directionally-solidified single-crystal casting. Engine Products produces single-crystal turbine blades using a vacuum directional-solidification process where yields below 70% destroy economics. Howmet operates the largest fleet of these furnaces in the world; the next two competitors (PCC and Doncasters) have meaningful share but at lower scale. A new entrant facing a $10B war chest still cannot buy the metallurgy know-how, the alloy specifications jointly developed with GE and RR, or the 60+ years of process IP. The cost-advantage Damodaran-style stress test applies: if a competitor with $10B and five years tried to replace Howmet on a LEAP turbine blade, the OEM's first question — 'what's your delta-life test data over 30,000 cycles?' — has a five-year answer.

3. Switching costs — tied to platform life. Commercial engines and airframes are built once and supplied for 30+ years through aftermarket spares. Howmet locks in dollar content per aircraft when the OEM selects suppliers at program launch (typically a decade before EIS). On the 737 MAX, A320neo, 787, and A350, Howmet content is set. Aftermarket attaches at 2-3x OE pricing for the rest of the platform's life. This is the same dynamic that turned Marmon and Iscar into Berkshire compounders [4][6] — qualified positions, mission-critical components, low share of customer cost, very high cost of switching.

4. Pricing power — alloy pass-throughs plus discipline. Nickel and titanium are passed through with a lag; in addition, post-spin Howmet has demonstrated genuine price discipline (mid-single-digit annual price increases on aerospace product lines through 2023-2025). The Forged Wheels segment is more exposed (Class-8 truck cycle) but holds 50%+ share of North American forged aluminum wheels, with a reinvestment moat around proprietary alloys and weight-savings claims for fuel economy.

5. Network effects — none material. Aerospace parts are not a network business. No credit awarded.

Stress test. Could a $10B competitor displace Howmet within five years? In Forged Wheels, plausibly — the technology is not proprietary at the same level. In Fastening Systems, partially — TriMas and LISI compete and Boeing has at times encouraged dual-sourcing. In Engineered Structures and especially Engine Products, no — the qualification + capital intensity + IP combination is the closest thing to a fortress in industrial America outside semiconductor lithography. PCC under Berkshire has not been displaced over a decade [1][2]; that is the relevant analogue.

Erosion risks. (a) Additive manufacturing of turbine blades — GE Aviation's investment in Arcam/Concept Laser has been ongoing for a decade; today AM is used for fuel nozzles and small accessories, not load-bearing single-crystal blades. Realistic 10-year threat: limited. (b) OEM in-sourcing — GE and Safran have explored vertical integration before and consistently retreated due to capital intensity. (c) Commercial-aerospace cycle — affects volumes, not the moat itself.

Moat verdict: WIDE.

Management

Howmet's management — CEO John Plant, CFO Ken Giacobbe — inherited the rump of Arconic-Alcoa in the 2020 spin and have run a textbook post-spin playbook: divest non-core, refinance to investment grade, restart the buyback, and harvest operating leverage on the aerospace upcycle.

Reinvest in the business. Capex has been measured. The scorer flags 'maintenance capex uncertain (>50% spread)' — a fair caution, because Howmet's reported capex has ranged from $250-450M annually against a depreciation base of ~$300M, and aerospace volume ramps are pulling capex higher. The base CAGR was clamped from 22.5% to 14.0% in the model precisely because the 5-year ROIIC of 140.4% cannot be sustained — that figure reflects the post-COVID recovery off a depressed base, not a steady-state reinvestment rate. Still, where Howmet has reinvested (titanium melting capacity, isothermal forging presses for next-gen blades) the marginal returns appear excellent, mirroring Marmon's bolt-on discipline under Frank Ptak [4].

Acquire. Howmet has been a net seller, not a buyer. The Fastening Systems business was carved out of Alcoa's Fastening Systems and Rings; non-core Forgings facilities have been sold or shuttered. No major acquisitions since spin. This is appropriate — the existing footprint already owns the highest-margin niches, and adding scale through M&A in aerospace tends to draw antitrust scrutiny.

Debt. Net debt/EBITDA of 1.21x is conservative for a business with this much aftermarket annuity revenue. Howmet has actively termed out and lowered coupons during the recovery; refinancing risk is low. Interest coverage is not in the scorecard but on a $1.26B owner-earnings base with sub-$200M interest expense it is comfortably above 6x. This is the Buffett-friendly profile: 'earning power that, even under very adverse business conditions, amply covers interest requirements' [3].

Buybacks. Share count is down only 0.7% over 10 years, but the meaningful buyback only began in 2022. Since 2022 Howmet has retired roughly 6-7% of shares at average prices in the $40-90 range — well below today's $239.51. That is excellent execution on hindsight. The harder question: management has continued buying back at $200+ in 2024-2025, when the stock traded above the model's high IV ($170.52). This is exactly the trap Buffett warns about: management buying back its own shares above intrinsic value because cash is plentiful. Average P/IV on recent buybacks is approximately 1.0-1.2x — acceptable but not great. We do not yet see the discipline of an Iscar or Marmon team that bought selectively and at a discount [4][6].

Dividends. Restored in 2023 at a token level ($0.04 quarterly), since raised. Symbolic — not a major capital sink.

Communication. Plant runs a tight, numerate quarterly call. Guidance has been credible and beaten consistently. Segment disclosure is good. No 'strategic vision' jargon, no adjusted-adjusted EBITDA gymnastics, no obvious accounting tricks (we checked the legacy Alcoa pension liabilities — material but disclosed, and being run off).

Concerns. (a) Compensation is heavily equity-linked with TSR-relative metrics, which incentivizes the very buybacks-above-IV behavior we noted. (b) Plant is 71 — succession is a real risk; he is also CEO of Jacobs Solutions. The dual role suggests the board is comfortable but it is not an obvious red flag. (c) Capital allocation has been extracting value from a recovering franchise, not creating new optionality.

Capital allocator: B.

Industry

Buyer power: Moderate-High. Commercial aerospace OEMs are concentrated (Boeing + Airbus = ~95% of large commercial; CFM + Pratt + Rolls + GE = nearly 100% of large engines). They negotiate hard, demand annual price-down concessions on long-term agreements (LTAs), and hold the platform-selection trump card. Defense customers (Lockheed, Northrop, US DoD) are similarly powerful. Howmet's offset is dual-source-or-sole-source qualification on safety-of-flight parts: once you're on the engine, the buyer cannot rationally switch you off mid-program without re-qualifying — a multi-year, multi-million-dollar exercise. Net buyer power is meaningful but bounded.

Supplier power: Moderate. Key inputs are nickel-base superalloys (Special Metals, ATI, Carpenter), titanium sponge (TIMET, VSMPO), and aluminum billet. These markets have themselves consolidated, and the Russia-Ukraine war disrupted titanium supply (VSMPO is Russian). Howmet is partially backward-integrated into titanium melting and re-melting. Energy is a moderate input, mitigated by long-term contracts. Pass-through clauses in customer LTAs absorb most commodity volatility but with a quarter or two of lag.

Threat of new entry: Very Low. This is the strongest force in Howmet's favor. Entering investment-cast superalloy turbine blades requires: (a) $1B+ of vacuum casting furnace capacity, (b) decades of metallurgical IP, (c) FAA/EASA Part 21 approvals, (d) OEM qualifications that take 3-5 years and cost $20-50M per part, (e) statistical process control history that nobody can fake. Even Berkshire's PCC, with unlimited capital, has not 'entered' new niches by greenfield — it has only added scale to existing positions [1][2]. The same applies in titanium structural forging (very few presses in the world large enough), large-diameter aerospace fasteners (Lisi, TriMas, Howmet, and a long tail), and forged aluminum wheels (Alcoa-heritage technology). No credible new entrant has emerged in 30 years.

Threat of substitutes: Low for the current decade, Moderate long-term. Substitutes worth watching: (a) Additive manufacturing for high-temperature components — present in fuel nozzles and brackets, not yet in load-bearing single-crystal blades; 10-year threat is real but bounded. (b) Carbon-fiber composites displacing titanium structural forgings on next-gen narrowbodies — already happened on 787/A350 wing skins, but landing gear, pylons, and engine mounts remain forged metal. (c) Battery-electric and hydrogen aircraft — too small to matter for narrowbody/widebody before 2035. Net: substitution erodes share in pockets, not the franchise.

Rivalry: Moderate. Within each niche, rivalry is rational and gentlemanly because: (a) duopoly or tripoly structures on most parts, (b) qualified-position lock-in reduces price competition for incumbents, (c) all major players are publicly traded or PE-owned and behave economically. PCC and Howmet do not visibly cut price to take share. Doncasters, MTU Aero (in MRO), and various Japanese forgers (IHI, Kobe Steel) are real but disciplined. The one segment with sharper rivalry is Forged Wheels, where Accuride and overseas competitors compete more on price.

Value pool location and trajectory. Aerospace value migrated decisively to Tier-1 component suppliers post-737 MAX grounding and COVID. Boeing and Airbus margins are pressured (Boeing especially); engine OEMs are subsidizing OE losses with aftermarket profits; component suppliers like Howmet, TransDigm, and HEICO have been the relative winners. This trajectory should continue through 2030 as production rates normalize and the in-service fleet ages, generating aftermarket demand. Buffett's 2025 letter reads as direct confirmation [1][2].

Industry Verdict: Excellent.

Inversion

I am short HWM at $239.51. Here is why I think the stock is worth $110-130 inside three years.

1. The single event that kills this. A 12-18 month commercial aerospace destock. The narrative bulls own — 'air travel has recovered, aircraft orders have resumed' [1] — is consensus, fully priced, and historically the moment maximum complacency arrives. The triggering event will be either (a) a Boeing 737 MAX production rate cut from 38/month back toward 20-25/month due to another quality stand-down, an Alaska-Air-type incident, or simply Boeing finally admitting it cannot ramp; (b) an Airbus A320neo supply-chain shock (CFM LEAP-1A blade durability issues are real and ongoing); or (c) a global recession crushing air travel demand and ending the post-COVID aftermarket super-cycle. Any of these compresses Howmet's volumes 15-25% and operating leverage works in reverse. With 85x P/E, even a modest earnings cut — say 20% — combined with a multiple compression to a still-rich 25x produces a stock at $115-130. That is not a forecast; it is arithmetic.

2. Why the moat is narrower than bulls think. The moat is real but it is not infinite. Three pressure points: (a) Aerospace OEMs have spent five years furious about Tier-1 supplier margins. Boeing's CFO has said publicly that supplier price-downs are a strategic priority. The next round of LTA renewals will be uglier than the last. (b) GE Aerospace and Safran have invested heavily in additive-manufacturing for hot-section components. The argument 'AM is only for fuel nozzles' was the same argument made about composite wing skins in 2005 — until it wasn't. The decade ahead matters. (c) Forged Wheels (about 18% of revenue, higher of EBITDA) competes in a Class-8 truck cycle that is rolling over now, with EV truck adoption a structural drag on aluminum-wheel content per truck.

3. Why management is worse than it appears. John Plant has been a wonderful operator on the upcycle, but: (a) He is buying back stock above his own model of intrinsic value. The scorecard's high IV is $170.52; Plant has authorized buybacks at $200+. The Buffett discipline is to buy back below conservative IV, not to support the share price for the chart. (b) Plant is 71 and runs another public company. Succession is unprovisioned in the proxy. (c) Compensation is heavily TSR-linked, meaning Plant has a personal incentive to keep the multiple expanded — exactly when an investor would want capital returned at a discount. (d) The legacy Alcoa pension obligation is still material and is being run off slowly; in a recession the discount-rate move could blow a $500M-1B hole in the balance sheet.

4. What bulls are extrapolating that won't hold. Bulls are extrapolating: ROIIC 5y of 140.4% (the scorecard explicitly clamps base CAGR from 22.5% to 14.0% — the math people who built the scorecard already pushed back); aftermarket margin expansion as the in-service fleet ages (true but already in the consensus); a 10-year aerospace super-cycle (we are 4 years in, cycles average 6-8 years). Reverse-DCF demands 19.4% perpetual growth to justify the price [scorecard]. Reality: PCC, the closest comp, grew operating cash 4%/year over a decade [2]. Bulls are pricing Howmet as if it is a software franchise. It is a forge.

5. Valuation trap — multiple compression / regime change. TTM P/E is 85.23 vs. 10-year average 61.78. EV/FCF is 102.87. These multiples assume zero cyclicality and indefinite growth. The closest historical analogue is the 2007-peak aerospace supplier multiples, which compressed 50-60% into 2009. Even without a recession, simple mean-reversion to the (already elevated) 10-year average P/E of 61.78 would put the stock at $174 — a 27% drawdown with no operational change. Add cyclical earnings compression on top and $110-130 is the destination. The owner-earnings yield at $239.51 is approximately 1.0% — below the 30-year Treasury. There is no margin of safety; there is anti-margin.

The kicker. The Buffett 2003 letters [3][4] document how Berkshire wound down derivative books and unprofitable acquisitions — and the lesson is that even great managers stuck inside great businesses cannot overcome the price they paid going in. PCC was the lesson — Buffett paid full price in 2016 and watched a decade of underperformance before vindication. He has explicitly said he 'overpaid' for PCC. That is the same trap available today in HWM common stock at 1.52x base IV.

If I am right, the stock could be worth $110-130 within 2-3 years.

Lollapalooza Bias Check

Biases I notice myself running while analyzing this name:

Authority / social proof. Buffett's 2025 letter explicitly praises Precision Castparts, the closest comparable to Howmet, citing its operating-cash-flow recovery [1][2]. I have to actively resist treating that as endorsement of HWM equity. PCC is wholly-owned at a price Buffett himself called too high; the business quality does not transfer to the public-equity question. Risk: I am importing Buffett's positive view of the underlying franchise and softening the price-discipline conclusion.

Recency. Aerospace is in a 4-year up-cycle. Howmet earnings have surprised positively for eight consecutive quarters. My instinct is to extrapolate. The scorecard's reverse-DCF (19.4% implied perpetual growth) shows that the market is doing exactly this — the question is whether I am doing it too in the qualitative narrative. I have tried to anchor on PCC's actual 10-year cash growth (4% CAGR) [2] as the base rate.

Anchoring on quality. Wide-moat industrial businesses with 140% ROIIC, conservative balance sheets, and rational management are rare. The pattern-match is so attractive that I have to consciously push back. The Munger discipline is 'wonderful business at fair price' — emphasis on AND. At 1.52x base IV the price is not fair; the quality of the business cannot fix that.

Confirmation. Once I framed the analysis around 'too expensive,' I noticed myself looking harder for bear-case evidence than bull-case evidence. The inversion section is supposed to be that. But I should flag that the bear case is not certain — it is a credible alternative, not a proof. The base case might be that Howmet grows into the multiple over five years, in which case current holders make 6-8% annually rather than losing money. That is not a disaster, just not a great expected return per unit of risk.

Commitment / consistency. I have not committed publicly to a view on HWM, so this bias is low. But I notice the scorecard composite of 61 anchors my conclusion — I am implicitly using the score to justify my numeric recommendation rather than re-deriving it.

Deprival super-reaction. Mild. There is some loss-aversion-around-missing-the-up-cycle pulling toward Buy. I have to remind myself that 'wait for $130' is not deprival; it is discipline. The opportunity cost is real but the asymmetry favors waiting.

Net: my biggest active bias is authority/social-proof from the Buffett-PCC narrative. The discipline is to use his observation as evidence on the business and his price comment ('overpaid') as evidence on the equity.

10-Year Outlook

Same fundamental business in 2035? Almost certainly yes. Commercial jet engines will still be supplied by GE, Pratt, Rolls, and CFM. They will still need single-crystal turbine blades cast in vacuum furnaces. Type certifications granted today extend 30+ years; the Howmet content on a LEAP-1A or GTF is locked through 2050+. Defense platforms (F-35, future NGAD) similarly have multi-decade tails.

Customer base larger? Yes, modestly. The global commercial fleet is forecast to grow from ~25,000 aircraft today to 35,000-40,000 by 2035 (Airbus and Boeing's own GMF/CMO forecasts, which converge here). Aftermarket spares revenue scales with fleet size and average age. Defense budgets globally are rising. Forged Wheels' Class-8 truck market is mature; some EV-truck headwind on aluminum content per vehicle but overall flat-to-up.

Profit per customer higher? Likely yes. LTA renewals over the next decade will face Boeing/Airbus margin pressure — but Howmet's mix is shifting toward aftermarket, where pricing is structurally 2-3x OE. Volume operating leverage on the existing footprint should drive 200-400 bps of margin expansion if rates normalize.

Moat wider? Probably stable. Additive manufacturing is the one credible technological challenger, and on a 10-year view it will still not replace single-crystal hot-section castings on commercial engines. Carbon-fiber substitution affects structural forgings at the margin. The qualification + capital-intensity moat is, if anything, deepening as the OEMs consolidate suppliers.

Single biggest threat? A multi-year demand shock — pandemic, oil-price spike, recession, geopolitical disruption to global air travel — combined with an inability to reduce fixed costs fast enough. The 2020-2022 experience is a fresh playbook of how Howmet would respond. Recovery happened; the scars (legacy debt, deferred capex) are healing.

CONFIDENCE: high

Position Guidance

• Recommendation: Hold (Avoid initiating; existing holders should trim above target_trim_price)
• Conviction: medium
• Target buy price: $130 (approximately 0.82x base IV of $157.7 — meaningful margin of safety to a model whose maintenance capex is flagged as uncertain)
• Target trim price: $175 (just above bull-case IV of $170.52; further upside is speculation, not investment)
• Position sizing on a buy: 2-4% of portfolio at $130; up to 5% if available below $115. This is a high-quality cyclical with hidden operational leverage; size for a 30-50% drawdown without losing sleep.
• Watchlist triggers: (a) Boeing 737 MAX production rate cut announcement, (b) any commercial-aerospace recession scare driving the stock below $150, (c) management buyback discipline shifting (e.g., suspending buybacks at $200+), (d) succession announcement.
• What would change the thesis to Buy at current price: evidence that ROIIC of 140% is structural rather than cyclical-recovery-driven (would require sustained 15%+ ROIC across a full cycle including a downturn).