STLD 22 min read

Steel Dynamics, Inc. (STLD) — Investment Tree v1

Stage 7 final essay. Bilingual companion: tree_v1_zh.md.

Date: 2026-08-02 · Anchor price: $251.26 · Market cap: ~$36.01B · Trailing P/E: 23.16x

Archetype: A tariff-driven cyclical earnings surge funding an unproven second-metal growth platform — no clean fit in the Phase 1 archetype library; closest internal analogs are this corpus's NEM/GS/MS "cyclical, evidence-thin, genuinely undecided" cohort, blended with a genuine Pre-Inflection Growth question (Aluminum Dynamics) that those three pure-cyclicals don't carry. Proposed new archetype: "Windfall-Funded Platform Transition" (see frameworks.md).

SOURCE QUALITY: GENERATE mode (Tier B/C) throughout. SEC EDGAR direct fetch (data.sec.gov / www.sec.gov Archives) returned HTTP 403 from the build sandbox on every attempt — no 10-K, 10-Q, or 8-K was directly read this cycle. All company-specific figures are WebSearch-aggregated from named primary filings as surfaced through secondary sources (company IR excerpts, PR Newswire, Benzinga/Seeking Alpha transcripts, MacroTrends, StockTitan). Peer multiples in Section VI/peers.md are Tier C (not a live quote-feed pull) — a materially weaker evidentiary basis than this corpus's NEM/GS/MS precedent, flagged explicitly rather than presented as resolved. R2 verification against the primary FY2025 10-K and Q2 2026 10-Q is required before this tree is treated as fully resolved — see the open-questions list in dashboard.md.

0. Company Fundamentals — what Steel Dynamics is and how it earns

Figures FY2025 / Q1-Q2 2026 unless noted.

What it is & how it earns. Steel Dynamics is a vertically integrated electric-arc-furnace (EAF) steel producer and metals recycler operating four segments: Steel Operations (flat-rolled and long-product EAF mills, 72% of FY2025 net sales), Metals Recycling (OmniSource ferrous/nonferrous scrap), Steel Fabrication (joists and deck for nonresidential construction), and — new as of 2026 — Aluminum Operations (Aluminum Dynamics, a flat-rolled aluminum mill in Columbus, MS). Revenue is driven by shipment volume × realized selling price, with steel selling prices currently elevated by Section 232 tariff protection. FY2025 net sales were $18.2B; Q2 2026 net sales were $6.09B (record quarterly, +33% YoY).

Unit economics. STLD's EAF metal-spread model (selling price minus scrap/alloy input cost) carries lower capital intensity and faster capacity flex than blast-furnace integrated competitors — high operating leverage on the upside when selling-price realization outpaces input-cost inflation, which is the mechanism behind Q2 2026's EPS surge (diluted EPS $3.69 vs. $2.01 a year earlier, on 33% revenue growth). Management has flagged that 75-80% of Q2 2026 flat-rolled business was still tied to lagging (below-spot) contracts, implying continued price pass-through into Q3.

Cash-flow anatomy.

Q2 2026 diluted EPS narrowly missed the $3.71 consensus by $0.02 while revenue beat consensus (~$5.68B) by ~$410M (evidence STLD-2026-08-02-Q2-001) — a beat-on-top-line, near-miss-on-bottom-line quarter, not a clean beat-and-raise.

Balance sheet & capital allocation. Cash $769.9M at FY2025 year-end (+31% YoY); total liquidity >$2.2B. FY2025 capital returns: $901M share buybacks + $291M dividends (a combined $1.19B against ~$1.4B OCF — a high payout ratio in a peak-earnings year). Quarterly dividend raised 6% to $0.53/share, carried through Q1 and Q2 2026. Total debt, net debt, and leverage ratio were NOT resolved this session — flagged prominently throughout this tree as the single largest capital-structure gap (primer.md R2 item #4).

What drives it. Two forces in active tension: (1) Section 232 tariffs (steel/aluminum import tariffs raised toward 50% via the June 2026 proclamation) — a single, executive-branch, historically-reversible policy lever that is the entire proximate cause of 2026's earnings surge; (2) the ~$2.9B Aluminum Dynamics platform, physically complete (third and final cold mill commissioned July 2026) but commercially unproven, guided to eventually contribute $650-700M in through-cycle EBITDA. The tree resolves whether the tariff windfall is financing a durable category shift (steel cyclical → multi-metal platform) or whether the market is simply paying a peak multiple for peak, policy-dependent earnings.

I. One-sentence verdict

STLD's 23.16x trailing P/E — well above the historical 8-12x through-cycle norm for domestic steel producers — requires the market to believe either that 2026's tariff-driven earnings surge is close to durable, or that some portion of the multiple already reflects unstated credit for the Aluminum Dynamics diversification bet; this session's evidence cannot resolve which is true (7 of 15 leaves are honestly ⊗, concentrated exactly on the crux — an internally contradictory utilization guidance figure for the aluminum ramp, CONTRADICTION-001), and the valuation math is unambiguous regardless: at a 0.42× asymmetry ratio, today's price offers unfavorable risk/reward for a fresh entry, even though durability (18/25) clears the long-term-hold threshold on the strength of a genuine — if unproven — reinvestment runway.

II. Company snapshot

Steel Dynamics, Inc. (Fort Wayne, IN; CEO/Chairman Mark D. Millett, co-founder) is a domestic EAF steel major and one of the "Big Three" (alongside Nucor and Cleveland-Cliffs) benefiting from 2026's Section 232 tariff escalation. FY2025 revenue $18.2B, record 13.7M-ton steel shipments; Q1-Q2 2026 results show a two-quarter earnings acceleration (Q1 operating income +73% YoY; Q2 net income +79% YoY). Alongside the steel-cycle windfall, STLD has committed ~$2.9B to Aluminum Dynamics, a flat-rolled aluminum platform in Columbus, MS, physically completed July 2026 but still loss-making (losses narrowing) as of the most recent quarter. Steel Fabrication backlog — the strongest forward-demand signal in the company — is up ~45% YoY with visibility into 2027. STLD completed the acquisition of the remaining 55% of New Process Steel (its own largest flat-rolled customer) in December 2025, continuing a nearly two-decade pattern of adjacency-only bolt-on M&A (OmniSource, 2007, being the template).

III. The five facts that drive everything

1. Q1-Q2 2026 earnings acceleration is a two-quarter pattern, not a single-quarter anomaly — Q1 operating income +73% YoY, Q2 net income +79% YoY, on record shipments both quarters. ✅B
2. STLD trades at 23.16x trailing P/E — well above the historical 8-12x through-cycle norm for domestic steel — with sell-side consensus lopsidedly bullish (9 buy / 1 sell, avg target $272.38, range $221-310). ✅C (market-data snapshot)
3. Section 232 tariffs escalated to ~50% on steel/aluminum imports via a June 2026 proclamation; domestic steel imports fell ~40% YoY to the lowest level since the 2008 financial crisis — the proximate cause every source names for the current earnings level. ✅B
4. Aluminum Dynamics is a ~$2.9B committed program, physically complete (third cold mill, July 2026) but still loss-making, with internally inconsistent utilization guidance for exiting 2026 (40-50% vs. 75% — CONTRADICTION-001, unresolved this session). ⊗ (the crux fact — resolution pending)
5. No sell-side source gathered this session builds a blended steel+aluminum sum-of-the-parts valuation for STLD — every price-target action and trade-press framing discusses STLD purely in domestic-steel-cyclical terms. ✅B — this is the direct evidence for the structural-blindness mispricing mechanism.

IV. The H-0 thesis

H-0 (one sentence): STLD's above-historical-norm 23.16x trailing multiple is not fully explained by the tariff-driven steel cycle alone, and the market has not yet decided whether to credit the multiple to a durable category shift (single-commodity steel cyclical → funded multi-metal platform) or reprice it back toward historical steel norms once the tariff tailwind normalizes.

Mispricing taxonomy: Structural blindness × Category (per mispricing.md, primary). Sell-side coverage-universe classification (GICS steel sub-industry, "domestic steel major" peer set) has not been updated to reflect that Aluminum Dynamics is building a second, structurally distinct metals platform inside the same company — a classification lag that is a feature of how coverage assignment works (re-peer-grouping happens only after a segment forces it), not a considered judgment the reclassification is premature.

Secondary mechanism: Timing gap × Lifecycle. The market broadly agrees on direction (aluminum will eventually ramp; tariffs will eventually normalize to some degree) but not horizon — and the two horizons racing against each other (aluminum-ramp-to-materiality vs. tariff-tailwind-durability) is the tension the contradiction surfaces. If the tariff cushion fades before Aluminum Dynamics reaches materiality, the funding relationship that makes the category-shift bull case work breaks down.

3 falsification conditions (per h0_thesis.md, unchanged from Stage 0-2 — none have fired this cycle):

• FF1 Aluminum Dynamics utilization stalls below the low end of guidance (sustained sub-40% Columbus utilization through end of 2026) for two or more consecutive quarters → falsifies the "funded category shift in progress" reading
• FF2 A Section 232 proclamation rollback or material carve-out expansion occurs within 12 months, and steel-segment margins compress back toward pre-2026 levels within two quarters → falsifies the "durable re-rating" bull reading
• FF3 Sell-side coverage explicitly re-peer-groups STLD alongside diversified multi-metal industrials within 18 months, with a published blended valuation model → this is a resolution of H-0 (the positive-case exit condition), tracked alongside the falsification conditions

V. Tree — five branches

H-0: STLD's 23.16x multiple is not fully explained by the tariff-driven
     steel cycle alone -- unresolved whether the market is crediting a
     durable category shift (steel cyclical -> multi-metal platform) or
     will reprice toward historical steel norms once the tariff wall
     normalizes. 7 of 15 leaves are honestly ⊗, concentrated on the
     crux (Aluminum Dynamics utilization contradiction).
│
├── L1A — Tariff-durability trajectory  ⚠️C mixed, mostly not yet testable
│   ├── 1.1 Carve-out expansion since June 2026 proclamation      ⊗ not yet testable (window too short)
│   ├── 1.2 2018-2019 historical base rate                        ⚠️C partial (persistence-with-erosion, not a clean cliff)
│   └── 1.3 Real-time import-volume trend                          ⊗ insufficient evidence
│
├── L1B — Steel-cycle earnings normalization  ⚠️B net positive
│   ├── 1.1 Lagging-contract mechanism fully worked through?       ⊗ not yet testable (Q3 2026 pending)
│   ├── 1.2 Fabrication backlog accelerating/plateauing            ✅B strongly supported (38%→45% YoY, accelerating)
│   └── 1.3 FY2025-2026 capital-return cadence continuing          ⚠️B partial (FY2025 confirmed; 2026 buyback $ not yet disclosed)
│
├── L1C — Aluminum Dynamics ramp execution (CRUX)  ⊗ genuinely unresolved
│   ├── 1.1 Reconciled Columbus utilization trajectory             ⊗ CONTRADICTION-001 — unresolved internal contradiction
│   ├── 1.2 Segment operating-loss narrowing rate                  ⚠️C partial (directionally positive, rate unquantified)
│   └── 1.3 Product qualification → customer commitments           ⊗ insufficient evidence
│
├── L1D — Category/peer-group re-rating mechanics  ⚠️B mechanism confirmed, unresolved outcome
│   ├── 1.1 Any sell-side blended SOTP model published?            ✅B confirmed NO — structural-blindness mechanism intact
│   ├── 1.2 Capability transfer: scrap/EAF vs. aluminum-metallurgy ⚠️B partial (input side transfers; go-to-market side doesn't)
│   └── 1.3 SOTP valuation implication if guidance is hit           ⊗ deferred to Stage 4 (see Section VI/VII)
│
└── L1E — Capital allocation & balance-sheet capacity  ⚠️B net positive, key gap
    ├── 1.1 Total debt / net leverage / remaining capex             ⊗ insufficient evidence (single largest gap in the tree)
    ├── 1.2 Bolt-on M&A track record (OmniSource, New Process Steel) ✅B strongly supported — nearly 2 decades disciplined
    └── 1.3 30-40% OCF-decline stress test                          ⚠️B partial (illustrative math reassuring, not primary-sourced)

Total: 3 ✅ / 5 ⚠️ / 0 ✗ / 7 ⊗ across 15 leaves
H-0 verdict: GENUINELY UNRESOLVED, ~48% confidence

VI. Key findings

Finding 1 — The crux question (aluminum ramp) is exactly where the evidence is thinnest

CONTRADICTION-001 — two irreconcilable Columbus utilization figures for exiting 2026 (40-50% vs. 75%) — sits directly on top of the single most consequential open question in the entire tree. Every other branch's confidence is downstream of this one unresolved fact: if Aluminum Dynamics is genuinely on-track (75%), the funded-platform-transition bull case has real legs; if it's badly behind (40-50%), the entire "category shift" reading collapses back to a pure cyclical-peak story. This session could not resolve it — R2 (Q3 2026 earnings call or 10-Q MD&A) is mandatory before the next material verdict revision.

Finding 2 — The structural-blindness mechanism is confirmed real, but that alone doesn't tell you which way it resolves

leaves.md L1D 1.1 confirms with high confidence that no sell-side source has built a blended steel+aluminum valuation model for STLD — the classification-lag mechanism mispricing.md describes is genuinely operative. But confirming the mechanism exists is not the same as confirming the market is underpricing STLD — it's equally consistent with a market that is correctly ignoring an unproven, pre-commercial segment. This distinction is the entire reason H-0 lands at ~48% confidence rather than higher.

Finding 3 — The strongest evidence in the tree is behavioral, not financial

The three ✅B leaves (fabrication-backlog acceleration, structural-blindness-persists, and the bolt-on-M&A track record) are all about demonstrated, already-disclosed corporate behavior — the same pattern this corpus's other GENERATE-mode cyclical builds (NEM, GS, MS) show. Backlog trends and M&A discipline are easier to verify from press releases than commodity-cost mechanics, segment financials, or forward policy trajectories — a real strength, but one that leaves the harder financial questions (segment P&L, debt/leverage, utilization) unresolved.

Finding 4 — Risk/reward at $251.26 is unfavorable, and more sharply so than this corpus's other cyclical names

Per implied_prob.md: Bull +19.4% ($300) vs. Bear -46.3% ($135) — an asymmetry ratio of 0.42×, worse than NEM (0.77×), GS (0.94×), or MS (0.94×). The mechanism is mechanical, not a judgment call: STLD's 23.16x sits much further above its own historical 8-12x through-cycle norm than those peers' starting multiples did, so even a deliberately softened Bear case (14-15x, not a full cliff back to 8-12x) still produces a large percentage downside simply because there's more room to fall.

Finding 5 — Durability clears the long-term-hold threshold, but on a genuinely different basis than this corpus's other cyclicals

durability_test.md scores 18/25 — above NEM's 16/25 — specifically because STLD has a real, funded, already-underway second-platform reinvestment (Aluminum Dynamics) plus a nearly-20-year bolt-on M&A track record, neither of which a pure harvest-mode miner or bank carries. But this is explicitly a qualified Selective-Hold: the moat-trajectory (Q2) and capital-allocation (Q3) scores are both capped by the same unresolved facts (aluminum execution, debt/leverage) driving the low H-0 confidence. Durability and valuation point in different directions here — a genuinely interesting, non-trivial combination this tree does not paper over.

VII. Three valuation scenarios

(See scenarios.md for full detail.)

Probability-weighted 12-month expected value: ~$239.75 — -4.6% below spot. Asymmetry ratio 0.42× (unfavorable, and sharply so) per implied_prob.md. Flat 25/50/25 prior used — no documented basis this session for a customized deviation.

VIII. Triggers and red flags

(Full detail in triggers_redflags.md.)

Triggers (Bull-case fires):

• T1 (~Oct 2026): CONTRADICTION-001 resolves toward ~75% Columbus utilization — the single highest-priority resolving event in the tree
• T2 (~Q4 2026/Q1 2027): Aluminum segment loss narrows to near-breakeven
• T3 (any time): Sell-side publishes a blended steel+aluminum SOTP model
• T4 (~Oct 2026): Fabrication backlog holds/accelerates a 3rd consecutive quarter
• T5 (ongoing): Section 232 tariff coverage holds stable or expands
• T6 (~Oct 2026): Primary-source debt data confirms conservative leverage

Red flags (Bear-case fires):

• RF1 (~Oct 2026 - Feb 2027): CONTRADICTION-001 resolves toward 40-50% AND segment loss fails to narrow
• RF2 (any time): Tariff carve-outs expand materially with visible margin compression
• RF3 (any quarterly print): Fabrication backlog decelerates
• RF4 (~Oct 2026): 2026 buyback cadence pulls back materially from FY2025 pace
• RF5 (~Oct 2026): Primary-source debt data shows material leverage
• RF6 (any time): Sell-side reaffirms pure-steel framing with a lowered target

IX. Long-term holdability verdict

durability_test.md scores 18/25 — above the 17-point Selective-Hold threshold. No escalation warning fires under the CLAUDE.md long-term-investability workflow, and no fatal flag is present (Q3 3/5, Q4 4/5, balance-sheet survivability unconfirmed-but-not-negative). This IS a genuine Selective-Hold candidate on durability grounds — driven specifically by a reinvestment runway (Aluminum Dynamics + a nearly-20-year bolt-on M&A pattern) that distinguishes STLD from this corpus's harvest-mode cyclicals (NEM) or capital-return-dominant financials (GS, MS).

What this means in practice: durability and current-price valuation point in different directions. STLD clears the bar for a long-term-hold-eligible business — but implied_prob.md's 0.42× asymmetry ratio means today's price is not offering favorable entry terms for that business. This is not a contradiction; it is the honest output of a tree that separates "is this a durable business" (Section IX) from "is this a good price to buy it at" (Section VII) rather than collapsing the two into a single verdict.

Position-sizing recommendation

• Current (pre-CONTRADICTION-001-resolution): 0% — WATCH. The valuation asymmetry (0.42×) argues against a fresh entry regardless of durability, and the crux uncertainty (aluminum ramp) is genuinely unresolved, not just cautiously priced.
• Post-T1 confirmation (utilization resolves toward 75%, ~Oct 2026): Consider a 1-2% starter — durability plus a resolved crux would materially improve the risk/reward, though the tariff-durability question (L1A) would remain open.
• Post-T1+T2+T6 confirmation (utilization, loss-narrowing, AND conservative leverage all confirm): Scale to 2-3%, capped by the archetype's cyclical-macro-sensitive classification, not a compounder-style 5%+ allocation.
• Post-RF1 or RF2 (either alone): Avoid initiating; do not average into cost/multiple compression.

Correlated-exposure note (per K.3.4)

Load-bearing macro factor: Section 232 tariff policy durability + domestic steel-cycle pricing power. cycle_exposure: uncorrelated to the AI-capex factor overlay (NVDA/TSM/AMD/AJNMY/ASML/AVGO/MU cluster) — no cap-adjustment coordination needed with that bucket on the primary driver. Caveat (per assumptions.md C2, carried forward, not dropped): Steel Fabrication's backlog strength (the smallest of STLD's four segments) is partially data-center/AI-capex-driven and not yet decomposed — if held alongside a large AI-capex-cluster position, this secondary linkage is worth monitoring even though it doesn't change the primary uncorrelated band.

X. Historical analogue

The 2018-2019 Section 232 tariff episode (Trump administration, first term). The direct historical precedent for 2026's tariff wall: steel/aluminum Section 232 tariffs were imposed in 2018, and — per leaves.md L1A 1.2's training-knowledge recall (Tier C, not independently re-sourced this session) — the headline policy proved durable in aggregate (steel tariffs remained continuously in force from 2018 through today, across two administrations) while eroding at the margins via negotiated country-specific quotas and carve-outs (Canada, Mexico, EU, South Korea all received exceptions within 12-24 months of the original proclamation). This is precisely the "persistence with gradual erosion, not a clean cliff" pattern assumptions.md A1 argues neither the bull nor bear narrative gathered this session engages with directly. If 2026 follows the same arc, STLD's tariff-driven earnings level neither collapses on a single date (the bear case's cleanest scenario) nor holds at a perfectly constant level indefinitely (the bull case's cleanest scenario) — it likely erodes slowly, country carve-out by country carve-out, which is a genuinely harder scenario for either side of this tree to get a clean vindication or refutation from.

A second, internal analogue: STLD's own Sinton, TX flat-rolled steel mill ramp. Though not independently re-sourced to primary-source precision this session, STLD's own recent history includes at least one prior instance of ramping a major new mill (Sinton) to full capacity — the company's stated "flexible mini-mill capacity additions" moat claim (per primer.md Section 5) rests partly on this track record. If Aluminum Dynamics follows a similar execution arc, the CONTRADICTION-001 utilization gap may simply reflect normal early-ramp variability rather than a genuine execution problem — but this is an inference from an unverified internal precedent, not confirmed evidence, and should be weighted accordingly.

XI. When H-0 fails

Scenario 1: CONTRADICTION-001 resolves toward the 40-50% utilization reading, and the segment loss fails to narrow meaningfully by Q4 2026 (RF1/FF1). The market reads this as confirmation the aluminum bet is behind schedule at exactly the moment the tariff cushion funding it is most needed. scenarios.md's Bear case ($135) likely understates the downside in this specific scenario, since a confirmed execution failure would also cast doubt on management's capital-allocation credibility (Q3 durability grade) simultaneously — a compounding effect neither this tree's Bear case nor the bull/bear narratives gathered this session fully model (per assumptions.md C1's inversion test).

Scenario 2: A Section 232 proclamation amendment expands carve-outs materially, with visible margin compression within two quarters (RF2/FF2). Falsifies the "durable re-rating" bull reading independent of Aluminum Dynamics' progress. Given the 2018-2019 historical analogue (Section X), a gradual version of this scenario (incremental carve-out expansion) is arguably more likely than a sharp single-event rollback — meaning this risk may manifest as slow multiple compression over several quarters rather than a single sharp repricing event, which is harder to act on decisively than a clean red-flag trigger.

Scenario 3: A sell-side house explicitly re-affirms the pure-steel-cyclical framing with a lowered target, with no blended model appearing (RF6). Would suggest the market is actively resolving away from the category-reassignment bull case, not merely lagging it — a more negative signal for H-0 than continued silence, since it implies analysts have looked at Aluminum Dynamics and found it wanting rather than simply not yet covered it.

XII. Investment Scorecard (per MANUAL_en.md Part K.6)

15-question scorecard (analytical-tree Q-list, Format B per K.3.5)

Verdict tally (M1 evidence-tier suffixes per K.3.6): 10 ✅(B) · 2 ⚠️(B) · 3 ⚠️(C) · 0 ✗ — Q1✅B, Q2✅B, Q3✅B, Q4✅B, Q5⚠️B, Q6✅B, Q7✅B, Q8⚠️C, Q9⚠️C, Q10✅B, Q11✅B, Q12✅B, Q13✅B, Q14⚠️B, Q15⚠️C.

K.3.5 Weighted-score derivation

Applying the 4-tier weighting from MANUAL_en.md §K.3.5 (verdict values: ✅ = 1.0, ⚠️ = 0.5, ✗ = 0.0):

xii_score = 31.0 / 39 = 79%. This matches the INDEX_META declaration above — per CLAUDE.md's xii_score drift-check discipline, the body computation is authoritative.

79% = moderate-buy-with-sizing band (65-85%), toward the lower-middle of that range. The two Critical-tier ⚠️s (Q9 unresolved debt, Q14 contested moat) are exactly the two findings that also cap the durability score's Q3/Q2 grades below a higher tier — internally consistent with Section IX, not a separate contradictory signal. Note the meaningful gap between xii_score (79%, structural quality) and h0 (48%, thesis confidence) — per MANUAL_en.md's convention, this gap names real headroom: STLD's underlying business scores well on the standard 15-question checklist, but the specific H-0 thesis (category reassignment via Aluminum Dynamics) remains genuinely unresolved at the evidence level, a wider gap than this corpus's NEM precedent (81% vs 64%, a 17-point gap) — STLD's gap is 31 points (79% vs 48%), reflecting a thesis test that is unusually evidence-thin even relative to structural quality.

Scorecard summary

Final verdict: WATCH — 0% at current price, contingent 1-2% starter on CONTRADICTION-001 resolution

The 2-minute pitch:

"Steel Dynamics trades at 23.16x trailing earnings — well above the historical 8-12x steel-cycle norm — on the back of a two-quarter tariff-driven earnings surge (Section 232 tariffs pushed import competition to 2008-crisis lows). Layered on top is a ~$2.9B aluminum-platform bet, physically complete as of July 2026 but still loss-making, that no sell-side analyst has yet built into a blended valuation model. The bull case: the tariff windfall is financing a genuine category shift from single-commodity steel cyclical to diversified multi-metal industrial, and the market's 23x multiple already reflects partial, unstated credit for that shift. The bear case: this is a peak-multiple-on-peak-earnings setup funded by a single reversible policy lever, and the aluminum bet's own utilization guidance is internally contradictory (one source says 40-50% exit-2026 utilization, another says 75% — they can't both be right). This session's research could not resolve which reading is correct — 7 of 15 tree leaves are honestly marked insufficient-evidence, concentrated exactly on that contradiction. What IS resolvable: the valuation math is unfavorable regardless (0.42x asymmetry — the Bear case has much further to fall than the Bull case has to rise), so this is a WATCH, not a buy, until Q3 2026 earnings (~October) resolve the utilization question one way or the other."

Risk types most relevant (per MANUAL_en.md Part K.4):

• Regulatory/policy risk (Section 232 tariff reversibility — the single largest identified risk)
• Execution risk (Aluminum Dynamics ramp, CONTRADICTION-001 unresolved)
• Cyclical risk (steel-price mean reversion as lagging contracts normalize)
• Capital-structure risk (total debt/leverage genuinely unconfirmed this cycle)
• Valuation risk (23.16x well above historical norm; unfavorable asymmetry at current price)

"When NOT to buy" anti-pattern check (per MANUAL_en.md Part K.5):

• ✅ NOT buying on a single news headline — the thesis explicitly waits for a dated catalyst (Q3 2026 earnings, ~October)
• ✅ NOT buying on momentum alone — the valuation math (0.42× asymmetry) argues against a fresh entry independent of any price-momentum read
• ❌ Online hype: institutional/sell-side-driven commentary (9 buy/1 sell), not a retail-meme name
• ❌ Has "AI" attached: no direct AI narrative — the AI-capex linkage (fabrication backlog, assumptions.md C2) is a minor, secondary, unquantified factor, not the thesis driver
• ⚠️ "It looks cheap because of strong fundamentals": worth naming explicitly — record shipments and accelerating backlog are real and positive, but they do not by themselves justify paying 23x for earnings whose durability is explicitly the open question
• ⚠️ Brand/incumbency bias: STLD is a well-regarded, long-tenured "Big Three" name — this tree's WATCH verdict is grounded in the specific asymmetry math and the unresolved CONTRADICTION-001, not a reputational discount

Net: 0 hard anti-pattern flags, 2 soft cautions (fundamentals-justify-the-price reasoning, incumbency familiarity) — both explicitly addressed above, not ignored.

Last updated 2026-08-02. Source quality: Tier B/C (GENERATE mode) throughout; peer multiples in peers.md are Tier C (not live-verified), a real gap relative to this corpus's NEM/GS/MS precedent. K.3.6 evidence-strength suffix convention applied to all leaf and scorecard verdicts. Next refresh: Q3 2026 earnings (~October 2026) — the single highest-priority date, resolves CONTRADICTION-001.

"Don't read news; update your tree." — 90s.PM.Investing