Wolfspeed

Phase A — Understand the business

Lens 1 · Company Overview

Wolfspeed is the former Cree, Inc. (same SEC registrant, CIK 895419), reborn as a pure-play silicon-carbide (SiC) compound-semiconductor manufacturer. It exited LEDs (Cree LED sold to SMART Global, 2021) and RF (sold to MACOM, RTP fab transferred July 25, 2025) to bet the entire company on SiC. The "hardware" coverage bucket undersells it: this is a capital-intensive semiconductor-fab and crystal-growth business whose product is the wide-bandgap material that lets power electronics switch at higher voltage, temperature and frequency than silicon.

Two product lines inside one reportable segment:

• Power Products — $414M FY2025 (~55%): SiC MOSFETs, diodes and modules for EV traction inverters, industrial/energy, and increasingly AI-datacenter power.
• Materials Products — $343M FY2025 (~45%): SiC bare and epitaxial substrates/wafers sold to other device makers (Renesas, Infineon, etc.). Historically Wolfspeed's crown jewel and the world's largest merchant SiC substrate supplier.

Total FY2025 revenue $757.6M. Contract structure: a mix of merchant wafer supply (some long-term, take-or-pay-style — the Renesas 10-year agreement being the extreme case) and design-win-driven device sales. No single customer exceeded 10% of consolidated revenue in FY2025, though AR is concentrated (two customers = 26% and 12% of receivables at June 29, 2025). Majority of revenue is international.

The strategic spine is a vertically integrated 200mm (8-inch) SiC footprint: the Mohawk Valley Fab (Marcy, NY) — the world's first 200mm SiC device fab, in revenue production since late FY2023 — fed by the new John Palmour Manufacturing Center (Siler City, NC), a 200mm materials factory substantially completed in late FY2025. The legacy 150mm Durham, NC device fab is being closed and Durham repurposed to materials.

Lens 2 · Supply Chain

Upstream → Wolfspeed → end market, named at each node:

• Inputs: silicon-carbide crystal growth is largely in-house (the moat) — Wolfspeed grows SiC boules, slices wafers, and grows epitaxial layers itself. External inputs: high-purity graphite, silicon/carbon source powder, specialty gases, and — critically — 200mm SiC crystal-growth and wafering equipment (a chokepoint; few tool vendors, long lead times). Electricity and skilled crystal-growth labor are real constraints.
• The company: substrate/epi (Durham + Siler City JP Center) → device fab (Mohawk Valley, NY) → modules/packaging.
• Merchant substrate customers (Materials): Renesas (10-yr agreement, now also a ~35% owner — see Lens 9), Infineon, and other device makers who buy Wolfspeed wafers to make their own SiC devices. This is the node under greatest pricing pressure (see Lens 3/13).
• Device customers (Power): EV/automotive Tier-1s and OEMs; industrial/energy; and the emerging AI-datacenter power channel (hyperscaler power-shelf / solid-state-transformer suppliers) plus GE Aerospace (10kV SiC MOU, June 2026).
• Financing "supply chain": an unusual and load-bearing node — Apollo-led lender group (senior secured), Renesas ($2.06B deposit turned equity/debt), and the U.S. Department of Commerce / CHIPS Act ($750M preliminary) are as central to the story as any silicon input.

Single-source dependency: Wolfspeed is the single source for much of its own material — vertical integration is both moat and concentration risk. The chokepoint that broke the company was not a supplier but capital: it built 200mm capacity (Mohawk + JP Center) ahead of demand and could not fund the ramp.

Lens 3 · Competitive Advantages (moats)

Real, durable moats:

• Crystal-growth process IP and 25+ years of SiC know-how — yield on 200mm boules is a hard-won, tacit-knowledge advantage. Wolfspeed remains a 200mm first-mover and demonstrated the first single-crystal 300mm SiC wafer (Jan 2026) — a genuine technology lead.
• Vertical integration — substrate-through-module control is rare; most rivals buy wafers (often from Wolfspeed).
• Only US-scaled, greenfield 200mm SiC footprint — a policy/onshoring moat (CHIPS, defense, GE Aerospace).

Weak or eroding moats:

• Merchant substrates are commoditizing. Chinese 6-inch SiC wafer spot prices have collapsed to ~$200 (from ~$400), and Chinese suppliers are scaling 8-inch and dumping 6-inch inventory. Wolfspeed's Materials line ($343M) sits directly in this price war.
• Bargaining power is inverted. A company that just wiped out its equity and burned its largest customer ($1.7B Renesas loss) does not have pricing power over customers, and its lenders (first-lien notes with minimum-liquidity covenants) hold the whip hand.

Net: the materials/crystal-growth technology moat is real; the commercial moat (pricing, bargaining) is thin. The bet is that 200mm/300mm cost leadership out-runs Chinese 6-inch commoditization.

Lens 4 · Segments

Single reportable segment; product-line and geographic splits:

Successor-period run-rate has stepped down, not up: Q3 FY2026 (three months ended March 29, 2026) revenue was $150.2M, versus $185.4M in the year-ago (Predecessor) quarter — ~-19% YoY. The only accelerating sub-vector management names is AI-datacenter applications, +~30% sequentially off a small base. Geography: majority international; no clean Successor geographic table is disclosed post-emergence. The segment story is contraction in the legacy EV/industrial base, a nascent AI-datacenter green shoot.

---

Phase B — Measure performance

Lens 5 · Earnings Result (latest print — Q3 FY2026, quarter ended March 29, 2026, Successor)

All unless noted:

• Revenue $150.2M — roughly the midpoint of guidance; -19% YoY vs $185.4M Predecessor.
• Gross loss $(40.0)M → GAAP gross margin ~-27% (cost of revenue $190.2M > revenue). Non-GAAP gross margin -21%. The company is selling SiC below cost — a screaming underutilization signal at Mohawk Valley.
• Operating loss $(114.3)M; net loss $(119.9)M; basic LPS $(3.05) on 39.3M weighted shares.
• Adjusted EBITDA $(62)M → annualized ~-$248M; adj. EBITDA margin ~-41%.
• Liquidity: $1.2B cash + short-term investments ($695.1M cash + $469.7M ST investments = $1,164.8M). Refinanced ~$476M of first-lien debt in the quarter, cutting total debt ~$97M and annual interest ~$62M.
• Guidance Q4 FY2026: revenue $140–160M, gross margin still negative.
• Balance-sheet flags: inventory down (write-offs $29.1M Successor stub); PP&E fresh-start-written-down from $3,916.5M (Predecessor) to $717.1M (Successor) — an accounting admission that the capex was value-destructive; intangibles stepped up $23.8M → $409.2M on fresh-start revaluation.

Read: a still-shrinking, still-deeply-loss-making quarter dressed up by a fixed balance sheet. The print that matters is the negative gross margin, not the debt refinancing.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts are on the research-layer shelf (transcripts=0) — this lens is. Tone has pivoted hard over the last ~4 quarters:

• Pre-petition (FY2025 calls): survival, "strengthening the balance sheet," restructuring, CHIPS, going-concern doubt.
• Post-emergence (Q1–Q3 FY2026, CEO Robert Feurle): "new era," "financially stronger," and a deliberate re-narration from EV to AI datacenters — 300mm breakthrough, 10kV MOSFET, TOLT packaging, 3.3kV modules, a Silicon Valley data-center team, GE Aerospace MOU.

What they stopped saying: aggressive EV-volume ramp targets and Mohawk utilization dates. What they started saying: "AI," "data center," "300mm," "grid." The sentiment shift is real but is narrative substitution — the growth engine that justified the capex (EV) is being quietly swapped for one (AI-datacenter SiC power) that is earlier and unproven at scale.

Lens 7 · Comps

SiC/power-semi peers. Multiples are , as-of ~July 2026; where unsourced, n/a. Wolfspeed's own earnings-based multiples are n/m (negative EBITDA and EPS).

Wolfspeed EV/Sales derivation: net debt = total debt $1,720.5M − liquidity $1,164.8M = ~$556M; EV ≈ $2.32B market cap + $0.56B = ~$2.88B; TTM revenue ≈ Q4 FY25 $197.0M + FY26 9-mo $515.5M = ~$712.5M → EV/Sales ~4.0x TTM (~4.8x on the ~$600M forward run-rate). The tell: WOLF trades at roughly STMicro's EV/Sales multiple despite negative gross margin and declining revenue — the market is paying a profitable-peer sales multiple for an unprofitable, shrinking turnaround. The five leaders (Infineon, STM, Wolfspeed, onsemi, ROHM) control >90% of a ~$5.3B (2026) SiC market growing ~10-11%/yr.

Lens 8 · Stock-Price Catalysts

Pre-emergence price history is void (old equity cancelled). Post-emergence (Successor common, ~48.3M shares), the drivers of >5% moves are:

• Restructuring milestones: plan confirmation (Sept 8, 2025), emergence (Sept 29, 2025), CFIUS clearance + Renesas equity issuance (Jan 29, 2026).
• AI-datacenter narrative: repeated June 2026 rallies on 3.3kV modules, the data-center team, GE Aerospace MOU, and the 300mm story ("WOLF stock jumps/surges").
• Small-float mechanics: ~48M-share float + heavy retail interest = momentum-driven volatility, not fundamental repricing. Third-party quotes even disagree materially (one source cited ~$63 vs another ~$36 for July 2026), a hallmark of a thin, contested post-reorg tape.

Pattern: the market reacts to narrative and technical milestones (AI, CFIUS, 300mm), not to earnings — because earnings are uniformly bad. That is the profile of a story stock, and it cuts both ways.

---

Phase C — Judge people & books

Lens 9 · Management

• CEO Robert Feurle — appointed May 2025, ex-Micron (memory-industry operator), parachuted in to run the reorg and the pivot. Predecessor Gregg Lowe (who drove the debt-funded 200mm build) was ousted Nov 2024; Thomas Werner served as executive chairman in the interim. Track record on this asset is <18 months — the turnaround is unproven under him.
• Tenure & skin in the game: essentially a new board and management post-emergence, appointed by creditors. Insider ownership is not the story — creditor ownership is (Renesas + former noteholders now hold the company; insider-transactions.csv not on shelf).
• Capital-allocation history (prior regime): catastrophic. Roughly $3.1B+ of capex in FY2024–25 alone (net $2,095.5M FY24 + $1,031.0M FY25) built capacity into a demand air-pocket; the assets were then fresh-start-impaired to $717M and the company wiped out its equity.
• Capital allocation (new regime): disciplined by necessity — capex cut to ~$68M/half, debt bought back, cost focus. Right instincts, tiny runway to prove them.
• Red flags: the Renesas $2.06B deposit functioned as disguised customer financing that detonated in the restructuring (Renesas took a ~$1.7B loss and became a forced ~35% owner) — an aggressive financing structure that blew up. Pre-petition securities-disclosure litigation is outstanding (Lens 10).
• Archetype: professional-manager turnaround team, not founders. Appropriate for a restructuring; the open question is whether they can also invent a growth story (AI/300mm) fast enough.

Lens 10 · Forensic Red Flags

Income statement / balance sheet / cash flow risks:

• Negative gross margin (-27% GAAP) — the core forensic fact. Cash cost of revenue exceeds revenue; margin is hostage to utilization, not accounting.
• Fresh-start accounting (ASC 852) — Successor numbers are discontinuous with history; the $3.9B→$0.7B PP&E write-down and the $23.8M→$409.2M intangible step-up reset depreciation and amortization, flattering future gross margin optically (lower D&A) even if cash economics don't change. Watch for non-cash tailwinds masquerading as operating improvement.
• Non-GAAP vs GAAP gap — non-GAAP gross margin -21% vs GAAP -27%; SBC and restructuring add-backs are material. Adjusted EBITDA $(62)M excludes real cash items.
• Cash-vs-earnings: operating cash burn $(126.4)M over the Successor stub; PIK interest ($21.8M) and derivative mark-to-market ($(87.8)M swing) make reported net loss noisy. The convertible-note conversion features and Renesas warrant are liability-classified derivatives — expect quarterly non-operating volatility.
• Dilution machine: 73M-share Share Reserve, Renesas warrant (4.94M @ ~$23), pre-funded warrants, and 2L converts mean basic 48.3M shares understates fully-diluted count materially (toward ~90-100M) — every valuation-per-share is a moving target.

Regulatory findings (required):

• SEC Litigation Releases / AAERs: none. Verified via SEC EDGAR EFTS (LR + AAER), 2021-07-10 → 2026-07-10 — 0 findings.
• Item 3 / Legal Proceedings (company's own disclosure): a putative securities class action, Gary Zagami v. Wolfspeed, Inc., over past public disclosures, is pending (consolidated; a June 2025 motion to transfer to the M.D.N.C. was briefed) — the standard post-collapse shareholder suit.
• Non-SEC enforcement (web): no material FTC/DOJ/FDA/CFPB action surfaced for Wolfspeed as of 2026-07-10.
• Summary: no accounting-fraud enforcement on record; the live legal exposure is ordinary securities-class-action tail risk from the pre-bankruptcy disclosures, not a going regulatory action.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (FY2027–FY2029; FYE ~late June)

Built bottom-up from the latest actuals + guidance. No forecast.ts logged (watchlist/unattended breadth run). Every line unless cited; base case: EPS stays negative across the entire horizon — this name is not valued on near-term EPS.

• Revenue: FY26 run-rate ~$600M (Q4 guide $140–160M ×4).
• Bear: EV/industrial stays weak, China commoditizes Materials → FY27 ~$580M, FY28 ~$620M.
• Base: modest AI-datacenter + auto recovery → FY27 ~$680M, FY28 ~$820M, FY29 ~$1.0B.
• Bull: Mohawk ramps + AI-datacenter SiC inflects → FY27 ~$760M, FY28 ~$1.1B, FY29 ~$1.5B.
• Gross margin (the whole ballgame): -27% today. Fixed-cost absorption means margin is a step function of volume. Base: -10% FY27 → +8% FY28 → +20% FY29. Bull: breakeven by mid-FY27. Bear: still negative through FY28.
• EPS: on ~48–64M basic shares (more fully diluted), base case ≈ $(6) to $(8) FY27, $(2) to $(4) FY28, ~breakeven-to-slightly-negative FY29. Bull case reaches GAAP breakeven FY29; bear case never does in-horizon. Interest ($120–130M/yr on ~$1.7B debt, partly PIK) and D&A (reset lower by fresh-start) are the swing lines below gross profit.
• Cash runway: ~$1.2B liquidity vs ~$390M/yr total cash burn ($126M opex + $68M capex per half, annualized) → ~2–2.5 usable years before the $250–500M minimum-liquidity covenant floor. Adequate only if gross margin inflects or new capital/CHIPS arrives before the 2030 senior notes.

The base call as a scoreable statement (not logged): WOLF FY2027 GAAP gross margin remains negative (< 0%) — p ≈ 0.60 — resolves ~2027-06-30. The valuation debate reduces to the timing of the gross-margin zero-crossing.

Lens 12 · Bull vs Bear

Bull case. A de-risked balance sheet (debt -74%, interest -60%, $1.2B liquidity, positive equity) buys time for the only US-scaled, vertically integrated 200mm SiC pure-play to inflect. Fresh-start wiped $3.2B of PP&E depreciation drag, so incremental volume at Mohawk drops through at high contribution margin — massive operating leverage off a written-down asset base. The 300mm first-mover lead and the AI-datacenter power TAM (800V HVDC, solid-state transformers, GE Aerospace, hyperscaler power) are a credible second act with secular tailwinds; +30% sequential AI-datacenter growth is the seed. CHIPS ($750M) + ~$1B tax credits are unpriced optionality. If utilization crosses breakeven, a $2.3B EV re-rates violently on a small float.

Bear case (permanent-impairment risks). (1) The core is commoditizing — merchant SiC substrates (45% of revenue) face Chinese 6-inch at ~$200 and 8-inch scaling; Wolfspeed may never earn an acceptable margin on Materials again. (2) Negative gross margin + declining revenue is the worst quadrant; the AI-datacenter pivot may be too small and too late to fill Mohawk before the runway thins. (3) Dilution + overhang — 73M reserve shares, Renesas's ~35% burned-strategic stake, and ex-creditor holders who want out cap the equity. Pre-mortem (18 months out, thesis broken): Q4-FY26/FY27 gross margin stays deeply negative, AI-datacenter revenue is still <10% of sales, cash falls toward the covenant floor, WOLF raises equity into a weak tape and dilutes the small float — the "fixed balance sheet" narrative dies on the same rock (demand, not leverage) that caused the first bankruptcy.

Are multiples too high? Yes, on any earnings basis (n/m) and rich on sales (~4x EV/Sales for a shrinking, loss-making name vs profitable peers at ~4.8x). The stock is priced for the bull inflection.

Contrarian view (what the market refuses to see): consensus treats emergence as the catharsis — the balance sheet was never the disease, it was the symptom. The disease is *structurally negative unit economics in a commoditizing material during an EV air-pocket. Chapter 11 cured the leverage and left the actual problem — under-absorbed 200mm capacity against soft demand — fully intact. The AI-datacenter story is the market's permission slip to ignore that.

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

Dismantling the bull. What structurally breaks the model: SiC's value proposition is efficiency at a price premium; Chinese oversupply is collapsing that premium at the substrate level, and EV — the volume application that justified 200mm — is decelerating. Revenue concentration: the Materials line (~$343M) is a merchant business being commoditized, and Renesas is simultaneously the anchor substrate customer and the ~35% owner that just lost $1.7B — a captive, resentful counterparty, not a growth account. Most dangerous underestimated competitor: not STMicro/Infineon (also long 200mm) but the Chinese substrate complex (e.g. TanKeBlue, SICC) that will sell 8-inch at cash cost to take share and force Western consolidation. Worst capital-allocation history: the entire $3B+ 200mm build funded with debt + the Renesas deposit into a demand air-pocket — the reason equity was wiped. What must hold for today's price: gross margin inflects positive within ~6–8 quarters and AI-datacenter SiC becomes a real, ordered revenue line — two independent bets. If growth disappoints 20–30%: at ~$580M revenue with margin still negative, the equity is a levered call on a covenant-constrained turnaround; a dilutive raise near the $250–500M liquidity floor is the base case, not the tail. Single permanent-impairment scenario: SiC substrate ASPs fall faster than Wolfspeed's 200mm cost curve descends, Mohawk never reaches breakeven utilization, and the company becomes a serially dilutive, covenant-managed also-ran — plausible, not remote.

Lens 14 · Management Questions (ordered by information value)

1. At what Mohawk Valley utilization / quarterly device revenue does GAAP gross margin cross zero, and what is the specific path and date to get there?
2. AI-datacenter revenue: what was the dollar figure last quarter (not just "+30% sequential"), and what is the FY27 order book behind it?
3. What is your fully-diluted share count including the Renesas warrant, all 2L converts, pre-funded warrants and the Share Reserve — and the realistic FY27 exit count?
4. How low can merchant SiC substrate ASPs go before you exit or reprice the Materials business against Chinese 8-inch?
5. What are the minimum-liquidity covenants on the New Senior Secured Notes, and how many quarters of current burn separate you from them?
6. Will you need to raise equity before FY2027, and under what liquidity trigger?
7. What is Renesas's intended holding period and board role, and is there a lock-up or standstill on its ~35% stake?
8. Status of the CHIPS $750M PMT and ~$1B tax credits post-administration-change — signed, contingent, or dead?
9. What is the 300mm commercial timeline — is it a real revenue product this decade or a materials-science flag-plant?
10. What is the FY2027 capex plan, and is any of it demand-contingent vs committed?
11. Which EV/automotive design-wins convert to volume in FY27, and which have slipped or cancelled since FY25?
12. What structural cost-downs (yield, 200mm boule size, epi throughput) close the gap to Chinese cash costs?
13. How do you defend Materials gross margin when your largest wafer customer (Renesas) is now also your largest shareholder?
14. What is the GE Aerospace MOU worth in committed dollars and over what timeframe?
15. If gross margin is still negative in four quarters, what is Plan B — sale, further restructuring, or business-model change?

---