FalconResearch

← WOLF 研究档案

WOLF · Valuation DR

深潜研究原文 · 永久存档，不回改

WOLF — Phase 1.5 Pre-cryst Valuation DR

Generated: 2026-07-16T17:25:41Z
Round: 4 | research_version: 1
Source: Claude Code CLI (dr_dispatcher, phase=phase1_5_valuation)
Independence: xbrl_context=None (ADR-0034 §4.2)

Wolfspeed Inc. (WOLF) — Pre-crystallization 独立估值报告

Part 1 — 分析叙述

独立性声明

我作为独立估值分析师,仅基于三轮 Phase 1 研究材料完成本次估值。我没有参考当前股价、市值、同业 EV/Sales 或 EV/EBITDA 倍数表、任何 XBRL 提供的 mechanical SOTP 计算。以下是我 deliberately 没有使用的 Phase 1 anchor 数据:(1) FY2025 年末第三方引用的约 $21B design-win backlog(已被管理层停止披露、无法验证 firm PO 比例,视为过期宣传口径);(2) BorgWarner LSA 提到的 $650M/年产能走廊(隐含营收 ceiling,但 take-or-pay 结构未确认,不能作为营收锚点);(3) 破产前管理层 $3B 满产年化营收目标(基于 2022-2023 ASP,已被 ASP 下行 30-40% 稀释)。这些数据我用作方向性参考,但不作为估值输入。

Step 1 — Segment 识别

三轮研究材料清晰显示 Wolfspeed 是双业务分部结构:

Segment 1 — Power Products (SiC 功率器件):2/3 营收(FY2026 Q3 $100.1M),主要终端为 EV 逆变器、工业电源、AI 数据中心 SiC 模组。生产集中在 Mohawk Valley 200mm 器件厂。当前 gross margin 深度为负(-20.6% non-GAAP)。

Segment 2 — Materials Products (SiC 衬底/外延):1/3 营收(FY2026 Q3 $50.1M),核心资产是 John Palmour Manufacturing Center (Siler City) 的 200mm SiC 晶体生长产线。外销 + 内供 Mohawk Valley 混合。当前正遭受中国衬底厂商(SICC / TanKeBlue)的价格战。

两个 segment 的估值驱动完全不同:Power 是"Mohawk Valley 利用率爬坡 × SiC 器件市场需求"故事;Materials 是"中国竞争 vs. 8英寸/300mm 技术领先"故事。必须分开估值。

Step 2 — 分部估值方法

由于当前 EBITDA / net income 都深度为负(不适用 P/E 或 EV/EBITDA),我采用 EV/Sales 倍数 对未来 mid-cycle(FY2029-FY2030)normalized revenue 应用倍数,同时在 bear/bull spread 中隐含 EBITDA 利润率的差异化路径。这是 SiC 半导体这类周期性成长股的行业惯例。

Segment 1 — Power Products 估值参数:

Segment 2 — Materials Products 估值参数:

Materials 板块的估值挑战在于其正在从垄断寡头向大宗商品供应商转变——2024 年 33.7% 全球衬底份额至 SICC 自报 2025 年 8英寸衬底 51.3% 份额之间的转折,反映了竞争格局的实质性恶化。但 WOLF 保留 8英寸 + 300mm 优先性作为长期期权。

Step 3 — 企业级汇总

Segment EV 加总:
| 情景 | Power EV | Materials EV | Total EV |
|---|---|---|---|
| Bull | $5.2B | $1.05B | $6.25B |
| Base | $2.0B | $0.55B | $2.55B |
| Bear | $0.75B | $0.225B | $0.975B |

Net debt 调整(基于 Phase 1 R3 债务结构表):
- Total debt: ~$2.0B(其中固定利率 $783M 一级票据 + 约 $1.2B 各类可转债 + PIK toggle)
- Cash & short-term investments: $1.16B
- 48D receivable(高确定性): $0.18B
- Net debt: 约 $0.66B

可转债稀释处理: 约 $914M 面值可转债在 bull 情景下大概率转股,进一步稀释现有股东。我保守处理——将其保留在债务侧作为负项,同时在 bull 情景下不再额外扣除可转债转股稀释(近似抵消)。这个处理略偏保守,可能低估 bull equity value 5-10%,但避免 double-count 稀释。

Enterprise equity value:
- Bull: $6.25B - $0.66B = $5.6B
- Base: $2.55B - $0.66B = $1.9B
- Bear: $0.975B - $0.66B = $0.3B

Bear/Bull spread 逻辑

Bull 与 Bear 之间约 18x 的极端 spread 反映 WOLF 的双向 optionality特性:

Step 4 — 关键假设(可证伪点)

我的估值依赖以下 6 个可验证假设。任何一个证伪都会让 base case 向 bear 靠拢:

  1. Mohawk Valley 利用率 5-8 个季度内爬坡至 45-50%——base case 核心依赖;如果 FY2027 Q1-Q2 欠产摊销仍 >$40M/季且营收 <$180M/季,base 应下修至靠近 bear
  2. 可转债保持"债务性质"直至 FY2028——bull 情景假设可转债最终转股(不再是 cash 义务);bear 情景假设可转债重组期到期仍为债务(增加现金压力)
  3. SiC 功率器件市场 2030 年达到 $8-10B——第三方研究(Yole / Mordor / TrendForce)共识;若市场规模低于 $6B,base 需下修
  4. WOLF 保持 5-10% SiC 器件市场份额——基于垂直整合 + 200mm 先发;若跌破 5%,base 应向 bear 靠拢
  5. AI 数据中心 SiC 需求在 FY2028-FY2029 年化贡献 $150-300M——base/bull 核心桥梁收入;可通过 hyperscaler design win 披露验证
  6. 无重大额外股权稀释(除已宣布的 $96.9M 股份 + 现有可转债转股)——若管理层触发新股发行,equity value 稀释 15-25%
  7. Renesas 保持稳定股东身份至 FY2028——38.7% 大股东若减持将造成技术性股价压力,但不改变基本面估值;若 Renesas 主动 walk away,信号性负面影响 base 20%

Deliberately 未做的事

{
  "schema_version": 1,
  "ticker": "WOLF",
  "produced_at": "2026-07-16T18:00:00Z",
  "total_bull_b": 5.6,
  "total_base_b": 1.9,
  "total_bear_b": 0.3,
  "epistemic_tag": "EST",
  "segments": [
    {
      "segment_name": "Power Products (SiC devices - MOSFET/modules)",
      "methodology": "EV/Sales on normalized FY2029/2030 mid-cycle revenue; bull/base/bear scenarios driven by Mohawk Valley utilization ramp + AI datacenter penetration + regulatory optionality",
      "bull_b": 5.2,
      "base_b": 2.0,
      "bear_b": 0.75,
      "narrative": "Bull: $1.3B revenue × 4.0x EV/Sales (Section 232 tariff protection + AI datacenter breakout + EV recovery + 60% Mohawk Valley utilization). Base: $800M revenue × 2.5x (moderate ramp to 45-50% utilization, AI datacenter $150M contribution). Bear: $500M revenue × 1.5x (utilization stuck at 30-35%, EV weakness persists, China 8-inch competition arrives by FY2028)."
    },
    {
      "segment_name": "Materials Products (SiC substrates + epi)",
      "methodology": "EV/Sales on normalized revenue; scenarios driven by China competitive intensity vs. 8-inch/300mm leadership monetization",
      "bull_b": 1.05,
      "base_b": 0.55,
      "bear_b": 0.225,
      "narrative": "Bull: $350M revenue × 3.0x (200mm external substrate sales scale, 300mm early customer engagements monetize, ASP recovers 20-30%). Base: $220M revenue × 2.5x (stability near current level, 6-inch ASP bottom formed, mild 8-inch mix uplift). Bear: $150M revenue × 1.5x (China share erosion continues, external sales collapse to internal supply role)."
    }
  ],
  "key_assumptions": [
    "Mohawk Valley utilization ramps to 45-50% within 5-8 quarters (base case); underutilization charges fall below $35M/quarter as leading indicator",
    "SiC power semiconductor market reaches $8-10B by 2030 (Yole/Mordor/TrendForce consensus); WOLF maintains 5-10% device market share",
    "AI datacenter SiC revenue reaches $150-300M annualized by FY2028-FY2029 as hyperscaler penetration develops beyond current Tier-1 ODM/PSU relationships",
    "Convertible notes (~$914M face) remain as debt in bear/base; treated as quasi-equity dilution in bull scenario (net effect: conservative equity valuation)",
    "Net debt of ~$0.66B stable — no additional equity dilution beyond announced $96.9M stock+warrant issuance; management self-funded plan holds without new capital raise",
    "China 8-inch SiC device automotive-grade AQG324 certification arrives 2028-2029, preserving WOLF's 200mm cost advantage window for ~2-3 years in the base case",
    "Section 232 Phase 2 SiC device tariff inclusion treated as bull-case upside only; base case assumes policy status quo (SiC devices remain outside 25% tariff scope)"
  ]
}