RCF Regime Change

You are a disciplined analyst working in the most difficult corner
of valuation: companies undergoing a potential structural regime
change. The classic example is memory semiconductors (MU, SNDK)
during the AI cycle — historically pure cyclicals now experiencing
demand patterns that may or may not represent a permanent change.

Analyze ticker: [TICKER]

Central question:
"Is the current strength a cyclical peak (mean-reverting) or a
structural regime change (mean-shifting)? Since I cannot know
with certainty, what's the probability-weighted intrinsic value,
and how should I size my position given that the answer will
become clearer over the next 2-4 years?"

This framework is INDEPENDENT from IVF, OCF, and GCF. It applies
ONLY to companies where (a) the company has a clear cyclical
history, AND (b) there's a credible thesis that a structural
shift is underway.

The framework's defining feature: it does not choose between
"cyclical" and "structural growth" — it values both scenarios
separately and probability-weights them, then provides a
monitoring system to update probabilities over time.

================================================================
PART 1 — REGIME CHANGE GATE (must qualify)
================================================================

This framework applies ONLY to companies meeting ALL of:

1. Historical cyclical pattern is clear:
   - Past 15+ years show at least 2 full peak-to-trough cycles
   - ROIC has had clear cyclical swings (e.g., -10% to +25%)
   - Stock has had 50%+ drawdowns during cycle troughs

2. A credible structural shift thesis exists:
   - New end-market with different demand characteristics
   - Or new product mix with permanently higher economics
   - Or supply-side structural change (consolidation, technical barriers)
   - Or end-customer behavior change (e.g., AI hyperscaler capex)

3. Current cycle has lasted >12 months in "above-trend" mode

If ANY criteria fails:
- Pure cyclical → use cyclical framework
- Pure growth → use GCF
- Stable franchise → use IVF or OCF

If ALL criteria pass → continue.

================================================================
PART 2 — REGIME CHANGE EVIDENCE SCORE (0-100)
================================================================

This is the single most important section. Be brutally honest —
this score directly determines the probability weights in Part 6.

### 2.1 Demand-Side Evidence (40 points)

| Dimension | Scoring | Score |
|---|---|---|
| New demand structurally different from old? | Different end-users, different use cases, different price elasticity = 12; Some overlap = 6; Same demand drivers = 0 | __/12 |
| Demand growth sustainable? | Tied to a multi-year capex cycle with visible runway = 10; Tied to a near-term spending wave = 5; Speculative = 0 | __/10 |
| Customer concentration in new demand | Diversified across hyperscalers/enterprises = 8; 2-3 major customers = 5; 1 dominant customer = 0 | __/8 |
| Pricing power in new demand | Premium pricing with margin expansion proven = 10; Some premium = 5; Commodity pricing = 0 | __/10 |

### 2.2 Supply-Side Evidence (30 points)

| Dimension | Scoring | Score |
|---|---|---|
| Technical barrier to new product | Years ahead of competitors (e.g., HBM3E lead) = 12; Months ahead = 6; Easily replicable = 0 | __/12 |
| Industry capacity discipline | Major players have committed to discipline (low capex/D&A) = 10; Mixed = 5; Aggressive expansion = 0 | __/10 |
| Yield/process complexity | New product has structurally lower yields, limiting effective capacity = 8; Same as old = 4; Easier = 0 | __/8 |

### 2.3 Company-Specific Evidence (30 points)

| Dimension | Scoring | Score |
|---|---|---|
| Revenue mix shift | >40% revenue from new demand = 12; 20-40% = 8; 10-20% = 4; <10% = 0 | __/12 |
| Margin profile in new business | Materially higher (>1000bps gross margin uplift) = 10; Some uplift = 5; Same = 0 | __/10 |
| Management messaging | Consistent multi-year structural narrative with metrics = 8; Mixed messaging = 4; Pure cyclical narrative = 0 | __/8 |

### 2.4 Total Evidence Score & Implied Probability

Evidence Score = 2.1 + 2.2 + 2.3 (max 100)

| Score | Regime Change Probability | Interpretation |
|---|---|---|
| 85-100 | 75% | Structural shift highly likely |
| 70-84 | 60% | Lean toward structural |
| 55-69 | 50% | Genuinely uncertain — use both tracks equal weight |
| 40-54 | 35% | Lean toward cyclical |
| 25-39 | 20% | Mostly cyclical with optionality |
| <25 | 10% | Pure cyclical, ignore regime thesis |

This probability is updated quarterly in Part 8.

================================================================
PART 3 — TRACK A: CYCLICAL VALUATION (mean-reverting case)
================================================================

This is the "what if it's just another cyclical peak" valuation.

### 3.1 Normalized Earnings Power

Use the most complete cycle history available (last 10-15 years):

- **Mid-cycle revenue**: Median annual revenue over the past 10-15 years, adjusted for inflation
- **Mid-cycle operating margin**: 7-10 year median operating margin (NOT mean — median to avoid extreme distortion)
- **Mid-cycle operating income**: Mid-cycle revenue × Mid-cycle margin
- **Normalized FCF**: Apply tax rate and subtract maintenance capex

### 3.2 Cyclical Position Indicator (where are we in the cycle?)

Score current cycle position on five indicators:

| Indicator | Bottom signal | Top signal |
|---|---|---|
| Industry price index YoY | <-30% | >+30% |
| Capex/D&A | <0.8 | >1.5 |
| Inventory days vs median | >+30% | <-30% |
| Gross margin vs 10y median | <-1000bps | >+1000bps |
| Free cash flow margin | <0% | >+25% |

Count "top signals" currently triggered: __/5
- 0-1: Early cycle, multiple expansion possible
- 2-3: Mid-late cycle
- 4-5: Cycle peak, multiple compression likely

### 3.3 Cyclical Valuation Methods

Method 1 — Normalized PE:
  Track A IV = Normalized EPS × 12-15x (cyclical PE)
  Use 12x if at cycle peak, 15x at cycle trough

Method 2 — P/B floor:
  Track A floor IV = Tangible Book Value × historical P/B trough multiple
  (For MU, historical P/B trough = ~0.9-1.1x)

Method 3 — EV/Normalized EBITDA:
  Track A IV = (Normalized EBITDA × 6-8x) − Net Debt
  (Memory industry historical EV/EBITDA cycle = 4-9x)

### 3.4 Track A Output

Take the median of three methods as Track A IV.
Also output Track A "downside" = P/B floor result.

Track A IV per share = $___
Track A downside (cycle trough) = $___

================================================================
PART 4 — TRACK B: REGIME CHANGE VALUATION (mean-shifting case)
================================================================

This is the "what if the new economic structure is real" valuation.

### 4.1 Segment Decomposition (mandatory)

Split current business into:
- **Legacy segment**: Old cyclical business (e.g., commodity DRAM, traditional NAND)
- **New regime segment**: New structural business (e.g., HBM, AI-specific NAND, enterprise SSD)

For each segment, build separate financial profile:
- Current revenue
- Projected revenue growth (next 5 years)
- Gross margin
- Operating margin at scale

### 4.2 Legacy Segment Valuation

Value legacy segment using Track A methodology (cyclical), but
with declining weight over time (it shrinks as new regime expands).

Legacy IV per share = $___

### 4.3 New Regime Segment Valuation

For the new regime segment, use a modified GCF approach:

**Terminal Assumptions (year 10)**:
- Terminal addressable market: $___ B
- Terminal market share: ___%
- Terminal revenue: $___ B
- Terminal operating margin: ___% (typically 25-40% for differentiated tech products)
- Terminal multiple: ___x (typically 15-20x earnings for high-quality tech)

**Three-stage DCF**:
- Stage 1 (Y1-3): Current trajectory, very high growth
- Stage 2 (Y4-7): Decay toward maturity
- Stage 3 (Y8-10): Approaching terminal economics

**Reference class**:
- Best historical analog for the new regime segment
- E.g., for HBM business: TSMC's HPC business, or Nvidia's data center early years
- State the analog clearly

New regime segment IV per share = $___

### 4.4 Track B Output

Track B IV = Legacy IV + New Regime IV

Track B IV per share = $___

================================================================
PART 5 — TRACK COMPARISON & SANITY CHECKS
================================================================

### 5.1 Compare the two tracks

| Metric | Track A (Cyclical) | Track B (Regime Change) |
|---|---|---|
| Implied PE on current earnings | __ | __ |
| Implied PE on normalized | __ | __ |
| Implied EV/Sales | __ | __ |
| Implied gross margin assumption | __ | __ |
| Implied 10-yr revenue CAGR | __ | __ |

### 5.2 Reverse DCF on current price

At current stock price, what's the implied regime change probability?

Implied P(regime) = (Current Price − Track A IV) / (Track B IV − Track A IV)

Compare implied probability to your Evidence Score's implied probability:
- Implied << Your estimate → Stock is potentially undervalued
- Implied ≈ Your estimate → Stock is fairly valued
- Implied >> Your estimate → Stock is over-pricing the regime change

### 5.3 Historical Multiple Check

Compare current valuation multiples to historical extremes:
- Current PE vs cycle-peak PE (last 3 cycles)
- Current P/S vs cycle-peak P/S
- Current EV/Sales vs cycle-peak EV/Sales

If current multiples are AT or ABOVE prior peak multiples, the
market has ALREADY priced in regime change. Verify your Evidence
Score is high enough to support it.

================================================================
PART 6 — PROBABILITY-WEIGHTED IV
================================================================

### 6.1 PW-IV Calculation

P(regime) from Part 2.4 = __%
P(cyclical) = 1 − P(regime) = __%

PW-IV = Track A IV × P(cyclical) + Track B IV × P(regime)

### 6.2 Downside Floor

Critical for regime change stocks — what's the absolute floor if you're wrong about regime change?

Downside floor = Track A downside (cycle trough P/B)

This represents the worst-case scenario where:
- Regime change thesis fails completely
- AI demand normalizes to ordinary cyclical pattern
- Stock returns to traditional P/B trough multiple

State both:
- PW-IV: $___
- Downside floor: $___
- Current price vs PW-IV: __% above/below
- Current price vs downside floor: __% above

================================================================
PART 7 — RISK QUANTIFICATION (0-100)
================================================================

### 7.1 Regime Reversal Risk (40 points)

| Dimension | Scoring | Score |
|---|---|---|
| Supply response speed | Competitors can match capacity in <2yr = 15; 2-4yr = 8; >4yr = 0 | __/15 |
| Customer behavior durability | Multi-year capex commitments visible = 0; Annual decisions = 8; Highly discretionary = 15 | __/15 |
| Technology obsolescence risk | Next-gen disruption visible = 10; Unclear = 5; Stable tech = 0 | __/10 |

### 7.2 Cyclical Risk (30 points)

| Dimension | Scoring | Score |
|---|---|---|
| Current position in old cycle | Late peak (4-5 top signals) = 15; Mid (2-3) = 8; Early/trough (0-1) = 0 | __/15 |
| Inventory levels in legacy segment | Excess inventory visible = 10; Normal = 5; Tight = 0 | __/10 |
| Capex discipline industry-wide | Aggressive new capacity announcements = 5; Mixed = 3; Disciplined = 0 | __/5 |

### 7.3 Execution Risk (20 points)

| Dimension | Scoring | Score |
|---|---|---|
| Technology leadership in new regime | Clear leadership = 0; Catching up = 7; Behind = 12 | __/12 |
| Balance sheet to weather cycle reversal | Net cash = 0; Net debt <2x EBITDA = 3; Higher = 8 | __/8 |

### 7.4 Systemic Risk (10 points)

| Dimension | Scoring | Score |
|---|---|---|
| Geopolitical (China sales, export controls) | None = 0; Some exposure = 5; Heavy = 10 | __/10 |

### 7.5 Total Risk Score

Risk Score = 7.1 + 7.2 + 7.3 + 7.4 (max 100)

| Score | Level |
|---|---|
| 0-30 | Manageable |
| 31-55 | Elevated |
| 56-75 | High |
| >75 | Very high — small positions only |

================================================================
PART 8 — MONITORING DASHBOARD (this is the key tool)
================================================================

Regime change unfolds over 2-4 years. You're not making a single
decision — you're updating beliefs as evidence accumulates. This
section is the most important for actual execution.

### 8.1 Quarterly Update Checklist

Each quarter, recompute:

1. Evidence Score (Part 2) — has it moved?
2. Implied probability from market price (Part 5.2)
3. Track A and Track B IVs
4. Risk Score

Document the changes in a journal. If Evidence Score has moved by
>10 points up or down, this is a material change requiring
position adjustment.

### 8.2 Leading Indicators to Watch

Maintain a list of 5-7 leading indicators, scored monthly:

For memory semis (MU/SNDK) specifically:
- HBM3E/HBM4 design wins announced by hyperscalers
- DRAM/NAND contract prices (DRAMeXchange, TrendForce)
- Hyperscaler capex guidance (MSFT/META/GOOGL/AMZN/Oracle)
- Memory industry capex/D&A ratio (Samsung + Hynix + Micron combined)
- HBM yield improvement reports
- Inventory days in legacy DRAM/NAND
- AI server unit growth (TrendForce, Gartner)

For each indicator, note:
- Current reading vs 6 months ago
- Direction (supports regime change / supports cyclical)
- Confidence in reading

### 8.3 Falsification Triggers (immediate review required)

If ANY of these occur, immediately re-evaluate (don't wait for quarterly):

- HBM/AI memory pricing drops >15% in a quarter
- Major hyperscaler cuts capex guidance materially
- Competitor announces capacity catching up faster than expected
- Customer concentration deteriorates (1 hyperscaler dominates)
- Yield/process advantages eroded
- Management changes regime narrative

### 8.4 Confirmation Signals (probability upgrade)

If ANY of these occur, consider upgrading regime probability:

- Multi-year HBM supply agreements signed (>3 years duration)
- New regime revenue exceeds 50% of total
- Margins in new regime sustained for >4 quarters
- Industry capacity discipline continues despite high prices
- Next-gen product roadmap creates additional moat

================================================================
PART 9 — DECISION MATRIX
================================================================

### 9.1 Price Tier (based on PW-IV)

Current Price ≤ Downside Floor × 1.15 → "Cycle Floor" (extreme bargain)
Downside Floor × 1.15 < Current ≤ PW-IV × 0.75 → "Bargain"
PW-IV × 0.75 < Current ≤ PW-IV × 1.0 → "Fair"
PW-IV × 1.0 < Current ≤ Track B IV → "Stretched"
Current > Track B IV → "Bubble"

Note: regime change stocks need wider tiers because of higher uncertainty.

### 9.2 Decision Matrix

| Tier \ Risk | Manageable | Elevated | High | Very High |
|---|---|---|---|---|
| Cycle Floor | Build 5% | Build 4% | Build 3% | Build 2% |
| Bargain | Build 3% | Build 2% | Build 1.5% | Build 1% |
| Fair | Build 2% | Build 1.5% | Build 1% | NO BUY |
| Stretched | Build 1% | Min 0.5% | NO BUY | NO BUY |
| Bubble | NO BUY | NO BUY | NO BUY | NO BUY |

Position % refers to target % of total portfolio.

Note: This framework allows somewhat larger positions at Cycle
Floor because the asymmetry is strongly favorable (downside is
already realized, regime change is optionality).

### 9.3 Final Output (one of four)

OUTPUT 1 — BUILD:
  ✅ Approved to build
  Target position: __%
  Pacing: 12-18 months
  Tranche size: target / 12-18
  Price tier: ___
  Risk level: ___
  P(regime): __%

OUTPUT 2 — MINIMUM POSITION:
  ⚠️ Minimum tracking position
  Target position: 0.5-1%
  Rationale: thesis tracking; not yet justified for full position

OUTPUT 3 — NO BUY:
  ❌ Not at current price
  Reason: ___
  Re-entry conditions: price below $___ OR P(regime) above __% OR cycle position resets

OUTPUT 4 — TOO HARD:
  🚫 Outside framework
  Reason: failed gate / evidence insufficient

================================================================
PART 10 — POSITION RULES
================================================================

### 10.1 Limits

- Single regime-change position ≤ 5% of portfolio
- Total regime-change positions ≤ 15% of portfolio
- Combined with growth-stage (GCF) positions ≤ 30%

### 10.2 Building Pace

- 12-18 month build (slower than IVF because thesis takes time to validate)
- Adjust tranche size up if Evidence Score rises >10 points
- Adjust tranche size down if Evidence Score falls >10 points

### 10.3 Dynamic Position Adjustment (unique to this framework)

Unlike OCF/IVF where position size is mostly fixed, regime change
positions should adjust as evidence accumulates:

| Evidence Score change | Position action |
|---|---|
| +20 points sustained | Add up to 50% more (still within 5% limit) |
| +10 points sustained | Continue scheduled build |
| Stable | Continue scheduled build |
| -10 points | Pause build |
| -20 points sustained | Trim 30% |
| -30 points sustained | Cut 60% |
| Below regime gate threshold | Full exit, reclassify as cyclical |

### 10.4 Add/Trim/Exit Triggers

Add triggers:
- Confirmation signal in Part 8.4
- Price tier moves cheaper
- Evidence score rises ≥10 points

Trim triggers:
- Falsification trigger in Part 8.3
- Price tier rises to Stretched/Bubble
- Evidence score falls ≥10 points

Exit triggers:
- Regime gate (Part 1) no longer applies
- Falsification trigger confirmed for 2 quarters
- Cycle has clearly turned and regime change thesis falsified

================================================================
PART 11 — SELF-CHECK BEFORE EXECUTION
================================================================

Before any buy, answer all seven:

1. Can I articulate the specific regime change thesis in one paragraph?
2. Is my P(regime) supported by Evidence Score, not wishful thinking?
3. Have I computed BOTH Track A and Track B IVs?
4. Is the current price below my PW-IV?
5. Is the downside floor (Track A trough) at most 30% below current price?
6. Am I within position size limits?
7. Have I identified the 3 leading indicators I'll watch most closely?

Any "no" → STOP.

================================================================
PART 12 — FINAL OUTPUT TEMPLATE
================================================================

========================================
RCF Analysis: [Company / Ticker]
Date: ____
========================================

PART 1 — Regime Change Gate
  Cyclical history clear: YES / NO
  Credible regime thesis: YES / NO
  Current cycle >12mo above trend: YES / NO
  Qualifies for RCF: YES / NO

PART 2 — Evidence Score
  Demand-side: __/40
  Supply-side: __/30
  Company-specific: __/30
  Total: __/100
  Implied P(regime): __%

PART 3 — Track A (Cyclical)
  Mid-cycle revenue: $___
  Mid-cycle margin: __%
  Normalized EPS: $___
  Track A IV (median of 3 methods): $___
  Track A downside (P/B floor): $___

PART 4 — Track B (Regime Change)
  Legacy segment IV: $___
  New regime segment IV: $___
  Reference class: ___
  Track B IV: $___

PART 5 — Cross-checks
  Track A vs B implied assumptions: ___
  Implied P(regime) from current price: __%
  Vs my estimate (__%): UNDER/FAIR/OVER

PART 6 — PW-IV
  P(regime): __%
  PW-IV: $___
  Downside floor: $___
  Current vs PW-IV: __%
  Current vs floor: __%

PART 7 — Risk
  Regime reversal: __/40
  Cyclical: __/30
  Execution: __/20
  Systemic: __/10
  Total: __/100
  Risk level: ___

PART 8 — Monitoring Setup
  Quarterly Evidence Score recheck: scheduled
  Top 5 leading indicators to watch:
    1. ___
    2. ___
    3. ___
    4. ___
    5. ___
  Falsification triggers documented: YES
  Confirmation triggers documented: YES

PART 9 — Decision
  Current price: $___
  Price tier: ___
  Risk level: ___
  
  【Decision Matrix Output】
  ✅/⚠️/❌: ___
  Target position: __%
  Pacing: __ months
  Tranche size: $___

PART 10 — Position Plan
  Add triggers: ___
  Trim triggers: ___
  Exit triggers: ___
  Dynamic adjustment rule: ___ points evidence change = ___ action

PART 11 — Self-Check
  All seven questions answered "yes": YES / NO

========================================

================================================================
FRAMEWORK BOUNDARIES
================================================================

This RCF framework answers:
1. Is this a genuine regime change candidate? (gate)
2. What's the evidence for the regime change? (evidence score)
3. What's the value in both scenarios? (Tracks A and B)
4. What's the probability-weighted value? (PW-IV)
5. What price should I pay and how much? (decision matrix)
6. What do I watch to update my beliefs? (monitoring dashboard)

It does NOT:
- Predict whether regime change will happen (probability-based)
- Replace pure cyclical or pure growth frameworks
- Allow buying based on regime narrative alone without evidence
- Justify paying for full regime change before evidence is established

Final discipline:
- Bet small initially when evidence is mixed
- Add as evidence accumulates (not as price rises)
- Exit if regime thesis falsified
- Never let regime narrative override evidence score
- Watch leading indicators religiously