Revenue expanded 4.5x in under 18 months—the largest cyclical ramp in Micron's modern history and unsustainable. The prior similar spike (FY22 at $30.8B) collapsed 50% to $15.5B in FY23, with gross margins going negative and net income swinging from +$8.7B to -$5.8B. The stock fell ~47% peak-to-trough.

• FY22 revenue: $30.8B
• FY23 revenue collapsed to $15.5B (50% drawdown in a single year)
• Net income swung from +$8.7B to -$5.8B
• Stock fell ~47% peak-to-trough
• FY23 gross margin went negative
• Current revenue expanded 4.5x in less than 18 months

↩ rebuts: “Structural margin expansion argument (80%+ gross margins as 'FAANG-tier')”

LTAs are a price ceiling, not a price floor. They lock volume through the contract term but not the renewal price. When LTAs come up for renewal in 2028-2030, if DRAM supply normalizes from Samsung, SK Hynix, CXMT, and Micron's own $31B CapEx all ramping simultaneously, renewals happen in a buyer's market. Contract prices may include market reset clauses. LTAs work both ways—hyperscalers can take delivery early and build inventory to renegotiate lower.

• LTAs signed at 'peak pricing' per Trefis
• Hyperscalers have incentive to take delivery early and build inventory if demand softens
• Renewals in 2028-2030 face buyer leverage if supply normalizes
• Contract language on price locks not public

↩ rebuts: “LTA moat thesis (customers paying deposits through 2030 equals locked-in prices)”

The PEG of 0.16 reflects analyst extrapolation bias at cycle peaks, not fundamental undervaluation. Forward P/E of 6.74x requires EPS to grow from $44.28 TTM to ~$159—a 3.6x increase and 20x the prior cycle peak EPS of $7.75 in FY22. If forward EPS misses by 30%, forward P/E goes to 9.6x; by 50%, to 13.5x. If the cycle rolls and EPS approaches FY23's -$5.34, the forward P/E framework breaks entirely.

• Trailing P/E: 24.21x
• Forward P/E: 6.74x
• TTM EPS: $44.28
• Implied forward EPS: ~$159
• FY22 peak EPS: $7.75
• FY23 EPS: -$5.34
• Analysts were forecasting EPS of $40 in early 2022, then revised to a loss within nine months

↩ rebuts: “PEG of 0.16 indicates deep undervaluation”

CXMT threat dismissed in one paragraph is intellectually lazy. Even if CXMT only competes in commodity DRAM and mobile (25-30% of Micron's revenue), that's enough to compress blended margins by 20% on standard DRAM pricing. The path from mobile DRAM to DDR5 and HBM is engineering iteration, not fundamental research. The Chinese government has explicitly identified semiconductors as strategic priority. Samsung rose 3.3% on CXMT news—implicit read that incumbents are losing ground.

• Samsung rose 3.3% on CXMT news
• CXMT produces 24Gb mobile DRAM (entry state, not end state)
• Government policy support is explicit and strategic
• 25-30% of Micron's revenue base in commodity DRAM and mobile
• YMTC already competing in NAND retail channels and enterprise SSD

↩ rebuts: “CXMT is a non-threat because it lacks HBM at scale and won't displace advanced node business”

Burry and Shkreli are early warning systems, not contra-indicators. Burry made his career identifying structural regime shifts before consensus. Shkreli's thesis ('if memory prices drop, it's all over') is the entire bear case in one sentence. When both go public with the same directional view in the same 60-day window, that's pattern recognition. Analyst upgrades are failing to lift the tape—Citi raised to $1,300 and the stock dropped 2%, indicating distribution into strength.

• Citi raised to $1,300; stock dropped 2%
• Stock has compounded 280% YTD
• Burry shorting AI-exposed names with stretched valuations
• Shkreli thesis: 'if memory prices drop, it's all over'

↩ rebuts: “Burry is a credit analyst not a memory-cycle expert; Shkreli flipped within two months”

The technical setup is a bear flag, not a coiling breakout. Pattern: high at $1,254.81 (June) → crashed 41% to $737.88 in 30 days → rallied 45% off the low to $1,071.88, now consolidating just below the prior breakdown level. Bear flag measured move implies targets of $700-800. The 50 SMA has been flattening between $926-$945 since late August (bearish). RSI at 62 showing bearish divergence vs. 66 peak. Bollinger upper-band test at $1,082 rejected twice—resistance not support.

• June high: $1,254.81
• July low: $737.88 (-41%)
• Rally off low: $1,071.88 (+45%)
• 50 SMA flat between $926-$945 since late August
• RSI at 62, declining from 66 peak
• Bollinger upper band at $1,082
• Bear flag measured move target: $700-800

↩ rebuts: “Consolidation is 'coiling for a breakout'; MACD crossover and 50 SMA flattening are bullish”

Annualized CapEx of ~$31B funded by peak earnings is the textbook late-cycle trap. CapEx is sticky—once committed to capacity expansion, Micron must sell that capacity at profitable prices for years. When earnings normalize from peak (80%+ gross margins) to 30-40% gross margin range, CapEx-to-CF ratio goes from 1:3 to 1:1.5 or worse. The company faces: (1) cut CapEx (signals cycle end, triggers stock collapse), (2) take on debt (compresses balance sheet), or (3) issue equity (dilutes EPS). No fourth option.

• Annualized CapEx: ~$31B
• Annualized operating cash flow: ~$100B
• Current coverage ratio: 3-to-1
• Receivables: $31B (4x YoY)
• FY23 gross margin went negative
• FY23 net income: -$5.8B

↩ rebuts: “CapEx coverage of 3-to-1 is fine”

Macro is late-cycle tightening, not Goldilocks. Recession odds at 8% is the floor, not low—recession odds rise rapidly once they start (2008 started at 5%). Fed at 'higher for longer' is deliberately restrictive; the last three recessions started with Fed funds at restrictive levels and inflation above target. AI bubble odds at 11% is complacency—bubble calls go from 10% to 50% in weeks. Taiwan at 4% is binary tail risk—if odds go to 20%, Micron gaps down 30%. Gold falling, USD rising, yields rising means real financial conditions tightening. MU has a 2.22 beta—gets hit twice as hard as the market.

• Recession odds: 8%
• AI bubble odds: 11%
• Taiwan risk: 4%
• MU beta: 2.22
• Current environment: gold falling, USD rising, yields rising

↩ rebuts: “Recession odds at 8% is 'low'; AI bubble at 11% is 'easing'; Fed policy is supportive”

Earnings is asymmetric to the downside for three reasons: (1) Bar is impossibly high—consensus baking in continued DRAM pricing strength, HBM growth, and AI capex continuation; any in-line guidance will be sold as 'cycle is peaking.' (2) Customer commentary is the tell—any hyperscaler hint at inventory build, demand softening, or renegotiation flexibility triggers 10-15% gap down. (3) Options market pricing elevated implied vol—market knows something could go wrong. When IV is elevated into a binary event with bullish consensus, asymmetry favors bears.

• Implied vol elevated into MU earnings
• Analyst upgrades failing to lift tape
• Any in-line guidance will be sold as cycle peaking

↩ rebuts: “Hold through late-September earnings print”