When: In-line/miss + cautious guide + AR widening

Then: Trim 20-25%

When: Daily close below $924

Then: Exit position; structural thesis has failed pre-print

HOLD (with half-size allocation, stop moved to $1,003, and 25-30% trim into strength)

• TTM operating margin of 80.4% is not a durable steady state; HBM allocations can be renegotiated, capex at cycle peak risks oversupply within 12-18 months
• Forward EPS of $159 requires revenue to continue at ~$40B+ per quarter with margins holding; consensus estimates are structurally upward-biased at cycle peaks and vulnerable to downward revision
• Memory peer Sandisk down 23% from 52-week high signals market differentiating AI-exposed names; Micron has no peer group to hide in if AI thesis wobbles
• Stock at 64.5% above 200-day MA and 86% of 52-week range with 2.22 beta and 4.5% daily ATR creates no margin of safety for additions
• Forward PE of 6.79x with PEG of 0.16 is a warning sign at cycle peak, not a buy signal; trailing PE of 24.4x shows multiple compression from exploded earnings rather than undervaluation
• Stop loss at $924 too loose for a position up ~27% from $850 entry; tighten to $1,003 VWMA dynamic support
• Trim 25-30% of position into strength before binary window; maintain meaningful exposure for breakout but reduce magnitude of potential drawdown

→ vs aggressive: 80.4% operating margin is presented as the new steady state, but DRAM/NAND still trade on cyclical dynamics and historically high-margin quarters compress rapidly when supply catches demand

→ vs aggressive: $159 forward EPS used to justify conviction is exactly the consensus number that gets cut when guidance softens; cycle peaks see downward revisions after, not before, misses

→ vs aggressive: Single-day relative strength versus broader tape does not constitute leadership when direct peer Sandisk is down 23% from its high

→ vs aggressive: MACD bullish and RSI 63 are presented as conditions to add, but 64.5% above 200-day MA and 86% of 52-week range represent an extended position with no buffer for a soft guide

→ vs aggressive: Stepping aside before earnings is framed as destroying alpha, but it is actually a one-way framing; you can hold a smaller core position, trim into strength, and redeploy after a clean print

→ vs aggressive: Forward PE of 6.79x and PEG of 0.16 are called compelling valuations, but the denominator (forward EPS growth) is consensus-biased at cycle peaks and will compress if guidance moderates

Hold half-size position, tighten stop to 1,027 (10 EMA), trim 10-15% into strength, consider diversification

• Aggressive thesis correct on momentum, fundamentals, and technical structure; conservative thesis correct on late-cycle dynamics, entry point with 27% unrealized gain, and binary window asymmetry
• 80.4% TTM operating margin is peak-cycle data, not evidence of structural re-rating—it is the dependent variable being explained, not the proof
• HBM structural premium is legitimate thesis but unproven; 80.4% margin is exactly what peak cycle would produce regardless of structural change
• Relative strength in Micron vs. Sandisk (down 23% from 52-week high) signals AI differentiation but also shows broader memory complex rolling over—peers led down in prior cycles
• 159 dollar forward EPS estimate compounds on top of peak quarter; only needs rate of growth moderation from 73% QoQ to 10-15% to trigger revisions, not sequential revenue decline
• 1,003 stop (VWMA) too tight: only 1.6 ATR below price in a name averaging 4.5% daily moves; will trigger on routine volatility
• Original 924 stop anchored to real structural level; new working stop at 1,027 (~1 ATR below price) respects volatility while protecting most of 27% gain
• Half-size in 2.22 beta name is already meaningfully defensive starting posture; 30% trim reduces to 35% of normal allocation, conceding bulk of upside
• FY2023 analog flawed: no HBM supercycle, no AI capex wave, and 16 billion debt vs. current net cash position with 26 billion liquidity
• Aggressive analyst wrong that full size equals conviction—it is exposure; conservative analyst wrong that 30% trim is discipline—it is de-risking past meaningful upside capture

→ vs aggressive: Uses 80.4% TTM operating margin as evidence for structural re-rating, but this is the dependent variable; peak margins are expected at cycle peaks regardless of structural change, making this circular reasoning

→ vs aggressive: Relative strength argument ignores Sandisk down 23% from 52-week high—a pure-play memory company; broader memory complex rolling over signals cycle exhaustion, not just AI differentiation

→ vs aggressive: 159 dollar forward EPS extrapolates from peak quarter; only mild growth rate moderation from 73% QoQ to 10-15% would trigger sell-side revisions without any revenue decline needed

→ vs conservative: 1,003 stop only 1.6 ATR below price; in a 4.5% daily average move name, this triggers on noise not thesis invalidation—transaction cost dressed as risk management

→ vs conservative: 30% trim of half-size position reduces to 35% of normal semiconductor allocation, conceding most upside to avoid drawdown that may not materialize

→ vs conservative: FY2023 analogy wrong: starting conditions fundamentally different—today net cash with 26 billion liquidity vs. 16 billion debt then, no HBM cycle or AI capex wave