When: Controlled pullback to $990-1,010 zone with basing behavior post-earnings

Then: Enter 50% of intended size

When: MU reaches $1,042 on push into Sept 30 catalyst

Then: Trim 15% of existing holdings — not aggressive's 0% or conservative's 20-25%

No new capital deployment pre-earnings; smaller pullback sizing if entry taken (33% not 50%); tighter stops (950 not 933); more aggressive trim (20-25% not 10-15%); defer all new positions until after Sept 30 earnings print

• 86% gross margins represent a multi-decade extreme for a memory manufacturer; prior peaks in 2017, 2018, and early 2022 each preceded 30-50% drawdowns over 6-12 months
• DSO deteriorated 4x from ~58 days to 236 days despite proportional revenue and AR growth, signaling collection stress or channel stuffing
• Stock is 2.24 points above upper Bollinger Band with declining 50-day SMA, creating an overextended rather than melt-up setup
• Forward PE of 6.63 and PEG of 0.15 are only attractive if forward estimates hold; estimates at cycle peaks are historically optimistic
• Institutional bullishness is already priced in; Goldman and BofA flips moved the stock 2% on the day, exhausting the marginal buyer
• Quarterly FCF of $17.6 billion is flattering: $11.7 billion in receivables growth consumed operating cash flow, masking true cash generation
• Risk envelope requires 1.5x ATR stop: at ATR 53.75, stop should be 947; current 933 stop is insufficient for beta 2.22 volatility profile
• September 30 earnings creates binary catalyst risk too significant for new capital deployment; 15-25% drawdown possible on in-line print

→ vs aggressive: The cycle timing argument is flawed: structural shifts don't eliminate cycles, they change amplitude and duration; 86% margins, $10.9B CIP, and 236-day DSO signal late-cycle positioning, not mid-cycle

→ vs aggressive: Pullback entry math ignores binary gap-down risk: Sept 30 earnings could take stock to 850 before any bounce, making the 'miss 12%' framing incomplete

→ vs aggressive: 25-40% upside thesis doesn't price in margin normalization: if gross margins compress to 60-65% from 86%, earnings revise lower and multiple compresses

→ vs aggressive: DSO 'scaling not stuffing' argument is weakest point: 3.7x revenue growth paired with 3.7x AR growth masks a 4x deterioration in collection days

→ vs aggressive: Institutional flow alignment is already exhausted: when Goldman, BofA, and Susquehanna all flip constructive simultaneously, the marginal buyer has been priced in

→ vs neutral: The neutral framework of wanting confirmation above 1,042 is not 'how you underperform'—it's a 1.4% validation threshold that correctly identifies overextended momentum with declining intermediate trends

HOLD with 50% sizing on breakout above 1,042 or pullback to 990-1,010, stop at 950, trim 15% at 1,042, no new capital before earnings

• HOLD base case correct; maintaining existing exposure
• New capital deployment: 50% of intended size on breakout above 1,042 with confirmation day, OR 50% on pullback to 990-1,010 showing basing behavior
• Stop loss at 950 regardless of entry type (approximately 1.5 ATR below current close)
• Trim 15% of existing holdings toward 1,042 if reached pre-earnings to lock in gains and create dry powder
• No new capital in size before September 30 earnings due to binary event risk; reassess post-print
• DSO deterioration from 58 to 236 days is a real signal requiring monitoring over next two quarters
• FCF generation is real but quality flattered by 11.7 billion change in receivables; monitor collection trends
• HBM supply dynamics genuinely different from 2017-2018 due to advanced packaging constraints, not just wafer starts
• Institutional flow convergence (Goldman, BofA, Susquehanna) supports higher floor but creates sharp reversal risk when narrative shifts
• Beta 2.22 means magnitude of adverse moves is the primary risk; stop placement at 950 gives appropriate room to work

→ vs aggressive: Confirmation bias: dismissing every bearish signal as irrelevant or proof of outdated frameworks is not analysis, it's bullish confirmation bias

→ vs aggressive: 50 SMA is declining, not rising, undermining momentum continuation argument based on historical NVDA patterns

→ vs aggressive: 933 stop too tight given gap-down risk on binary event; asymmetric payoff calculation ignores left tail scenario where stock opens at 880

→ vs aggressive: Upper Bollinger Band tags exhibit selection bias; 50% size on breakout gives participation without overcommitting

→ vs conservative: DSO deterioration 4x in a year is real signal but FCF generation is also real; balance sheet transformation is genuine

→ vs conservative: Institutional flow concentration argument would have been wrong on Tesla 2020 and NVDA 2023; momentum narratives sometimes persist for years

→ vs conservative: 33% size leaves significant money on table if stock rips 20% on beat and raise; sizing conflates two different risks

→ vs conservative: Waiting until after earnings completely sidelines through the catalyst; overcorrection in risk management

→ vs conservative: 20-25% trim at 1,042 too aggressive given stock already broken out and trading at highs; 15% trim is sufficient protection

→ vs conservative: HBM supply constrained by advanced packaging, not wafer starts, making supply dynamics genuinely different from historical cycles