When: Beat-and-raise on FY27 guidance

Then: Tactical add back; repairs forward multiple concern

TRIM exposure to 60-70% (from current level) ahead of earnings event; maintain core position for structural thesis; wait for pullback to 50 SMA zone or confirmed earnings beat-and-raise before adding back

• Forward P/E of 28x assumes analyst EPS forecasts materialize; if growth decelerates to 15% from 30%, P/E expands to 32x on lower numbers, eliminating margin of safety
• Receivables jumped 29% QoQ versus 15% revenue growth, indicating potential DSO extension and customer caution on end demand
• Inventory build from 4.0 to 4.28 billion contradicts tight upcycle narrative—small build at cycle peak is early warning sign of demand slowdown
• Technical: failed resistance breaks at 347-348, bearish intraday reversal candle, ATR contraction at top of range sets up potential volatility expansion to downside
• 10 EMA at 327 is untested support; if it fails, next level is 307 on 50 SMA representing 7.9% drop from current levels
• High-beta (1.85) name at 33x book value exposed to multiple compression when long rates rise; 10-year at 2002 highs, 30-year at multi-decade highs, mortgage rates at 7.58%
• Polymarket data shows 96% probability of zero Fed cuts in 2026, creating unfavorable cost-of-capital environment for customer fab buildouts
• Earnings guidance matters more than Q4 print itself; if FY27 guidance comes in at 15-20% versus 30% Q3 growth, stock will sell off on guide not beat
• Institutional news flow silence on $417B mega-cap ahead of earnings signals analysts keeping powder dry, not smart money loading
• Risk-reward math: 20% drawdown from $334 equals $67/share loss versus 12% gain of $40; asymmetry favors downside
• Stock at 58x trailing earnings with 235% 52-week range creates vulnerability to negative catalysts (China export controls, customer push-outs, memory capex cuts)

→ vs aggressive: Treats analyst EPS forecasts as facts rather than projections that get revised downward at cycle peaks; forward P/E of 28x is not a margin of safety if consensus cuts estimates

→ vs aggressive: Dismissed receivables spike as normal timing; extending DSO at 58x trailing earnings requires confirmation before adding exposure, not assumption of benign explanation

→ vs aggressive: Framed inventory build as healthy (growing slower than revenue); at cycle peak, inventory build signals orders landing faster than consumption—the first warning sign of a top

→ vs aggressive: Called $358 stop level 'meaningless' but it's a re-entry trigger for trimming; consolidation at resistance after rejection is uncertainty, not bullish consolidation

→ vs aggressive: Dismissed 10-year at 2002 highs as 'background noise'; high-beta, high-multiple cyclical is prototypically most exposed to multiple compression when long rates rise

→ vs aggressive: Claims AI capex supercycle is immune to macro; 2008 and 2018-2019 downcycles prove WFE cycle cuts first and deepest during recessions

→ vs aggressive: Said sentiment silence means opportunity; actually signals information vacuum where analysts keep powder dry until print—conservative move is to reduce exposure, not increase

→ vs aggressive: Asymmetry claim (8-12% upside vs 15-25% downside) is closer to reality than admitted; math clearly favors reducing exposure ahead of binary event

Trim 30-40% of position immediately; hold 50% core through earnings; reserve 20% to add on a flush toward 10 EMA zone (327) or trim more aggressively if 10 EMA fails on a daily close. Use 10 EMA at 327 as the real stop/invalidation level, not 358.

• 28x forward P/E embeds ~105% EPS growth consensus, which is a margin of error rather than margin of safety; any earnings miss would expand the multiple on lower numbers
• 29% QoQ receivables jump on 15% revenue growth is an ambiguous DSO signal that should not be interpreted as a healthy upcycle without confirmation
• Intraday reversal from 345.80 to 333.89 (~$12) at resistance is a technical warning shot; ATR compression at top of range often precedes violent expansion in high-beta names
• 358 level is a re-entry trigger, not a stop loss; 10 EMA at 327 is the true invalidation of the structural uptrend on a daily close basis
• Base case is favorable: record Q3 FY26 across revenue, gross margin, operating margin, and net income; memory upcycle is early-to-middle innings; risk of soft guide is 25-30%, not 50%
• 60-70% trim is capitulation given stock remains above 10 EMA and 50 SMA, structural trend intact, and company is 65% ROE net-cash mid-upcycle franchise
• Absence of institutional news flow, retail chatter, and catalysts three weeks from a binary event is degraded information environment, not a bullish or bearish signal
• If stock closes above 358 with conviction, retest of prior highs becomes likely and trimmed portion should be added back

→ vs aggressive: The 28x forward P/E is a market-implied expectation, not objective safety; the multiple compression already happened and earnings must now meet the new bar to justify it

→ vs aggressive: The 29% QoQ receivables jump on 15% revenue growth is ambiguous working capital signal; calling it shipment timing and moving on glosses over a real DSO extension risk

→ vs aggressive: Framing the 12-dollar intraday reversal at resistance as bullish consolidation misreads the tape; ATR contraction at range highs often precedes violent expansion in 1.85 beta names

→ vs conservative: 60-70% trim is capitulation, not risk management; stock still above 10 EMA, 50 SMA rising, 200 SMA at 276 well below, and structural trend intact

→ vs conservative: Earnings risk is 25-30%, not 50%; record Q3 results, seasonal Q4 strength, and early-to-middle innings memory upcycle make meaningful guide shortfall unlikely

→ vs conservative: Sentiment silence is neither bullish nor bearish signal; it is degraded information environment requiring position reduction, not elimination