When: Confirmed breakout above $518.58 with above-average volume

Then: Momentum confirmation that sequence of lower highs ($584→$540→$518) has been broken

REDUCE initial tranche to 30-35% of target; deploy remainder contingent on technical confirmation (50 SMA reclaim at 511.73 on above average volume) or cleaner pullback setup (460-470 zone with RSI confirmation); cap total position at 50-60% of standard semis weighting

• Forward P/E of 31.2 with PEG of 1.01 offers zero margin of safety; 90% delivery of analyst expectations triggers hard repricing
• Beta of 2.49 makes stock inherently asymmetric in BOTH directions; potential downside worse than 14% if tape turns
• Inventory climbed to $8.47B; demand signal unclear (supply catching up vs. shipments outpacing orders)
• FCF dropped 39% with capex of $808M against operating cash flow of $2.37B creating underutilized capacity risk if Helios ramps slower
• Stock trading below 50 SMA (511.73), 10 EMA, and Bollinger middle band simultaneously; three layers of overhead resistance
• RSI neutral at 47, MACD negative; setup is choppy bounce, not confirmed trend resumption
• No Fed cuts currently priced in means sticky inflation could trigger growth multiple compression; 2.49 beta amplifies downside in that scenario
• Nvidia deploying $500B war chest to finance customer deployments and lock in multi-year commitments, making AMD switching costs prohibitively expensive
• Recent dilution at 482 could represent permanent value destruction if stock corrects to 400 on slower Helios ramp
• Goodwill from Xilinx sits at $25.5B; impairment review in softer AI environment poses balance sheet risk
• Stop at 415 appropriate; mental stop at 434 (Bollinger lower band) should invalidate thesis
• 8% recession probability is not zero; 2.49 beta means drawdown roughly double the S&P in that scenario

→ vs aggressive: The 31.2x forward P/E with 2.49 beta is not asymmetric upside; it's zero margin for error with amplified downside. Beta cuts both ways.

→ vs aggressive: The inventory build to $8.47B and 39% FCF decline with capex of $808M suggests potential underutilized capacity if Helios ramps slower, risking write-downs.

→ vs aggressive: Three technical resistance layers (50 SMA at 511.73, 10 EMA, Bollinger middle band) are being ignored; deploying 60-65% now risks catching a falling knife.

→ vs aggressive: Paying 10-15% more for confirmation is better risk management than full deployment into a stock below multiple moving averages ahead of earnings.

→ vs aggressive: Nvidia's $500B war chest is financing customer deployments and locking in commitments, making AMD's credible challenger status not equivalent to share gains.

→ vs aggressive: The $8.8B net cash and $8.8B LTM FCF ignore $25.5B in Xilinx goodwill vulnerable to impairment in softer AI environment.

→ vs aggressive: Framing dilution as 'opportunistic' ignores that issuing shares at 482 locks in permanent value destruction if stock corrects to 400 on slower ramp.

BUY (phased entry: 50% initial, add zones at 460-470 and 434-440/breakout)

• 50% initial deployment recommended, down from aggressive 60-65% but up from conservative 30%
• Add zone 1: 460-470 range receives 20% of total position (tightened from 25%)
• Add zone 2: 434-440 or breakout above 511.73 receives 30% of position (weighted toward confirmation)
• Phillip Securities price target of 755 (56% upside) cited as fundamental anchor; 14% downside to 415 stop defines risk
• Technical setup fragile: price below 10 EMA (484.15), Bollinger middle (494.20), and 50 SMA (511.73) with MACD negative
• July 28th panic drop from 552 to 454 and subsequent failure at 518 resistance cited as confirmation risk
• Beta of 2.49 means failed reclaim from 511 could gap to mid-460s rapidly
• July 429 low unconfirmed as higher low; V-shape bounce may form double bottom like NVDA in 2022-23
• Fundamentals strong: revenue 7.7B to 11.5B YoY, gross margin 39.8% to 53.8%, operating income negative 134M to positive 1.99B
• Capex ramping from 282M to 808M in Q2 26; FCF dropped 39% QoQ; inventory at 8.47B creates dilution risk if funded at stretched multiple
• Portfolio constraint: total semis exposure capped at 50-60% of standard weighting
• Breakout trigger at 511.73 must be on above-average volume to avoid failed reclaim
• Conservative analyst's 30% initial sizing 'leaves money on the table' given positive expected value math
• Higher-for-longer macro risk acknowledged but Polymarket shows 86% no cuts priced; sizing for base case with defined stop recommended

→ vs aggressive: Dismissal of 50 SMA reclaim at 511.73 as 'retail behavior' is wrong; with 2.49 beta after 18% drawdown, technical confirmation is foundational not optional

→ vs aggressive: The 'knife already fell' argument ignores second-leg tests common in high-beta semis; 429 low unconfirmed as higher low; NVDA did this multiple times in 2022-23

→ vs aggressive: Net cash and FCF framing ignores capex ramp to 808M, 39% FCF QoQ decline, and 8.47B inventory; issuing equity at 480 to fund Helios destroys value if stock corrects to 400

→ vs aggressive: Claim that waiting for reclaim means 'paying 10-15% more' ignores the risk of failed reclaim gapping to mid-460s and wiping out 5-7% of oversized tranche

→ vs conservative: Framing AMD as 50-50 coin flip conflates whisper number beats with fundamental revenue ramp stalls; even 70-80% execution supports higher multiple than current

→ vs conservative: 30-35% initial cap misses upside participation if reclaim happens cleanly; position too small to materially contribute even if AMD hits 755

→ vs conservative: Higher-for-longer sizing ignores Polymarket 86% no cuts priced and 8% recession probability; sizing every position for tail scenarios sacrifices long-term returns

→ vs conservative: Zero margin of safety framing undersizes genuinely high-conviction asymmetric setup with 56% upside and defined 14% stop loss