When: Pullback to $580-605 zone with bullish reversal AND MACD reset

Then: Add 0.5% of portfolio

When: Pullback to $562 (10 EMA) with confirmed higher-low AND RSI reset 50-55

Then: Add 0.5% of portfolio

When: Pullback to $510-535 (50 SMA zone) with confirmed higher-low

Then: Add 0.5-1% of portfolio, raise total sizing toward 3-4%

When: Close <$429.56 (July 29 capitulation low)

Then: Opens $338-360 reload zone at full conviction sizing

HOLD - Tighten trailing stop to $605-$610, trim 10-15% into rallies toward $680-$700, do not add until a confirmed close above $650 with volume or pullback holds above the 50-day moving average with the 10-year Treasury yield stabilizing below 4.5%.

• Revenue up 50% year-over-year, gross margins expanding 1,400 basis points, operating margins inflecting from negative to 17.3%
• AMD crossed a one trillion dollar market cap, stock up ~300% over the past year
• Trading at highest forward P/E among major chipmakers while posting lowest operating margin in peer group
• Goodwill of ~$25.5 billion on balance sheet (~30% of total assets), largely from Xilinx acquisition, vulnerable to impairment if growth narrative stumbles
• Inventory at $8.5 billion and rising
• Beta of 2.48 makes stock vulnerable in tightening macro environment
• ATR of ~$24.71 (about 4% daily move)
• OpenAI warrants: 160 million shares at $0.01 strike are dilutive upon exercise
• 10-year Treasury yield at 2007 highs creates mechanical multiple compression via higher discount rates
• Polymarket read of 96% probability of zero Fed cuts in 2026
• RSI plateauing at 72-74 for four sessions while price makes marginal new highs is momentum divergence, not coiled spring
• AMD is the challenger, not the incumbent, in AI accelerator market; Nvidia's CUDA ecosystem is a structural moat not yet cracked by AMD
• Stop at $585 is too generous; 7% drawdown is only ~1.8 standard deviations for a stock with beta of 2.48

→ vs aggressive: Operating leverage conclusion overstates margin of safety; market has priced everything in at a one trillion dollar valuation with highest forward P/E and lowest operating margin in peer group

→ vs aggressive: Nvidia at 300 and Apple at 30 comparisons are survivorship bias; AMD is the challenger, not the incumbent, and challengers get punished harder when sentiment turns

→ vs aggressive: Macro dismissal is incomplete and dangerous; higher discount rates mechanically compress multiples, mortgage rates near 7.5% will eventually bleed into consumer spending and enterprise IT budgets, and easy-money tailwind is gone with 96% probability of zero Fed cuts in 2026

→ vs aggressive: Valuation framework ignores $25.5 billion goodwill vulnerable to impairment and $8.5 billion rising inventory on balance sheet

→ vs aggressive: Risk-reward math understates downside; ATR of $24.71 means 7% drawdown to stop is only 1.8 standard deviations, not a tail event for a stock with 2.48 beta

→ vs aggressive: OpenAI warrants framing as $100 billion embedded option ignores dilution mechanics; 160 million shares at $0.01 strike will dilute existing shareholders when exercised

→ vs aggressive: Lisa Su at state dinner does not change competitive position against Nvidia, does not accelerate MI350 ramp, and does not close software gap

→ vs aggressive: Stop at $585 is too generous given volatility profile; should be tightened to $605-$610

→ vs aggressive: Should trim into $680-$700 zone, not wait for $720; locking in profits at $700 is risk management, not lack of conviction

→ vs aggressive: Do not add on weakness to $584; probability of seeing $584 in next 60 days is meaningfully higher than credited; wait for either confirmed close above $650 with volume or pullback holding above 50-day moving average with 10-year yield below 4.5%

HOLD with trailing stop at $590-$595; no add now; opportunistic add of up to 25% of position only on pullback to $580-$605 zone with trend intact; trim 20-25% only into $710-$730; no trim below $700; Q3 earnings is next hard catalyst check

• Operating leverage is real and structural: revenue up 50% YoY, gross margins expanding 1,400 bps, operating margins from negative to 17%
• PEG of 0.64 reflects growth that has actually shown in financials; server CPU TAM tripling to $211 billion by 2030 gives multiple years of runway
• BofA target of $720 anchored on structural growth thesis, not this quarter's print; bull case not exhausted at $630
• Macro backdrop is a real headwind: 10-year at 2007 highs, 30-year at 2004 highs, 96% probability of zero Fed cuts in 2026, arguing for higher discount rate on long-duration assets
• AMD has beta of 2.48 trading at 40x forward earnings; equity market marginal buyer is sensitive to discount rate environment
• RSI plateauing at 72-74 while price makes marginal new highs is a yellow flag: momentum sustained but not accelerating, easy money has been made
• AMD ATR of $24.71 with beta of 2.48 means stock can swing 4-6% weekly as routine behavior; $585 downside is a normal Tuesday, not a tail event
• BofA upgrade was just issued with credible structural thesis; trimming at $680 implies thinking BofA is wrong, which is fragile justification
• AMD is $1 trillion position with beta 2.48, partially correlated with NVDA, AVGO, MRVL, and SMH; 30-40% portfolio concentration in AI semis amplifies risk
• Q3 earnings in late October is next hard catalyst: revenue below $12.5B or Q4 guidance below $13B would be thesis-breaker; clean beat and raise would justify holding through year-end

→ vs aggressive: Treating macro backdrop as noise is dangerous for a stock with beta of 2.48 trading at 40x forward earnings; one week's price action does not establish a macro regime

→ vs aggressive: Computing expected value based on point estimates ($720 target vs $585 stop) ignores path dependency; actual distribution is much wider than 2-to-1 setup suggests

→ vs conservative: Stop at $605 sits roughly 1.0 ATR below current price, tighter than 1.5-2.0 ATR required for momentum names; high probability of getting whipsawed on routine volatility

→ vs conservative: Trimming 10-15% at $680 is too timid and too early; should trim only into upper third of target zone, not lower end

→ vs conservative: Waiting for $650 confirmed breakout or $504 pullback is effectively 'don't add' dressed up as disciplined patience; 50-day MA at $504 requires 20% drawdown in confirmed uptrend with unlikely 10-year stabilization below 4.5%