When: Close above $623 (200-day MA) with volume >20M

Then: Add to position with higher conviction

When: Price above $600 on volume with capex decelerating AND margin expanding above 35% AND AI monetization evidence emerging

Then: Add to position

HOLD with bias toward trimming into strength rather than adding on weakness; sell into rallies at $590-$620; stop at $510

• Price below 50-day moving average at $593 and 200-day at $624; 19% drawdown from July peak confirms trend reversal
• Q2 revenue grew $13.3B YoY but operating income declined due to costs surging faster, representing negative operating leverage
• Operating margin compressed from 43% to 30.9% (called 'trough' in report)
• Q3 2025 earnings collapsed to $1.05 per share due to tax event exposing thin margin for error
• OCF of $31.9B is up 25% year-over-year
• Q2 FCF was $1.7B against $30.1B in capex; capex consuming 94.5% of operating cash flow in a single quarter
• TTM FCF of $21.6B is roughly half the prior run rate
• Net debt rose by $51B in the past year with $29.9B in new long-term debt issued specifically to fund AI infrastructure buildout
• PEG ratio of 0.82 is backward-looking and assumes 30% EPS growth rate persists; if growth revises down to 12%, valuation becomes expensive
• Total debt stands at $112B
• Stock beta of 1.24 amplifies downside risk
• Polymarket data shows $520 strike at 30% probability (up 20 percentage points in one week) with $620 upside collapsed to 5% and $500 at 10%
• Bernstein $800 price target is a marketing document, not independent signal; sell-side targets exist on all major stocks
• California AG lawsuit involves $200B claim and is still pending; FTC antitrust appeal is being revived
• Walmart $89B selloff represents warning about retail and CPG ad spend—Meta's largest ad verticals
• AI security incident (reported exploitation of Meta AI model) creates trust issue for enterprise monetization thesis
• FTC reviving monopoly charges with regulators pushing for structural remedies, possibly forced divestiture of Instagram or WhatsApp
• Trader's expected value of $616 represents only 12% above current price using probability-weighted scenarios
• Conservative analyst assigns 5-10% probability to $800 bull case vs aggressive analyst's 25%

→ vs aggressive: Technical indicators are not 'lagging noise'—multiple aligning bearish signals (death cross, MACD negative, RSI below 40, price below both 50-day at $593 and 200-day at $624) with real-time 19% drawdown confirm a sustained trend reversal, not market noise

→ vs aggressive: Q2 fundamentals show negative operating leverage: $13.3B revenue growth but costs surged faster; management's capex-over-buybacks choice is a bet the market is not rewarding, not 'capital allocation discipline'

→ vs aggressive: FCF argument is intellectually dishonest: Q2 FCF of $1.7B against $30.1B capex means 94.5% of OCF consumed in one quarter; TTM FCF at $21.6B is half prior run rate; net debt up $51B

→ vs aggressive: PEG of 0.82 assumes 30% EPS growth persists; if growth revises to 12% due to AI capex non-monetization or ad budget compression, valuation becomes expensive, not cheap

→ vs aggressive: Bernstein $800 target is sell-side marketing, not independent signal; Corning fiber deal reflects their business model, not Meta-specific thesis; teen lawsuit dismissal ignores $200B California AG trial and revived FTC appeal

→ vs aggressive: Polymarket interpretation is backwards—$520 at 30% probability and $620 upside collapsed to 5% is the market pricing meaningful downside risk, not a contrarian bullish signal

→ vs aggressive: Risk-reward math of 2:1 to 6:1 uses optimistic assumptions; bull base raised from 35% to 40% and bull upside from 20% to 25% without justification—confirmation bias, not analysis

→ vs aggressive: Recession probability of 8% is base case, not bullish tailwind; Walmart's $89B selloff directly threatens CPG and retail ad spend (Meta's largest verticals) with 2-3 quarter lag; aggressive analyst didn't address this at all

→ vs aggressive: AI security incident (reported exploitation of Meta AI model) is a structural trust risk for enterprise monetization via Muse Code—ignored by aggressive analyst

→ vs neutral: Agrees trader HOLD is correct but would push holders to trim into rallies at $590-$620 rather than add; 50-day at $593 and 200-day at $624 are resistance levels/exits, not upside targets

HOLD with refined execution: maintain current position weight, keep stop at $510 with mental stop at $520 if volume expands on weakness, wait for Q3 earnings catalyst before adjusting exposure

• The trader's probability-weighted expected value of $616 (only 12% above current price) already leaned bullish without further probability adjustments
• When capex consumes 94.5% of operating cash flow and company issues $30 billion in new long-term debt, marginal cost of capital is rising even if balance sheet remains strong
• Polymarket data showing 30% odds of hitting $520 means 70% odds of NOT hitting it—wide tails indicate elevated risk in both directions
• 50-day moving average at $593 is a statistically likely bounce zone after 19% drop tagging lower Bollinger Band at $534, not an automatic rejection zone
• Bernstein's $800 target implying 45% upside indicates at least one institutional research desk sees the asymmetric setup
• Corning's multi-billion-dollar forward purchase agreement with Meta signals conviction in the capex trajectory—board approval and capital commitment required
• FTC structural remedies timeline is years (2027-2028), not quarters (Q3 2026), making full tail risk pricing premature
• Decision tree: If capex decelerates from $30 billion quarterly run rate, margin expands above 35%, and AI monetization materializes, add conviction above $600 on volume confirmation; if capex continues and FCF stays compressed, trim toward 50% of current weight
• The setup is genuinely bifurcated—revenue growth is exceptional, gross margin intact, and valuation compelling, but FCF is compressed, trend is broken, and macro could pressure ad budgets

→ vs aggressive: The 'waiting for confirmation means paying for confirmation' argument is how people justify catching falling knives. Optimizing for bull case while treating downside as afterthought violates asymmetric risk-adjusted return principles.

→ vs aggressive: Probability adjustment from 35% to 40% bull base and 20% to 15% continued decline was intellectually dishonest—moving probabilities to make the math work is confirmation bias, not analysis.

→ vs aggressive: Dismissal of Polymarket data as 'reactive not predictive' doesn't hold up—$520 strike probability jumping 20 percentage points in one week represents informed capital repositioning.

→ vs conservative: Calling 50-day and 200-day moving averages 'resistance to exit into' is too rigid. Markets mean-revert after 19% drops tagging lower Bollinger Band—prescribing trim-into-strength at oversold levels is premature capitulation.

→ vs conservative: Dismissal of institutional signals as 'marketing' is reductive. Bernstein's $800 target represents real work on revenue models and AI monetization pathways by sophisticated research teams.

→ vs conservative: Corning deal analysis was reductive. Multi-billion-dollar supply agreements require board approval and forward-looking judgment—buyers don't sign such deals unless they believe deployment will occur.

→ vs conservative: Pricing full FTC structural remedy tail risk (potential divestiture of Instagram or WhatsApp) into near-term decisions overweights distant risks over near-term catalysts.