When: Pre-FOMC event risk (September 16-17)

Then: Sell $680+ calls, buy $560- puts

When: Close at $598 (defensive) or $545 (capitulation)

Then: Trim or exit per stop-loss discipline

HOLD with patience; do not add aggressively at current levels; wait for pullback to $611-$622 with stop at $598 for better risk-reward entry

• Q2 FCF declined 80% year-over-year to $1.75 billion versus $9 billion a year ago, with risk of going negative if capex continues at $25-30 billion range
• Total debt doubled to $112 billion in one year; interest expense tripled to $783 million per quarter (~$3 billion annualized)
• Buyback pause signals management prioritizes AI capex over capital returns
• Price action shows distribution pattern: opened $653.60, ran to $663.50, closed $644 (down 21 points)
• RSI rolled from 70.4 to 66.4 after touching overbought; MACD histogram expanded for 7 straight sessions before reversal
• Close pinned to upper Bollinger band—a statistically stretched condition
• JPMorgan's $820 target represents street high estimate (27% upside from $644), not the base case
• Muse requires access to inboxes, calendars, and payments—higher trust threshold than prior launches
• C-suite Form 4 open-market sells at $650 indicate informed positioning, not noise
• META rallied from $539 to $644 (20%) during restrictive Fed regime; next hawkish surprise could trigger similar drawdown
• September 16-17 FOMC is a binary event risk that aggressive analyst is dismissing
• Position already up 21%—trimming is disciplined portfolio management, not predicting the top

→ vs aggressive: AWS analogy is a category error—AWS had proven unit economics and customer wins; Meta has $30 billion quarterly capex for products with unproven commercial returns and no monetization roadmap

→ vs aggressive: Fortress balance sheet framing is misleading—net debt of $68B against $110B EBITDA ignores that total debt doubled and interest expense tripled to $783M quarterly

→ vs aggressive: The $820 JPM target is the street's high outlier, not the central case; using it as the base case is how investors get hurt

→ vs aggressive: Survivorship bias on Meta's past launches is not a forward-looking strategy; Muse's trust requirements (inboxes, calendars, payments) are fundamentally different

→ vs aggressive: Textbook reversal reading is dangerous—price action IS a distribution pattern with RSI cooling from overbought and Bollinger band compression

→ vs aggressive: Trim recommendation is not mathematically indefensible—it's basic portfolio discipline after a 21% run; aggressive analyst's 3% drawdown assumption ignores potential to 50 SMA at $600 (7-10%)

→ vs aggressive: Insider selling is not noise—C-suite Form 4 open-market sales at $650 are informed positioning from those with most business visibility

→ vs aggressive: Fed doesn't matter argument is backwards—META's rally during restrictive regime means hawkish dot-plot or tightening surprise could trigger drawdown similar to July's drop to $539

Hold core position; drop aggressive trim; do not add at stretched levels; hedge with options for event risk; add in tranches on confirmed pullbacks after FOMC and Q3 print clear; respect macro headwind on multiple expansion

• Aggressive analyst's AWS analogy is flawed—Amazon had paying enterprise customers from day one; Meta has top-three App Store ranking with no disclosed monetization path, no ARPU data, no retention curve
• Fortress balance sheet is correct: net debt of $68B against $110B in EBITDA is comfortable
• Insider selling at round-number price level like $650 after $100B one-day market-cap move and major product launch is 'informed timing', not noise
• $820 price target is single bank's high estimate, not street median—building position sizing on outlier targets is flawed
• Buying at upper Bollinger band with cooling RSI into FOMC meeting is not how to compound capital safely
• Sept 16-17 FOMC is a real binary event even with 93% odds of no cuts; risk is hawkish dot-plot surprise
• 10-15% trim into $660-$665 sacrifices too much convexity (25-30% institutional-validated upside target) for little hedging value
• Sell short-dated calls against position to monetize premium for downside protection instead of trimming
• Do not add aggressively on flush to $611-$622; add in tranches instead
• Macro headwind: 93% odds of no cuts means structurally restrictive liquidity backdrop; 8% recession probability is low but not zero; hot PPI and surging long-end yields work against multiple expansion
• Base case is grind higher with intermittent volatility, not straight-line rally to $820

→ vs aggressive: AI capex assumption is not preordained to work—distribution and monetization are two completely different problems; conflating them leads to value traps

→ vs aggressive: Using $820 as central case is problematic—that's single bank's high estimate, not street median; building on outlier targets is building on sand

→ vs aggressive: Adding at every pullback to $611-$622 is not disciplined; should add in tranches with defined stops, not full position on first flush

→ vs conservative: Trim math framing is flawed—$600 base case is worst-case scenario, not base case; trimming gives away too much convexity

→ vs conservative: Trimming core position is wrong tool for event hedging; should use options instead to preserve long-term compounding exposure