When: Daily close below $578 (VWMA) on volume

Then: Move to Underweight and reassess before $533-$544 retest

When: Close below $545

Then: Structural capitulation - exit position

HOLD at reduced conviction with ATR-based position sizing, hard stop discipline, and pre-earnings trim to underweight given binary risk-reward

• 200-day SMA declining at $622-$625 creates supply wall of long-term holder exits; failed tests at $624.80 high and $613.48 close confirm downtrend structure, not coiled spring
• Q2 FCF of $1.7B derived from $31.9B operating cash flow minus $30.1B capex; TTM FCF compressed from $46B+ to ~$21B (50%+), threatening dividend and buyback sustainability
• Interest expense jumped 225% from $241M (Q2 2025) to $783M (Q2 2026); annualized exposure exceeds $3B as $30B new debt matures, representing leveraged AI infrastructure speculation
• Downside scenarios of 13-25% drawdown to $533-$544 based on $19.73 ATR and lower Bollinger band tests; asymmetric risk is holding through resistance into failed breakout
• Regulatory precedent: $17B child safety settlement signals material ongoing liability; algorithmic feed targeting threatened across EU/US jurisdictions
• Breakout base rate argument ignores declining 200-day SMA, VWMA $35 below price, and compressing operating margins—conditions that historically produce failed breakouts
• ATR of $19.73 means one-sigma daily moves exceed 3% of share price; binary Q3 earnings catalyst makes full-size position unacceptable risk-reward

→ vs aggressive: 200-day SMA dismissed as 'rearview mirror' but it is structural resistance used by institutional desks and algorithmic systems; declining 200-day strengthens its role as resistance, not weakens it, because long-term holders are sitting on losses creating supply wall

→ vs aggressive: FCF argument is 'dangerously wrong': $130B annual capex with no demonstrated near-term monetization path; Meta's ad stack cannot justify frontier model capex if Polymarket shows 0% probability of AI model leadership and Muse launch popped 1% then faded

→ vs aggressive: Net debt to EBITDA of 0.6x is misleading; 225% interest expense increase is happening before new debt fully accrues, with annualized exposure to exceed $3B with no defined exit if AI capex disappoints

→ vs aggressive: 'Coiled spring' framing is unfalsifiable and ignores asymmetric downside of 13-25% drawdown to $533-$544; trim-then-buy-back plan is rationalization, not disciplined risk management

→ vs aggressive: Australia regulatory dismissal ignores EU digital services regulation, US antitrust scrutiny, and precedent that $17B settlement signals material ongoing liability exposure threatening ad targeting moat

→ vs aggressive: Breakout base rate argument exhibits survivorship bias; historical success rates do not apply when breakout occurs below declining 200-day SMA with VWMA $35 below price and compressing margins

HOLD at 85-90% of benchmark weight with partial collar hedge before earnings; maintain mechanical breakout trigger at $625 and breakdown trigger at $578

• Position sizing should be 80-90% of full weight due to ATR of $19.73 and binary earnings catalyst within weeks
• Failed test at $624.80 closing at $613.48 represents real rejection that the aggressive analyst underweights
• TTM FCF has compressed from over $46 billion to roughly $21 billion - more than 50% drop
• Debt has doubled from $49.6 billion to $112.3 billion with interest expense rising from $241 million to $783 million annualized
• Regulatory precedent (Australia, EU digital services, US antitrust) on algorithmic feeds is underappreciated risk
• Pre-emptive trimming before earnings is market timing dressed as risk management - the answer is hedging, not position reduction
• Capex question is genuinely uncertain binary - position must be sized so either outcome is tolerable
• Neither aggressive nor conservative presents quantification for their conditional probability arguments
• Aggressive analyst's framing of $30.1 billion capex as 'operating dominance' is faith statement, not analysis
• Polymarket 0% probability on AI model leadership is a real market signal that should not be dismissed

→ vs aggressive: FCF defense is 'a faith statement, not an analysis' - TTM FCF compressed from over $46 billion to roughly $21 billion (more than 50% drop), not acknowledged in bull case

→ vs aggressive: Calling the debt picture 'fine' while it more than doubles from $49.6 billion to $112.3 billion is 'convenient framing' - annualized interest expense will reach $3 billion plus once all debt matures into run rate

→ vs aggressive: Claims 200-day SMA is 'rearview mirror' but it represents real supply pressure from long-term holder cost basis - professional capital transacts billions at these levels

→ vs conservative: 'Trim to underweight before earnings' essentially concedes the trader is wrong to maintain benchmark weight - that's discretionary timing dressed as discipline

→ vs conservative: Breakout base rate discussion lacks quantification - makes conditional probability argument without presenting data, making it 'caution' not analysis

→ vs conservative: Pre-emptive trimming is 'a recipe for systematically selling low and buying back high' - if Meta prints clean beat you could see 5-8% gap up

→ vs conservative: Aggressive analyst's Llama and Muse development and $130 billion annual capex explicitly contradict the claim that Meta doesn't need to lead AI models