When: Above $284.02 on volume >75M shares

Then: Move to full target weight

When: Pullback to 50 SMA ($247.91) on elevated volume with RSI cooling to 45-50

Then: Initiate 25% of target sizing

When: 50 SMA retest from below as new support (3+ daily closes above)

Then: Add additional 25% of target sizing

HOLD or slight trim; wait for 10 EMA reclaimed AND MACD cross back up with positive histogram AND close above $270 before adding

• TTM free cash flow of $3.2 billion funds roughly 50x annual capex ($170 billion), creating structural FCF negativity during largest infrastructure build
• $59.5 billion single-quarter debt issuance at 57% YoY total debt growth is a leverage profile destroying equity holders if assumptions break
• $486 billion net PPE becomes stranded capital if AI workloads disappoint, Anthropic fails to monetize, or hyperscaler pricing cracks 5%
• Bezos $4 billion Form 144 sale filed August 3rd at $284 stock price is informative signal, not just mechanical pre-arrangement
• Technical momentum showing distribution: MACD rolled over 4 sessions, price lost 10 EMA, 6 consecutive lower highs and lower lows after 19.8% single-day spike
• 50 SMA at $247.91 is untested support, not a floor—if it breaks, next reference is 200 SMA at $237.52 with nothing between except air
• Forward EPS $10.38 vs trailing EPS $12.43 signals Street pricing in lower earnings power; forward P/E 25.5x vs trailing 21.3x demands growth to materialize against shrinking denominator
• Conservative analyst's ATR of $8.54 means 6.4% stop at $50 SMA equals ~$15.90, less than two days of ATR and will get run by noise
• Oil up 86% YTD margin pressure not yet fully reflected in Q3/Q4 cost base—Q2 margin expansion is timing artifact
• 52-week low at $196 is 26% from current; gap fill at $235.50 is 11.5% down in days, not weeks, if momentum breaks
• Beta 1.45 amplifies any hawkish Fed surprise, oil spike, or AI demand wobble through concentrated single-name exposure

→ vs aggressive: Capex 'war chest' framing ignores 50:1 FCF-to-capex ratio and echoes telecom fiber 2000 and shale 2014 narratives that destroyed capital

→ vs aggressive: 32% upside to consensus target is misleading—sell-side targets lag market by months and get marked down AFTER stocks fall 20%

→ vs aggressive: 10% downside to 200 SMA is constructive scenario floor, not base case; 52-week low at $196 is 26% down and gap fill at $235.50 is 11.5% in days

→ vs aggressive: 6.4% stop loss at $247.91 is not 6.4% risk if level fails—it becomes a stop-run triggering much wider subsequent stops

→ vs aggressive: Survivor bias argument on Bezos selling is pattern matching through rose-colored lens; every crash was preceded by Form 144 sales too

→ vs aggressive: 55 RSI 'launch-pad territory' ignores that RSI stalling in mid-50s after post-earnings gap and fade in mega-cap names typically produces multi-week base or leg down to retest breakout

→ vs aggressive: Oil absorption argument extrapolates Q2 print where oil pressure hadn't fully hit cost base—this is timing error for Q3/Q4 peak shipping season

→ vs aggressive: 50% target sizing recommendation at $265 in $2.86 trillion name with 1.45 beta, $4 billion insider overhang, $170 billion capex commitment, and fragile technicals is concentrated event risk, not defined risk

→ vs neutral: Neutral HOLD framework is right but should be tightened—want 10 EMA reclaimed AND MACD cross up AND close above $270 before adding, and start trimming if $247.91 breaks with conviction

HOLD at benchmark weight

• HOLD at benchmark weight is the correct endpoint, not a starting point for further defensive action
• Realistic measured move is 15-20% upside, not the full 32% consensus; reward-to-risk is approximately 1.5-to-1, not 3-to-1
• Wait is an active position with positive expected value, not a passive decision representing missed opportunity
• Upside trigger tightened to confirmed close above $270 with MACD histogram turning positive for two sessions
• 50 SMA is a high-conviction add level at 25% of target sizing (smaller than aggressive's 50%), with another 25% available on successful retest from below as new support
• 62 analysts at zero sell ratings with Ackman endorsement and 8% recession odds support benchmark weight positioning
• Post-earnings gap-and-fade pattern will resolve via higher-low re-accumulation or flush to gap retest—both constructive; 50 and 200 SMAs are rising and structure is intact
• Bezos $4 billion filing near $284 creates soft bearish supply pressure over 4-8 weeks but is not a regime-changing signal
• Oil at 86% YTD is a known known and risk monitor, not a decision variable at current levels
• 50 SMA likely to be tested within 2-4 weeks; that test will provide the signal to act
• Cost of waiting is approximately 0% to -3%, while cost of aggressive add being wrong is a 6-10% drawdown
• Capex $170 billion is strategically correct in current macro regime with 8% recession odds and record equity highs, but becomes a liability if macro changes

→ vs aggressive: Assumes volatility compresses on entry and ATR of $8.54 from a name with beta 1.45 won't compress just because you decide to buy; stop at $247.91 must travel through $256 territory and will likely be tested in first 3-5 days in a normal post-earnings fade pattern

→ vs aggressive: Confuses patience with missed opportunity; 'wait' is actually a position with optionality that has positive expected value in this specific setup

→ vs conservative: Implying the 50 SMA carries zero weight is wrong when price, momentum stops, and Fibonacci clusters all agree at that level; it has high coincidence factors even without confirmation

→ vs conservative: Framing HOLD as 'capital preservation with optionality' subtly shifts the question from 'do we add?' to 'should we be at benchmark at all?'—the second question has a different answer that supports benchmark weight, not trimming

→ vs conservative: Pricing in tail risks as if they were base case; trimming from benchmark on a $3 trillion mega-cap with 62 analysts at zero sell ratings fights every non-idiosyncratic signal available

→ vs conservative: Prescription of slight trimming is overkill for a setup that is at worst neutral with positive macro tailwinds fully in place