When: Daily close below $820

Then: Full exit of position

SELL with 15-20% trim (not 25-40%), collar hedge (sell December 1,020 calls, buy 850 puts), pre-committed re-entry at 960 with RSI above 45 and bullish MACD divergence, exit below 876 closing basis

• Current price 911.93, down 22% from late June high, with confirmed downtrend across all timeframes
• HOLD position equates to hidden long with full downside exposure; 876 stop zone implies 4% mechanical loss before merit arguments
• If 876 fails, measured move targets: 850, then 800, then retest of 750 prior demand zone
• 39.9 billion in deferred advances are contracted backlog from utilities and hyperscalers, not accounting optics—real multi-year annuity stream
• Q1'26 operating margin of 1.9% was distorted by acquisition with doubled goodwill to 14.4 billion; other quarters show improving trajectory: Q2'25 4.2%, Q4'25 5.5%, Q2'26 5.9%
• Naked puts at 850 strike too expensive given implied vol north of 50 and ATR of 47; collar structure more capital-efficient
• Initial trim should be 15-20%, not 25-40%, to preserve optionality and avoid signaling conviction into thin tape
• MACD stretched at negative 23; 200 SMA at 876 is 50/50 bounce setup
• Pre-committed re-entry trigger: reclaim 960 on rising volume with RSI above 45 and bullish MACD divergence
• Pre-committed exit trigger: below 876 on closing basis triggers next incremental trim or full exit

→ vs neutral: HOLD is not a neutral stance but a hidden long position with full downside exposure at 911.93, with 876 stop zone below implying 4% automatic loss, and 750 retest risk if 876 fails; framing inaction as prudence in a confirmed breakdown is backwards

→ vs aggressive: 25-40% trim is oversized; 15-20% preserves more optionality and avoids signaling conviction into a thin tape

→ vs aggressive: Naked long puts at 850 strike are too expensive given implied vol north of 50 and ATR of 47; premium decay will erode recovery gains; a collar structure (sell December 1,020 calls to fund 850 puts) is more capital-efficient

→ vs aggressive: Customer advances/contracted backlog of 39.9 billion dismissed as accounting optics cannot simultaneously justify sell thesis and be ignored when assessing cash flow and downside floor

→ vs aggressive: Q1'26 1.9% operating margin cited as proof of structural failure ignores acquisition integration costs (goodwill doubled to 14.4 billion) and Q1 seasonality; other quarters show improving margin trajectory

Trim 25-30% of position in two tranches (half at 935-955 zone, half at 975-993 zone); hedge residual with put spread (buy 870 puts, sell 820 puts, 3-4 month expiration); re-enter half on daily close above 960 with above-average volume, second half on higher-low above 970 with 10 EMA sloping up; exit 50% more below 876, exit fully below 820.

• SELL is correct; HOLD is indefensible with lower band being tested and no reversal signals
• Forward P/E of 36x is rich; Q1'26 operating margin of 1.9 percent is likely distorted by acquisition that doubled goodwill to 14.4 billion
• 39.9 billion in customer advances represents real contracted backlog from credit-worthy counterparties
• Aggressive analyst's 60-80x normalized OCF is selectively pessimistic; adjusted FCF is roughly 30-35x
• Conservative analyst's 15-20% trim is too small; doesn't meaningfully protect against 200 SMA break scenario
• Conservative analyst's collar (sell 1020 calls/fund 850 puts) caps upside at 12% when recovery probability is 25-30%
• Put spread costs roughly one-third of naked put premium, defines maximum risk at 820 strike, preserves upside
• Re-entry should use two-stage triggers (960 close + volume, then 970 higher-low) rather than Conservative's stacked three-filter approach
• This is risk re-allocation, not a directional bet; goal is 60-75% of benchmark weight involvement in AI power thesis

→ vs aggressive: FCF normalization strips working capital inflows but leaves outflows in place; customer advances are structural, not one-time benefit

→ vs aggressive: Catastrophizing to 750 is stress test written as forecast; ignores 200 SMA at 876 support and stretched MACD mean reversion

→ vs aggressive: Naked 850 puts cost 4-6% of position value with ATR of 47 and IV above 50; not free insurance, it's a second short position

→ vs conservative: 15-20% trim is half-measures given confirmed breakdown signals; reduces beta by only ~10% on portfolio weight basis

→ vs conservative: Collar structure sells 1020 calls capping upside at 12% when rally probability is 25-30%; put spread is more capital-efficient