When: Q1 FY27 earnings print showing FCF stabilization alongside maintained revenue growth

Then: Upgrade to Overweight

When: Confirmed Maia 300 benchmarks demonstrating competitive perf-per-watt

Then: Upgrade to Overweight

When: Daily close below 10 EMA at ~$483 on heavy volume signaling trend exhaustion

Then: Downgrade to Underweight

When: FY27 guidance baking in continued FCF compression below $60B

Then: Downgrade to Underweight

When: Azure growth deceleration below 18% signaling AI capex ROI disappointment

Then: Downgrade to Underweight

When: Daily close below 50 SMA at ~$411

Then: Break structural thesis entirely

HOLD

• Revenue grew 8.7% sequentially with 34% ROE, but the 42% two-month run means this is already priced in
• RSI 71 showing bearish divergence (was 79 ten days ago), MACD histogram contracting from 12 to 5 indicating momentum decay
• ATR at 14.40 implies daily swings of $15-17 in both directions
• Accounts receivable jumped $20.9B sequentially (+35%) in Q2 — flagged as a watch item for potential DSO elongation
• Capex more than doubled YoY from $17B to $35B per quarter, compressing FCF and raising execution risk
• Forward P/E 21x and PEG 1.65x leaves limited margin for error if 31% EPS growth doesn't materialize
• OpenAI $1T IPO catalyst is speculative with execution uncertainty; SoftBank faces $40B loan deadline in 2027
• Crowding risk: institutional consensus on bullish side often signals easy money already made
• Historical drawdown of 36% in eight months (Oct 2025-Jun 2026) is recent context not to be ignored
• 86% YTD oil rise and 86% probability of no Fed cuts cap multiple expansion from here

→ vs aggressive: The argument that 'RSI can stay overbought for weeks' is selectively applied — RSI falling from 79 to 71 while price moves sideways is textbook bearish divergence, not healthy consolidation

→ vs aggressive: Smart money accumulation narrative conflates Ackman (one manager with one style) and speculation on Berkshire 13F filings with actual confirmed institutional backing — crowding is a risk, not a safety net

→ vs aggressive: Waiting for $483 is not 'missed opportunity' — patient capital beats aggressive capital because a pullback of 10-15% crystallizes real losses if you chase at 52-week highs

→ vs aggressive: The OpenAI $1T IPO is not a clean catalyst — SoftBank's $40B loan deadline creates tension that could resolve in multiple ways including messy outcomes

→ vs aggressive: The 21x forward P/E is only reasonable if 31% EPS growth materializes, but doubled capex ($17B to $35B/quarter) raises the bar for AI monetization success

HOLD core position; stage additions in $470-490 zone with RSI 55-60 or volume-confirmed 50 SMA test near $411; rebalance to target allocation if meaningfully overweight; do not add at $497

• The institutional accumulation story (JPM target hikes, Ackman reaffirmation, Berkshire speculation, IREN deployment) is real and anchored in tangible data
• MACD histogram contraction from 12 to 5 during sideways price consolidation at highs resembles late-stage bull run patterns rather than healthy continuation
• 21x forward P/E with 31% expected EPS growth is reasonable for AI-exposed cloud franchises with these unit economics; PEG of 1.65 is not a sell signal in a market where S&P 500 average PEG exceeds 2
• Closing 15 branch offices in China is operationally insignificant for a company doing $90 billion per quarter
• $1 trillion OpenAI valuation is speculative journalism, not an imminent event; MSFT stake crystallization would follow multi-year equity accounting process
• Do not make automatic trim mandatory — rebalancing toward target allocation is situational, not mechanical
• $483 entry should be framed as a $470-490 ZONE, not a precise entry point; accept opportunity cost if stock powers through $506
• Use elevated ATR (2-3% daily swings now normal) to size positions appropriately

→ vs aggressive: Calling MACD histogram contraction from 12 to 5 'healthy consolidation' is confirmation bias, not analysis — this pattern resembles late-stage bull runs, not continuation patterns

→ vs aggressive: The $1 trillion OpenAI IPO is seekingalpha.com speculation, not a crystallizable event; timing and structure of any MSFT stake realization are completely unknown

→ vs conservative: 21x forward P/E with 31% EPS growth is not rich — comparable AI-exposed cloud peers trade higher; PEG of 1.65 is reasonable, not a sell signal in a market where S&P 500 average PEG exceeds 2

→ vs conservative: Closing 15 branch offices in China is operationally insignificant for a company doing $90 billion per quarter; mixing the China narrative into the bear case dilutes stronger arguments

→ vs conservative: Receivables spike deserves more scrutiny than initially given — $20.9 billion sequential jump is unusual and warrants attention at next earnings call, though the matching deferred liability increase is somewhat reassuring