When: Major OCI deal, sovereign cloud announcement, Wisconsin timeline resolution, or clean Q2 FY27 earnings with capex guidance moderating

Then: Re-evaluate for potential re-entry or upgrade

When: Close >$155 on >20M volume with MACD histogram positive turn

Then: Re-evaluate for potential re-entry or upgrade

MEASURED TRIM OF 20-30% WITH ATR-RESPECTING INVALIDATION POINT BELOW $132.60, STAGED OVER 10-15 SESSIONS

• Recommended trim of 20-30% (vs aggressive 30-50%), not full exit or hold
• ATR of $6.88 should be used to set stops, not 1x ATR above entry at $138.50
• Hard invalidation point below recent swing low of $132.60 with buffer
• Staged execution over 10-15 sessions (vs aggressive 5-10), allowing tape to confirm or invalidate bear case
• Neither full bear nor full bull case resolved; position for actual uncertainty
• Current price at $138.50 reflects known risks (94% Fed pause odds, 202bp 10Y yields)
• 50 SMA resistance at $143.15 key level; break below $132.60 on rising volume confirms bear case
• Macro beta of 1.73 means stock could snap back 20% if 10Y reverses from current highs

→ vs aggressive: Stop loss at $146 is only 1x ATR above $138.50 entry—will trigger on normal volatility in a stock with $10-30 daily ranges; trim size of 30-50% sells into oversold conditions missing mean reversion

→ vs aggressive: 5-10 session execution window forces premature decision before tape confirms bear case

→ vs neutral: Hold ignores $23.7B FCF burn, 56% YoY debt service cost increase, 57% decline from high, gross margin compression from 73% to 60%, and $19.9B equity raise diluting shareholders

→ vs neutral: Read-across from Accenture +23% and IBM +5% is not clean—Oracle is $156B indebted entity funding $55B annual capex with rising debt service costs in hawkish Fed environment

UNDERWEIGHT (target 60-70% of benchmark weight; trim 25-35% of existing position over 8-12 sessions with conditional acceleration framework)

• Revenue grew from low single digits to 29.5% YoY in most recent quarter
• Operating margin expanded from 27.5% in FY23 to 33.5% in FY26
• ROE is at 41%
• FCF burning $23.7B
• Debt service up 56% YoY
• Share count diluting after $19.9B equity raise
• Gross margin compressing structurally from 72% to 60%
• Stock has made lower highs and lower lows for 12 months
• ATR is nearly $7
• Swing low printed at $132.60
• Volume on down days heavy: 56M shares September 10, 80M September 11
• Stock hasn't closed below $132.60 yet (no confirmed technical structure break)
• MACD has been negative with negative histogram
• Beta is 1.73
• 10-year yields at 2,000 (2.00%) highs
• Polymarket odds at 94% for Fed pause
• Agrees with trimming; neutral hold is too passive given capital structure deterioration
• Considers IBM as alternative with similar hybrid cloud thesis but healthier balance sheet and lower capex intensity
• Re-entry triggers: (1) FCF turning positive on trailing twelve-month basis, (2) 10-year yields rolling over below recent highs with Fed dovish pivot, (3) ORCL-specific catalyst such as sovereign cloud deal, OCI consumption growth surprise, or guidance update resolving FCF uncertainty

→ vs aggressive: Trim of 30-50% is too large; trimming 50% of position down 57% from cost basis crystallizes a 45% loss from cost at $138

→ vs aggressive: Stop placement at $146 (1x ATR) is too tight for a stock with intraday ranges of $10-30 during volatile September sessions; one bad CPI print or Fed rumor triggers stop-out

→ vs aggressive: Execution timeline of 5-10 sessions too fast; trim should not begin before confirming technical structure break

→ vs conservative: Token trim of 20-30% leaves 70-80% of position exposed to continued rate shock, sovereign cloud deal risk, or OCI renegotiations despite identifying all those risks as serious

→ vs conservative: Waiting for close below $132.60 is too reactive; by then investor gives up another 6% before stop triggers

→ vs conservative: Dismissal of macro rate risk for 1.73 beta stock is incorrect; equity market hasn't fully discounted sustained 10-year yields above 5% impact on AI infrastructure monetization thesis

→ vs conservative: Argument that there's no marginal information to act on is wrong; trend itself is continuous stream of marginal information (lower highs/lows, MACD negative, heavy distribution volume)