Maintain HOLD; trim 25-30% of position into current strength; tighten stop loss to $141-$142; do not add above $148 until MACD turns positive and price reclaims 20 SMA; avoid adding within two weeks of Q3 earnings

• Stock has run 87% from entry, now extended with 12 sessions of negative MACD histogram widening from -0.21 to -0.98 in 11 days
• Volume of 31M shares (3x average) on down close near resistance signals distribution, not ordinary profit-taking
• 15.4x forward P/E is not cheap relative to embedded assumptions—requires successful pipeline conversion, smooth $10B acquisition integration, and Keytruda biosimilar offset by 2028
• Q3 earnings (late October) present binary catalyst risk with 0.23 beta irrelevant on earnings day; a miss to $9.30 or single-digit revenue growth guidance could compress multiple 10-15%
• Keytruda LOE in 2028 (24 months) is the true binary event not yet priced in—if Welireg, MK-1084, or Daiichi deals disappoint, multiple faces compounding compression
• Wait for confirmed close above 20 SMA on volume before adding—this is disciplined risk management, not buying tops

→ vs aggressive: Being 'aggressive about being right' in an 87% run with negative MACD and a binary catalyst 5 weeks away is not conviction—it's enthusiasm seeking permission to add to a drawdown

→ vs neutral: The 15.4x forward P/E appears cheap only in absolute terms; relative to the probability-weighted pipeline assumptions embedded in it, the multiple offers no margin of safety

HOLD. Do not trim 25-30% into strength; do not add above $148 with deteriorating momentum. Define a binary decision tree: add on pullback to $143-$145 with MACD histogram flattening on declining volume below 20M shares; trim 15-20% only if price closes below $137 on volume above 25M shares. Consider a protective collar (sell call $158-160, buy put $138-140) to hedge the Q3 earnings binary event while preserving the high-conviction thesis.

• September 18 volume print of 31M shares (3x average daily volume) on a down close at a retraced level is a legitimate distribution footprint—described as a 'yellow flag, not a red one'—but calling it 'ordinary churn' is 'intellectually lazy'; comparable volume events occurred near the August peak and early-2025 low
• 15.4x forward earnings is not cheap in context—the stock has already re-rated 87% to reach that multiple; achieving $170 target requires multiple expansion to ~17.8x, which requires earnings upgrades above $9.55 consensus rather than multiple re-rating in the current macro environment with zero Fed cuts priced
• Stop loss at $137 (50 SMA) is the correct structural level for the medium-term uptrend; tightening to $141-142 (lower Bollinger band) would result in whipsaw given ATR of $3.26 and routine intraweek noise
• Fundamentals remain solid: 75% gross margins, 34% operating margins, $15B FCF, 1.2x Net Debt/EBITDA, growing dividend at 5-6% annually, CHMP opinion just dropped, Auxier pipeline endorsement
• Do not add within two weeks of Q3 earnings given binary nature of the event with deployed multiple; deteriorating short-term momentum makes adding into overhead resistance poor risk management
• Protective collar (selling call ~$158-160, buying put ~$138-140) is recommended to monetize elevated implied volatility, define downside through earnings, and preserve upside participation
• Next real decision point is Q3 earnings in late October; current consolidation represents an 8% correction from the August high

→ vs aggressive: The September 18 volume print of 31M shares (3x average) on a down close is not 'ordinary churn'—it is a legitimate distribution footprint comparable only to events near the August peak and early-2025 low; calling it ordinary is intellectually lazy

→ vs aggressive: The 15.4x forward multiple already represents an 87% re-rating; the path to $170 requires multiple expansion to ~17.8x, which requires earnings upgrades above the $9.55 consensus, not multiple re-rating in a macro environment with zero Fed cuts priced

→ vs aggressive: Adding up to $149 into the same overhead resistance where 31M shares just distributed is too aggressive; the risk-reward at that level does not justify the timing risk given deteriorating momentum

→ vs conservative: Fundamentals are not breaking—gross margins at 75%, operating margins at 34%, FCF of $15B, Net Debt/EBITDA of 1.2, growing dividend, CHMP opinion, and Auxier endorsement; trimming 25-30% implies franchise impairment when none exists

→ vs conservative: Moving the stop from $137 (50 SMA) to $141-142 (lower Bollinger band) will result in being shaken out on routine noise within two weeks with very high probability; the 50 SMA at $137 is structural support, not a volatility level

→ vs conservative: Selling 25-30% right now is an overreaction to a 2-week MACD wobble; the stock has already corrected 8% from the high, and trimming before a confirmed breakdown is fear-driven rather than disciplined