REDUCE to approximately 50% benchmark weight. Wait for 50 SMA reclaim, MACD crossover confirmation, and clarity on recession tail risk or credit cost outlook before increasing exposure.

• Earnings yield of 5% offers essentially zero equity risk premium versus 5% risk-free rate (10-year Treasury)
• FY2025 net income guidance flat year-over-year at $10.83 billion versus $11.31 billion TTM
• 18 percentage point gulf between actual earnings trajectory (4.4% sequential pickup) and consensus forecast (22% growth)
• Price at 327.78 is below 200 SMA at 335.27 and 50 SMA; confirmed downtrend with negative MACD momentum
• VWMA converging with price indicates selling pressure exhausting, not a buy signal; MACD histogram still negative and expanding
• Polymarket data shows 94% probability of zero Fed cuts in 2026, fully pricing the rate tailwind
• Recession probability at 10% and rising; asymmetric risk not yet priced
• A 1% increase in net charge-offs on $220 billion loan book equals $2.2 billion pre-tax earnings impact
• NII grew from $9.9 billion to over $17 billion in three years, but this tailwind is now fully priced
• AXP fell from $363 to $320 (11.7% drawdown) while Berkshire Hathaway remained a holder
• Forward PEG of 1.3 based on 22% consensus growth extrapolation that remains contestable
• Conservative analyst does not advocate capitulation; 50% benchmark weight preserves franchise-quality optionality

→ vs aggressive: The equity risk premium argument misreads the concept; ERP is compensation for equity-specific risk, not a function of normalized growth; AXP's 5% earnings yield offers essentially zero premium for absorbing consumer credit cycle exposure on a $220 billion loan book, recession tail risk, premium card margin compression, and contested forward growth.

→ vs aggressive: FY2025 flat net income of $10.83 billion is not a victory lap; TTM is $11.31 billion with only 4.4% sequential pickup, creating an 18-percentage-point gulf versus 22% consensus forecast that requires everything to go right in an environment where multiple things are actively going wrong.

→ vs aggressive: VWMA convergence is a precondition for a base, not a base itself; it is not a buy signal but barely a neutral signal; MACD histogram is still negative and expanding with price below both 50 and 200 SMA; bottom-fishing before confirmation is indistinguishable from being early and wrong.

→ vs aggressive: The macro argument smuggles in a wrong conclusion; 94% probability of zero Fed cuts is already priced, not new information; the 10% recession probability is the asymmetric risk not priced; if it rises to 20-25%, credit losses on the $220 billion loan book would devour the NII tailwind.

→ vs aggressive: Berkshire Hathaway owning AXP does not make it immune to multiple compression or drawdowns; AXP fell 11.7% while Berkshire was still a holder; institutional ownership slows decline and provides marginal bid but does not prevent price declines.

→ vs aggressive: The American Banker piece on competitors' rising costs and limited consumer pool compresses category economics; the best position in a compressing category is still a position in a compressing category.

Hold at or modestly below benchmark weight (75-80%); set re-evaluation triggers at $339.50 (add back to benchmark if closed above on volume) or $319 (trim to 50% if broken below on volume); do not short; do not add aggressively

• TTM net income of $11.31 billion; 4.4% pickup cited by aggressive analyst represents marginal sequential improvement off depressed base, not annualized growth
• Roughly 18 percentage point gap between actual trajectory and consensus expectations for forward EPS
• Technical basing signals (VWMA convergence, RSI neutral, higher low) are early signs of stabilization but unconfirmed
• Price below both 50 and 200 SMA with MACD negative and expanding
• NII tailwind operates at full strength in near term given 10-year yields at 5% with no cuts priced in
• Credit cost pressure lags 12-18 months after unemployment rises, favoring aggressive analyst near-term but conservative analyst over 12-24 months
• Berkshire provides structural floor but did not prevent 11.7% drawdown from highs
• American Banker premium card headwind is structural, slow-burn risk to underweight not exit over
• Current sell-at-337 with stop-at-345 trade is poorly calibrated with 8 dollar risk window and risks stop-out at precisely the level where trend would turn constructive

→ vs aggressive: Equity risk premium argument not dismantled; restating bullish case louder is not rebuttal; TTM net income of $11.31 billion with 4.4% pickup is marginal sequential improvement off depressed base, not acceleration

→ vs conservative: Selling at 337 with stop at 345 is poorly calibrated short bet; if AXP reclaims 50 SMA at 339 and runs toward prior highs, forces stop loss at exactly the level where trend would turn constructive