When: Close at or below $326 on volume

Then: Hard stop—exit all positions

Underweight; do not initiate; trim on rallies into 333-342 supply zone; reassess on confirmed close above 358

• Price below 200 SMA with lower highs since July; MACD histogram expanded from negative 0.65 to negative 1.34 in two weeks
• RSI at 35.74 with no bullish divergence confirmed; Bollinger band breakdown today
• Net debt surged from $1.87B to $13.88B year-over-year, a 7x increase inflating the 34% ROE through aggressive buybacks
• NII declined sequentially from $4.69B in Q1 2026 to $4.65B in Q2 2026; Q2 2026 EPS growth was 11% year-over-year versus 22% forward consensus
• Affirm's 36% GMV growth and Mastercard's Klook APAC push contest AXP's premium cross-border franchise, the highest-margin revenue line
• Berkshire's $576M dividend is a hold signal, not a buy signal; Greg Abel keeping AXP in the core equity book is not the same as adding to the position
• Entry at 320-325 with stop below 307 offers roughly 1:1 risk-reward given ATR of 6.32 and Feb 27 capitulation low at 307.09 already tested on heavy volume
• Confirmed close above 358 required before reassessing; 200 SMA at 336 must be reclaimed with volume to change technical thesis

→ vs aggressive: The aggressive analyst's claim that 'every bearish indicator is already priced in' is pattern recognition bias run in reverse; MACD histogram expanding from negative 0.65 to negative 1.34 in two weeks, RSI at 35.74 with no bullish divergence, and confirmed Bollinger band breakdown indicate a coherent bearish regime where the path of least resistance remains lower

→ vs aggressive: The 34% ROE cited is inflated by aggressive buybacks; true organic ROE on a stable share count is meaningfully lower as net debt went from $1.87B to $13.88B year-over-year (7x increase)

→ vs aggressive: 22% forward EPS growth is what bulls need, not what bulls have; three-year EPS CAGR is 16% and Q2 2026 came in at 11% year-over-year, with analyst consensus having a documented tendency to be revised lower into earnings

→ vs aggressive: NII is not accelerating—Q2 2026 at $4.65B was a sequential decline from Q1 2026's $4.69B; the aggressive analyst is reading a peak as a runway

→ vs aggressive: Affirm is moving aggressively into card products, targeting small business and installment lending—exactly the segments AXP has been growing; dismissing Affirm as 'BNPL, a different model' ignores quantified competitive moves in AXP's most profitable categories

→ vs aggressive: Berkshire receiving the $576M dividend is a hold signal, not a buy signal; Greg Abel keeping AXP in the core equity book is not the same as adding, and a Q3 13F trim is a real possibility given concentration discipline and the move into Alphabet

→ vs aggressive: The entry at 320-325 with stop below 307 is not asymmetric—it's roughly 1:1 risk-reward with ATR of 6.32, risking 17 points to make 18-35 points, and the hard stop is placed at a level already tested on the heaviest volume in the dataset

→ vs aggressive: 'Every bullish catalyst is still ahead' is a forecast, not a fact; the chart disagrees until price reclaims the 200 SMA at 336 with volume, MACD histogram stops expanding negative, and RSI establishes a higher low above 40

HOLD existing core positions; do not initiate new longs below 336; do not initiate shorts outright; trim modestly 10-15% on rallies toward 333-340 (not 25-33%); set re-entry trigger at 336 with contracting MACD histogram (not 358); use defined-risk options for tactical trades rather than directional equity positions

• 34% ROE is inflated by buybacks; net debt expanded from 1.87B to 13.88B in one year
• Aggressive analyst's buy-at-320-325 with stop-at-307 places stop below Feb 27 capitulation low of 307.09 on 9.8M shares (heaviest volume in dataset)
• NII trajectory overstated: Q2 2026 NII of 4.65B was sequential decline from Q1 2026's 4.69B, not acceleration
• Conservative trim plan of 25-33% into 333-342 has execution risk; rallies in downtrends are often sharp and short-lived
• Conservative re-entry trigger at 358 requires stock to move 10% above current before reassessment, missing early reversion
• Q2 was a record quarter; fundamentals have not deteriorated
• Macro backdrop is supportive: hawkish Fed protecting NII, 60% soft landing consensus, Berkshire endorsement, 5.8B annual buybacks
• ATR of 6.32 means directional equity positions carry meaningful risk; defined-risk options structures are appropriate
• Catalyst window: Q3 earnings (late October), FOMC meeting (mid-September), Berkshire 13F (mid-November)
• First reasonable add point is price reclaiming 200 SMA at 336 with contracting MACD histogram, not 358

→ vs aggressive: 34% ROE is not a trump card; net debt expansion from 1.87B to 13.88B represents deliberate leverage build to fund buybacks, and calling Q2 2026 NII of 4.65B 'acceleration' ignores that Q1 was 4.69B (sequential decline)

→ vs aggressive: Stop placement at 307 is in 'no man's land' between Feb 27 capitulation low (307.09 on 9.8M shares) and June 3 low at 299; if 307 fails on volume, position is 25 points in the hole

→ vs conservative: Re-entry trigger at 358 (10% above current) means waiting until stock has already moved 10% to re-enter; this is 'opportunity cost disguised as prudence'

→ vs conservative: Underweight or short position in this environment fights fundamental tailwinds: accelerating earnings, supportive macro (hawkish Fed, soft landing), 2% annual buyback, Berkshire holding

→ vs conservative: Pair trade (long other quality financials vs short AXP) has merit only as relative-value hedge for existing holders, not as standalone directional bet