When: Confirmed close above $63.83-63.96 (10 EMA / Bollinger middle) with RSI recovery above 50 AND MACD histogram turning positive

Then: Second half add (full Hold-graded position)

HOLD

• Three consecutive sessions of accelerating lower closes: August 14 at 65.42, August 18 at 64.79, August 19 at 62.82, August 20 at 61.78 (near session low of 61.77)
• MACD below signal line at 0.44 versus 0.81 with negative histogram of negative 0.37
• RSI at 39.86 is described as weak and near oversold but not below 30, explicitly not an automatic buy signal per the technical report
• Short-term borrowings increased five-fold from 7.6 billion at year-end 2025 to 37.3 billion at June 30, 2026
• Total debt increased from 68.4 billion to 96 billion (increase of 27.6 billion)
• Net debt increased from 21.5 billion at year-end 2025 to 29.5 billion at end of Q2
• Price below 50-day SMA (62.11), 10-day EMA (63.83), and Bollinger middle band (63.96); above 200-day SMA (55.13)
• Five layers of resistance before retesting highs: 62.10, 50-day SMA (62.11), 10-day EMA (63.83), Bollinger middle band (63.96), upper band (65.81)
• Traders stop identified at 55.13, representing 11% downside from current levels
• Number one risk to monitor per macro report: Bessent's intervention holding, failure would reignite AFS mark pain
• Data gaps flagged: no information on credit quality, regulatory capital, CRE exposure, or asset-quality trends
• Bullish case requires Polymarket 85% no Fed cuts holding, Bessent support holding, Fed holding, and consumer holding despite Walmart warning

→ vs aggressive: The 5.6% pullback is not noise; it is a confirmed three-session breakdown with accelerating closes at session lows and negative MACD momentum confirming the price action

→ vs aggressive: RSI at 39.86 is explicitly described in the technical report as not an automatic buy signal; calling it a coiled spring is hopeful narrative construction without price confirmation

→ vs aggressive: The 85% Polymarket no-cut probability does not resolve the funding risk; Bessent already stepped in once this week, indicating fragility, not confidence

→ vs aggressive: The asymmetric risk-reward math fails in actual chart context; there are five layers of resistance before the upside versus a clear 11% path to 55.13 stop on downside

→ vs aggressive: The bullish case requires confidence in credit quality, Bessent support, Fed holding, and consumer strength simultaneously—too many conditional dependencies for a high-conviction trade

HOLD

• Current prices: Aug 14 at 65.42, Aug 18 at 64.79, Aug 19 at 62.82, Aug 20 at 61.78 (closing at 61.77, a tick off session low)
• MACD at 0.44 versus signal at 0.81, with negative histogram at negative 0.37
• RSI at 39.86, above 30 but below 50
• 50-day SMA at 62.11, 200-day SMA at 55.13, price 6.65 above 200-day SMA
• Lower Bollinger Band at 62.10, upper Bollinger Band at 65.81, middle band at 63.96
• 10-day EMA at 63.83
• Q2'26 record EPS with 20% YoY net income growth
• 26% tangible book value growth over 18 months
• 8.3% loan growth
• Short-term borrowings increased 5x from 7.6 billion to 37.3 billion in one quarter
• Cash position of 66 billion
• Tier 1 add trigger: price reclaims 62.10-62.11 zone on daily close with above-average volume (half tranche)
• Tier 2 add trigger: price reclaims 63.83-63.96 zone with RSI recovering above 50 and MACD crossing back above signal (another half tranche)
• Tier 3 add trigger: price sustains close above upper Bollinger Band at 65.81 with follow-through (full position, stop moves to 61.77)
• Stop maintained at 55.13 throughout
• Three explicit monitoring items: funding-mix trajectory in Q3 earnings, credit quality deterioration or deposit flight, sustainability of Bessent's bond market intervention

→ vs aggressive: The 5.6% pullback is not 'noise' — three consecutive sessions of accelerating lower closes with confirmed technical breakdown (MACD below signal, RSI below 50, close below 50-day SMA and lower Bollinger Band) represents a tactical correction, not a rounding error. The funding-mix shift (5x increase in short-term borrowings to 37.3 billion) is a real yellow flag that cannot be waved off. The asymmetry argument ignores the path: five layers of resistance to 65.42 versus open air to the 200-day SMA at 55.13.

→ vs conservative: Saying 'we are not adding' should be conditional, not indefinite paralysis. Fundamentals have strengthened (record Q2'26 EPS, 20% YoY net income growth, 26% TBV growth). The SVB comparison is a stretch — USB has 66 billion in cash, is well-capitalized, and the funding increase appears opportunistic rather than a deposit run. Treating a yellow flag as a binary exit trigger is overfitting to one data point.