When: Close below $1,000 on expanding volume

Then: Move toward Underweight

When: Q3 2026 earnings below $18 with deteriorating backlog commentary

Then: Move toward Underweight

When: Pullback to $999-$1,010 with volume capitulation

Then: Initiate new longs

When: Clean close above 50 SMA at $1,040 on >1.8M volume with successful retest

Then: Initiate new longs

When: Q3 2026 earnings confirming forward EPS trajectory with stable comp discipline

Then: Initiate new longs

Hold core position; deploy new capital to cash/short-duration Treasuries; collect 1.93% dividend yield; wait for one of three confirmations before adding: (1) successful retest of 200 SMA at 941 on heavy volume, (2) clean close above 1,058 on above-average volume followed by successful retest, or (3) Q3 earnings confirming forward EPS trajectory with stable comp ratios.

• Payout ratio at 117% leaves no cushion for revenue misses; consensus EPS of 74.03 is a Wall Street estimate, not a guarantee
• Q2 revenue declined 22% sequentially from Q1 peak; Q2 2026 of ~13.5B represents year-over-year decline vs Q2 2025 at 14.6B
• Stock stuck in 1,020-1,055 range for six weeks with roughly average volume on rejections; not accumulation, equilibrium
• RSI oscillating between 40-55 for weeks; definition of stock without a trend, not a reset ready to launch
• GS up 38.6% over past year and 9.2% above 200-day moving average; Anthropic upside already partially priced in
• P/B of 2.84x justified by 17% ROE vs peers at 10-12%, but ROE inflated by buybacks; 30-40% of recent EPS growth from share count reduction
• Comp ratio set annually based on full-year revenue; using Q1 peak revenue (17.2B) to argue comp discipline is backwards
• Downside trigger (hot CPI) is a known event this week; Anthropic IPO is a hopeful event six weeks out
• Lower Bollinger Band at 1,000 already broken intraday on September 1
• Aggressive tactical entry below 1,015 with stop at 940 implies ~7.5% downside risk vs 10-12% upside; probability-weighted outcome favors waiting

→ vs aggressive: Consensus EPS of 74.03 is a Wall Street estimate, not a guarantee; Q2 revenue declined 22% sequentially and payout ratio could reach 130-140% if earnings disappoint

→ vs aggressive: Five Bollinger Band rejections at 1,058 show roughly average volume with no higher lows; this is equilibrium, not a coiled spring or accumulation

→ vs aggressive: RSI at 49 is neutral after six weeks of oscillation between 40-55; not the same as neutral RSI after V-shaped recovery

→ vs aggressive: GS already up 38.6% over past year and 9.2% above 200-day MA; IPO mandates can fall through and past left-lead wins (Twitter, Facebook, Snowflake) didn't always gap on news

→ vs aggressive: 8% recession probability means 92% chance of being fine right until the 2008 recession was priced at less than 10% probability 'until it wasn't'

→ vs aggressive: P/B 2.84x justified by peak-cycle 17% ROE; ROE inflated by buyback-driven denominator shrinkage; tangible book multiple compressed below 1.0x in 2022-2023 downturn

→ vs aggressive: Comp ratio uses Q1 peak revenue (17.2B) to argue discipline; comp ratio set annually against full-year revenue which will be much lower if Q2 13.5B continues

→ vs aggressive: Q2 2026 at ~13.5B is year-over-year decline vs Q2 2025 at 14.6B; aggressive analyst mischaracterizes this as 'tracking higher' when it is tracking slightly lower

HOLD with covered call overlay strategy: maintain core position, sell covered calls at 1,060-1,075 strikes (December or March expiry) against the core position to collect premium, keep new capital in short-duration Treasuries, and await one of three confirmed triggers before deploying additional capital.

• Anthropic catalyst is real but timing uncertainty and probability weighting (60% landing as left-lead, 30% hot CPI) don't favor initiating new positions today
• Q1 standalone comp ratio of 11% vs 18.8% earnings growth is selectively optimistic—annualizing Q1 revenue of 17.2B gives 68.8B, but full-year likely closer to 60B
• 117% payout ratio is a flex lever, not an automatic red flag—banks can slow buybacks if Q3 revenue softens
• Technical setup lacks evidence of coiled base—volume has been roughly average with no higher lows since July
• The balanced path is a covered call overlay: monetizes implied volatility in a range-bound stock, cushions downside if 1,000 breaks, captures upside if 1,060-1,075 breaks
• Breakout trigger should require close above 50 SMA at 1,040, not just above Bollinger band—reclaiming the 50 SMA is higher-conviction
• GS at 1.28 beta and 2.84x book is not low-volatility; consider pairing with lower-correlation exposure like KRE regional bank short

→ vs aggressive: Buy-on-weakness-to-1,015 with 940 stop is a bet on catalyst timing, not defined-risk; if Anthropic lands at 60% probability and hot CPI at 30%, expected value doesn't favor initiating new positions today

→ vs aggressive: Comp ratio argument uses Q1 seasonal peak (17.2B) to annualize 68.8B, but full-year revenue likely closer to 60B given Q2 around 13.5B, materially changing the math

→ vs conservative: "Priced in" argument cuts both ways—if 93% no-cut is priced in, surprise cut is also priced out, making unexpected easing asymmetrically bearish for GS

→ vs conservative: 2008 recession example is survivorship bias—for every 2008 there are a dozen periods where consensus held; can't argue consensus is wrong only when it disagrees with your view

→ vs conservative: Not trimming into rallies leaves free money on the table; trimming 5-10% into 1,000-1,058 band tests and re-adding on weakness is mathematically profitable across cycles