Berkshire Hathaway's Q2 trim signals sector rotation away from credit-cycle exposure, not benign rebalancing

• Q2 2026 Berkshire added $3.0B in Alphabet, initiated Lennar and Delta, initiated Constellation Energy
• COF trimmed alongside Nucor, Kroger, and Constellation
• Every new position (Alphabet: AI infrastructure, Lennar: housing late-cycle play, Delta: travel rebound, Constellation: AI/power demand) is a secular growth or rate-cut-beneficiary name
• Capital One is a credit-cycle-exposed consumer lender
• Bank of America — a money-center bank — remains a top Berkshire holding
• COF ran 124.7% in three years vs. Apple's run
• Reddit returned ZERO mentions across wallstreetbets, stocks, and investing

↩ rebuts: “Bull claim that Berkshire trim is 'just position sizing' and 'means nothing'”

July operating data invalidates Q2 recovery signal as Q2 was the peak of a seasonally strong quarter and a low-base effect

• Q2 2026: Revenue +4.1% QoQ, EPS +41.6% QoQ, NII +1.9% QoQ
• BofA flagged slower card growth in July operating update
• Q2 2025 had a $4.277B net loss with only $425M of tagged unusual charges
• YoY 'recovery' is partly a base effect from a one-time charge
• BofA is a Buy-rated house and covers credit card space most closely
• Net loans grew 2.4% QoQ to $434.5B ($10B)
• Tangible Book Value declined 2.7% QoQ to $65.85B ($107.35/share)
• Textbook risk-on growth at the expense of capital quality

↩ rebuts: “Bull claim that Q2 acceleration proves the card growth recovery is real”

$47.94B of goodwill (42.1% of equity) is not noise — it's a concentrated bet that the Discover deal works, and goodwill does NOT amortize

• Goodwill + Intangibles Q2 2026: $47.943B
• Goodwill + Intangibles Q1 2026: $44.589B
• Goodwill + Intangibles YoY (vs. 2024): +194.8%
• Total Equity Q2 2026: $113.793B
• Total Equity Q1 2026: $112.261B
• Total Equity YoY: +86.9%
• Goodwill as % of Equity Q2 2026: 42.1%
• Goodwill as % of Equity Q1 2026: 39.7%
• Q2 goodwill rose by $3.35B sequentially — merger accounting still being finalized
• If Discover synergies underperform, the $48B gets written down in a discrete, headline-blowing charge
• P/TBV of 2.02x with ROE of 9.03% = paying ~22x ROE-multiple-to-TBV
• For a healthy large-cap bank, that ratio should be ~1.5–2.0x
• COF is at top of P/TBV range while ROE is at the bottom

↩ rebuts: “Bull claim that goodwill bump is just purchase accounting and TBV decline is noise”

MACD shows confirmed bearish cross with expanding negative histogram — not a 'pause,' a momentum deterioration pattern

• MACD line: +2.72 (down from peak)
• Signal line: +3.89 (above MACD line)
• Histogram: −1.17 and expanding negative
• Three consecutive days of deteriorating momentum
• Signal has crossed MACD line and gap is widening in bearish direction
• Price rejected at $226.54 (Aug 14) — upper Bollinger band is $227.04, less than $1 from rejected high (double-resistance tap that failed)
• RSI at 53.61 — neutral but declining
• VWMA at $219.38, just 1% above price — market's marginal cost basis is current price
• Upper Bollinger at $227.04 is less than 4.5% above price
• Lower Bollinger at $209.40 is 3.6% below price
• Bull's 'asymmetric' trade is actually +4.5% to resistance vs. −3.6% to support — near-symmetric with slight upside lean

↩ rebuts: “Bull claim that MACD cross is 'just a pause' and to 'buy the dip'”

Forward EPS estimate of $24.014 is fantasy-dependent — requires flawless Discover execution over 18 months

• Forward EPS estimate: $24.014
• Trailing EPS: $18.14
• Q2 2026 annualized run-rate: $4.73 × 4 = $18.92
• Q2 sequential growth rate: +41.6%
• Requires Q3 2026 EPS +13.3% sequentially despite slower card growth and July deceleration
• Requires Q4 2026 EPS +1.1% sequentially in seasonally weak quarter
• Requires 2027 average EPS +22% above current run-rate to absorb full-year Discover synergies
• If COF misses by 10% on forward EPS, multiple re-rates from 9.05x to 10.1x trailing — zero multiple expansion

↩ rebuts: “Bull claim that 9.05x forward P/E is attractive and 'normalized'”

Higher-for-longer rates is the most dangerous assumption — COF has largest subprime/near-prime credit card book and consumers are weakening

• COF has the largest subprime and near-prime credit card book among major U.S. issuers
• Higher-for-longer rates = subprime borrowers hit maximum debt service ratios within 12–24 months
• UMich consumer sentiment: 49–52 (recession-warning level)
• Polymarket prices PCE inflation re-accelerating
• Only 8% recession odds — but 87% no-cuts means Fed has no cushion to cut if credit cracks
• If charge-offs spike 100–200 bps, ROE drops to 5–6%
• Stock multiple goes to 6–7x when dividend coverage and buyback capacity disappear
• Upside scenario: EPS roughly flat, stock range-bound $200–$230
• Downside scenario: charge-offs spike, NIM compresses, buybacks suspended, stock re-rates to 6–7x trough earnings = $130–$150 (-30% to -40% from $217)

↩ rebuts: “Bull claim that 'higher-for-longer = NIM stays wide and credit losses don't blow up' as a tailwind”

Capital return story is a mirage — $2.747B in buybacks barely moved the share count due to dilution mechanics

• Q2 net income: $3.020B
• Q2 common capital returned: $3.247B ($2.747B buybacks + $500M dividends)
• Common shares outstanding declined only 0.4% sequentially (613.485M vs. 615.857M)
• COF spent $2.747B on buybacks and share count barely moved
• Average buyback price was high (close to $200+ per share)
• Stock-based compensation, RSU vesting, and tax withholding are eating buyback at almost 1:1 ratio
• Mechanical capital return illusion — cash leaves building but share-count benefit is rounding-error small
• 108% of net income return is the high-water mark, not sustainable run-rate
• In credit downturn, buyback pace gets cut first

↩ rebuts: “Bull claim that 108% of net income returned proves strong capital return story”

Regulatory and legal overhang is real multi-quarter/multi-year drag, not 'noise'

• COF previously paid $390M FinCEN penalty (prior enforcement action regulators can use as precedent)
• 385 Trump account closures subject to active litigation — discovery, depositions, document production
• BofA July note flagged credit concerns simultaneously with AML story gaining airtime
• Regulators correlate enforcement priorities — when bank is in news for AML, examiners tighten scrutiny on capital adequacy, stress testing, consumer compliance
• Multi-quarter legal expense
• Multi-year reputational drag
• Elevated risk of follow-on regulatory action
• Grinds through compliance costs and limits strategic flexibility for 18–36 months

↩ rebuts: “Bull claim that Trump-debanking/AML story is political noise”

Valuation is mid-range with downgrade risk — 11.97x trailing P/E with 9.03% ROE is stretched relative to peers

• Historical median P/E for diversified credit-services issuers (10-yr): ~10.5x trailing
• COF trailing P/E today: 11.97x
• COF trailing P/E relative to own 5-year average: ~17–18x trailing during normal periods (post-2021 normalization)
• COF at 11.97x trailing and 9.05x forward — at lower end of recent range but only if forward EPS is delivered
• ROE: 9.03%
• P/B: 1.29x
• Tangible book multiple: 2.02x
• Mid-pack valuation for mid-pack return profile with above-average goodwill exposure and credit-cycle leverage

↩ rebuts: “Bull claim that 9x forward P/E is 'cheap'”

Technical setup has clear asymmetry to the downside — entire buy zone is less than 4% below current price with momentum already broken

• Price rejected at $226.54 on Aug 14
• Upper Bollinger: $227.04 (less than $1 from rejection)
• Lower Bollinger: $209.40 (currently $217.21 — only 3.6% below)
• 50 SMA: $209.04 (key support, only 3.9% below)
• 200 SMA: $205.95 (structural floor, only 5.5% below)
• Bull identified $209–$212 as 'most attractive risk/reward' zone — that's less than 4% below current price
• No high-conviction buy zone — tight range where downside is fully mechanical and upside is conjecture
• Bull's own analyst says 'don't buy here, wait for a 4% pullback' — that's not a BUY rating, that's a HOLD with a watch order
• If price breaks below $209 with volume, MACD histogram is already negative and expanding
• 200 SMA at $205.95 is last line of defense
• Below that, next support is May 2026 low at $180.13 — a -17% move

↩ rebuts: “Bull technical strategy of 'buy the dip in the $209 zone'”