JPMorgan's Trading Boom Is the Only Floor Under the Banks This Print — Provisions Just Confirmed the Trade

Trading desks across the five largest U.S. banks are projected to ring up roughly $38B in Q3 2026 revenue, down about 19% sequentially from Q2's ~$47B record haul but still up about 13% versus Q3 2025. JPMorgan Chase is the only firm giving the Street a bullish line item — co-president Doug Petno told an industry conference on Sep. 15 that the bank expects investment banking fees and Markets revenue both up a mid-to-high-teens percentage year-over-year, implying Markets revenue of roughly $10.2B to $10.6B versus $8.9B in Q3 2025. That sits well above the $7.9B consensus Bank of America has guided to — Moynihan flagged sales and trading as flat and investment banking fees down 10% or more — and above the FICC-skewed caution Goldman Sachs CEO David Solomon telegraphed on Sep. 16, when he said the FICC unit would run softer than equities.

The credit-cycle trade hinges on whether the consumer-borrower cracks under the highest 10-year yield since 2002 and a Fed that just delivered its first rate hike in more than three years. The most recent datapoint in this series comes from Wells Fargo's Q2 2026 print — the firm reported a provision for credit losses that included a modest allowance build, with the consumer net loan charge-off rate falling to 0.74% (annualized) from 0.78% the prior quarter, driven by lower auto and credit card net charge-offs. A year earlier, the same firm's Q3 2025 charge-off rate had fallen to 0.73% from 0.81%, with the $681M provision running36% below the $1.07B set aside in Q3 2024 — so the trend of falling credit costs is now stretching across four consecutive quarters.

Reuters's Oct. 8 preview quoted analysts expecting "no signs of deterioration in the credit portfolio" in the upcoming print, and the prior quarter at Citigroup showed the same shape: a $109M Markets provision (with a $127M allowance build) and a $242M Banking provision against a strong $1.5B in investment banking revenue, +44% year-over-year. The contrast between strong trading/IB lines and modest provisions is exactly the gap the print-day thesis rests on.

The market is walking into this morning after a brutal stretch for bank stocks — the KBW Nasdaq Bank Index has fallen 9.7% over the past month against a S&P 500 up about 1.5%, with the six largest lenders shedding roughly $270B in market capitalization. The driver has been the move higher in long-end rates, not the Q2 results themselves: JPMorgan Chase's Q2 net income of $21.2B ($7.70 reported / $6.14 adjusted EPS) was the highest quarterly profit in U.S. banking history, but it carried a $4.6B one-time gain on the tender of the bank's 18.6 million Visa Class B-2 shares — a number that will not repeat. The clean read of underlying earnings power is the adjusted figure, and that prints roughly 14% above the Wells Fargo Q3 2025 comparable and roughly 7% above the Goldman Sachs Q3 2025 baseline.

Goldman Sachs and Citigroup are the two clearest expression trades on the Markets-versus-everything-else theme. Goldman ran Q2 2026 net revenues of $20.34B (+39% YoY), with equities trading up 72% YoY on the way to a record $20.98 reported EPS — but the equities surge was set against a FICC franchise that management now expects to be the weaker leg into Q3. The implied bar is that Q3 2026 net revenue falls back toward the $15B-to-$17B area, with the FICC print the swing factor.

Citigroup, by contrast, has spent the past four quarters compounding a Markets franchise that crossed $7B in Q2 2026 (+17% YoY) while Banking grew 34% on Investment Banking +44% (DCM +65%, ECM +92%). The $30B multi-year buyback program authorized at the 2026 Investor Day gives the print a clear per-share tailwind — net income is on track for double-digit YoY growth in four of five businesses, with RoTCE running 13% firm-wide. The Q3 setup for Citigroup is the highest-quality top-line beat in the group, even if the absolute EPS dollar figure remains below the Goldman Sachs print.

For a market that has already paid down the sector's earnings expectations by9.7% in a month, the bar to clear is less about absolute number-beating and more about whether the trajectory of the prior two quarters — JPMorgan Chase's record $21B Q2 profit, Goldman Sachs's record $20.34B Q2 revenue, Citigroup's $24.8B Q2 revenue, Wells Fargo's $33.6B Q2 revenue — has the second leg in it. The forward earnings multiple on JPMorgan Chase has compressed about 11% this year even as the share price has gained only about 3%, according to TIKR; that valuation reset means the Q3 print has more latitude to surprise positively than the year-to-date tape suggests.

The most-cited risk in the Reuters Oct. 8 preview was that higher long-end rates would weaken credit quality. The print-day evidence cuts against that: Wells Fargo's consumer net charge-off rate has fallen in four straight quarters, Citigroup's allowance build in Markets was a modest $127M against a $7B revenue base, and no analyst pre-print forecast called for a meaningful provision step-up. The cleanest single signal that the cycle has turned would be a Wells Fargo Q3 2026 provision that meaningfully exceeds the $681M Q3 2025 print, paired with a consumer charge-off rate that breaks back above 0.80% — neither of which is the base case into this morning.

The structural read is that banks have spent the three years since the 2023 regional crisis shortening duration on their securities portfolios and tightening underwriting on consumer loans. That shows up in the Wells Fargo net-charge-off trajectory and in the Citigroup credit-cost discipline, and it means the 10-year yield at multi-decade highs is not the same setup it was in 2022-2023 — there is less unrealized loss exposure sitting in held-to-maturity books. The print-day story is therefore not whether the cycle has turned (it hasn't, by the data) but whether the trading-desk tailwind is enough to offset the multiple compression from rates.

Trading desks alone cannot explain the ~9.7% sector pullback — the sell-side and buy-side both see the trading print as already-discounted. The unresolved questions are whether JPMorgan Chase's Q3 2026 Markets revenue can hit the high end of Doug Petno's mid-to-high-teens YoY guidance against a tough $8.9B Q3 2025 base, whether Goldman Sachs can defend its record Q2 equities level, and whether Citigroup's $30B buyback authorization converts into a buyback-execution tailwind in the EPS line. The trading-side variance is real but bounded; the credit-cycle narrative has more room to surprise if the Wells Fargo provision print cracks higher.

The cumulative read: trading-led earnings beats are already in the tape from Q2, and the question is whether the trading story keeps delivering into Q4 — which depends on the FICC cyclical signals and the equity-market volatility that powered Goldman Sachs's record Q2. The credit cycle is benign by every datapoint that has printed, but the multiplier on that benign signal is a sector that has already de-rated 11% on the multiple. The print-day trade is whether JPMorgan Chase can clear the high end of its own guidance and whether Wells Fargo's charge-off rate can extend its four-quarter downtrend.