The banks beat the quarter and downgraded the year. Only one of those is the signal.

TL;DR

• JPMorgan Q1 2026: net income +13% YoY to US$16.5B; trading revenue a record US$11.6B; Dimon language "increasingly complex set of risks." 2026 NII guidance trimmed to US$103B from US$106B.
• Bank of America: beat consensus on trading and wealth; loan growth underwhelmed.
• Wells Fargo: missed on NII and efficiency; charge-off rate ticked up to 0.51%.
• Citi: beat on IB and markets; narrative still organisational rather than earnings-driven.
• The trading line was a volatility dividend off the Iran war, private-credit stress, and SaaS dislocation. None of those are earnings engines — they are weather.
• The NII downgrades are the signal. Deposit betas are finally showing up; loan growth is not.

What happened

Week of 14 April, the four US money-centre banks reported Q1 2026. Four headlines landed in order:

• JPMorgan beat on every line. EPS US$5.87 against US$5.31 consensus. Trading revenue was a record. Dimon's prepared remarks contained the phrase "increasingly complex set of risks" — a phrase he has used in six of the last eight quarterly calls, and the phrase's frequency is itself worth noticing.
• Bank of America beat, led by trading and wealth fees. Consumer banking was flat. Loan growth came in at the low end of guidance.
• Wells Fargo missed on NII. The stock was down 3.4% the day of the release. Efficiency ratio deteriorated to 64.8%.
• Citi beat on markets and investment banking; the organisational restructuring narrative continued to dominate the call.

JPMorgan's 2026 NII guidance moved from US$106B to US$103B. That is a US$3 billion downward revision mid-year, roughly equivalent to the bank's entire Q1 credit card net revenue. The stock went up 1.2% on the print and the beat; the guide-down is what the sell-side quietly noted afterwards.

What it actually means

Here is the thing about Q1.

The banks made an enormous amount of money trading. They made it trading because the Iran ceasefire cracked, private credit gated, SaaS multiples collapsed by 22% year-to-date, and the yield curve steepened by 40 basis points over three weeks. Bank trading desks do well when clients need to move, and clients moved a lot in Q1 2026.

This is usually where the commentary says "trading is a lumpy line" and moves on. The commentary is correct but incomplete. A better framing: record trading revenue is a sign of volatility, and volatility is a sign that someone else is losing money. The banks are the counterparties, not the protagonists. The trading beat tells you the market had a hard quarter; it does not tell you the banks had a durably good one.

Meanwhile, the non-trading lines told a different story.

Net interest income, the actual core engine of a bank's earnings power, was weaker than the headline numbers suggested. JPMorgan cut guidance. Wells missed outright. BofA was flat. The mechanics are straightforward: deposit betas are finally catching up. Banks held deposit rates below market through 2024 and most of 2025 on the assumption that deposits were sticky. In Q1 2026, deposit cost of funds ticked up 14 basis points sequentially at JPM and 19 at Wells — the largest single-quarter moves in three years. Simultaneously, loan growth came in at 2.1% annualised across the group, the slowest pace since 2023.

Banks earn on the spread between the two numbers. The spread narrowed.

The second thing worth noticing: credit costs are normalising, not spiking. Net charge-offs across the group ran at 0.49% — up from 0.41% a year ago, down from the peak scare of 0.62% in 2024. This is not a crisis. It is a return to 2018-level normal, which after a decade of sub-normal credit costs feels like a problem.

The Dimon framing — "increasingly complex set of risks" — is load-bearing. He named, in order, private-credit stress, geopolitical conflict, commercial real estate, and AI-cannibalisation of borrower revenue. That last one is a CEO of a global bank publicly naming the same transmission channel Blue Owl's investor letter named last week.¹ When JPM and Blue Owl are reading from the same script, it is a script.

Hype deconstruction

"The banks are resilient." They are. This is priced. The KBW Bank Index is up 9% YTD, ahead of the S&P. The question is not whether the banks are resilient — they are — but whether the specific earnings source that produced the Q1 beat is repeatable. It isn't.

"Record trading is a positive signal for the economy." It's the opposite. Record trading is the economy's intermediaries getting paid for other people's discomfort.

"Credit is deteriorating." Overstated. Charge-offs at 0.49% are textbook normal. The previous decade of 0.30% charge-offs was abnormal. Calling the current rate "deterioration" anchors against the wrong baseline.

"Dimon is bearish." Mildly underlined. Dimon's language is deliberately calibrated — he runs a global systemically important bank, and his call commentary moves bond prices. What he actually said was that the risk environment is more complex, not that the economy is weak. Those are different claims.

Stakeholder landscape

Cross-layer implications

• Regional banking. Regionals report end-April. They lack the trading desks that carried the money centres; the deposit-beta story will dominate. Expect 2–3 NII misses and at least one efficiency-ratio disappointment.
• Commercial real estate. CRE charge-offs at JPM rose to 1.9% from 1.4% year-ago. The 2024 CRE scare did not go away; it got absorbed and is now bleeding through slowly. Two-to-three-year tail risk.
• Consumer credit. Card charge-offs flat at 3.1%. Affordability pressure visible but contained. Watch Q2.
• M&A advisory. Investment banking fees up 16% YoY across the group — the clearest confirmation that the Q1 M&A rebound is real, not a valuation illusion.
• AI. JPM disclosed US$2.1B in AI-related technology spend for 2026, up 28% YoY. The capex is concentrated in trading (where it has measurable P&L), fraud detection (measurable), and client service (not measurable). The mix is a tell about which AI investments are delivering and which are still on faith.
• Wealth management. Wealth AUM inflows ran at 6.3% annualised across BofA and JPM. This is the quiet shift: wealth is becoming a larger share of earnings than commercial banking at both institutions, and the boards know it.

What this means for you

If you are a bank-equity investor. H2 2026 EPS revisions are coming down. The trading line is not repeatable; the NII guide-down is. Rotate from the beat-on-trading names (JPM, BAC, C) to the ones with cleaner deposit-franchise optionality (regional aggregators that have absorbed deposit-beta earlier and look cheaper into H2).

If you are a corporate treasurer. Lock funding now. The NII downgrades mean banks are tightening credit appetite in the 12-month window, not the 3-month. Extending revolvers in Q2 is cheaper than renegotiating them in Q4.

If you are a CFO evaluating deposits. Your bank is finally competing for your cash. Treasury yields aside, negotiated deposit rates are 25–40bp better than they were in January. This window closes when deposit betas top out — probably Q3.

If you are on a bank credit-committee staff. The borrower-revenue-disruption risk Dimon named is already being expressed in covenant tightening on new middle-market loans. Expect internal policy updates before the end of Q2.

If you are an economist. Loan growth at 2.1% is the single most informative number in the earnings package. It says the private sector is not borrowing to grow. That is a leading indicator worth more than the headline GDP print.

Uncertainty ledger

• Deposit beta peak. The current betas are catching up to where they should have been 12 months ago. Whether they overshoot in Q2–Q3 is the most important unknown in the package.
• Credit normalisation vs deterioration. The delta between those two interpretations is roughly 15 basis points of charge-offs. Q2 data will tell.
• AI capex productivity. JPM's US$2.1B is spread across measurable and unmeasurable projects. The 2027 budget cycle will be the first real test.
• Trading-line repeatability. Whether Q2 produces a similar volatility environment is unknowable. The base rate for consecutive record trading quarters is low.
• Dimon succession timing. Not an earnings item. But the increasing complexity in Dimon's language is the kind of signal that historically precedes announcement windows.

Bottom Line

The banks beat the quarter on trading and downgraded the year on NII. The beat is a gift from volatility; the downgrade is a confession about deposit costs and loan demand. Only one of those is the signal. The interesting line in JPM's call was not the EPS — it was Dimon naming AI-cannibalisation of borrower revenue as a risk, which is the same risk Blue Owl named in writing last week. When money-centre banks and private-credit funds start reading from the same script, the script has moved from "view" to "consensus." That repricing has not finished.

Written in the tradition of — M.

Sources

• Tier 1: JPMorgan Q1 2026 10-Q and earnings call transcript; Bank of America Q1 2026 10-Q; Wells Fargo Q1 2026 10-Q; Citigroup Q1 2026 10-Q; Federal Reserve H.8 release
• Tier 2: Bloomberg; Financial Times; Wall Street Journal; KBW industry reports