Newsletter · · Ashutosh Agarwal

Fed Hikes and Bank Stocks Get Hammered as BofA Warns on Fees - Banks & the Rate-Cut Cycle - Week of September 18, 2026

Banks & the Rate-Cut Cycle for the week of September 18, 2026: the Fed raised rates a quarter-point to 3.75%–4.00% and signaled more to come, yet banks were the worst sector on the week as Bank of America warned investment-banking fees are down 10% with trading flat, JPMorgan said its fees are up mid to high teens, and a flattening curve undercut the asset-sensitivity story.

Banks & the Rate-Cut Cycle

Week of September 18, 2026: Fed Hikes and Bank Stocks Get Hammered as BofA Warns on Fees

"Banks are getting knocked about off the back of this decision. They've had some difficult guidance from the likes of Bank of America more recently, too." Bloomberg Intelligence, moments after the Fed's decision

For months this newsletter has carried a small irony in its name. It is called "Banks & the Rate-Cut Cycle," but the Fed has not been cutting; it has been threatening to go the other way. This week it finally did. On Wednesday, September 16, Chair Kevin Warsh's Federal Reserve raised interest rates for the first time in three years, a quarter-point move to a range of 3.75%–4.00%. And then something that should confuse a casual observer happened: the banks, which usually cheer higher rates, had their worst week in months.

That contradiction is the whole story this week, so it is worth slowing down to explain it.

A quick translation of two terms that recur below. Net interest income (NII) is what a bank earns on its loans and bonds minus what it pays savers on deposits; net interest margin (NIM) is that same gap expressed as a percentage. The big U.S. banks are what analysts call asset-sensitive: when short-term rates rise, the interest they earn on loans tends to climb faster than the interest they pay depositors, which widens the margin. So in theory, a rate hike is good news for bank profits.

In theory. This week the theory ran into two problems at once: the shape of the bond market turned unfriendly, and one very large bank told investors its Wall Street businesses are cooling off. The result was a sharp, and revealing, sell-off.

TL;DR

  • The Fed hiked, and it was hawkish. Warsh raised rates a quarter-point and signaled this is not a one-off. There is, in his words, "considerable more accommodation to be removed." Markets now expect a string of hikes, not one and done.
  • Bank stocks fell anyway, the worst sector on the week. Bank of America down about 7%, Goldman Sachs down nearly 7%. The trigger was not the Fed alone; it was BofA CEO Brian Moynihan warning that its investment-banking fees are down 10% and trading will be flat.
  • The bull and bear split cleanly between two CEOs at the same conference. JPMorgan's Doug Petno said dealmaking is robust and its fees are up "mid to high teens." Moynihan, a day earlier, said the opposite about his bank. Which one is the tell for the group is now the debate.

What's new

Ranked by how much each should move a bank book, not by when it aired.

1. Bank of America's CEO spooked the whole group: investment-banking fees down 10%, trading flat. The single biggest driver of the week came not from the Fed but from a conference stage. On Closing Bell (2026-09-14), reporting from the Barclays Global Finance Conference, the show relayed that Bank of America CEO Brian Moynihan said the bank's investment-banking revenue will fall in the current quarter, that quarterly sales-and-trading revenue would be "flat compared to last year," and that "investment banking fees are down 10 percent." Those comments dragged JPMorgan, Citigroup and Goldman Sachs lower with it.

Erica Najarian of UBS, a sell-side bank analyst, explained on the same program why the market punished BofA so hard, and it was not really about the deal slowdown everyone already expected:

"They said in July, look, we're going to generate 300 to 400 basis points of operating leverage... And then the CEO said in September that they're going to revert back to their original 200 to 300 basis points guide."

In plain terms: in July, BofA promised its revenue would outgrow its costs by a healthy margin; in September it quietly walked that promise back. Najarian added that consensus had penciled in "$18.3 billion of expenses for the third quarter. And Mr. Moynihan said $18.6 billion. And that's too big of a gap." The stock fell below $60. Her verdict was measured but pointed: "not necessarily a franchise issue," but a "communication issue" that needs addressing. Why it matters: this is the first hard, company-specific data point in weeks, and it is a negative one. It reframed a story that had been all tailwinds (the record 2026 deal boom this newsletter chronicled last week) into a question: have we seen "peak trading"? As Najarian put it, "maybe investment banking can't go up in perpetuity... a pretty strong reality check."

2. JPMorgan said the exact opposite, and that is the debate in a nutshell. One day later, on CNBC's "Fast Money" (2026-09-15), the same conference produced the mirror image. JPMorgan co-president Doug Petno said the bank "is seeing robust deal activity and expects Q3 investment and banking and trading fees to each be up in the mid to high teens." JPMorgan shares closed higher while the rest of the group struggled. Trader Karen Finerman drew the obvious conclusion:

"It sort of makes Brian Moynihan's comments from yesterday look very specific to them."

Why it matters: this is the crux. If Petno is right, BofA stumbled on its own and the fee engine at the best-run franchises is still humming. If Moynihan is the honest bellwether, the second quarter was the peak and every capital-markets desk, including JPMorgan's, glides lower from here. Two CEOs, one conference, opposite signals. Your view on the group hinges on which you believe.

3. The hike was hawkish, and it flattened the curve, the unfriendly part for margins. On Bloomberg Intelligence (2026-09-16), Stephanie Roth of Wolfe Research called it "a very believable delivery of a relatively hawkish hike," and flagged the line that mattered most: the Fed described this as merely "removing accommodation," which she said was "one of the single most important lines... certainly not a one and done." Markets heard a campaign, not a gesture. The program's hosts watched the reaction in real time: "Goldman... drop is some four or five percentage points this afternoon."

Here is the mechanical problem for banks. A hike lifts short-term rates, but the long end of the bond market (where the 10-year Treasury sits above 5%) did not rise to match; the gap between short and long rates narrowed, or "flattened." Banks like a steep curve, because they borrow short (deposits) and lend long (mortgages, commercial loans), and earn the spread between the two. Flatten that curve and you squeeze the very margin the hike was supposed to help. Why it matters: the textbook says "rates up, banks win." The fine print says "only if the curve cooperates." This week it did not.

4. Even the constructive analysts now hedge on credit and loan demand. On Closing Bell (2026-09-16), CNBC's Leslie Picker laid out the second worry cleanly. The big banks have "two levers" (their Wall Street businesses, trading and investment banking, and their lending margin) and both now face questions:

"You have the potential for stickier inflation, higher for longer interest rates, a potential hiking cycle... what that means for loan demand, what that means for credit quality are all kind of big questions."

Why it matters: a single quarter-point hike does nothing to borrowers. A cycle of them, on top of a 5% ten-year, eventually cools loan demand and tests credit quality, and that is the scenario the dot plot now hints at.

The debate

Bull NII case: the asset-sensitivity math still works. The clearest statement of the bull case came from a bank analyst, not a pundit. On Closing Bell (2026-09-17), Barclays' Jason Goldberg walked through it plainly:

"As the Fed hikes, you'll see deposit costs rise. But also assets are repricing higher... the majority of banks remain asset-sensitive. So those assets will reprice quicker, which is actually constructive for net interest margins."

He argued a "25 basis point hike, 50 basis point hike is probably not enough" to break the economy, and that loan demand is "quite robust" outside of mortgages: commercial lending driven by the AI investment cycle, a pullback in private-credit competition, and even commercial real estate "returning" as office fears subside, with credit-card borrowing still resilient because "consumers have jobs and they continue to spend." His punchline on the stocks: valuations are "still... in line with historical average, despite the fact... it's a better-than-average historical environment," so there is "more to go."

Bear NIM case: the good news may already be priced, and the levers are cooling. Goldberg himself supplied the caveat: "all else equal. All else is not always equal." The bear case does not argue margins collapse; it argues three things quietly turned against the group this week. The curve flattened rather than steepened. The fee engine, the part that does not care about NIM at all, just took its first real warning shot from Moynihan. And RBC strategist Lori Calvasina, an admitted financials overweight, told Fast Money the sector had gotten "expensive... enough for us to call it out," a "yellow flag," with the "capital markets group" the stretched piece and "regional banks looking fine." When a bull is trimming and the two Wall Street levers are cooling into a hawkish Fed, "priced for perfection" starts to look like the risk.

The textbook says higher rates help banks. This week taught the fine print: only if the curve steepens, the borrower stays healthy, and the fee machine keeps running. This week, at least one of those wobbled.

Stocks in play

JPMorgan (JPM). Bull: the cleanest, best-sourced read of the week, straight from co-president Doug Petno: robust deal activity and Q3 fees up "mid to high teens," with the stock rising while peers fell. Scale, a fortress balance sheet, and a fee book that is still growing make JPM the group's safest way to own the hike. Bear: if Moynihan, not Petno, is the honest signal on the industry, even JPMorgan's fee momentum fades, and a flattening curve caps the NII upside the hike was supposed to deliver. Next catalyst: Q3 results in mid-October, which will confirm whether "mid to high teens" survived contact with a cooling September.

Bank of America (BAC). Bull: as the most rate-sensitive of the giants, BofA has the most NII torque to a Fed that holds or hikes, and Najarian was explicit that this looks like a communication stumble, not a broken franchise; a stock below $60 at a discount multiple to JPMorgan and Goldman is priced for the bad news. Bear: management's own credibility took the hit this week: a July operating-leverage promise walked back in September, investment-banking fees down 10%, trading flat, and a $300 million miss versus consensus on expenses. Down 7% on the week for a reason. Next catalyst: Q3 earnings, where Moynihan has to prove the guide-down was conservatism, not a leak.

Wells Fargo (WFC). Bull: the classic asset-sensitive, deposit-rich franchise, and the one with the least Wall Street exposure to the fee wobble that hit BofA and Goldman. A hawkish, higher-for-longer Fed is precisely the backdrop for the profitability climb CEO Charlie Scharf laid out last week (a raised return target of 17–18%, with delinquencies still improving). Bear: a flattening curve pressures the spread that is the core of Wells' earnings, and with less capital-markets fee income to offset it, the NIM story has to carry more weight than at JPMorgan or Citi. Next catalyst: Q3 results proving the profitability target is a print, not a promise.

Citigroup (C). Bull: Citi's markets and underwriting desks read directly off the constructive side of the tape: Petno's "robust" deal activity, and a league-table backdrop where the bulge bracket is still capturing the vast majority of M&A fees. A cheap valuation and an improving-returns story remain the setup. Bear: Citi sits squarely in the businesses Moynihan just warned about, and its shares were among those "dragged lower" by his comments; if the IPO and equity-capital-markets pipeline is cooling, Citi feels it on both the fee line and the lending spread. Next catalyst: Q3 proof that return on tangible equity keeps grinding higher despite the September chill.

Read-throughs

  • Super-regionals (USB, PNC, TFC). These three are the purest test of the week's central worry. They lack the deep capital-markets fee engines of JPMorgan and Citi, so their earnings lean hardest on the lending spread, which means the flattened curve is a sharper headwind for them than for the money-center banks. The offset is Goldberg's read that "commercial loan demand is quite robust" and that banks are "returning" to commercial real estate as office fears fade, a genuine positive for regional loan books. And Calvasina's note that "regional banks looked fine" on valuation, while the capital-markets names looked stretched, argues the relative value has shifted toward exactly this group.
  • Deposit competition. Goldberg was blunt that "as the Fed hikes, you'll see deposit costs rise." A higher-for-longer world keeps cash and short-term Treasuries paying savers handsomely, which keeps the fight to hold deposits expensive. That is a bigger drag for the regionals, whose funding is less sticky and more rate-sensitive, than for the money-center banks with the cheapest, stickiest deposit bases.
  • Capital-markets and lending fees. This is the loudest and most contradictory signal of the week. Moynihan says fees down 10% and trading flat; Petno says fees up mid-to-high teens. The league-table backdrop, via the Market Maker podcast (2026-09-14), still favors the giants: Goldman Sachs sits atop M&A with "$1.4 trillion worth of deals done, picking up almost $3 billion worth of fees... a 33% increase," followed by JPMorgan and Morgan Stanley. And executives at the Barclays conference, per Leslie Picker, still described "CEO confidence being strong, boding well for mergers and acquisitions... and loan demand holding up." The pipeline looks full; the argument is over how fast it drains.
  • CRE and consumer credit. The direct commentary stayed constructive: Goldberg pointed to commercial real estate recovering as office fears subside and to resilient credit-card borrowing on the back of a healthy job market. The caveat is forward-looking, from Picker: a genuine hiking cycle, layered on a 5% long-end, is what would eventually slow loan demand and test credit quality. No cracks yet; the watch item is whether the cycle Warsh hinted at ever arrives.

What changed vs last week

Last Friday's issue was titled "The bankers finally spoke, and they sound confident." Two bank CEOs and a team of JPMorgan dealmakers had turned an abstract macro thesis into concrete, sourced optimism: Charlie Scharf raising Wells Fargo's return target to 17–18% with credit still improving, and JPMorgan's equity bankers describing a record 2026 issuance year. The September 16 meeting was framed as a coin-flip "hike or hold."

What changed: the coin landed on a hike, and it was hawkish, not the gentle "one and done" some hoped for. Warsh signaled more to come. And the confident operator chorus cracked: the same conference that produced last week's optimism produced Moynihan's warning this week, and the fee-boom narrative met its first real counter-evidence.

What was contradicted: last week's cleanest bull point, a record, still-accelerating capital-markets year, is now openly contested. Petno restated it; Moynihan undercut it. The "best year since 2021" story did not die, but it stopped being consensus.

What carried over, and hardened: the higher-for-longer, asset-sensitive framework is fully intact; Goldberg's NII math is the same as it was a week ago. What is new is the recognition that asset sensitivity only pays if the curve steepens and the fee levers keep running, and this week neither cooperated, which is why banks were the worst-performing sector even as the Fed did the "bank-friendly" thing.