Newsletter · · Ashutosh Agarwal

Bank Stocks Slide as the Yield Curve Flattens Before Earnings - Banks & the Rate-Cut Cycle - Week of October 2, 2026

Banks & the Rate-Cut Cycle for the week of October 2, 2026. Podcast synthesis on why a flattening 2-year/10-year yield curve is erasing the benefit of the Fed's first hike in three years, the AI scare hitting bank shares since Meta's Muse launch, a regional CEO's doubled loan pipeline and sticky deposits, the bull-versus-bear NII debate, and what Q3 earnings starting October 13 could settle for JPMorgan, Bank of America, Wells Fargo and Citigroup.

Banks & the Rate-Cut Cycle

Week of October 2, 2026: Bank Stocks Slide as the Yield Curve Flattens Before Earnings

"Yes, rate hikes have historically been good for banks, but a flatter yield curve is not good for banks."

Liz Thomas, on the RiskReversal Pod

Bank stocks are supposed to like rate hikes. Banks earn more on loans when rates rise, and they usually take their time passing that on to savers. So it should be a puzzle that, two weeks after the Fed raised rates for the first time in three years, bank stocks are one of the worst places in the market.

It isn't a puzzle once you look at the shape of interest rates. Short-term rates, which drive what banks pay depositors, have risen faster than long-term rates, which drive what banks earn on loans and bonds. That gap is close to vanishing. Add a fresh worry that AI assistants could pull customers away from banks, plus a slower calendar for big stock offerings, and you get a group that has gone from leader to laggard in about a month.

The good news: the banks themselves report in under two weeks. JPMorgan, Wells Fargo and Citigroup go first on October 13.

Three terms we use a lot.

  • Net interest income (NII) is what a bank earns on its loans and bonds, minus what it pays depositors.
  • Net interest margin (NIM) is that same gap, shown as a percentage.
  • The yield curve is the line connecting short-term and long-term interest rates. When it is "steep," long-term rates are well above short-term ones, which is good for banks. When it is "flat," the two are close together, which squeezes them.

TL;DR

  • The curve is the problem, not the hike. Strategist Liz Thomas says the gap between 2-year and 10-year Treasury yields hit "18 or 19 basis points" this week, close to flat. Banks borrow short and lend long, so a flat curve pinches margins. She named Bank of America as the big bank most exposed.
  • AI fear is hitting bank stocks. Since Meta launched its Muse AI agent this month, CNBC's Power Lunch reported Citi "down about 5 percent," JPMorgan "down seven," Wells Fargo "about double digits," and regional banks down "about 8 percent."
  • A regional CEO pushes back on both worries. Valley National CEO Ira Robbins says his commercial loan pipeline has "doubled" to $4 billion in a year and that people pick banks "based on relationships... Not so much on interest rates." Q3 results, starting October 13, will show whether the big banks see the same thing.

What's new

Ranked by how much each should matter to someone who owns bank stocks, not by date.

1. A flatter yield curve is eating the benefit of higher rates

On RiskReversal Pod: "U.S. Bonds Are Trading Like an Emerging Market with Liz Thomas" (September 28), market strategist Liz Thomas (Pundit) gave the clearest explanation of why the Fed's hike has not helped bank stocks.

"Historically, rate hike cycles have been good for banks," she said. They usually mean a bank's "net interest margin rises." But "the issue this time is that as the long end of the curve has risen, the short end of the curve has risen faster." The gap between 2-year and 10-year yields "hit a low this week of, I want to say, 18 or 19 basis points." (A basis point is one-hundredth of a percentage point, so 18 basis points is 0.18%.) "That's not inverted, but it's much closer to inverted than it was a couple months ago."

Then she named a name:

"Bank of America has a lot of exposure to lending and consumer banking. So if it is lending on the long end, borrowing on the short end, meaning trying to pay out for deposits on the short end, its net interest margin gets squeezed and they have a disproportionate exposure to that."

She compared it to Goldman Sachs, which has "disproportionate exposure to deal flow." In other words, each big bank has a different weak spot, and the flat curve finds Bank of America's.

Thomas added that banks were "in the bottom 10 of industry groups, down about 3% over the trailing five-day period." She watches banks as a signal for the whole market: "I don't like seeing banks roll over here because that starts to confirm for me that the broadening out that we've been excited about in this market for so long is stopping."

Why it matters. This is the key question for Q3 results. Last week, Chris Whalen said deposit costs had started rising at Wells Fargo and some regionals. Thomas explains why that hurts more than usual: the long-term rates banks lend at are not rising fast enough to make up for it. If the curve stays this flat, expect banks to be more cautious about their 2027 interest-income outlooks.

2. A regional bank CEO says loan demand is booming, and deposits are not going anywhere

The most useful operator voice of the week came from outside our seven banks. On Power Lunch: "OpenAI DevDay, Jamie Dimon's Op-Ed, Valley Bank CEO Interview 9/29/26" (September 29), Ira Robbins, CEO of Valley National (Operator/Insider), pushed back on almost every worry hanging over bank stocks.

On loan demand. "We have an unbelievable pipeline today at Valley. Our pipeline's $4 billion versus about $2.5 billion just a year ago." When the anchor asked, "The pipeline almost doubled?" Robbins said simply: "Doubled." The growth is in business lending: "even though there's been inflation, even though there's been rise in interest rates, there's still value out there. And we're seeing people investing in equipment, investing in working capital." Consumer lending is a different story: "Residential's flat. Auto's flat... There's very little, if any, refinance activity."

On whether AI will drain deposits. Asked whether AI agents could move customers' cash to wherever rates are highest, Robbins said: "Absolutely incorrect." His argument is history plus math. Money market funds "came in in the 1970s and took the rate-sensitive deposits out of banks" decades ago. What is left is small and sticky. He said the median checking account at Valley holds about $4,400, and across the country it's "only $8,000 sitting in a checking account." His summary:

"People choose banks based on relationships, based on location, based on consumer and digital mobility and what the ease of that is. Not so much on interest rates."

On credit. Business clients look healthy: "Our balance sheets of our commercial clients are strong... There's a lot of liquidity in the market." Consumers are split: "you have those that have a lot of money that are still continuing to spend... And those that don't, you know, significant credit card debt."

On regionals versus the giants. Because "most of our revenue comes from the commercial client," Robbins expects regional banks' revenues to "far out do better than what some of the other money center banks are that have a larger dependency upon the consumer."

Why it matters. Two of this month's biggest fears are that deposits will flee and loan demand will dry up. Robbins is a CEO talking his book, but the $2.5 billion-to-$4 billion pipeline is a specific number from a bank operator, not a forecast. If the big banks report similar business-loan growth on October 13, the "no loan demand" story weakens. Watch for this read-across in the super-regionals in particular (see below).

3. The AI scare: bank stocks are trading on Meta's Muse

A new force is moving bank stocks, and it isn't interest rates. Power Lunch's anchors laid out the damage since Meta launched Muse, its AI agent: "Take a look at Citi. Down about 5 percent. JP Morgan down seven. Wells Fargo, Morgan Stanley and Goldman Sachs have fallen about double digits since the launch... The regional banking sector has lost about 8 percent of its value."

On The Rundown: "Mortgage Rates Hit Highest Level in 2 Years, Meta Tests 'Human Concierge' for Muse" (September 23), the host (Pundit) described one bad day: "The financial sector was the worst performing sector in the S&P yesterday, dropping 2%. Some big names like Charles Schwab fell 6%, while JP Morgan and Bank of America fell 3%. And the reason for that is AI. The fear now is that AI agents like Meta's Muse could disrupt everything from wealth management and insurance to payments and banking." The host's own take was skeptical: "I personally don't think that banks are going to be that impacted, but I guess we'll see."

Why it matters. The fear is that an AI assistant could shop around for the best savings rate, card or loan on a customer's behalf, wearing away the customer loyalty that keeps bank funding cheap. Nothing in the banks' numbers shows that yet. But the share-price moves are real, and they compound the rate pressure. Robbins' "Absolutely incorrect" is the operator rebuttal; Q3 deposit data is the evidence that will settle it.

4. The bearish case from pundits: "Any kind of bank stock"

Several generalist commentators turned sharply negative this week.

  • On The KE Report: "Joel Elconin - Can the 'Buy the Dip' Mentality Survive Rising Yields and Sector Breakdown?" (October 1), Joel Elconin (Pundit, Pre-Market Prep Show) said "financials are in freefall" because banks "have to have a higher payout on, you know, the money that they have sitting in the bank. That's not good. That squeezes margins." On loans: "Is loan demand going to be increasing in a rising interest rate environment? Absolutely not. So, the banks are in a load of trouble here as rates continue to go higher." That is the exact opposite of what Robbins is seeing.
  • On Schwab Network: "William Lee's Big Picture Outlook on Fed, AI Financing & Private Credit Risks" (September 28), William Lee (Pundit, chief economist at Global Economic Advisers and a former Citigroup chief U.S. economist) focused on private credit, meaning loans made by investment funds rather than banks. "It's not the size of private credit that matters. It's the opaqueness and how it is intricately interwoven into the banking system." Banks are tied in through "a lot of credit lines and what they call net asset value loans" to those funds. When the funds hit trouble, "you immediately draw on the bank credit line." He compared it to the collapse of the hedge fund LTCM in 1998, which "worked itself so intricately into the banking system" that it "sucked in the entire financial system." Asked which asset class he'd be most cautious holding, he answered: "Any kind of bank stock."
  • On RenMac Off-Script: "Beneath the Bull" (October 2), RenMac's strategists (Pundit) said they "started to see oversold conditions develop in investment banks" and "some of the diversified regional banks. So the curve and rates started to hit financials." They added a useful nuance: "at some point those higher rates can actually be good for financials, right? But you have to cross the Rubicon and it becomes bearish."

Why it matters. "Oversold" means stocks have fallen so fast that a bounce becomes more likely. When generalists line up this bearishly two weeks before earnings, the bar for the banks to beat gets lower. That cuts both ways: real margin damage would confirm the gloom, but steady numbers could spark a sharp recovery.

5. Dimon on inflation: "I don't bet on" it coming down

On Squawk on the Street: "9AM HOUR: Nasdaq Extends Record Run, AI Trade Goes 'Meta,' Paramount Settlement 9/22/26" (September 22), CNBC played a clip of JPMorgan CEO Jamie Dimon (Operator/Insider). He warned that "re-militarization, infrastructure spending, AI spending, you know, ongoing government deficits" will keep pushing inflation up, so "part of that inflation number may be the die's already been cast. It isn't about what the Fed does next, next raise." His bottom line: "I'm hoping that inflation stays here and starts to come down. I don't bet on that. I think there's a chance it won't. And it may even go up a little bit."

The same program's hosts summed up the other side, relaying Bank of America CEO Brian Moynihan's consistent view of "resiliency out of the consumer in nominal terms," citing "credit card balances" and "deposit data" that "continues to show growth." They added that they have been hearing the same from Wells Fargo CEO Charlie Scharf: "we've been hearing it from all the bankers."

Why it matters. If the head of the largest U.S. bank thinks inflation could rise from here, he is not planning for rate cuts anytime soon. That supports the "higher for longer" case for bank interest income, but it also means more pressure on deposit costs and more strain on weaker borrowers. Dimon also appeared on Bloomberg Talks: "JPMorgan's Dimon, Gov. Whitmer and Ford's Farley Talk the Michigan Lift" (September 29), where JPMorgan announced a commitment of "$1 billion of debt capital to support Michigan manufacturing sector over the next decade," as reported on Bloomberg Daybreak: US Edition (September 30). That was a policy appearance; he gave no new figures on interest income or credit.

The debate

Bull NII case: higher for longer still pays. The bull case rests on three supports. First, big banks still earn more as rates rise, because loans and bonds reprice upward over time while much of their deposit base, especially small checking accounts, barely moves. Robbins' point that the median Valley checking account holds about $4,400 is the core of this argument: small balances rarely chase rates. Second, Dimon's warning that inflation "may even go up a little bit" implies rates stay high, giving that repricing more time to work. Third, loan demand may be better than feared. Robbins' pipeline "doubled" to $4 billion, driven by companies "investing in equipment, investing in working capital." On ETF Spotlight: "Positioning Portfolios for the Midterm Elections" (September 28), an ETF strategist (Pundit) added a capital tailwind: "new capital rules were released in March, and they're likely to become effective at the end of the year. And that should have the potential of freeing up capital on bank balance sheets." The strategist noted that in Q2, "99% of financial firms either beat or exceeded their expectations."

Bear NIM case: the curve has turned against banks. The bear case got sharper this week. Thomas' "18 or 19 basis points" between 2-year and 10-year yields means the main engine of bank profits, borrowing short and lending long, is barely turning. Last week Whalen said deposit costs were already rising at Wells Fargo and at least two regionals. Put the two together and margins get squeezed from both ends. Elconin's "where's the loan demand?" argument says volume won't make up the difference. Lee adds a risk that sits outside the usual margin math: bank credit lines to private-credit funds could be drawn suddenly if those funds hit trouble. And the AI scare, if it turns out to be even partly right, would wear away the cheap, loyal deposits the bull case depends on.

The bull case says banks earn more on loans before they pay more on deposits. The bear case says that only works if long-term rates are well above short-term ones, and right now they barely are.

Where we come out. The bear case won the week on stock prices, but on evidence it is still mostly pundit opinion. The only operator data point, Robbins' doubled pipeline and sticky deposits, leaned bullish. The flat curve is real and is the single biggest risk to Q3 and Q4 margin outlooks. Bank of America, by Thomas' reasoning, is the most exposed of the big four. Q3 results will tell us whether the gloom is priced in.

Stocks in play

JPMorgan (JPM).

  • Bull: Dimon's inflation view points to rates staying high, and JPMorgan has the cheapest, stickiest deposits of the group. Co-president Doug Petno's guidance from two weeks ago, investment-banking and trading fees up "mid to high teens" in Q3, has not been withdrawn. The $1 billion Michigan manufacturing commitment is a small sign of confidence in business lending.
  • Bear: the stock is down about 7% since Meta's Muse launch, per Power Lunch, and it has the most to lose if the market starts to believe AI threatens big-bank customer relationships. With the curve near flat, even JPMorgan's margin won't grow much.
  • Next catalyst: Q3 results on October 13. Watch whether management changes its 2026 or early 2027 interest-income outlook given the flatter curve.

Bank of America (BAC).

  • Bull: Moynihan has been, in CNBC's words, "consistent on this for a long time" about consumer strength, and card and deposit balances "continue to show growth." The stock has already absorbed last month's fee warning, so expectations are low.
  • Bear: Thomas named Bank of America as the big bank with "disproportionate exposure" to a flat curve, because so much of its business is consumer lending funded by short-term deposits. It also fell 3% on the AI scare day, per The Rundown, and Robbins' point that money-center banks have "a larger dependency upon the consumer" applies here most of all.
  • Next catalyst: Q3 results on October 14. The margin outlook for Q4, and how deposit costs moved after the hike, are the numbers that matter most.

Wells Fargo (WFC).

  • Bull: Wells has the least reliance on deal fees of the four, and CEO Charlie Scharf is among "all the bankers" CNBC's hosts said are pointing to consumer strength.
  • Bear: Wells is down "about double digits" since Muse launched, per Power Lunch, the worst of the big four on that measure. Last week Whalen named it first among the banks whose deposit rates are rising. For a bank that depends mainly on its lending margin, a flat curve plus rising deposit costs is the most direct hit in the group. No podcast this week offered new Wells-specific numbers.
  • Next catalyst: Q3 results on October 13. Look for whether rising deposit costs were a one-quarter blip or the start of a trend.

Citigroup (C).

  • Bull: Citi has held up best of the four since Muse launched, "down about 5 percent," per Power Lunch. Its cheap valuation leaves room for a positive surprise if trading and banking fees hold up.
  • Bear: Thomas' point about the OpenAI IPO being "pushed... out into 2027" hurts the banks that lean on capital markets, and Citi is one of them. RenMac's note that investment banks have gotten "oversold" shows how quickly sentiment on that business has turned. No podcast this week discussed Citi's own results or targets.
  • Next catalyst: Q3 results on October 13. The question is whether Citi's profitability keeps improving despite a slower end to the quarter for deals.

Read-throughs

  • Super-regionals (USB, PNC, TFC). No podcast this week discussed U.S. Bancorp, PNC or Truist by name. The read-across from Valley's Robbins is direct, though, and mostly encouraging. Like Valley, these three earn much of their money from business customers. A doubled commercial pipeline and "strong" commercial deposit balances at a regional peer is a good sign for their loan growth. The warning signs point the other way: Power Lunch said regional banks have lost "about 8 percent" since Muse launched, and on Forward Guidance: "The Bond Market Pain Isn't Over | Weekly Roundup" (September 25), the hosts (Pundit) noted the regional bank ETF, KRE, "literally is sitting on the 200," meaning its 200-day moving average, a widely watched trend line. They added that Fed Chair Kevin Warsh "specifically called out regional lending," so "the stressors that you're going to see are going to manifest in regional lending. Maybe it's CRE. Maybe it's private credit." USB and PNC report October 15, Truist October 16.
  • Deposit competition. This week's debate is about whether deposits are loyal or lazy. Robbins says loyal: customers pick banks "based on relationships, based on location," and small checking balances rarely chase rates. The AI bears say an AI agent could change that by doing the rate-shopping for customers. Meanwhile, rates elsewhere are high and rising: The Rundown cited the 30-year fixed mortgage rate at "7.12%," the highest "since May of 2024," with the 10-year Treasury "at 4.966%." When savers can get close to 5% in Treasuries, banks have to pay up to keep larger, rate-sensitive balances.
  • Capital-markets fees (deal pipeline, trading). The tone softened. Thomas said OpenAI pushing its IPO to 2027 has investors worried "we're not going to get as much capital markets activity for these big banks as we thought." The Squawk on the Street strategist noted the banks have given up "how much of a lead" they had this year, after "the second quarter was kind of a wild level of everything going right... the capital markets levered banks in particular." In other words, Q2 set a very high bar. On the positive side, Dimon talked up "the power of our capital markets," saying "you can raise tons of capital in debt markets and preferred markets." Last week's Moynihan comment that the deal "pipeline... is full" still stands. The question is when it unclogs.
  • CRE and consumer credit. No podcast offered new commercial real estate loss figures for the seven banks. On consumers, the operator view is that the split is widening: Robbins described customers with "a lot of money" still spending, and others with "significant credit card debt." Housing activity is weak, with Robbins saying "the price of the homes have come down" and there's "very little, if any, refinance activity." Forward Guidance flagged CRE as one place regional stress could show up. Lee's warning about bank credit lines to private-credit funds is a newer risk category to watch on the Q3 calls.
  • Capital and regulation. The ETF Spotlight strategist expects new capital rules, released in March, to take effect at year-end and free up "capital on bank balance sheets." That would support buybacks in 2027 even if margins are flat.

What changed vs last week

Last Friday's issue, "A quiet week, one loud warning: deposit costs rising," centered on Chris Whalen's comment that Wells Fargo, Fifth Third and Huntington were seeing deposit rates rise "after six quarters of falling interest expense," plus a fuller reading of Brian Moynihan's Bloomberg interview.

What's new: two new drivers. The first is the flattening yield curve, with the 2-year/10-year gap at "18 or 19 basis points" per Thomas. Last week the bear case was about rising deposit costs. This week it is about rising deposit costs and long-term rates that are not keeping up. The second is the AI scare: Meta's Muse launch has become a real force in bank share prices, with JPMorgan down about 7% and Wells Fargo down "about double digits" since the launch, per Power Lunch.

What changed: the loan-demand debate now has an operator data point. Last week Henry Peabody of GMO argued that private credit was taking the best corporate loans from banks. This week Robbins reported a commercial pipeline that "doubled" to $4 billion. That is one regional bank, not the industry, but it is the first hard number on loan demand since the hike.

What was contradicted: the pundit claim that loan demand can't grow while rates rise. Elconin said "Absolutely not"; Robbins said his business borrowers are "investing in equipment, investing in working capital." Both can't be right across the industry, and Q3 loan growth will settle it.

What carried over: the "higher for longer" view got another backer. Last week Moynihan's economists forecast three hikes and inflation above target into 2027-28. This week Dimon said he does not "bet on" inflation coming down. Still missing: any hard numbers from the big banks themselves. Those arrive starting October 13.