Bank Stocks After the Fed’s Surprise Rate Hike

The Federal Reserve raised its benchmark rate by a quarter point on September 16, 2026, to a range of 3.75% to 4%, its first increase since 2023. Sixteen of the eighteen officials at that meeting now project at least one more hike before year end. If you were holding bank stocks for the rate-cut story that dominated headlines earlier this year, that story is on hold. The mechanics that decide what a bank earns just flipped back the other way.

I keep coming back to bank stocks for a simple reason: they are cheap most of the time and they pay well. The market treats every one of them as a coiled disaster regardless of the individual balance sheet. That skepticism is annoying to own through and useful to buy into. What changed this month is not the skepticism. It is the direction rates are moving, and that direction decides which mechanic in a bank’s income statement does the work.

My thesis: a fresh hiking cycle, even a small one, helps banks with cheap, sticky deposit bases and punishes the ones that spent the last two years paying up for deposits they gathered when everyone assumed cuts were coming, not another hike.

Net interest margin runs on who reprices first

A bank borrows money, mostly from depositors, then lends that money out at a higher rate. The gap between what it earns on assets and what it pays on funding, as a share of those assets, is net interest margin. It is the main profit engine for most banks. Small moves in it swing earnings a lot, because a bank’s balance sheet is built to amplify that spread.

When the Fed cut rates earlier in this cycle, loan yields on new and repricing loans drifted down while a lot of deposit costs, especially term deposits opened during the prior high-rate period, took longer to follow. That squeezed margin at banks slow to cut what they paid savers. Now that the Fed has reversed into a hike, the same lag works in reverse: loan yields on new originations move up quickly, and a bank that can hold the line on deposit rates a little longer captures the difference. The banks that built the cheapest, most patient deposit franchises during the cutting phase are the ones positioned to benefit first from the hike.

Deposit betas decide who keeps the difference

Deposit beta is the share of a rate change a bank passes through to its depositors. A beta near 1 means a bank raises what it pays savers almost dollar for dollar with the Fed. A beta near zero means it barely moves its own rates and keeps the difference as margin. Retail-heavy banks with a lot of low-cost checking and savings balances tend to run low betas, because that money is sticky. Banks that lean on brokered deposits, CDs or high-yield online savings accounts to fund growth tend to run high betas, because that money is rate-shopped and will leave for a better rate elsewhere.

A high-beta bank gets squeezed fastest when rates fall, since it has to cut what it pays quickly to protect margin. It gets squeezed again on the way back up if its funding costs were already tracking the market closely and jump right back with the Fed. A low-beta bank is closer to a one-way beneficiary of a hike: its funding cost barely moved down during the cuts and barely needs to move up now, while its asset yields reprice higher regardless.

The last hiking cycle, in 2022 and into 2023, is the reference point here. Betas across the industry ran unusually low for an unusually long stretch back then, which is a big part of why margins expanded as much as they did before that cycle’s later strains caught up with a handful of banks. Whether this smaller, more isolated hike repeats that pattern, or whether depositors who got a hard lesson in 2023 about shopping for yield reprice faster this time, is an open question. I would not assume the old betas hold just because the direction of rates looks familiar.

History does not repeat exactly here. It rhymes enough to matter, and not enough to trade on blindly.

Rising rates cut the other way on bond portfolios

A hike is not free for banks either. Every bank holds a portfolio of Treasuries and mortgage-backed securities funded against its deposits, and when rates climb, the market value of those older, lower-yielding securities falls. Banks that classify that portfolio as available-for-sale run that decline straight through their capital ratios, even before a single bond is sold. That mechanic, an unhedged bond book losing value faster than a bank’s capital could absorb it, is what turned Silicon Valley Bank’s balance sheet into a solvency problem in March 2023, once depositors asked for their money back faster than the securities could be sold without locking in the loss.

That is the risk a hike reintroduces to this sector, and it deserves more than a footnote. A bank with a short-duration securities book and a broad, diversified deposit base barely notices a quarter-point move in the value of what it holds. A bank that stretched for yield with a longer-duration book during the years rates sat near zero, and that leans on a smaller number of large, flighty depositors, notices right away. That is the second reason deposit franchise quality matters more than the headline margin math this quarter. The institutions with sticky, low-cost deposits also tend to be the ones that never needed to reach for duration to make a securities book pay for itself in the first place.

The yield curve stopped fighting the sector

The other piece is the shape of the curve. As of September 21, the two-year Treasury yielded 4.76% and the ten-year yielded 4.96%, a spread of about 20 basis points, according to the St. Louis Fed’s own data series. That is narrow, but it is positive, which matters because banks borrow short and lend long. A bank earns more on that maturity transformation when the curve is upward sloping than when it is flat or inverted, which it was through much of 2022 to 2024.

A 20-basis-point spread is not generous. It is a lot better than zero or negative, and it means new long-duration lending, mortgages, commercial real estate, long fixed-rate business loans, gets written at a real premium over short-term funding costs again, rather than at a spread that barely covers overhead. If the curve steepens further as the hiking cycle proceeds, that premium widens further, and banks with more of their book in longer-duration assets benefit more from each additional basis point of steepening than banks that stayed short.

JPMorgan, Bank of America and SoFi sit differently now

I have written about Bank of America’s rate exposure and about why JPMorgan trades near a record using pieces of this same framework, and this month’s reversal changes the weight I would put on each. A bank like Bank of America, which built a large base of retail checking and savings deposits over decades, tends to run a lower beta than a bank still growing that base through promotional rates, which is closer to the position SoFi has been in while it builds out its deposit franchise under a bank charter. A hike helps the first kind of bank faster and more cleanly than it helps the second, at least until the second has had a full cycle to prove its deposit costs do not just track the market up and down in lockstep.

None of that makes the second kind of bank a bad business. It changes which one earns the margin tailwind first, and margin timing is most of what separates a good quarter from a mediocre one in this sector. A bank building a deposit franchise from scratch is playing a longer game than a single rate cycle, and a hike arriving before that franchise has matured is simply an awkward moment in a story that was never going to be finished this year anyway.

Benchmark	Value	As of
Fed funds target range	3.75%-4.00%	September 16, 2026
2-year Treasury yield	4.76%	September 21, 2026
10-year Treasury yield	4.96%	September 21, 2026
2s10s spread	about +20 bps	September 21, 2026

The deposit-cost number that settles this

I do not think the next Fed meeting is the number to watch here. It is deposit cost data in third-quarter results, when most large banks report in mid-October. If deposit costs at retail-funded banks keep drifting down even as the Fed hikes, that confirms the pricing discipline built up during the cutting phase carried through, and the low-beta thesis holds. If deposit costs snap back up nearly as fast as the Fed’s own rate did, betas are more symmetric than management presentations tend to admit, and the margin tailwind I am describing here arrives smaller and later than the stocks may already be pricing.

The case against all of this is straightforward: if the projected second hike does not happen and the Fed pivots back toward cuts in 2027, the entire hiking-helps-margins argument reverses again, and the deposit-heavy banks I would favor in a hiking quarter become the wrong ones to hold relative to banks with shorter-duration, more rate-sensitive asset books. Eighteen months is a long time in a cycle that has already reversed once this year.

Analysis and opinion only, not investment advice. Figures come from the Federal Reserve’s own policy page (federalreserve.gov) and CNBC’s coverage of the September 16 decision (cnbc.com); Treasury yields are from the St. Louis Fed’s FRED database and were checked on September 23, 2026.